Ulsab engineer report complete

Engineering Services, Inc.

Porsche Engineering Services, Inc.

ULSAB Program Phase 2

Final Report

to the

Ultra Light Steel Auto Body

Consortium


Ultra Light Steel Auto Body

Member Companies

AceraliaAK SteelBethlehemBHP SteelBritish SteelCockerill SambreCSNDofascoHoogovensInlandKawasaki SteelKobeKrakatauKrupp HoeschLTV SteelNational SteelNippon SteelNKK

POSCOPreussagRouge SteelSIDERARSIDMARSOLLACSSABStelcoSumitomoTataThyssenUS Steel GroupUSIMINASVSZVOEST-ALPINEWCIWeirton

Table of Contents - Page 1


ULSAB Final Report Table of Contents

Preface

1. Executive Summary

2. Phase 2 Introduction

2.1. Phase 2 Program Goal

2.2. Phase 2 Design and Analysis

2.3. Demonstration Hardware (DH)

2.4. Scope of Work

2.5. Materials

2.6. Testing of Test Unit

2.7. Phase 2 Program Timing

3. ULSAB Phase 2 Package

3.1. General Approach

3.2. Package Definition

3.2.1. Vehicle Concept Type

3.2.2. Exterior Dimensions

3.2.3. Interior Dimensions

3.2.4. Main Component Definition

3.2.5. Underfloor Clearance

3.2.6. Seating Position

3.2.7. Visibility Study3.2.7.1. Horizontal and Vertical Obstruction

3.2.7.2. A-Pillar Obstruction

3.2.8. Gear Shift Lever Position

3.2.9. Pedal Position

3.2.10. Bumper Height Definition

3.3. Package Drawings



4. Styling

4.1. Approach

4.2. 2-D Styling Phase

4.2.1. Sketching

4.2.2. Clinic

4.2.3. Electronic Paint

4.2.4. Styling Theme Selection

4.3. 3-D Styling Model

4.3.1. Surface Release

4.4. Rendering

5. Design and Engineering

5.1. Phase 2 Design and Engineering Approach

5.2. Design and Engineering Process

5.3. ULSAB Phase 2 Design Description

5.3.1. Parts List – Demonstration Hardware

5.3.2. ULSAB Structure Mass

5.3.3. ULSAB Demonstration Hardware Mass

5.3.4. Mass of Brackets and Reinforcements – Phase 2

5.3.5. ULSAB Structure Mass Comparison Phase 1 – Phase 2

5.3.6. DH Part Manufacturing Processes

5.3.7. Material Grades

5.3.8. Material Thickness

5.4. Detail Design

5.4.1. Weld Flange Standards

5.4.1.1. Weld Flanges for Spot or Laser Welding

5.4.1.2. Scalloped Spot Weld Flanges

5.4.1.3. Locator, Tooling and Electrophoresis Holes

5.4.2. Design Refinement



6. CAE Analysis Results

6.1. Selected Tests for CAE

6.2. Static and Dynamic Stiffness

6.2.1. Torsional Stiffness

6.2.2. Bending Stiffness

6.2.3. Normal Modes

6.3. Crash Analysis

6.3.1. AMS Offset Crash

6.3.2. NCAP 100% Frontal Crash

6.3.3. Rear Crash

6.3.4. Side Impact Analysis

6.3.5. Roof Crush (FMVSS 216)

6.4. CAE Analysis Summary

7. Materials and Processes

7.1. Material Selection

7.1.1. Material Selection Process

7.1.2. Definition of Strength Levels

7.1.3. Supplier Selection

7.2. Material Specifications

7.2.1. General Specifications

7.2.2. Material Classes7.2.2.1. Mild Steel Definition

7.2.2.2. High Strength Steel Definition

7.2.2.3. Ultra High Strength Steel Definition

7.2.2.4. Sandwich Material Definition

7.2.3. Material Documentation

7.3. Tailor Welded Blanks

7.3.1. Selection of Welding Process

7.3.2. Weld Line Layout

7.3.3. Production Blank Layout



7.4. Hydroforming

7.4.1. General Process Description

7.4.2. Benefit for the Project

7.4.3. Forming Simulation (Review)

7.4.4. Tube Manufacturing

7.4.5. Process Steps for Rail Side Roof

7.4.6. Results

7.5. Hydromechanical Sheet Forming



7.5.3. Process Limitations

7.5.4. Results

8. Parts Manufacturing

8.1. Supplier Selection

8.2. Simultaneous Engineering

8.3. Part Manufacturing Feasibility

8.4. Quality Criteria

9. DH Build

9.1. Introduction

9.2. Joining Technologies

9.2.1. Laser Welding

9.2.2. Spot Welding

9.2.3. Active Gas Metal Arc Welding (MAG)

9.2.4. Adhesive Bonding

9.3. Flexible Modular Assembly Fixture System

9.4. Design of Assembly Fixtures

9.5. DH Build

9.5.1. Assembly Team

9.5.2. Build of the Test Unit

9.5.3. Build of DH #2 to DH #13



9.6. Quality

9.6.1. Body Quality Control Team

9.6.2. Quality Control Measurements of DHs

9.7. Conclusion

10. Testing and Results

10.1. Scope of Work

10.2. Targets

10.3. Static Rigidity

10.3.1. Test Setup10.3.1.1. General

10.3.1.2. Static Torsion

10.3.1.3. Static Bending

10.3.2. Results10.3.2.1. Static Torsion


10.4. Modal Analysis

10.4.1. Test Setup

10.4.2. Results

10.5. Masses in Test Configuration

10.6. Summary

11. Economic Analysis

11.1. Introduction

11.2. The Process of Cost Estimation

11.2.1. Overview

11.2.2. Cost Model Algorithm Development

11.2.3. General Inputs

11.2.4. Fabrication Input

11.2.5. Assembly Input

11.3. Cost Model Description



11.4. ULSAB Cost Results

11.4.1. Overall Cost Results

11.4.2. Cost Breakdown for Fabrication

11.4.3. Cost Breakdown for Assembly

11.4.4. Cost Analysis for New Technologies and Materials

11.4.5. Sensitivity Analysis

11.5. Body Structure – Comparative Study

11.5.1. Overview

11.5.2. Assumptions

11.5.3. Overall Results

11.6. Conclusion

NOTE: The cost models may be found on the Porsche ULSAB

Phase 2 CD ROM Version 1.0.2.

12. Summary of Phase 2 Results

ULSAB Final Report Appendix Table of Contents

NOTE:

The following information is located on thePorsche ULSAB Phase 2 CD ROM Version 1.0.2.

1. Parts Book

1.1. Exploded View1.2. Index – Parts Book Sheets1.3. Parts Book Sheets1.4. Index – Parts book, Brackets & Reinforcements1.5. Parts Book Sheets – Brackets & Reinforcements



2. Part Drawings

2.1. Exploded View2.2. Parts List – Sorted by Part Number2.3. Parts List – Sorted by Material Grade2.4. Part Drawings

3. Typical Sections

3.1. Overview Illustration3.2. Index – Typical Sections3.3. Typical Section Sheets

4. Assembly

4.1. Assembly Tree4.2. Index –Weld Assemblies4.3. Weld Assembly Drawings4.4. Assembly Sequence Illustrations4.5. Index – Bolted and / or Bonded Assemblies4.6. Assembly Drawings, Bolted and / or Bonded Parts4.7. Assembly Illustrations – Bolted and / or Bonded Parts

5. Package Drawings

5.1. Side View5.2. Plan View5.3. Front & Rear View


6.1. Assembly System Data6.2. Stamping Process Sheets

Preface - Page 1


Preface

In 1994, the steel industry, through the Ultra Light Steel Auto Body Consortium(ULSAB), commissioned Porsche Engineering Services, Inc. (PES) to conduct aconcept phase (Phase 1) of the ULSAB project to determine if a substantially lightersteel body structure could be designed.

In September 1995, worldwide auto industry attention was focused on the studywhen the results of Phase 1 were announced. The results also affected the growthof the ULSAB Consortium to 35 member steel companies, representing 18 nationsworldwide.

Encouraged by the results of Phase 1, the ULSAB Consortium once againcommissioned PES to continue with Phase 2, the validation of the Phase 1concepts, culminating in the build of the demonstration hardware. Phase 2 provedthat the weight reduction, predicted in Phase 1, could be achieved. The use of highstrength steels, tailor welded blanks, hydroforming and laser welding in assemblywere particular challenges to overcome in Phase 2. ULSAB Consortium memberscommitted themselves to supplying all steel materials, as well as the tailor weldedblanks and raw materials for hydroforming, for all parts to be manufactured.

The focus of Phase 2 was the same as in Phase 1, i.e., weight reduction withoutcompromising safety or structural performance. Without altering the aggressivetargets for mass and structural performance, the safety requirements wereincreased in Phase 2 in response to growing industry and government concern forincreased auto safety. It was imperative to keep up with safety requirementchanges that occurred during the course of the program, which ran from spring1994 to spring 1998. As a result, it was necessary to analyze the ULSAB structurefor offset crash behavior. With this new challenge, and valuable input gathered indiscussions with OEMs during the presentation of Phase 1 findings, PES and theULSAB Consortium commenced Phase 2.

Preface - Page 2


Phase 2 ended in Spring 1998 with the debut of the ULSAB demonstration hardwareand will prove the Phase 1 concept to be not only feasible, but that performancetargets will be exceeded by 60% for torsional rigidity, 48% for bending rigidity and50% for the normal mode frequency. Relative to the benchmark average, massreduction remained at 25%, while crash analysis showed excellent results for theselected crash analysis events, including the offset crash.

As a result of Phase 2, the use of high strength steels in the ULSAB demonstrationhardware structure has now increased to 90% relative to its mass. The trendtoward using high strength steel and new technologies to reduce body structuremass while improving safety, can be seen already in recently launched cars. Thenew Porsche Boxster, for example, uses 30% high strength steel, as well as tailoredblanking, hydroforming and laser welding in assembly.

Cost analysis in Phase 1 was conducted by IBIS Associates under contract to theULSAB Consortium. In Phase 2, a more detailed cost analysis study wasconducted, under the supervision of PES with the support of ULSAB consortiummember companies. With the detailed information provided with the conceptvalidation in Phase 2, a new cost model was created and the cost to produce theULSAB structure was analyzed. The results show that it is possible to reduce themass of body structures without cost penalty.



Chapter 1 - Page 1



Ultra Light Steel Auto Body (ULSAB) Phase 2

Introduction

On behalf of an international Consortium of 35 of the world’s leading sheet-steelproducers from 18 countries, Porsche Engineering Services, Inc. (PES) in Troy,Michigan, was responsible for the program management, design, engineering, andthe building of the demonstration hardware (DH). In addition, PES conducted theeconomic analysis study for the Ultra Light Steel Auto Body (ULSAB) program.

Program Goal

The goal of the ULSAB program was to develop a light-weight body structure designthat is predominantly steel. This goal included:

• Providing a significant mass reduction based on a future reference vehicle• Meeting functional and structural performance targets• Providing concepts that will be applicable for future car programs

Program Structure

In order to achieve the above-mentioned goals the program was structured in threephases:

• Phase 1 Concept Development (paper study)• Phase 2 Concept Validation (build of demonstration hardware)• Phase 3 Vehicle Feasibility (total vehicle prototype assembly and

evaluation)

Chapter 1 - Page 2


Phase 1 – Concept

In September 1995, the results of Phase 1 were published. In this phase, theULSAB program concentrated on developing design concepts for light-weight bodystructures and validating crashworthiness. Based on benchmarking data, theperformance of a future reference vehicle was predicted and the structuralperformance targets for the ULSAB structure, excluding doors, rear deck lid, hoodand front fenders were established. Because the ULSAB program focuses on massreduction, a much more aggressive target was set for mass than for the otherstructural performance targets. These targets were:

For the concept validation, the following crash analysis was performed in Phase 1:

• NCAP, 100% frontal crash at 35 mph• Rear moving barrier crash at 35 mph (FMVSS 301)• EEVC, side impact crash at 50 km/h (with rigid barrier)• Roof crush (FMVSS 216)

The analytical results of Phase 1 were:

ULSAB Future Reference Performance Targets* Vehicle Prediction Mass 200 kg 250 kg

Static torsional rigidity 13000 Nm/deg 13000 Nm/deg

Static bending rigidity 12200 N/mm 12200 N/mm

First body structure mode 40 Hz 40 Hz * All targets were set for body structure with glass, except the target for mass

m

m

m

[

Performance Phase 1 Results*

Mass 205 kg

Static torsional rigidity 19056 Nm/deg

Static bending rigidity 12529 N/mm

First body structure mode 51 Hz

*Structural performance results were calculated with glass; the mass excludes glass

Chapter 1 - Page 3


With the exception of mass, the results exceeded the targets. Mass was calculated at205 kg and slightly above the aggressive target of 200 kg.

An independent cost study indicated that, based on a North Americanmanufacturing scenerio, the Phase 1 concept could cost less to produce thancomparable current vehicle structures. This result, based on the relatively low levelof detail of the ULSAB Phase 1 concept, indicated that a light weight structure couldmake substantial use of high strength steel, tailor welded blanks, laser welding inassembly, and hydroforming, and end up in the cost range of structures of similarsize using a more conventional approach at a higher mass.

Phase 2 - Validation

The Phase 1 design concept and its structural and crash performance resultshaving had a relatively low mass, provided an excellent foundation for Phase 2 ofthe ULSAB program. Based on the success of this Phase 1 paper study, and thepositive recognition by OEMs around the world, the ULSAB Consortiumcommissioned PES to undertake Phase 2 starting in November 1995.

The overall goal of Phase 2 was the validation of Phase 1 results, culminating in thebuild of the ULSAB demonstration hardware structure. The tasks andresponsibilities of Phase 2 for PES, besides the program management, were tomanage the necessary detail design, engineering, crash analysis, materialselection, design optimization for manufacturing, supplier selection for parts and toassemble, test and deliver the demonstration hardware to the ULSAB Consortium.In addition, PES was responsible for a detailed cost analysis based on the Phase 2detailed design.

Chapter 1 - Page 4


Crash Analysis

During the course of the ULSAB program after the start in Spring 1994, the publicdemanded increased vehicle safety, and governments reacted with newrequirements for crashworthiness. Therefore, the decision was made prior to thebeginning of Phase 2, to analyze and to design the ULSAB structure for offsetcrash. This would enhance the credibility of the results. The AMS (Auto MotorSport) 50% offset frontal crash at 55 km/h was considered the most severe test atthat time and would represent the structural requirements an offset crash demands.This test was then added to the Phase 1 previously selected crash analysis events.For side impact crash analysis, a deformable barrier was used instead of the rigidbarrier as used in Phase 1.

The following crash analysis was performed in Phase 2:

• AMS, 50% frontal offset crash at 55 km/h• NCAP, 100% frontal crash at 35 mph (FMVSS 208)• Side impact crash at 50 km/h (96/27 EG, with deformable barrier)• Rear moving barrier crash at 35 mph (FMVSS 301)• Roof crush (FMVSS 216)

All crash calculations indicate excellent crash behavior of the ULSAB structure,even at speeds that exceed federal requirements. The front and rear impacts wererun at 5 mph above the required limit, meaning 36% more energy had to beabsorbed in the frontal impact. The offset crash also confirmed the overall integrityof the structure. The roof crush analysis validated that the federal standardrequirement was met, partialy due to the hydroformed side roof rail concept design.

Package

At the start of Phase 2, as a result of various discussions with OEMs during thepresentation of Phase 1 results, the ULSAB package was re-examined. In order tomake the results of Phase 2 more credible, the decision was made not to considersecondary mass savings. This resulted in significant changes in several areas ofthe body structure.

Chapter 1 - Page 5


The relatively small engine specified in Phase 1 was replaced by an average size3-liter V6, necessitating a complete redesign of the front-end structure, including arevised front suspension layout and subframe design. The rear suspension alsowas revised and the rear rails redesigned accordingly. Essentially, the wholestructure was redesigned, from front to rear bumper, but it still maintained thestructure features as developed in Phase I, such as the side roof rail and thesmooth load flow concept of front and rear rails into the rocker.

Styling

Using the revised package and the adjusted body structure design, styling theULSAB was the next challenge. Styling became necessary to create the surfacesfor the body side outer panel with its integrated exposed rear quarter panel, thewindshield, the backlight and the roof panel. The styling concept for thegreenhouse had to consider, in order to integrate, the side roof rail, as well as theoverlapping upper door frame concept. This door concept was chosen mainly forcosmetic reasons; to cover the visible weld seams, in the upper door opening areaof the body side outer panel which were caused by the tailor welded blank design ofthe body side outer panel. For the overall styling approach, the decision was madeto create a neutral, not too futuristic or radical, more conservative styling.

Styling was the first major milestone in Phase 2 and was performed entirely bycomputer-aided styling (CAS).

Design and Engineering

After the exterior styling was created, the package was then optimized and thedesign modified accordingly. The implication of any design change was assessedby modifying the Phase 1 static analysis model. Design changes resulting as anoutcome of the analysis were then incorporated into the styling and the package.With the performance targets met, styling and the Phase 2 package were frozen,and with a more detailed Phase 2 design, a new shell model for the structuralperformance analysis was created. Static analysis was then used to optimize the

Chapter 1 - Page 6


Phase 2 design until the requirements were met and new crash analysis modelswere built. In the process of design optimization, which included material grade andthickness selection, both static analysis and crash analysis were performed withconstantly updated models until the targets were met.

Throughout this process, simultaneous engineering provided input from the tool andpart suppliers, as well as from steel manufacturers, to ensure the manufacturingfeasibility of the designed parts. As a result of the simultaneous engineeringprocess, only minor design and tool changes were needed after the drawings werereleased. When the first part set was completed, a workhorse (test unit) was built.The validation of the test unit lead to further part optimization and, finally, to thebuild of demonstration structures.

Suppliers

At the start of the detail design process in Phase 2, suppliers for stamped andhydroformed parts were selected in order to be included in the simultaneousengineering process. Among the selection criteria were quality, experience, skillsand location. Supplier flexibility and their willingness to explore new manufacturingmethods, utilizing material grades rarely used in these applications and to “push theenvelope” in the application of tailor welded blanks or in hydroforming technologies,were as important in the selection process as their cost competitiveness.

Steel Materials

••••• Steel Grades

Perhaps the most important factor in meeting the targets for mass andcrash performance is high strength steel. More than 90% of the ULSABstructure utilizes high strength and ultra high strength steel. High strengthsteels are applied where the design is driven by crash and strengthrequirements. Ultra high strength steels with yield strength of more than550 MPa are used for parts to provide additional strength for front and sideimpact. High strength and ultra high strength steel material specificationsrange from 210 to 800 MPa yield strength with a thickness range from

Chapter 1 - Page 7


0.65 to 2 mm. With the restriction of lower elongation, different formingcharacteristics and greater spring back of high strength steels, materialsupplier support combined with forming simulations were important factorsin meeting the challenges for the development of manufacturable partdesigns.

••••• Steel Sandwich Material

The use of steel sandwich material has contributed to considerable masssavings. The sandwich material is made with a thermoplastic(polypropylene) core, with a thickness of 0.65 mm and is layered betweentwo thin steel skins, each with a thickness of 0.14 mm and yield strength of240 MPa for the spare tire tub and 140 MPa for the dash panel insert. Thesteel sandwich shares many of the same processing possibilities of sheetsteel, such as deep drawing, shear cutting, drilling, bonding, and riveting.However, it cannot be welded. Parts manufactured from steel sandwichmaterial can be up to 50% lighter than those made of sheet steel withsimilar dimensional and functional characteristics. The spare tire tubmade of steel sandwich material is a pre-painted module that is pre-assembled with the spare tire and repair tools. The module is dropped intoplace and bonded to the structure during the final assembly of the vehicle.

Another application of sandwich material is the dash panel insert, which isbolted and bonded into the body structure, during final vehicle assembly.

Tailor W elded Blanks

Tailor welded blanks enable the engineers to accurately locate the steel within thepart precisely where its attributes are most needed, while at the same time allowingfor the elimination of mass that does not contribute to performance. Other benefitsof tailor welded blanks include the use of fewer parts, dies and joining operations,as well as improved dimensional accuracy through the reduction of assembly steps.Nearly half (45%) of the ULSAB demonstration hardware mass consists of partsmanufactured using laser welded tailored blanks.

Chapter 1 - Page 8


The best example of tailor welded blank usage is the body side outer panel. Itemploys a fully laser welded tailored blank with different thicknesses and grades ofhigh strength steel. Careful placement of the seams in the tailor welded blank iscritical in order to minimize mass and facilitate forming. This consideration wasespecially important in the body side outer panel because of its complexity and size,its use of high strength steels and the integration of the rear quarter panel with itsClass A surface requirement. Mass reduction and the elimination of reinforcementswere key goals in the development of this one-piece design. The consolidation ofparts reduced mass and assembly steps.

Hydroforming

••••• Tubular Hydroforming

The use of hydroforming should be considered as one of the mostsignificant manufacturing processes applied in the ULSAB program for partmanufacturing. The hydroformed side roof rail represents a significantstructural member in the ULSAB structure. The side roof rail distributesloads appearing in the structure during vehicle operation, and in the eventof an impact, distributes loads from the top of the A-pillar along the roofinto B and C-pillar and then into the rear of the structure. The hydroformedside roof rail reduces the total number of parts and optimizes availablepackage space. The raw material used to manufacture the side roof rail isa laser welded, high-strength steel tube 1 mm thick with an outsidediameter of 96 mm and a yield strength of 280 MPa. The design wasoptimized and analyzed for feasibility using forming simulation.

••••• Hydromechanical Sheet Forming

The use of hydromechanical sheet forming was chosen for the roof panelfor mass reduction reasons. This process provided the opportunity tomanufacture the roof panel at a thinner material thickness and still achievea work-hardening effect in the center area, where the degree of stretch isnormally minimal and an increased material thickness is needed to meetdent resistance requirements. With hydro-mechanical sheet forming, this

Chapter 1 - Page 9


work-hardening effect is achieved by using fluid pressure to pre-stretch theblanks in the opposite direction towards the punch. This plastic elongationcauses a work-hardening effect in the center area of the blank. In thesecond step, the punch forms the panel towards controlled fluid pressureand because there is no metal-to-metal contact on the outer part surface,excellent part quality is achieved. The ULSAB roof panel is manufacturedin 0.7 mm high strength steel with a yield strength of 210 MPa.

Tooling

All tools for stamped parts are “soft” tools made of materials such as kirksite andbuilt to production intent standards. Tools used for hydroforming are “hard” toolsmade of steel. In both cases, part manufacturing tolerances and quality standardswere the same as those used in high-volume production.

DH Assembly

••••• Joining Technologies

For the final assembly of the ULSAB structure, four types of joiningtechnologies were applied. Spot welding is used for joining the majority ofparts. Laser welding became necessary to join the hydroformed side roofrail to its mating parts. In addition, the rails in the front end structure arelaser welded for improved structural performance. Laser welding in bodystructure assembly is already being used in mass production by manyOEMs. The active gas metal arc welding (MAG) process, with itsdisadvantages, such as slow welding speed and relatively large heatimpact zones, was kept to a minimum and used only in locations with noweld access for spot or laser welding. Bonding is used to join the sandwichparts that cannot be spot or laser welded into the structure. For the joiningof the DH, about one-third fewer spot welds and significantly more laserwelding is employed than for conventional body structures.

Chapter 1 - Page 10


••••• Assembly Sequence

For the DH build, the assembly sequence uses two stage body sideframing. The assembly sequence includes underbody assembly, body sideassemblies, roof and rear panel assemblies. All DHs were built in a singlebuild sequence.

••••• Assembly Fixtures

To assemble the DH, a modular fixture system was used. The fixtureswere developed in a CAD system and the positions of locator holes werethen incorporated into the parts design.

DH Testing

Testing was performed on the ULSAB test unit structure to validate its structuralperformance and mass. Included were tests for static torsion rigidity, static bendingrigidity, modal analysis and mass in various configurations, including some bolt-onparts. Testing was performed at Porsche’s Research & Development Center inWeissach, Germany. Physical testing for crash was not part of the ULSAB programin Phase 2 and may be performed in a possible Phase 3, after the necessarycomponents are built and/or assembled into the ULSAB structure.

Economic Analysis

With the detailed information created in Phase 2 of the ULSAB program, the costsof parts and assembly of the body structure were analyzed. Under the managementof a PES’ team, and with support from the ULSAB Consortium members, aneconomic analysis group, comprising of analysts from the Massachusetts Instituteof Technology (MIT), IBIS Associates and Classic Design, a detailed cost modelwas constructed that includes all aspects of fabrication and assembly. This costmodel will enable the automotive OEMs to calculate ULSAB cost based on theirown manufacturing criteria. Considering that the focus of Phase 2 was on massreduction and not on cost savings, the result of this cost analysis is quiteremarkable. It confirms that significant mass reduction of the body structure, in

Chapter 1 - Page 11


comparision to the benchmark vehicle average mass, was achieved with the use ofsteel with no cost penalty.

Summary/Conclusion

Throughout Phase 2, timely execution of the program was critical. All partsdesigned and released to our suppliers and all tooling and assembly of the first testunit have been on schedule. With the data acquired from the validation of the firsttest unit and subsequent testing, parts were refined and design optimization wasperformed. Refined parts were then used to build the demonstration hardware.

Based on the testing of the demonstration hardware, the ULSAB structure shows

Performance* Target Results

Mass 200 kg 203 kg

Static torsional rigidity 13000 Nm/deg 20800 Nm/deg

Static bending rigidity 12200 N/mm 18100 N/mm

First body structure mode 40 Hz 60 Hz

*Structural performances are test results with glass. ULSAB structure mass without glass

[

m

m

m

the following structural performances:

Achieving these results in a timely manner could only be achieved by utilizing theteam approach that involved all parties in the early stages of the ULSAB program.A close working relationship with the ULSAB Consortium members and thecommitment of our suppliers and their enthusiasm for the program helped to meetthe challenge of manufacturing parts made of steel materials and combinations thathave not been commonly applied previously. This “pioneering spirit” was carried onby all members of the PES team, including designers and engineers. The ULSABprogram has explored the potential for mass reduction in the body structure usingsteel as the chosen material. State-of-the-art manufacturing and joiningtechnologies, such as laser welding in assembly and hydroforming as well ascommercially available materials, contributed to the success of the ULSABprogram. It proves that steel offers the potential for light weight vehicle designwhich contributes to the preservation of resources and the reduction of emissions.

Based on this experience, the steel industry should further intensify its dialogue andcooperation with OEMs to achieve their common goal of mass reduction oftomorrow’s vehicles, to protect the environment and to secure mobility of futuregenerations.



Chapter 2 - Page 1



2.1. Phase 2 Program Goal

The program goal of Phase 2 was the validation of Phase 1 results and the build ofdemonstration hardware.

Phase 1 was the concept phase and consisted of concept design and analysis. Thedesign was basic wire frame and surface data, without holes for drainage or locatorsfor assembly. The Phase 1 analysis, based on the design concept, was meshed inits basic form to reflect the surfaces of the structure.

2.2. Phase 2 Design and Analysis

The design in Phase 2 was a refinement of the Phase 1 design. It includes surfacedata, allowing for production of tools including principal location points (PLP) andholes for tooling, drainage and weld access. Additionally, refinement of the designfor manufacture and assembly (DFMA) was developed as the final designprogressed, with emphasis on mass production (more than 100,000 units per year).

The intention in Phase 2 was to continue the development of a “generic” structurethat takes into consideration manufacturing and assembly methods. With thedetailed design of the structural components, and assemblies, and with materialsselected, build specifications and the final assembly sequence were established.

Chapter 2 - Page 2


Computer Aided Engineering (CAE) continued in Phase 2 in conjunction with therefinement of the design. The analysis provided confirmation of the design as wellas structural and crash performance. The CAE analysis in Phase 2 included:

• Finite Element Model Modification• Structural Analysis consisting of:

w Massw Static Torsionw Static Bendingw Modal Analysis

Continuing development of crash simulation concentrates on:


All models were continuously updated to compare Phase 2 and Phase 1 results inorder to maintain the same performance standards.

2.3. Demonstration Hardware (DH)

The term “demonstration hardware” is used to emphasize that the body structure isnot a prototype but a legitimate representation of a production unit. Alldemonstration hardware components had to be fully tooled (soft tools for stampingand hard tools for hydroforming). All demonstration hardware was built in a singlebuild sequence. The completed structure had to be “clear-coat” painted forunrestricted view of the build and construction methods.

Chapter 2 - Page 3


2.4. Scope of Work

Porsche Engineering Services, Inc. in Troy, Michigan executed the program. TheDH build, testing and the CAE analysis was performed at the Porsche R & D Centerin Weissach, Germany. To achieve the targets for performance, timing and cost,the program responsibilities of PES included the following tasks:

• Program Management and Planning• Build Management for the Construction of the Demonstration Hardware• Build of Demonstration Hardware• Part Supplier/Manufacturer Evaluation and Selection• Component Structure Design and Engineering• CAE Analysis• Physical Testing of Test Unit• Economic Analysis Study• Final Program Report

2.5. Materials

The ULSAB Consortium member companies provided all material-specific datarequired to design, develop and construct the ULSAB body structure in Phase 2. Allmaterials used to manufacture parts for the DH build were provided to Porsche byULSAB Consortium member companies including the tailor welded blanks and rawmaterial (tubes) for the manufacturing of the hydroform side roof rail. In addition,the individual ULSAB Consortium member companies supported the program withdata related to material selection and tailor welded blank development, as well asforming simulation and circle grid analysis on selected parts in order to create afeasible part design.

Chapter 2 - Page 4


2.6. Testing of Test Unit

Physical testing was undertaken on the test unit to provide data and allowcorrelation of the CAE results with regard to:

• Mass• Static Torsion• Static Bending• Modal Analysis

Physical crash testing was not part of Phase 2. This could be executed in apossible Phase 3, with the necessary components, such as suspension, powertrain,and interior available.

2.7. Phase 2 Program Timing

Prior to the start of Phase 2 the program timing was established and the varioustasks assigned.

Based on this timeline the ULSAB Consortium established specific information releasedates to keep

Chapter 2 - Page 5


ULSAB Phase 2 Program Timeline

Task Name

Package RefinementStyling (CAS)Class A SurfacingDesign & Engineering

CAE AnalysisDesign ChangesCAE Analysis (Iteration 1)Design ChangesCAE Analysis (Iteration 2)Design ChangesCAE Analysis (Iteration 3)Release Long Lead ItemsToolingTest Unit BuildTestingDesign ChangesCAE ValidationTooling AdjustmentsDH Build

Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q21996 1997 1998

Economic Analysis


3. ULSAB Phase 2Package

Chapter 3 - Page 1


3. ULSAB Phase 2 Package

3.1. General Approach

Discussions with OEMs about Phase 1 findings provided valuable input andguidance for the more detailed Phase 2 package layout created at the start ofPhase 2. The Phase 2 package was defined as a modification of the Phase 1package without being too specific so the package findings could apply to morethan one body structure concept. The most important components, spacedefinitions and dimensions had to be considered by either defining them usingengineering judgment, or by using actual component dimensions. Furthermore,secondary mass savings were not considered in order to take a more conservativeand more credible approach. This is also reflected in component size and mass, aswell as in the crash mass used for the crash analysis.

3.2. Package Definition

The first step in the package phase was to define the vehicle concept type, exteriordimensions, interior dimensions and the main components. With these packagedefinitions, package drawings were revised and structural hard points defined.

3.2.1. Vehicle Concept Type

In Phase 2 the same concept type definition was used as in Phase 1, fivepassenger and four door midsize sedan.

Chapter 3 - Page 2


3.2.2. Exterior Dimensions

Ident.* Definition Measurements

W101 Tread - front 1560 mm

W102 Tread - rear 1545 mm

W103 Vehicle width 1819 mm

W117 Body width at SgRP - front 1767 mm

L101 Wheelbase 2700 mm

L103 Vehicle length 4714 mm

L104 Overhang - front 940 mm

L105 Overhang - rear 1074 mm

L114 Front wheel centerline to front SgRP 1447 mm

L123 Upper structure length 2631 mm

L125 Cowl point - X coordinate 2016 mm

L126 Front end length 1281 mm

L127 Rear wheel centerline - X coordinate 4295 mm

L128 Front wheel centerline - X coordinate 1595 mm

L129 Rear end length 654 mm

H101 Vehicle height 1453 mm

H106 Angle of approach 14°

H107 Angle of departure 15°

H114 Cowl point to ground 1001 mm

H121 Backlight slope angle 61°

H122 Windshield slope angle 59°

H124 Vision angle to windshield upper DLO 15°

H136 Zero Z plane to ground - front 112 mm

H138 Deck point to ground 1091 mm

H152 Exhaust system to ground 170 mm

H154 Fuel tank to ground 188 mm

H155 Spare tire well to ground 311 mm

*SAE J1100 Revised May 95

Chapter 3 - Page 3


3.2.3. Interior Dimensions


W3 Shoulder room - front 1512 mm

W4 Shoulder room - second 1522 mm

W5 Hip room - front 1544 mm

W6 Hip room - second 1544 mm

W7 Steering wheel center - Y coordinate 350 mm

W9 Steering wheel maximum outside diameter 370 mm

W20 SgRP - front - Y coordinate 350 mm

W25 SgRP - second - Y coordinate 335 mm

W27 Head clearance diagonal - driver 79 mm

W33 Head clearance diagonal - second 83 mm

W35 Head clearance lateral - driver 136 mm

W36 Head clearance lateral - second 132 mm

L7 Steering wheel torso clearance 418 mm

L11 Accelerator heel point to steering wheel center 412 mm

L13 Brake pedal knee clearance 573 mm

L30 Front of dash - X coordinate 1942 mm

L32 SgRP - second to rear wheel centerline 473 mm

L34 Effective leg room - front 1043 mm

L38 Head clearance to windshield garnish - driver 266 mm

L39 Head clearance to backlite garnish 21 mm

L40 Torso (back) angle - front 25°

L41 Torso (back) angle - second 25°

L42 Hip angle - front 93°

L43 Hip angle - second 86°

L44 Knee angle - front 118°

L45 Knee angle - second 88°

L46 Foot angle - front 78°

L47 Foot angle - second 113°

L50 SgRP couple distance 780 mm

L51 Effective leg room - second 894 mm

L52 Brake pedal to accelerator 48 mm

L53 SgRP - front to heel 832 mm


Chapter 3 - Page 4


3.2.3. Interior Dimensions (Cont’d)


H5 SgRP - front to ground 519 mm

H6 SgRP - front to windshield lower DLO 495 mm

H10 SgRP - second to ground 529 mm

H11 Entrance height - front 798 mm

H12 Entrance height - second 810 mm

H13 Steering wheel to centerline of thigh 67 mm

H14 Eyellipse to bottom of inside rearview mirror 40 mm

H17 Accelerator heel point to steering wheel center 645 mm

H18 Steering wheel angle 23°

H25 Belt height - front 446 mm

H26 Interior body height - front at zero Y plane 1011 mm

H27 Interior body height - front at SgRP Y plane 1220 mm

H29 Interior body height - second at SgRP Y plane 1033 mm

H30 SgRP - front to heel 245 mm

H31 SgRP - second to heel 303 mm

H32 Cushion deflection - front 49 mm

H33 Cushion deflection - second 66 mm

H35 Vertical head clearance - driver 75 mm

H36 Head clearance vertical - second 49 mm

H37 Headlining to roof panel - front 7 mm

H38 Headlining to roof panel - second 7 mm

H40 Steering wheel to accelerator heel point 468 mm


Chapter 3 - Page 5


3.2.3. Interior Dimensions (Cont’d)


H41 Minimum head clearance - driver 88 mm

H42 Minimum head clearance - second 21 mm

H49 Eyellipse to top of steering wheel 17 mm

H50 Upper-body opening to ground - front 1317 mm

H51 Upper-body opening to ground - second 1339 mm

H53 D-point - front to heel 137 mm

H54 D-point - center passenger - front to tunnel 105 mm

H55 D-point - center passenger - second to tunnel 43 mm

H56 D-point - front to floor 182 mm

H57 D-point - second to floor 72 mm

H60 D-point to heel point - second 19 mm

H61 Effective head room - front 1019 mm

H63 Effective head room - second 972 mm

H64 SgRP - front to windshield upper DLO 796 mm

H69 Exit height - second 743 mm

H70 SgRP - front - Z coordinate 631 mm

H71 SgRp - second - Z coordinate 641 mm

H75 Effective T-point head room - front 994 mm

H76 Effective T-point head room - second 932 mm

H77 Seatback height - front 868 mm

H78 Seatback height - second 781 mm

H94 Steering wheel to cushion - minimum 223 mm


Chapter 3 - Page 6


3.2.4. Main Component Definition

Component Description RemarksEngine V6 Average size

~3000 ccm

Engine Mounts Total of 3 2 on top of front rail

1 on subframe

Radiator Size .252 m With single fan

Single routing, Vol 2.8 catalytic converter

Exhaust System 1 catalytic converter, 21 ltr. muffler, LHS

1 muffler

Battery L x W x H 280mm x 170mm x 170 mm LHS front of engine

compartment

Drive Train Transverse front wheel drive

Transmission Automatic - manual G-shift for manual

included in package

Suspension Type, Front McPherson Mounted to front subframe

Suspension Type, Rear Twist beam With separate spring

shock absorber

Tire Size Front-Rear 195/60R15 Winter tires 185/60R15

Spare Tire Space saver Tub to fit full size tire

Fuel Tank volume ~65 ltr Located under rear seat

Fuel Filler On RHS Routing in package

Bumper Front-Rear Bolt-on Crash boxes included

Steering Rack & pinion Steering rack housing on

top of crossmember dash

Cargo Volume 490 ltr VDA method with 200 x 100

x 50 mm module

Hinges Similar to Porsche 911 / Boxster Weld through type

Head Lamps Part of front end module

Interior Front and rear seat concept In package drawing

Cockpit Basic concept with I/P beam In package drawing

Pedals Unit with integrated In package drawing

foot-parking-brake

2

Chapter 3 - Page 7


3.2.5. Underfloor Clearance

The underfloor clearance of a vehicle depends on the vehicle load. Thedetermination of the underfloor clearance relative to the road surface was crucial forthe body structure design, styling, selection of components and their positioning inthe vehicle structure. Underfloor clearance is defined as the summary of fivedifferent parameters. These are:

• Curb Clearance Front / Rear• Angle of Approach / Departure• Ramp Brakeover Angle• Oil Pan Clearance• Ground Clearance

To define these parameters, three vehicle positions, which then depended on threespecific load cases, needed to be determined. The three load cases applied to thevehicle were:

• Curb weight:The weight of a vehicle equipped for normal driving conditions. Thisincludes fluids such as coolant, lubricants and a fuel tank filled to aminimum of 90%. Also included are the spare tire, tool kit, and carjack.

• Design weight:Vehicle curb weight plus the weight of three passengers (68 kg each,with luggage 7 kg each) with 2 passengers in the front seat and 1passenger in the rear seat.

• Gross vehicle weight:Vehicle curb weight plus maximum payload (5 passengers plusluggage).

Chapter 3 - Page 8


ULSAB Data Number of Seats 5

Wheelbase 2700 mm Tires

Front 195/60-R15Rear 195/60-R15

PressureFront 2.5 barRear 2.5 bar

Using the ULSAB data and the weights of the three load cases, the road surfacepositions relative to the zero grid Z-plane and to the vehicle were calculated.

Calculation of Road Surface Positions Relative to the Vehicle

To determine the vehicle position relative to the road surface under these loadconditions, the vehicle is positioned relative to zero grid Z-plane.

Figure 3.2.5-1 ULSAB Vehicle Position Relative to Zero Grid Z-Plane

A B

X

Z

R1 R2 Ground

Distance from Static Tire Load Case Zero Grid Z-Plane Radius Weight

A (mm) B (mm) R1 (mm) R2 (mm)

Curb Weight 395 392 301 308 1350 kg Design Weight 413 417 301 305 1575 kg Gross Vehicle Weight 415 462 303 300 1850 kg

Chapter 3 - Page 9


With the road surface positions relative to the vehicle, the underfloor clearance wasdetermined.

Figure 3.2.5-4 Angle of Approach/Departure

Figure 3.2.5-3 Curb Clearance Front/Rear

Figure 3.2.5-2 Road Surface Relative to Vehicle

Design WeightGross Vehicle Weight

Curb Weight

Design WeightGross Vehicle Weight

190 mm170 mm

15º

Design Weight

14º

Chapter 3 - Page 10


Figure 3.2.5-7 Ground Clearance

Figure 3.2.5-6 Oil Pan Clearance

Figure 3.2.5-5 Ramp Breakover Angle

Gross Vehicle W eight

143 m m

185 mm

Design Weight

Gross Vehicle Weight

14º

Chapter 3 - Page 11


3.2.6. Seating Position

At first the 2-D manikins (spelling taken from SAE) were aligned in a comfortableseating position taking into consideration the angles between joints such as hip,knee, and foot. When the seating position was defined, verification was made thatthe operating parts like steering wheel, gearshift lever and pedal were in reach.This was important for ergonomic reasons. Two types of 2-D manikins were used:The small female, 5th percentile with a height of 147.8 cm; and the tall male, 95thpercentile with a height of 185.7 cm. (5th percentile means that 5% of thepopulation is smaller or equal in size and 95% is taller. 95th percentile means that95% of the population is smaller or equal in size and 5% is taller.)

For the dash panel layout the tall male 2-D manikin was used because it is moredifficult to reach, since the seat position of the taller person is more rearward than itis for a shorter person.

Figure 3.2.6-1 Distance to Operating Parts of the 5% Female and the 95% Male

Chapter 3 - Page 12


Accelerator Heel Point

SgR-Point

Eyellipse

Eyepoints V1, V2

Torso Line

Thigh Centerline

3.2.7. Visibility Study

3.2.7.1. Horizontal and Vertical Obstruction

For the study of horizontal, vertical and A-pillar obstruction of the driver’s visibility,the following positions needed to be defined:

• Seating Reference Point (SgRP)It was necessary to determine the seating reference point (SgRP) in orderto position the eyellipse (spelling taken from SAE) template and theeyepoints V1 / V2. For adjustable seats, the SgRP is defined as the hip-point (H-Point) relative to the driving seat in its most rearward position.The H-point is defined as the pivot center of the torso and thigh centerlines.

Figure 3.2.7-1 SgRP, Eyellipse, Eyepoints

• Eyellipse (SAE J941)The eyellipse is a tool to describe the vision of a driver. The template withthe eyellipse is positioned with its horizontal reference line 635 mm abovethe SgRP and with the vertical reference line through the SgRP. Twotypes of templates, with two eyellipses, take the different seat track travel

Chapter 3 - Page 13


Traffic Light Vision Angle min. 14ºWiperfield Angle 10ºTransparent Windscreen Area 7º Through V1 (77/649/EWG)

Horizont View Through V1

Steering Wheel Rim Obscuration 1º Through V2 (77/649/EWG)

Unobstructedd Vision 4º Through V2 (77/649/EWG)

Transparent Windscreen Area 5º Through V2 (77/649/EWG)

V1

V2

ranges into consideration. For the ULSAB vehicle, with a seat track travelof 240 mm, a template for seat track travel of more than 130 mm wasused.

• Eye Points V1 / V2 (RREG 77/649)The coordinates of the eye points V1 / V2 relative to the SgRP weredetermined by using the following dimensions:

Using vision lines through the eye points, the following vision areas are described:

Figure 3.2.7-2 Horizontal Vision

Point X Y Z

V1 68 -5 665

V2 68 -5 589

Chapter 3 - Page 14


3.2.7.2. A-Pillar Obstruction

In order to determine the A-pillar obstruction, points P1 and P2 have to bedetermined first. The coordinates for these points related to the SgR-point are:

The ULSAB structure has a seat track travel of 240 mm. Therefore the X-value hasto be corrected by -48 mm.

Since the torso back angle is 25 degrees, no further correction is necessary for theX-value and Z-value.

The new coordinates for the P-points are:

Figure 3.2.7-3 Vertical Vision

Point X Y Z

P1 35 mm -20 mm 627 mm

P2 63 mm 47 mm 627 mm

Vision Area B 17º (78/317 2.3/EWG)

Vision Area A 20º (78/317 2.2/EWG)

Vision Area B 17º (78/317 2.3/EWG)

Vision Area A 13º (78/317 2.2/EWG)

V1, V2

Y

X

Point X Y Z

P1 -13 mm -20 mm 627 mm

P2 +15 mm 47 mm 627 mm

Chapter 3 - Page 15


Two planes are cutting the A-pillar in an angle of 2 and 5 degrees. In the front mostintersection, the horizontal planes S1 and S2 cut the A-pillar (Figure 3.2.7-5).

627

mm

5º

2º

S1

S2

S1

S2

SgRP

Pm

Figure 3.2.7.2-5 Determination of the Sections S1 and S2

Figure 3.2.7.2-4 Distance of the P-Points Relative to the SgR-Point

P2

Pm

P1

SgRP

Horizontal Line

+15 mm

+47 mm

-20 mm

-13 mm

Y

X

Chapter 3 - Page 16


The sections in the plan view are shown in Figure 3.2.7-6.

Figure 3.2.7-6 Sections S1 and S2 in Plan View

The point P1 is necessary to determine the A-pillar obscuration for the left side (fora left hand drive vehicle). P2 is necessary for the right side. If P1 fulfills therequirements, it is not necessary to determine the obscuration for the right A-pillar,since the right pillar is farther away from the driver.

The template to determine the obstruction is shown in Figure 3.2.7-7.

Figure 3.2.7-7 Template for A-Pillar Obstruction

P1

α

E1

E2 104 mm

65 m

m

Section S2 OuterSection S1 Inner

P1

P2

Pm

V1, V2

S1

S2

Chapter 3 - Page 17


340 mm

220 mm

290 mm

P1

1º

Figure 3.2.7-8 Template in Position

The point P1 on the template is aligned to the point P1 on the drawing. The line“Section S2 Outer” is laid tangent to the most outer edge of the A-pillar section (S2),including trim, door frame and door seal. The second tangent line “Section S1inner” is laid to the most inner edge of the A-pillar section (S1), including trim, sealand dot matrix. (Figure 3.2.7-8).

3.2.8. Gear Shift Lever Postion

The position of the gearshift lever depends on the SgRP-position and on the torsoback angle. The position of the gearshift lever in the side view is shown in Figure3.2.8-1.

Figure 3.2.8-1 Distance of Gearshift Lever Relative to SgR-Point

Chapter 3 - Page 18


3.2.9. Pedal Position

3.2.10. Bumper Height Definition

ECE R42 for the bumper height definition requires a pendulum 445 mm above thecurb weight vehicle position and the design weight vehicle position. At the sametime an overlapping of 35 mm of the pendulum to the bumper is required.

Figure 3.2.9-1 Pedal Position Side Figure 3.2.9-2 Pedal Position Rear

Figure 3.2.10-1 Pendulum in the Extreme Height Position

A

D

B

C

201 mm

89 mm

203 mm

48 mm (Brake)

50 mm (Clutch)

98 mm

59 mm58 mm 53 mm

53 mm

Seating Reference Point

Chapter 3 - Page 19


• A: Lower edge of the pendulum in the most upper level to the curbweight vehicle position.

• B: Upper edge of the pendulum in the most lower level to the designweight vehicle position.

• C: Overlapping of the pendulum to the bumper in extreme highposition.

• D: Overlapping of the pendulum to the bumper in extreme lowposition.

A B C D

Front 467 mm 431 mm 91 mm 40 mm

Rear 467 mm 402 mm 89 mm 38 mm

Chapter 3 - Page 20


3.3. Package Drawings

Since package drawings are orthographic projections of the vehicle contour in sideview, plan view, front view and rear view, these views include all essential parts ofthe interior such as seats, seat position, seating reference point (SgRP), operatingparts and the door openings. To define the interior of the vehicle including the seatposition, visibility, and obstruction by the pillars, roof, hood and deck lid positionswere determined. It was also important to define positions of the steering wheel,pedals, and gearshift lever. Other criteria were visibility to the instrument panel, andhead clearance to the front, top and side. In the engine compartment, the engine,gearbox, exhaust system, radiator and battery were used in defining the space forthe structural members of the front body structure. Components such as the fueltank with the fuel filler system, the catalytic converter and exhaust system, andspare tire tub were also included in the package drawings. The package drawingswere the starting point for the Phase 2 design.

Chapter 3 - Page 21


Fig

ure

3.3.

1-1

Pac

king

Dra

win

g S

ide

Vie

w

3.3.

1.

Sid

e V

iew

Chapter 3 - Page 22


3.3.

2.

Pla

n V

iew

Fig

ure

3.3.

2-1

Pac

kage

dra

win

g P

lan

Vie

w

Chapter 3 - Page 23


3.3.

3.

Fro

nt a

nd R

ear

Vie

w

Fig

ure

3.3.

3-2

Pac

kage

Dra

win

g R

ear

Vie

wF

igur

e 3.

3.3-

1 P

acka

ge D

raw

ing

Fro

nt V

iew


4. Styling

Chapter 4 - Page 1


4. Styling

4.1 . Approach

The Phase 1 concept design of the ULSAB program did not account for any Class Asurfaces for the outer panels of the structure. To establish Class A surfaces inPhase 2, a complete styling of the ULSAB vehicle was necessary in order to createthe surfaces of the roof panel, body side outer panel, the back light and thewindshield. Styling also provided the major feature lines for the doors, deck lid,hood, fender and front and rear bumpers; these were needed for the development ofthe mating structural parts. For Phase 2, styling also gave the ULSAB structure aprofessional look and provided surfaces for further design studies in the future, i.e.on hoods, doors, deck lids, etc. The styling was developed electronically using CAS(computer aided styling), no clay models were used. With support from Porsche’sstyling studio, PES selected A. D. Concepts, a local source, to carry out thecomputer aided styling in a simultaneous engineering approach with PES. At thefirst team meetings of PES and A. D. Concepts, several elements of the stylingwere discussed with a view to creating a 3-dimensional styling model. Using thepackage drawings, important criteria such as overall vehicle proportions, visionlines, bumper locations and proposed cut lines were specified. After the initialmeetings, a clearly defined vehicle architecture had evolved.

4.2. 2-D Styling Phase

4.2.1. Sketching

In a team review of the first sketches, a neutral styling approach was chosen toensure the ULSAB styling model would not be too futuristic or radical. Traditionalsketching techniques were used along with the latest electronic paint sketchingsoftware from the Alias|Wavefront company entitled StudioPaint running on SiliconGraphics High Impact workstations. Many automotive design studios around theworld use this combination of hardware and software. The use of this tool for sucha project increased productivity and enhanced the overall styling presentation withprofessionalism and accuracy, producing tighter sketches and more realistic,achievable styling goals.

Chapter 4 - Page 2


Figure 4.2.1-1 Styling Sketches

Chapter 4 - Page 3


4.2.2. Clinic

In the first clinic, dozens of sketches were reviewed by the design and styling teamto determine which direction the styling would take prior to its presentation to theULSAB Consortium. With the best sketches selected, five separate side viewproposals and several different front and rear end treatments were developed.

4.2.3. Electronic Paint

In the studio, the CATIA package data was imported into a 3-D conceptual modelingsoftware, called CDRS, and a side view outline drawing was developed forsketching purposes. The drawing was imported into StudioPaint and the five, verydisciplined, side view sketch proposals (A-E) along with front and rear end sketchproposals were developed.

Figure 4.2.2-1 Side View Proposal

Chapter 4 - Page 4


4.2.4. Styling Theme Selection

The final styling theme selection was made during a meeting of the ULSABConsortium’s editorial group, together with PES and A. D. Concepts. In a secretballot, the editorial group members from all around the world selected stylingtheme A. With the selection of the specific front and rear end treatments for the 3-Dmodel, the 2-D phase of the ULSAB styling reached its conclusion.

Figure 4.2.4-1 Selected Styling Theme A

Figure 4.2.4-2 Styling Theme B

Chapter 4 - Page 5


Figure 4.2.4-4 Styling Theme D

Figure 4.2.4-3 Styling Theme C

Figure 4.2.4-5 Styling Theme E

Chapter 4 - Page 6


Figure 4.2.4-6 Selected Front View Proposal

Figure 4.2.4-7 Selected Rear View Proposal

Chapter 4 - Page 7


4.3. 3-D Styling Model

To create the 3-D styling model, the package data was imported into CDRS alongwith the selected theme drawing and then the first phase of the 3-D modelcommenced. Side view lines, created using 2-D spine curves, were developed torepresent the major feature lines of the vehicle. Typical sections at specific Xlocations were constructed. This data was reviewed by the design team to verifythe positions of these major curves.

The construction of the greenhouse, (the upper glass and roof surfaces of thevehicle), was started, transferring preliminary surfaces back and forth betweenCDRS and CATIA using an IGES translator. In the following Class A surfacingusing CATIA, only subtle design changes were made to the CDRS surface modeluntil both the styling and engineering teams were comfortable with the result. Therelease of the styling data by the styling team, in IGES file format, marked the firststep in the 3-D modeling phase.

Next, body side lines were constructed and surfaces were created. With the wheelopenings, and the front and rear stance developed, the model started to take shape.The team developed the best proposal for front and rear door cut lines and thisinformation was then incorporated into the CDRS styling model.

After the front and rear end surfaces were completed, shaded tile images of thesurface model were used to evaluate the forms. Highlight sections and surfacecurvature graphs were used to verify the aesthetic value of the model.

Chapter 4 - Page 8


4.4. Rendering

After the release of the surface model, the CDRS model was prepared forrendering. Model colors were selected in texture maps created to enhance theoverall appearance of the photo realistic rendering. Neutral backgrounds andspecific views were selected to create the first ULSAB styling images. Toincorporate subtle engineering changes in the model, the CDRS 3-D models wererevised and additional renderings were created. The models were enhanced furtherby the addition of texture maps for items such as license plate and rear windowdefrost. The 3-D model was imported back into StudioPaint 3-D to examine stylingchanges to the front and rear lamp treatments. These changes were thenincorporated into the CDRS 3-D model and the final renderings completed, whichconcluded the styling phase.

4.3.1. Surface Release

Prior to the official surface release, the styling was reviewed to establish the exactlocation of all cut lines and shut lines. Shaded tile model images, with highlightreflection lines, were created in CDRS to allow both styling and engineering todiscuss potential areas of concern. With the final release of the IGES surfacemodel, the 3-D modeling phase was complete.

Figure 4.3.1- 1 Surface Release

Chapter 4 - Page 9


Figure 4.4-1

Figure 4.4-2



Chapter 5 - Page 1



5.1. Phase 2 Design and Engineering Approach

After the package was revised and the styling frozen, the challenge in Phase 2 wasto maintain the structural performances, especially the mass, as analyzed in thePhase 1 concept. Further research into steel sandwich material led to additionalchanges in the Phase 2 design. Because of restrictions in size and application ofthe material, new design solutions had to be created to compensate for theadvantages in mass reductions using sandwich material as it was applied inPhase 1. The hydroformed parts were analyzed for manufacturing feasibility usingthe detailed design data created in Phase 2. The restrictions of the hydroformingprocess, in combination with the refinement of the design, led to different concepts,design adjustments, and new solutions to achieve the target for mass.Furthermore, the 50% off-set crash, an additional crash analysis introduced inPhase 2, significantly influenced the design of parts, the application of steel grades,the material thicknesses and in particular, the changes to tailor welded blanks.

Every change in the design process also had to be analyzed for its suitability forassembly and parts manufacturing. The design approach was driven by massreduction and created innovative results without allowing initial component costconsideration to limit options. The design also focused on a production volume ofmore than 100,000 units per year.

As well as concentrating on reaching the targets for performance and mass,importance was also placed on the reduction of assembly steps, the integration ofreinforcements, the use of tailor welded blanks, and the avoidance of metal arcwelding, wherever possible. Using the same design approach in both Phases 1and 2, it was possible to maintain low mass and high structural performances. ThePhase 1 design concept and approach, the flexibility of the concept and thepotential that it could be adjusted to various design tasks, were challenged inPhase 2 and ultimately justified.

Chapter 5 - Page 2


5.2. Design and Engineering Process

The design and engineering process used in Phase 2 is shown in the flow chart(Fig. 5.2-1). All through this process, a simultaneous engineering approach wastaken to find the best solutions to overcome the design and engineering challengesemerging in Phase 2.

Using the Phase 1 package and concept design as the starting point, Phase 2 thenrefines the package. This refined Phase 2 package was the basis for the firststyling layout, and in an interactive process, both were adjusted until theengineering requirements were met. The styling was frozen and the Phase 1 shellmodel was adjusted and analyzed using material thickness optimization to achieve

Figure 5.2-1 Design and Engineering Process

MeetsStatic

Targets

Material / ThicknessSelection,

Design Modification

MeetsStatic

Targets

Create / ModifyPhase 2

Crash Model

Meets Static/Crash

Targets

Parts Feasible

MeetsStatic / Crash

Targets

Build of First

Test Unit

Build of FinalDemonstration

Hardware

Steel Supplier& Part Supplier

Input

Create / ModifyPhase 2 Shell

Model

Modify Desi gnMaterial / Thickness

AdjustmentNo

Yes

Yes YesNo

Yes

Yes

No

No

No

Phase 1 Package/Concept

Design

Phase 2 Package

Refinement

CreateStylin g

Concept

Modify Package/

Stylin g / Design

Modify Phase 1Shell Model

Start

No

No

No

No

No

Yes

Yes

Yes

Yes

Yes

Chapter 5 - Page 3


the mass target while maintaining the structural performance goals. Together withthe selected suppliers and the Material Group of the ULSAB Consortium, the partdesign was discussed and the material thicknesses were selected. With thisinformation, the design was revised and the Phase 2 shell model created, analyzedand modified until all targets were met. New Phase 2 crash analysis models werebuilt and after the first analysis, design modifications, material grade and thicknessselection, further crash analyses were performed, until the results were satisfactory.With the revised design and material selection, the shell model was updated andthe static analysis performed. The crash and static analysis models wereconstantly updated as a result of information from tool, part and steel suppliers.This was repeated until all results were satisfactory. The design was then modifiedand the part drawings released to the suppliers. With the first part set delivered, atest unit was built and the tests following provided the results for static performanceand most importantly for mass. The design was enhanced and material substitutedas needed. The process of shell and crash model modifications and analysis wasperformed again to validate the design. After the final design was released to thesuppliers, parts were manufactured and the demonstration hardware built.

Part of this process included regular design review meetings (not shown in the flowchart) of the design and engineering team as well as design review meetings withthe demonstration hardware build team, engineers and analysts at Porsche R & DCenter in Germany. In these internal PES meetings, technical problems werediscussed and design directions decided in order to prepare for the demonstrationhardware build and meet established deadlines.

Chapter 5 - Page 4


5.3. ULSAB Phase 2 Design Description

Figure 5.3-1 ULSAB Demonstration Hardware

The ULSAB structure went through many adjustments and modifications in itstransition from the Phase 1 concept to its final design stage at the end of Phase 2.This was due to added crash performance requirements, package issues,manufacturing processes and material application limitations. The exploded view(see Fig. 5.3-2) shows the demonstration hardware in the final Phase 2 designstage with the exception of minor brackets and reinforcements. Bolt-on parts andcomponents, used in the analysis for crash performance, such as front and rearbumpers, engine, suspension, subframe, shock tower braces, tunnel bridge andfenders, are not considered part of the body structure and therefore are not shownin the exploded view. However, the structure is equipped with important bracketsand reinforcements. Because tailor welded blanks can eliminate reinforcements,fewer were required. Included in the demonstration hardware, as shown on theexploded view, are the bolt-on front-end module and the dash-panel insert, includingthe brake booster reinforcement.

Chapter 5 - Page 5


5.3.1. Parts List – Demonstration Hardware

The parts list (Fig. 5.3.1-1) corresponds directly with the exploded view of thedemonstration hardware (Fig. 5.3.1-2) and shows the part name and number, thematerial grade, and thickness and the mass of the manufactured part. Parts listedthat have two or more material thicknesses and grades indicate that this part ismade from a tailor welded blank. The mass of the parts listed, is taken from actualmanufactured parts, but does not represent an average of all parts manufactured.Therefore, the mass of the demonstration hardware can vary slightly in comparisonto the listed mass of the total number of parts.

Figure 5.3.1-1

Demonstration Hardware Parts List

Mate ria l Ma te ria l Actua lPa rt Grade T hickness Pa rtNo Pa rt Name (MPa) (mm) Mass (kg)

001 Assy Reinf Radiator Support Upper (Bolted on) 350 1.00 1.613

002 Reinf Front Rail Extension RH 350 1.00 0.485

003 Reinf Front Rail Extension LH 350 1.00 0.489

008 A Assy Rail Front Outer RH 350 1.50 3.013

B (Tailor Welded Blank) 350 1.60

C 350 2.00

009 A Assy Rail Front Outer LH 350 1.50 3.037


C 350 2.00

010 A Assy Rail Front Inner RH 350 1.50 5.470


C 350 1.80

011 A Assy Rail Front Inner LH 350 1.50 5.500


C 350 1.80

012 Rail Front Extension RH 350 1.40 2.096

013 Rail Front Extension LH 350 1.40 2.061

014 Bracket Roof Rail Mount Low er RH 350 1.20 0.153

015 Bracket Roof Rail Mount Low er LH 350 1.20 0.150

021 Panel Dash 210 0.70 5.830

022 Panel Dash Insert (Bolted on) Sandw ich 0.95 0.875

026 Member Dash Front 600 1.20 2.290

028 Panel Cow l Low er 210 0.70 1.272

032 Panel Cow l Upper 210 0.70 1.374

034 Assy Member Front Floor Support (2-Req'd) 800 0.70 1.290

038 Assy Reinf Floor Front Seat Rear Outer (2-Req'd) 280 0.80 0.120

040 Pan Front Floor 210 0.70 14.650

Chapter 5 - Page 6


Demonstration Hardware Parts List (Cont’d)

Figure 5.3.1-1

Mate ria l Ma te ria l Actua lPa rt Grade T hickness Pa rtNo Pa rt Name (MPa) (mm) Mass (kg)

042 A Panel Rocker Inner RH 350 1.30 6.490


043 A Panel Rocker Inner LH 350 1.30 6.625


045 Member Rear Suspension 280 0.70 1.344

046 A Assy Rail Rear Inner RH 350 1.00 5.250


C 350 1.60

047 A Assy Rail Rear Inner LH 350 1.00 5.240


C 350 1.60

048 A Assy Rail Rear Outer RH 350 1.00 2.527


C 350 1.60

049 A Assy Rail Rear Outer LH 350 1.00 2.565


C 350 1.60

050 Panel Spare Tire Tub (Bonded on) Sandw ich 0.96 2.107

055 Member Panel Back 210 0.65 1.305

057 Panel Back 140 0.65 2.502

060 A Panel Body Side Outer RH 210 0.70 15.780


C 280 1.30

D 350 1.50

E 350 1.70

061 A Panel Body Side Outer LH 210 0.70 15.650


C 280 1.30

D 350 1.50

E 350 1.70

062 Panel A-Pillar Inner Low er RH 350 1.00 1.365

063 Panel A-Pillar Inner Low er LH 350 1.00 1.375

064 Panel B-Pillar Inner RH 350 1.50 3.586

065 Panel B-Pillar Inner LH 350 1.50 3.586

066 Reinf B-Pillar Low er (2-Req'd) 350 0.90 0.830

068 Panel Wheelhouse Inner RH 210 0.65 1.931

069 Panel Wheelhouse Inner LH 210 0.65 1.923

070 A Panel Wheelhouse Outer RH 140 0.65 2.116


071 A Panel Wheelhouse Outer LH 140 0.65 2.194


Chapter 5 - Page 7



Figure 5.3.1-1

Mate ria l Ma te ria l Actua lPa rt Grade T hickness Pa rtNo Part Name (MPa) (mm) Mass (kg)

072 Rail Side Roof RH 280 1.00 4.700

073 Rail Side Roof LH 280 1.00 4.860

074 Panel A-Pillar Inner Upper RH 350 1.50 1.425

075 Panel A-Pillar Inner Upper LH 350 1.50 1.416

080 Panel Package Tray Upper 210 0.65 1.876

081 Panel Package Tray Low er 210 0.65 1.497

082 Support Package Tray RH 280 0.80 0.084

083 Support Package Tray LH 280 0.80 0.076

085 Panel Roof 210 0.70 8.680

086 Panel Front Header 280 0.70 0.813

087 Panel Rear Header 140 0.70 0.773

090 Member Pass Through (2-Req'd) 140 0.65 0.662

091 Member Kick Up 800 0.70 1.397

094 Reinf Radiator Rail Closeout RH (Bolted on) 350 1.00 0.567

095 Reinf Radiator Rail Closeout LH (Bolted on) 350 1.00 0.575

096 A Panel Skirt RH 140 2.00 3.457


097 A Panel Skirt LH 140 2.00 3.468


098 Panel Gutter Decklid RH 140 0.65 0.434

099 Panel Gutter Decklid LH 140 0.65 0.437

102 Support Panel Rear Header RH 140 0.70 0.098

103 Support Panel Rear Header LH 140 0.70 0.098

104 Rail Fender Support Inner RH 420 1.20 2.712

105 Rail Fender Support Inner LH 420 1.20 2.699

106 Rail Fender Support Outer RH 350 0.90 1.297

107 Rail Fender Support Outer LH 350 0.90 1.297

108 Reinf Front Rail RH 350 1.00 0.838

109 Reinf Front Rail LH 350 1.00 0.830

110 Plate Rear Spring Upper (2-Req'd) 350 2.00 0.526

115 Reinf Panel Dash Brake Booster (Bolted on) 350 1.00 0.464

116 Assy Bracket Rear Shock Absorber Mount RH 350 2.00 0.335

117 Assy Bracket Rear Shock Absorber Mount LH 350 2.00 0.339

120 Reinf Floor Front Seat Rear Center 350 1.20 0.250

122 Assy Reinf Rear Seat Inner Belt Mount (2-Req'd) 350 2.00 0.244

128 Bracket Member Pass Through Low er (2-Req'd) 350 1.00 0.056

130 Bracket Member Pass Through Upper Front 350 1.00 0.129

136 Reinf Panel Dash Upper 350 1.00 0.100

140 Pan Rear Floor 210 0.70 4.240

142 Assy Reinf Hinge Decklid (2-Req'd) 350 1.50 0.224

144 Reinf A-Pillar RH 350 1.50 0.229

Chapter 5 - Page 8


Mate ria l Ma te ria l Actua lPa rt Grade T hickness Pa rtNo Part Name (MPa) (mm) Mass (kg)

145 Reinf A-Pillar LH 350 1.50 0.230

152 Bracket Member Pass Through Upper Rear 350 1.00 0.145

164 Assy Closeout Fender Support Rail RH 350 1.00 0.115

165 Assy Closeout Fender Support Rail LH 350 1.00 0.115

170 Reinf Rail Dash RH 350 1.30 0.309

171 Reinf Rail Dash LH 350 1.30 0.312

172 Assy Reinf Cow l Low er 350 1.00 0.127

455 Assy Hinge Door Upper RH (2-Req'd) 280 - 0.515

456 Assy Hinge Door Low er RH (2-Req'd) 280 - 0.549

457 Assy Hinge Door Upper LH (2-Req'd) 280 - 0.515

458 Assy Hinge Door Low er LH (2-Req'd) 280 - 0.549

180 Bracket Trailing Arm Mount RH 350 2.00 0.333

181 Bracket Trailing Arm Mount LH 350 2.00 0.341

188 Brace Radiator (2-Req'd) (Bolted on) 350 0.80 0.250

190 Assy Reinf Seat Belt Retractor Rear (2-Req'd) 350 1.20 0.104

Total Mass of Parts 196.770


Figure 5.3.1-1

Chapter 5 - Page 9


Fig

ure

5.3.

1-2

ULS

AB

Pha

se 2

Exp

lode

d V

iew

* *

* *

**

* *

* S

ee A

ssem

blie

s 45

5 -

458

Chapter 5 - Page 10


5.3.2. ULSAB Structure Mass

For the Phase 1 concept, it was assumed that future average body structures wouldcontain approximately 12 kg of brackets and reinforcements. This number can vary,up or down, depending on the type of vehicle, i.e., front or rear wheel drive, and thepackage of components. Since the goal of the ULSAB program is to providesolutions for a generic concept, it was assumed in Phase 1 that the 12 kg forbrackets and reinforcements have to be considered in the calculation for mass togive the Phase 1 results more credibility. In Phase 1, the ULSAB structure wascalculated with a mass of 193 kg. With the 12 kg for brackets and reinforcements,the total mass equals 205 kg. In Phase 2, some of the brackets and reinforcementsare already welded into the structure. These are reflected accordingly in the massof the demonstration hardware and also included in the parts list. With therefinement of the Phase 2 package, minor brackets and reinforcements weredesigned (but not manufactured) and their mass was calculated to get a moreaccurate determination than the general assumption used in Phase 1. Thesebrackets and reinforcements represent a more generic, than detailed, selection.The selection was based on package information, chosen components andengineering judgment. It can be assumed that in a possible Phase 3, the number ofbrackets and reinforcements, and their actual mass when manufactured, can beinsignificantly higher or lower. This depends on the final component selection; theirposition in the structure and efforts made to minimize their mass. Also included inthe mass calculation are 100 weld studs. This also represents a generic number forthis type of structure and is based on engineering judgment. The calculated massof the ULSAB structure (Fig. 5.3.2-1) is the measured mass of the demonstrationhardware parts and the calculated mass of brackets and reinforcements shown inFig. 5.3.2-2 and Fig. 5.3.2-3. The ULSAB structure mass in Phase 2 is 203 kg, withthe variation assumed to be +/- 1%. This low variation is due to each part beingmanufactured from one coil of steel. The differences in sheet thicknesses betweencoils do not apply for the demonstration hardware, but would have to be consideredin mass production.

Figure 5.3.2-1 Definition of ULSAB Structure Mass

ULSAB = Mass of Demonstration + Calculated Mass of Brackets

Structure Mass Hardware (Parts) and Reinforcements

203.2 kg = 196.8 kg + 6.4 kg

Chapter 5 - Page 11


Figure 5.3.2-3

Figure 5.3.2-2

Designed Reinforcements not Manufactured butConsidered Part of the ULSAB Structure

Designed Brackets not Manufactured butConsidered Part of the ULSAB Structure

Part No Name Qty Calc Mass [Kg]

331 Bracket Exhaust Mount 2 0.060

332/333 Bracket Engine Mount RH/LH 2 0.528

334/335 Bracket Fender Mount Rear RH/LH 2 0.228

336 Bracket Battery Tray 1 0.412

337 Bracket Spare Tire Mount 1 0.089

338/339 Bracket Fuel Tank Mount Rear RH/LH 2 0.242

340 Bracket Front Tie Dow n Hook 2 0.236

341 Bracket Rear Tie Dow n Hook 2 0.236

342/343 Bracket Front Jack Support RH/LH 2 0.656

344/345 Bracket Rear Jack Support RH/LH 2 0.548

346 Bracket Plenum Support Center 1 0.445

N/A Weld Studs ~ 100 - 0.300

TOTAL 19 3.980

Part No Name Qty Calc Mass [Kg]

310 Reinf Hood Hinge Mount 2 0.086

311 Reinf Instrument Panel Beam Mount 2 0.134

312/313 Reinf Sub-Frame Front Mount 2 0.050

314/315 Reinf Sub-Frame Center Mount 2 0.116

316/317 Reinf Sub-Frame Rear Mount 2 0.418

318 Reinf Steering Rack Assembly Mount RH 1 0.032

319 Reinf Steering Rack Assembly Mount LH 1 0.041

320 Reinf Gear Shift Mount 1 0.271

321 Reinf Front Door Lock Striker 2 0.106

322 Reinf Front Door Check Arm 2 0.030

323 Reinf Rear Door Lock Striker 2 0.146

324 Reinf Rear Door Check Arm 2 0.028

325 Reinf Front D-Ring Adjustment 2 0.298

326 Reinf Rear Seat Cushion Mount 2 0.140

327 Reinf Rear Seat Latch 2 0.068

328 Reinf Rear Seat Back Mount Outer 2 0.278

329 Reinf Rear Seat Back Mount Center 1 0.035

330 Reinf Deck Lid Latch 1 0.136

TOTAL 31 2.413

Chapter 5 - Page 12


Mass of Brackets,

Reinforcements, Bolt-on Parts,

DH Mass = Body Structure Mass + Welded and Assembled to

the Body Structure

196.8 kg = 186.6 kg + 10.2 kg

5.3.3. ULSAB Demonstration Hardware Mass

The mass of the demonstration hardware is 196.770 kg. This reflects the totalamount of the mass of one complete part set, including brackets, reinforcementsand bolt-on parts, as measured.

In Phase 1, nearly all brackets and reinforcements were included in the theoreticalnumber of 12 kg and only a few were included in the Phase 1 concept design of thebody structure. With the level of detail design in Phase 2 and the refined package,it was now possible to design and finally manufacture most of these brackets andreinforcements and weld or bolt them to the demonstration hardware. It was not thetask in Phase 2 of the ULSAB program to design and to manufacture all bracketsand reinforcements and therefore, the approach to concentrate only on theimportant ones was taken.

The mass of these manufactured brackets, reinforcements and bolt-on parts isincluded in the demonstration hardware mass and listed in the parts list (Fig.5.3.1-1). The parts are shown on the exploded view (Fig. 5.3.1-2).

For easier identification, the extracted list from the parts list (Fig. 5.3.3-2, -3 to Fig.5.3.3-4) identifies these parts including their mass.

The mass of the demonstration hardware as shown in Fig 5.3.3-1, consists of themass of the pure body structure and the mass of brackets, reinforcements, bolt-onparts manufactured and welded or assembled to the body structure.

Figure 5.3.3-1 Demonstration Hardware Mass Definition

Chapter 5 - Page 13


Part No Name Qty Mass [Kg]

038 Assy Reinf Floor Front Seat Rear Outer 2 0.120

110 Plate Rear Spring Upper 2 0.526

120 Reinf Floor Front Seat Rear Center 1 0.250

122 Reinf Rear Seat Inner Belt Mount 2 0.244

136 Reinf Panel Dash Upper 1 0.100

142 Assy Reinf Hinge Decklid 2 0.224

144 Reinf A-Pillar RH 1 0.229

145 Reinf A-Pillar LH 1 0.230

164 Assy Closeout Fender Support Rail RH 1 0.115

165 Assy Closeout Fender Support Rail LH 1 0.115

176 Hinge Base RH 4 0.650

177 Hinge Base LH 4 0.650

178 Hinge Stem 119 4 0.379

179 Hinge Stem 141 4 0.449

172 Assy Reinf Cow l Low er 1 0.127

190 Assy Reinf Seat Belt Retractor Rear 2 0.104

33 parts 4.512


116 Assy Bracket Rear Shock Absorber Mount RH 1 0.335

117 Assy Bracket Rear Shock Absorber Mount LH 1 0.339

180 Bracket Trailing Arm Mount RH 1 0.333

181 Bracket Trailing Arm Mount LH 1 0.341

4 parts 1.348


001 Assembly Reinf Radiator Support Upper 1 1.613

022 Panel Dash Insert 1 0.875

094 Reinf Radiator Rail Closeout RH 1 0.567

095 Reinf Radiator Rail Closeout LH 1 0.575

115 Reinf Panel Dash Brake Booster 1 0.464

188 Brace Radiator 2 0.250

7 parts 4.344

Brackets Manufactured and Welded to Structure

Reinforcements Manufactured and Welded to Structure

Figure 5.3.3-2

Figure 5.3.3-3

Figure 5.3.3-4

Bolt-On Parts Manufactured and Attached to Structure

Chapter 5 - Page 14


Figure 5.3.4-1 Mass Breakdown of Brackets, Reinforcements and Bolt-on Parts

5.3.4. Mass of Brackets and Reinforcements – Phase 2

The total mass of all brackets and reinforcements, (meaning the calculated mass ofdesigned, not manufactured parts) and bolted-on parts welded or assembled to thedemonstration hardware, amounts to 16.6 kg, and is included in the ULSABstructure mass of 203.2 kg.

6.4 kg

1.35 kg4.5 kg

4.35 kg

Calculated mass of brackets & reinforcements, not manufactured or part of the ULSAB Structure

Brackets welded to body structure

Reinforcements welded to body structure

Bolt-on parts assembled to body structure

Total Mass of Brackets, Reinforcements & Bolt-on Parts - 16.6 k g

Chapter 5 - Page 15


5.3.5. ULSAB Structure Mass Comparison Phase 1 – Phase 2

The comparison of the results of the ULSAB structure mass is shown in Fig.5.3.5-1. In Phase 2 the measured body structure mass has decreased with therefinement of the design, compared with the body structure mass as calculated inPhase 1.

The total calculated mass of 205 kg, as in the Phase 1 ULSAB structure, iscompared to the Phase 2 ULSAB structure mass of 203.2 kg, which includes theactual mass of the demonstration hardware plus the calculated mass of bracketsand reinforcements.

Figure 5.3.5-1 ULSAB Structure Mass Phase 1 - Phase 2

+ Offset crash

+ Package refinement

+ Styling

Assumed theoretical mass of brackets &

reinforcements

Calculated mass of brackets & reinforcements designed, not manufactured

+ Offset crash

+ Package refinement

+ Styling

Mass ofDemonstration

HardwareBod y structure

Mass

12 kg6.4 kg

193 kg 196.8 kg

205 kg

}

Phase 1 Phase 2

ULSAB Structure

Mass

203.2 kg 1%±

}Brackets, reinforcements& bolt-on parts included in demonstration hardware (10.2 kg)

{

Concept Validation

ULSAB Structure

Mass

Body Structure

Mass

Chapter 5 - Page 16


5.3.6. DH Part Manufacturing Processes

The ULSAB structure as developed during Phase 1 and refined in Phase 2 is ingeneral, a unibody design, with the exception of the hydroformed side roof rails.Stamping was the main manufacturing process considered for the parts design.

Relative to the body structure mass of 196.8 kg, 89.2% is the mass of all stampedparts.

The stampings can be divided into two groups; conventional stampings andstamped parts made from tailor welded blanks. 42.8% of the body structure massis represented by conventionally stamped parts and 44.9% is the mass of partsmade from tailor welded blanks. This relatively high percentage of tailor weldedblank stampings, relative to the body structure mass, is one good indication of howthe mass reduction was achieved. Especially if the use of high strength steels, inconnection with the tailor welded blanks, is put into consideration.

The hydroforming process is applied in the form of two processes:

• The tubular hydroforming process for the side roof rail manufacturing• The hydromechanical sheet forming process, for the roof panel

manufacturing.

The spare tire tub and the dash panel insert are designed to be manufactured fromsteel sandwich material, also using the stamping process.

Chapter 5 - Page 17


Figure 5.3.6-1 Manufacturing Process Relative to DH Mass

The mass of the stamped parts made from steel sandwich material is 1.5% relativeto the overall mass. 1.5% are miscellaneous parts, stock materials, such as tubes,or the forged hinge base of the weld through hinges.

The pie chart in Fig. 5.3.6-1 shows the mass distribution of the manufacturingprocesses relative to the DH mass.

The process used to manufacture the parts is shown in Fig. 5.3.6-2.

44.9% Tailor Welded Blank Stamping

42.8% Conventional Blank Stamping

4.4% Sheet Hydroforming4.9% TubularHydroforming

1.5% Steel SandwichMaterial Blank Stamping

1.5% Miscellaneous

ÒÒÒ 89.2% Stampings ÒÒ 9.3% Hydroforming Parts Ò 1.5% Misc.(Stock Material) Parts

Chapter 5 - P

age 18

Engineering S

ervices, Inc.

Figure. 5.3.6-2 ULSAB Manufacturing Processes of Demonstration Hardware Parts

Ò “Conventional Blank”, StampingÒ Tailor Welded Blank, StampingÒ Sheet, HydroformingÒ Tubular HydroformingÒ Sandwich Material Blank, StampingÒ Misc.(Stock Materials)

Part Manufacturin g Process

**

**

**

**

* See Assemblies 455 - 458

Chapter 5 - Page 19


5.3.7. Material Grades

The selection of the steel grades is a result of the need for good crash performanceand mass reduction.

In Phase 2, the utilization of high strength steel is 91%, relative to the DH mass(Fig. 5.3.7-1) of Phase 1.

The parts design had to consider the lower elongation, and together with the toolmanufacturer, the parts design was optimized to accommodate the different formingcharacteristics and greater spring back of high and ultra high strength steels. Thiswas most important for the design of the tailor welded blank stamped parts whichwhere different grades and thicknesses of high strength steels and combined intoone part.

High strength and ultra high strength steel material was used on parts contributingto the crash management of the structure, i.e. front rails, rear rails, rocker, etc. (Fig.5.3.7-2). With this approach, and in combination with tailor welded blanks, it waspossible to avoid the need for reinforcements and thus reduced the total number ofparts.

For mass reduction, steel sandwich material was applied in the spare tire tub andthe dash panel insert. Steel sandwich material contributes to 1.5% of the DH mass.

Due to the overall design, material specifications of steel sandwich material andrestrictions in its applications, such as low heat resistance and available size, thismaterial’s use was limited during Phase 2.

Chapter 5 - Page 20


Figure 5.3.7-1

2.7% - 420 MPa

2.5% - Ultra High Strength Steel > 550 MPa1.5% - Steel Sandwich

Material

13.5% - 280 MPa

7.6% - 140 MPa

45.1% - 350 MPa

27.1% - 210 MPa

Mild Steel 7.6%High Strength Steels 90.9%

Steel Sandwich Material 1.5%

Ò 140 MPaÒ 210 MPaÒ 280 MPaÒ 350 MPaÒ 420 MPaÒ > 550 MPa Ultra High Strength SteelÒ Steel Sandwich Material

Chapter 5 - P

age 21 Engineering S

ervices, Inc.

Ò 140 MPaÒ 210 MPaÒ 280 MPaÒ 350 MPaÒ 420 MPaÒ >550 MPaÒ Steel Sandwich Material

Figure. 5.3.7.-2 Material Grades of DH Parts

**

**

**

**

* See Assemblies 455 - 458

Chapter 5 - Page 22


7.6%

25.1%

0.8%3.0%

10.8%

4.2%

10.9%

2.1%

9.1% 8.4% 7.6%

3.0%4.4%

1.5% 1.5%

Material Thickness

0.65 0.70 0.80 0.90 1.00 1.20 1.30 1.40 1.50 1.60 1.70 1.80 2.00 Sandwich Misc

5.3.8. Material Thickness

The distribution of the used material sheet thicknesses relative to the DH mass isshown in Fig. 5.3.8-1. The majority of the mass (25%) is made from 0.7 mm sheetsteel. Parts with a large surface area such as the panel floor, the panel dash andthe panel roof are manufactured of high strength steel of this thickness, and areparts with secondary influence in crash performance.

All 1.3 mm thickness material is high strength steel with the yield strength rangingfrom 280 MPa (46%) to 350 MPa (54%). The parts made of 1.3 mm material usedin “conventional” stampings and tailor welded blank stampings have primaryinfluence on crash performance.

Since the demonstration hardware mass consists of 91% high strength steel, nearlyall parts are made from high strength steel sheets in a thickness ranging from0.65mm to 2.0mm.

Percent Distribution of Material Thickness Relative to DH Mass

Figure. 5.3.8-1

Chapter 5 - Page 23


5.4. Detail Design

PES executed an entirely paperless design using Computer Aided Design (CAD)and CATIA software for the detail design. With the involvement of part suppliers inthe United States and Europe, the Porsche R & D Center, in Germany, and thenecessary data exchange for the tool development and the design of the assemblyfixtures, this approach proved to be very efficient.

5.4.1. Weld Flange Standards

For the detail parts design it was important to define standards for the design of theweld flanges. The decision was made not to reduce the weld flange width for massreduction, which allowed the use of standard weld equipment for the demonstrationhardware assembly.

5.4.1.1. Weld Flanges for Spot or Laser Welding

For the design of parts to be spot welded, the flange length was designed to thePorsche standards shown in Fig. 5.4.1.1-1.

For the laser welding in assembly, the same standards were applied.

Figure 5.4.1.1-1 ULSAB Spot Weld Standards

Chapter 5 - Page 24


Figure 5.4.1.2-1 Part no. 81 Panel Package Tray Lower with Scalloped Flanges

The design is similar to the scalloped flanges used in production of the Porsche 911and Boxster. The second reason for scalloping weld flanges was to create two sheetspot welding where three sheet spot welding would have been applied, otherwise.Scalloped flanges were applied to parts not critical for sealing and not sensitive tocrash or durability. The mass reduction achieved with scalloped flanges on theselected parts, based on the calculated part mass equals 0.43 kg. (Fig. 5.4.1.2-4)

The flange geometry is shown in Fig. 5.4.1.2-2. The layout for a two sheet weldflange and a three sheet weld flange with scalloped flanges is shown in Fig. 5.4.1.2-3.

5.4.1.2. Scalloped Spot Weld Flanges

Scalloped flanges were used for mass reduction.

Chapter 5 - Page 25


Flange Geometry

Figure 5.4.1.2-2 Flange Geometry

Figure 5.4.1.2-3 Layout of 2 and 3 Sheet Weld Flanges

Two Sheet Weld Flange

Three Sheet Weld Flange

Chapter 5 - Page 26


Part Number

Part Name Calculated Part Mass[kg]

Calculated Part Mass withScalloped Flange [kg]

Mass Reduction[kg]

21 Panel Dash 6.180 6.140 0.040

28 Panel Cowl Lower 1.400 1.326 0.074

40 Pan Front Floor 15.934 15.892 0.042

45 Member Rear Suspension 1.486 1.440 0.046

55 Member Panel Back 1.450 1.424 0.026

68 Panel Wheelhouse Inner RH 2.141 2.110 0.031

69 Panel Wheelhouse Inner LH 2.141 2.110 0.031

81 Panel Package Tray Lower 1.700 1.594 0.106

140 Pan Rear Floor 4.330 4.298 0.032

0.428

5.4.1.3. Locator, Tooling and Electrophoresis Holes

Included in the detail part design are all locator holes for the assembly. All locatorholes needed for parts manufacturing and the holes necessary for theelectrophoresis of the body structure. After the location of the holes forelectrophoreses were first determined, they were then incorporated into the crashmodels and the crash analysis was performed to verify that their position did nothave any negative influence on the crash performance. After this verification, theholes were incorporated into the parts design.

Figure. 5.4.1.2-4 Mass Reduction with Scalloped Flanges

Chapter 5 - Page 27


5.4.2. Design Refinement

Phase 1 reflected a concept design. In Phase 2, the task was to make the designfeasible for manufacturing of the parts to maintain low mass and structuralperformances and also, to achieve the crashworthiness of the structure. In therefinement of the design, changes to the design concept were done for the followingreasons:

• Mass reduction• Manufacturing and tooling• Assembly• Material specifications• Crash performance• Package• Styling

The overview of design changes as shown in Fig. 5.4.2-1, names the parts or areasof the structure, the design change and the reason for the different solution orchange from Phase 1 to Phase 2.

Chapter 5 - Page 28


Overview of Major Design Changes in Phase 2Part Part / Location Description ReasonNo. Area of Change for Change

1 Fender Support RailHydroforming part was replaced with2 part stamping

Assembly, part manufacturing

2 Pan Front & Pan Rear Floor3 part front floor with sandwichmaterial tunnel deleted

Heat resistance of sandwich materialnot sufficient for bake hardeningprocess

3 Rear RailsSpring & shock absorber relocatedwith new rear suspension

Mass reduction, package

4 Front Rails Space between rails increased Package of bigger engine

Rear part of the front wheelhousedeleted

Mass reduction

5 Panel Skirt Redesigned, tailor welded blankPackage of new front suspension inconjunction with #4

Reinforcement shock tower deleted,integrated in new panel skirt

Mass reduction

6 Panel Spare Tire

Tub designed as separate modulefrom steel sandwich material and tobe bonded to the rear floor after finalassembly

Heat resistance of sandwich material,not sufficient for bake hardeningprocess

7 Package TrayRedesigned from 3 part to 2 partdesign roll formed member packagetray front deleted

Assembly

8Member Dash Front,Member Front FloorSupport, Member Kick-up

Material changed from high strengthto ultra high strength steel >550 MPayield strength

Front Crash, side impact crash

9 Panel Body Side OuterBlank configuration in tailor weldedblank with all blanks in high strengthsteels

Crash analysis, mass reduction

10 B-Pillar JointRocker inner extended upwards intoB-Pillar. B-Pillar lowerreinforcement modified

Side impact, crash assembly

11A-Pillar - Cowl - FenderSupport Rail-Hinge PillarJoint

Joint modified Assembly, revised fender support rail

12 Panel Back3 Piece design integrated into onepart

Mass reduction, assembly

13 Side Roof Rail Design refinements Manufacturing process - hydroforming

14 Bolt on Front End Welded Change in front end module concept

Figure 5.4.2-1



Chapter 6 - Page 1


Structural Performances Targets

Static torsion stiffness ≥ 13000 Nm/deg

Static bending stiffness ≥ 12200 N/mm

Normal modes (first modes) ≥ 40 Hz


6.1. Selected Tests for CAE

To verify that the ULSAB meets the targets set in the beginning of Phase 1, thefollowing tests were chosen for the static and dynamic stiffness.

For analytical crash testing the following tests were selected:


6.2. Static and Dynamic Stiffness

Based on CAD surface data the FE-Model (Figure 6.2-1) for the body in white wascreated. Because of the structure symmetry, only a half model with certainboundary conditions at the symmetry plane at y=0 for the static and dynamicstiffness simulations were used. The stiffness model consists in triangle andquadrilateral elements. To connect the different structure components, differentmethods were used. To connect laser welded parts in the FE-Model, the nodes ofthe flanges were equivalent. For spot welded areas the middle flange nodes areconnected with welding point elements. The weld point distance was with a point

Figure 6.1-1 Load cases and targets for static and dynamic stiffness

Chapter 6 - Page 2


Figure 6.2-1 FE-Model

distance of about 50 mm. The CAE configuration for the static and dynamicsimulations consist of the following parts:

• Welded Body Structure• Bonded Windshield and Back Light• Bonded and bolted Panel Dash Insert (Part-No. 022)• Bonded Panel Spare Tire Tub (Part-No. 050)• Bolted Reinforcement Panel Dash Brake Booster (Part-No. 115)• Bolted Braces Radiator (Part-No. 188)• Bolted Reinforcement Radiator Rail Closeout RH/LH (Part -No. 094/095)• Bolted Reinforcement Radiator Support Upper (Part-No. 001)• Bolted Tunnel Bridge Lower/Upper• Bolted Brace Cowl to Shock Tower Assembly

The stiffness model (per half model) consisted of:

• 54521 shell elements• 53460 nodes

The deformed shapes for the load cases torsion and bending are shown in theFigures 6.2.1-1 and 6.2.2-1. To view the stiffness distribution vs. the x-axis, thediagrams 6.2.1-2 (torsion) and 6.2.1-3 (bending) are used. The derivation vs. thex-axis for torsion (Fig. 6.2.1-3) and bending (Fig.6.2.2-3) as well as the strainenergy contour plots (Fig. 6.2.1-4 and Fig. 6.2.2-4) show the sensitive areas. Thecolored areas of the strain plots show the elastic energy, which is a result of the

Chapter 6 - Page 3


CAE Structural Performance

Static Torsional Stiffness 21310 Nm/deg

Static Bending Stiffness 20540 N/mm

CAE Mass* (with glass) 230.6 kg

CAE Mass* (without glass) 202.8 kg

First Torsion Mode 61.4 Hz

First Bending Mode 61.8 Hz

Front End Lateral 60.3 Hz

6.2.1. Torsional Stiffness

A load of 1000 N was applied at the shock tower front while the body structure wasconstrained at the rear center spring attachment in the lateral and verticaldirections.

Figure 6.2.1-1 Deformed Shape for Torsion

deformation stored in the structure, as internal energy. The deformed shape of thedynamic stiffness simulation, the normal modes are shown in the Figures 6.2.3-1 to6.2.3-3. The deformed frequency mode belongs to the normal modes mentioned inTable 6.2-2.

*Mass as in test configuration (Chapter 6, page 2), brackets and

reinforcements (6.4 kg) are not included (see Chapter 5, page 10)

Figure 6.2-2 Table of CAE Structural Performance

Chapter 6 - Page 4


500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500-0.05

-0.04

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

Longitudinal X-axis [mm]

Der

ivat

ion

of A

ngle

[de

g/m

m]

Support

Derivation of Torsion Angle

Shock TowerFront

Center, SpringAttachment Rear

500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500-0.01

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

Longitudinal X-axis [mm]

Ang

le =

ata

n (z

disp

/yco

or)

[deg

]

Support

Torsion Angle

Shock TowerFront


21310 Nm/deg

Figure 6.2.1-2 Torsion Angle vs. x-Axis

Figure 6.2.1-3 Derivation of Torsion Angle vs. x-Axis

Chapter 6 - Page 5


Figure 6.2.1-4 Strain Energy Contour Plot for Torsion

6.2.2. Bending Stiffness

The loads were applied to the center of the front seats and to the center of the twoouter rear seats. The measurements were taken under a load of F

b max = 4000 N

(4 x 1000 N).

Figure 6.2.2-1 Deformed Shape for Bending

Chapter 6 - Page 6


500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

Longitudinal X-Axis [mm]

Der

ivat

ion

of v

ertic

al

Z-D

ispl

acem

ent

[mm

] Support

Derivation of Vertical Z-Displacement

Shock TowerFront


Figure 6.2.2-2 z-Displacement vs. x-Axis, Bending

Figure 6.2.2-3 Derivation of z-Displacement vs. x-Axis, Bending

500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

Longitudinal X-Axis [mm]

Ver

tical

Z-D

ispl

acem

ent

[mm

]

Support

Vertical Z-Displacement

Shock TowerFront


20540 N/mm

Chapter 6 - Page 7


Figure 6.2.2-4 Strain Energy Contour Plot for Bending

Figure 6.2.3-1 Front End Lateral Mode

6.2.3. Normal Modes

Chapter 6 - Page 8


Figure 6.2.3-2 First Bending Mode

Figure 6.2.3-3 First Torsion Mode

Chapter 6 - Page 9


6.3. Crash Analysis

For three crash types of the ULSAB project, one common crash model wasgenerated. With this model the crash simulations were conducted:

• AMS 50% frontal offset crash at 55 km/h• NCAP 100% frontal crash FMVSS 208 at 35 mph• Side impact crash at 50 km/h (96/27 EG with deformable barrier)

For the rear crash (FMVSS 301) at 35mph only a half structure (Fig. 6.3.3-1) wasused. Fig. 6.3-1 shows the high level of detail for the FE-Model. To realize arealistic crash behavior of the simulation, all the spot welds and laser welded areaswere considered in the models. To analyze the crash behavior, all crash-relevantcar components were modeled, such as:

• Wheels with tire model• Engine and transmission• Steering system• Chassis system with subframe• Fuel tank• Bumper system including crashbox• Radiator with fan• Battery• Spare tire• Brake booster, ABS box and cylinder• Doors, front and rear without glass

The door concept used for all simulations was a typical two shell structure with aninner and outer panel, an upper door reinforcement and two high strength sideimpact beams at the front door and one side impact beam at the rear door.

A three point fixture with reinforcements at the hinges and the locks supported thedoors.

To reduce the model size for the roof crush analysis, the full model with reducedcontents was used (Fig. 6.3.5-1).

Chapter 6 - Page 10


A high level of detail of the surfaces, welding and mounting locations was necessaryto provide the resolution to be able to access the events. The LS-DYNA completefull model had 178386 elements and 174532 nodes.

Figure 6.3-1 Crash Analysis Model

Chapter 6 - Page 11


Curb Mass 1350 kg

Luggage 113 kg

Dummies 149 kg

Total Crash Mass 1612 kg

6.3.1. AMS Offset Crash

The AMS offset crash was defined in the year 1990 by the editor of the Germanautomotive magazine ‘Auto Motor Sport’ (AMS). The aim of this offset crash is tosecure the passenger compartment residual space. For this requirement a stiffpassenger compartment and a good energy absorption in the front structure isneeded. The initial velocity for the car is 55 km/h for the AMS crash.

The Offset barrier is a block with a 15 degree rotated contact area including twoanti-slide devices mounted on the contact surface. The left side of the car hits thebarrier with an overlap of 50%.

For actual crash tests AMS analyzes the following values:

• HIC-value (Head Injury Criterion)• Head, chest and pelvis acceleration• Maximum belt forces• Maximum femur forces• Dynamic steering deformation• Foot well intrusions• Door opening after test

Because the analysis did not include dummies, injury assessment could not bemade. Injury performance is greatly affected by the structural crash and steeringcolumn movement as well as by the knee bar design. Evaluation of passengercompartment intrusion can be made by looking at deformation in the foot well area(Fig. 6.3.1-4). Looking at the overall shape of the deformation (Fig. 6.3.1-2, -3 canassess structural integrity).

The vehicle mass was defined to be base curb weight plus two 50th percentile maledummies with 113 kg of luggage. The crash mass of the vehicle was set at1612 kg. The crash mass of the vehicle is calculated as follows:

Chapter 6 - Page 12


Figure 6.3.1-1 AMS Offset Crash Analysis Setup

The AMS Offset undeformed and deformed shapes are shown in Fig. 6.3.1-2 and6.3.1-3. The deformed shape in these figures is after 100 ms. The deformation inthe footwell area is shown in Fig. 6.3.1-4. The analyzed deformation is measured inthe foot well area where it is important to keep the deformations as low as possible,because of the injury of the passenger’s legs.

The internal energy absorption diagram in Fig. 6.3.1-5 gives an overview of theinternal energy absorbed in the parts subframe, bumper beam, crashbox, front railand fender side rail after 100 ms. The diagram in Fig. 6.3.1-6 shows the load pathfor the most important front structure components. The diagram shows the mainload path is the rail front. The fender side rail and the subframe have about thesame load level. The diagram, AMS Offset Crash Acceleration vs. Time (Fig.6.3.1-7) shows an average acceleration calculated from the rocker LHS, tunnel, androcker RHS. After the contact between AMS barrier and engine, a middleacceleration of about 25 g results in the passenger area. The Figure 6.3.1-8 showsthe function of the car deformation versus time. After about 90 ms the maximumdynamic deformation is reached.

Chapter 6 - Page 13


Figure 6.3.1-3 AMS Offset Crash Deformed Shapes of Longitudinals

Figure 6.3.1-2 AMS Offset Crash Deformed Shapes

t = 0 ms t = 100 ms

t = 0 ms t = 100 ms

Chapter 6 - Page 14


Figure 6.3.1-4 AMS Offset Crash Maximum Dynamic Foot Room Intrusion in mm

134 16

80

64146

92

40

39

9

36

8276

3360

102

Figure 6.3.1-5 AMS Offset Crash Internal Energy Absorption

Subframe

Bumper Beam

Crash Box

Rail Front

Fender S. Rail

0 10 20 30 40

Energy (kJ)

26.9

17.3

5.6

37.6

9.6

Chapter 6 - Page 15


Figure 6.3.1-6 AMS Offset Crash Typical Cross Section Forces

Figure 6.3.1-7 AMS Offset Crash Acceleration vs. Time

Subframe

Front Rail Ext.

Rocker

Rail Front

Fender S. Rail

0 20 40 60 80 100 120 140

Force (kN)

55

50

85

115

50

0 20 40 60 80 100-10

0

10

20

30

40

time [ms]

ax [g

]

Average Car Acceleration vs. TimeRocker LHS / Tunnel / Rocker RHS

-40

-30

-20

-10

0

+10

Chapter 6 - Page 16


Figure 6.3.1-8 AMS Offset Crash Deformation vs. Time

0 20 40 60 80 100-200

0

200

400

600

800

time [ms]

sx [m

m]

Car Deformation vs. Time

Chapter 6 - Page 17


In the following table (Fig. 6.3.1-9), the AMS crash events vs. time are explained:

Time (ms) AMS Offset Crash

12.00 Initial folding of longitudinal LHS

16.00 Initial folding of subframe

18.00 First buckling of rail upper in front of shock tower

36.00 Wheel LHS contacts barrier

40.00Engine contacts barrier, start of vehicle-rotation aroundz-axis

44.00

Deformable front end of the subframe totally deformed,stiffer rear end and the extension longitudinal LHS startsmoving rearwards and causes deformation in the frontfloor area, buckling of the longitudinal in the area of theshock tower

48.00 Second buckling of rail upper LHS behind the shock tower

52.00Buckling of the rear end of the subframe at the fixture onthe extension longitudinals

60.00Buckling of the brace cowl to shock tower LHS. Enginehits the steering gear.

68.00 Contact between gearbox-mounting and brake booster

70.00 Wheel LHS hits the hinge pillar

88.00 Maximum dynamic deformation reached

Figure 6.3.1-9 AMS Offset Crash Events

Chapter 6 - Page 18


This analysis shows good progressive crush on the barrier side (left), as well ascrush on the right, indicating transfer of load to the right side of the structure. Thistransfer means that the barrier side is not relied upon solely to manage the crashevent.

This transfer also contributes to the preservation of the occupant compartment.The intrusion of 146 mm into the footwell is minimal given the severity of this event.

The initial, early peak shown in the pulse graph should trigger air bag systems.

Peak deceleration of approximately 35 gs, a good result considering the severity ofthis event.

Chapter 6 - Page 19


Figure 6.3.2-1 NCAP 100% Crash Analysis Setup

6.3.2. NCAP 100% Frontal Crash

The conditions for the front crash analysis are based on several requirements. Inthe ULSAB program, the focus was on progressive crush of the upper and lowerload path, sequential stack up of the bumper, radiator, and powertrain, integritybetween individual components, A-pillar displacement, definition of the dooropening, uniform distribution of the load, toe pan intrusion, and passengercompartment residual space. These requirements contribute towards occupantsafety and the United State Federal Motor Vehicle Safety Standard, FMVSS 208.

The test sequence of the front crash analysis is set up to duplicate a 35 mph,National Highway and Traffic Safety Association (NHTSA) full frontal barrier test(Fig. 6.3.2-1).

Chapter 6 - Page 20


The NCAP 100% Frontal Crash undeformed and deformed shape is shown inFigures 6.3.2-2 and 6.3.2-3. The deformed shape in the figure is after 100 ms. Thedeformation in the footwell area is shown in Fig. 6.3.2-4. The analyzeddeformations are measured in the foot well area where it is important to keep thedeformations as low as possible, because of the injury of the passenger legs.

The internal energy absorption diagram in Fig. 6.3.2-5 gives an overview of theinternal energy absorbed in the parts subframe, bumper beam, crashbox, front railand fender side rail after 100 ms. The diagram in Fig. 6.3.2-6 shows the sectionforce for the most important front structure components. The diagram shows thatthe main load path is the rail front. The components, fender side rail and thesubframe have about the same load level. The diagram, NCAP Crash Accelerationvs. Time (Fig. 6.3.2-7), is an average of accelerations at the rocker LHS, tunnel,and rocker RHS. After the contact between barrier and engine it results a middleacceleration of about 29 g at the passenger area. The Figure 6.3.2-8 shows thefunction of the car deformation versus time. After about 68 ms the maximumdynamic deformation is reached.

Figure 6.3.2-2 NCAP 100% Crash Deformed Shapes

t = 0 ms t = 100 ms

Chapter 6 - Page 21


Figure 6.3.2-4 NCAP 100% Crash Maximum Dynamic Foot Room Intrusion in mm

Figure 6.3.2-3 NCAP 100% Crash Deformed Shapes of Longitudinals

58 51

85

7094

73

80

79

80

70

4045

5052

62

t = 100 mst = 0 ms

Chapter 6 - Page 22


Figure 6.3.2-5 NCAP 100% Crash Internal Energy Absorption

Figure 6.3.2-6 NCAP 100% Crash Typical Cross Section Forces

Subframe

Rail Upper

Rail Front

Crash Box

Bumper Front

0 10 20 30 40 50 60

Energy (kJ)

30

12.5

55.3

8

16

Subframe

Rocker

Rail Upper

Rail Front

Front Rail Ext.

0 20 40 60 80 100 120 140

Force (kN)

49

50

41

120

45

Chapter 6 - Page 23


Figure 6.3.2-7 NCAP 100% Crash Acceleration vs. Time

Figure 6.3.2-8 NCAP 100% Crash Deformation vs. Time

0 20 40 60 80 1000

200

400

600

800

time [ms]

sx [m

m]


0 20 40 60 80 100-10

0

10

20

30

40

time [ms]

ax [g

]Average Car Acceleration vs. Time

Rocker LHS / Tunnel / Rocker RHS

-40

-30

-20

-10

0

+10

Chapter 6 - Page 24


This analysis illustrates good progressive crush of the upper and lower structureand subframe. It shows peak deceleration of 31 gs, which is satisfactoryconsidering that this structure is designed with stiffer body sides to meet 50% AMSoffset crash requirements.

The pulse graph is sympathetic to current occupant restraint systems. It shows aconsistent rise to the peak of 31 gs then a smooth ride down to zero, indicating thatthe occupant would experience controlled restraint. The initial, early peak shouldtrigger air bag systems. Low intrusion at the footwell indicates that leg damage isunlikely.

Time (ms) NCAP Front Crash

12.00 Initial folding of longitudinal

16.00 Initial folding of subframe

21.00 First buckling of rails upper in front of shock tower

35.00 Engine contacts barrier

37.00Buckling of the rear end of the subframe at the fixture onthe extension longitudinals

50.00Rear end of longitudinals start to buckle behind thereinforcement (still stable)

51.00 Wheels contacts barrier


Figure 6.3.2-9 NCAP Front Crash Events

The following table (Figure 6.3.2-9) shows the NCAP crash events:

Chapter 6 - Page 25


6.3.3. Rear Crash

The conditions for the rear impact analysis are based on the United States RearMoving Barrier Test FMVSS-301. The test specifically addresses fuel systemintegrity during a rear impact. Automotive companies also include structuralintegrity and passenger compartment volume as additional goals for this test.

The impacting barrier is designed to represent a worst case rear crash (Fig. 6.3.3-1). The rear crash barrier is a rigid body with a mass of 1830 kg, making contact atzero degrees relative to the stationary vehicle. The Federal Standard identifies thatthe velocity of the rear moving barrier is 30 mph. The ULSAB program has raisedthe standard to 35 mph, which is 36% more kinetic energy of the moving barrier.

Evaluating fuel system integrity is done by representing a fuel tank system. Theadditional goals of passenger compartment integrity, residual volume, and dooropening after the test can be addressed by looking at the deformed shapes of thevehicle during the crash event. During the early stages of the impact, there shouldbe a little or no deformation in the interior. This sequence of events (Fig. 6.3.3-8)is necessary up to the time that the tires make contact with the barrier face andtransfer load to the suspension and the rear of the rocker panel.

For the rear crash a half structure model was used. The rear crash deformedshapes are shown in Fig 6.3.3-2. To analyze the rear passenger compartmentintegrity, Figure 6.3.3-3 shows that maximum dynamic intrusion in this area.

The diagram (Fig. 6.3.3-4) shows the energy absorption, and the cross sections ofthe main hood load paths are shown in Figure 6.3.3-5. Due to the results, the rearrail and the rocker were the most important hood paths of the rear structure.

The Rear Crash Acceleration vs. Time (Fig. 6.3.3-6) shows an average accelerationof the rocker RHS and the tunnel. Figure 6.3.3-7 shows the total car deformation, atapproximately 85 ms, the maximum dynamic deformation was reached.

Chapter 6 - Page 26


Figure 6.3.3-1 Rear Crash Analysis Setup

Chapter 6 - Page 27


Figure 6.3.3-2 Rear Crash Deformed Shapes

t = 0 ms t = 100 ms

t = 0 ms t = 100 ms

Chapter 6 - Page 28


Figure 6.3.3-4 Rear Crash Internal Energy Absorption (kJ)

Figure 6.3.3-3 Rear Crash Maximum Dynamic Room Intrusion (mm)

5

120

73

53

38

2

33

66

4

X

X

X

XX

X

X

X

X

Rear Rail

Crash Box Rear

Panel Rear Floor

Bumper Rear

0 5 10 15 20 25

Energy (kJ)

20.2

1.4

6.3

1.1

Chapter 6 - Page 29


Figure 6.3.3-5 Rear Crash Typical Cross Section Forces (kN)

Figure 6.3.3-6 Rear Crash Acceleration vs. Time

Rocker

Rear Rail

Rail Side Roof

Spare Wheel

0 10 20 30 40 50 60 70 80 90

Force (kN)

50

80

15

20

0 20 40 60 80 100-10

0

10

20

30

40

time [ms]

ax [g

]

Average Car Acceleration vs. Time

Chapter 6 - Page 30


0 20 40 60 80 1000

200

400

600

800

time [ms]

sx [m

m]


Figure 6.3.3-7 Rear Crash Deformation vs. Time

Chapter 6 - Page 31


Time (ms) Rear Crash

4.00 Initial folding of longitudinals rear

20.00 Spare tire contacts barrier

35.00 First buckling of crossmember rear suspension

40.00 Spare tire hits crossmember rear suspension

44.00 Buckling of the crossmember rear suspension

48.00Buckling of the rear end rocker at the connection tolongitudinal rear

52.00 Collapse of crossmember rear suspension

56.00 Buckling of the front end longitudinal rear


Figure 6.3.3-8 Rear Crash Events

The following table (Fig. 6.3.3-8) explains the rear crash events after impact:

This analysis shows that the structural integrity of the fuel tank and fuel filler wasmaintained during the event, so no fuel leakage is expected. The spare tire tubrides up during impact, avoiding contact with the tank.

Rear passenger compartment intrusion was restricted to the rear most portion of thepassenger compartment, largely in the area behind rear seat. This result is due togood progressive crush exhibited by the rear rail.

Chapter 6 - Page 32


6.3.4. Side Impact Analysis

The conditions for the side impact analysis are based on a European Side MovingBarrier Test. The European test specifically addresses injury criterion based ondisplacement data gathered from EUROSID side impact crash dummies.Automotive companies also include post-crash structural integrity and passengercompartment as additional requirements for this test.

The actual European side moving barrier uses a segmented deformable face whichcomplies with a required set of different load versus displacement characteristicsand geometric shape and size requirements. The barrier used in the analysis (Fig.6.3.4-1) conformed to the geometric requirements (i.e., ground clearance, height,width, bumper depth). The European specification requires the impacting barrier tohave a mass of 950 kg, making contact at ninety degrees relative to the vehiclelongitudinal axis. The center line of the barrier is aligned longitudinally with the frontpassenger ‘R-point’. The R-point is a car specific point which is defined by the seat/passenger location. The velocity of the side moving barrier at time of impact isdesignated to be 50 km/h.

Because the scope of analysis did not include side impact dummies, injuryassessment could not be made. Injury performance is greatly affected by interiortrim panel and foam absorber design as well as by structural crush. Evaluation ofpassenger compartment intrusion can be made by looking at door and B-pillardisplacements and intrusion velocities. Structural integrity can be assessed bylooking at the overall shape of the deformation, including any gross buckling of theB-pillar, rotation of the rocker rails, crush of the front body hinge pillar, folding of thedoor beams and door belts, and cross-car underbody parts such as the seatattachment members and the rear suspension cross member.

Chapter 6 - Page 33


The side impact undeformed and deformed shapes are shown in Fig. 6.3.4-2 and6.3.4-3, with the deformed shapes shown after 80 ms of impact.

During the early stage of the impact, the outer door structure crushes, the B-pillar isstable. As the impact progresses the rocker starts to buckle and causes also abulging of the floor section. At about 30 ms, the still stable structure of the B-pillaris moved by the barrier inside the car and therefore the roof starts to bulge. After40 ms the B-pillar develops an inward buckling. After about 64 ms the maximumdynamic deformation is reached.

For the injury performance, the intrusion velocities of the structural parts, whichcould come in contact with the passengers, are important. Figures 6.3.4-5 and6.3.4-6 show the intrusion velocities of typical points at the inner front door panel(No. 238) and the B-pillar inner (No. 235) (Fig.6.3.4-4).

The following Figures 6.3.4-2 and 6.3.4-3 show the deformed shape of the sidestructure:

Figure 6.3.4-1 Side Impact Crash Analysis Setup

Chapter 6 - Page 34


Figure 6.3.4-2 Side Impact Crash Deformed Shapes

t = 0 ms

t = 80 ms

Chapter 6 - Page 35


Figure 6.3.4-4 Side Impact Time History Node

Figure 6.3.4-3 Side Impact Crash Deformed Shapes of Side Structure

t = 0 ms t = 80 ms

No. 238

No. 353

No. 353

No. 238

Measured points for velocity Lower B-pillar enlarged

Chapter 6 - Page 36


0 50 100 150 200 250 300-2

-1

0

1

2

3

4

5

6

7

8

9

10

Y - Intrusion [mm]

Y -

Vel

ocity

[m

/s]

Velocity vs. IntrusionB-Pillar No 238

0 50 100 150 200 250 300-2

-1

0

1

2

3

4

5

6

7

8

9

10

Y - Intrusion [mm]

Y -

Vel

ocity

[m

/s]

Velocity vs. IntrusionDoor Inner Panel No 353

Figure 6.3.4-5 Side Impact Velocity vs. Intrusion at Node 353

Figure 6.3.4-6 Side Impact Velocity vs. Intrusion at Node 238

Chapter 6 - Page 37


The body side ring and doors maintained their integrity with only 248 mm ofintrusion. The velocity of the intruding structure was tracked to determine thedegree of injury an occupant may sustain. The maximum velocity was only8 meters per second. The event is considered complete when the deformablebarrier and vehicle reach the same velocity, in this case at 64 msec.

Time (ms) Side Impact

16.00 Buckling of the rocker in front of B-pillar

28.00 Buckling of the floor

35.00 Buckling of the roof

40.00 Buckling of the roof frame at the B-pillar

44.00 Buckling of the member kick up, still stable

48.00 Buckling of the brace tunnel


Figure 6.3.4-7 Side Impact Crash Events

The following table (Fig. 6.3.4-7) shows the side impact crash events:

Chapter 6 - Page 38


6.3.5. Roof Crush (FMVSS 216)

The conditions for the roof crush analysis are based on United States, FMVSS 216.This requirement is designed to protect the occupants in event of a rolloveraccident. The surface and angle of impact are chosen to represent the entirevehicle impacting the front corner of the roof.

The federal standard requires roof deformation to be limited to 127 mm (5 inches) ofcrush, and roof structure to support 1.5 times the vehicle curb mass or 5,000 lbs(22249 N), whichever is less.

For test purposes and repeatability, the complete body in white is assembled andclamped at the lower edge of rocker and the roof crush test is done in a quasi-staticforce versus displacement arrangement. In the computer analysis, the softwareprogram, LS-DYNA, requires that the roof crush be done in a dynamic, movingbarrier description as compared to the quasi-static test.

Figure 6.3.5-1 shows the undeformed shape of the FE-Model used for the roof crushsimulation. The shape of the structure after the limit of 127 mm deformation isshown in Figure 6.3.5-2.

The force versus displacement curve is shown in Fig 6.3.5-3. The peak force of36150 N is reached after a deformation of 72 mm of roof crush. Based on the curbmass of 1350 kg, the crush force of 19865 N is required for the federal standardsFMVSS 216. The analysis was continued to 127 mm (5 inches) of deflection inorder to determine the ability of the roof to sustain the peak load past 72 mm ofcrush. The analysis shows that the roof meets the peak load requirements and issteady and predictable.

Chapter 6 - Page 39


Figure 6.3.5-1 Roof Crush Undeformed Shape

Figure 6.3.5-2 Roof Crush Deformed Shape

Chapter 6 - Page 40


Figure 6.3.5-3 Roof Crush Deformation vs. Force

Analysis showed that 22.25 kN was reached within 30 mm of crush. The structureresisted the applied load all the way up its peak of 36.15 kN and continued tomaintain it quite well even after peak, when it dropped to about 28 kN at 127 mm.The load was well distributed through the A, B and C-pillars and down into the rearrail.

6.4. CAE Analysis Summary

For the AMS Offset crash test the overall deformation and intrusion are the criticalfigures. For the NCAP crash test, the critical figure is the vehicle crash pulse. Thetarget for the offset crash was to achieve low footwell intrusion. It is important toachieve a good balance between these two targets. The results of the crashanalysis show that for the ULSAB a good compromise has been found to fulfill theAMS as well as the NCAP frontal crash, considering the dependencies betweenthese two crash types.

To achieve the low footwell intrusion for the AMS crash a rigid front structure isneeded. A rigid front structure, however, means higher acceleration in the NCAP

0 25 50 75 100 125 150

-5

0

5

10

15

20

25

30

35

40

Deformation [mm]

For

ce [

N]

Force vs. Deformation

127

Chapter 6 - Page 41


test and results in higher HIC (Head Injury Criteria) values for the passengers, witha maximum footwell intrusion of 149 mm for the AMS Offset crash and a maximumacceleration of 30.4 g for the NCAP crash, the ULSAB structure shows a goodbalance in these criteria. The results also document the high safety standards ofULSAB, especially if one considers that the NCAP crash analysis was run at 5 milesabove the required speed of 30 mph and 36% more energy had to be absorbed.

The rear crash test requirements are addressing the fuel system integrity and lowdeformation in the rear seat area. The analysis shows no collapse of thesurrounding structure of the fuel tank, contact with the fuel tank itself or the fuel fillerrouting. Considering the fact that there was no rear seat structure the analysis alsoshows a low deformation of the rear floor. For the rear crash analysis in the ULSABprogram, the requirement was raised from 30 mph to 35 mph velocity of the rearmoving barrier, resulting in an increase of 36% of its kinetic energy.

In the side impact crash test, good performance means acceptable intrusion of theside structure at low intrusion velocity. For both criteria the ULSAB achievedsatisfactory results. The analysis shows a maximum intrusion of 250 mm and anintrusion velocity of 8 m/s at the inner door panel and the B-pillar. It is assumed thatin a fully equipped car the intrusion will be even lower.

For the roof crush test the Federal standard requires the roof deformation to belimited to 127 mm of crush and the structure to support 1.5 times the curb mass or5000 pounds, whichever is less. The force requirement of 19500 N was already metat 27 mm of crush. The continued analysis showed that the structure is steady andpeak load of 36 kN was met after 72 mm of crush. This result confirms the role theside roof rail plays as important part of the ULSAB structure.

The ULSAB crash analysis has shown that reducing the body structure mass usinghigh strength steel, in various grades and in applications such as tailor weldedblanks combined with the applied joining technologies in the assembly, such aslaser welding, does not sacrifice safety.

The goal was to maintain the high standards of state-of-the-art crash requirements,without compromising the ULSAB program goal to significantly reduce the bodystructure mass. The crash analysis of the ULSAB supports that this goal isreached.


7. Material & Processes

Chapter 7 - Page 1


7. Material and Processes

7.1. Material Selection

7.1.1. Material Selection Process

Based on ULSAB Phase 1 results, the body structure was redesigned in Phase 2 asdescribed in earlier chapters of this report. With respect to the new influences,such as crash requirements and styling, new calculations had to be made. Thecalculations concerning static behavior gave us a first indication of the sheet metalthickness needed. This is because performance is mainly related to sheet metalthickness and the design itself, and not to the strength of the material, because theE-modulus is very similar for all steel types. After the initial material selection, thefirst loop of crash calculations was performed. As a result, the material grades and/or the sheet metal thicknesses had to be adjusted.

Several iterations of the “Material Selection Process” (Figure 7.1.1-1) lead us to theoptimal strength/thickness level for each part. This procedure included amanufacturing feasibility check with our selected part suppliers. For the mostcritical parts, a forming simulation was performed simultaneously by the steelsuppliers. The results of these simultaneous engineering processes have beenimportant factors in successfully meeting the challenges of developingmanufacturable parts. Different criteria during the material selection process suchas formability, weldability, spring-back behavior, and static and dynamic propertieswere always taken into consideration.

Always having “Production Intent” in mind, the focus was on production-readymaterials, not on materials that are available only in laboratory scale. Generalmaterial specifications and the definition of the different material grades aredescribed in section 7.2 of this chapter.

Chapter 7 - Page 2


Figure 7.1.1-1

7.1.2. Definition of Strength Levels

In order to use the minimum variety of materials, every “master item” was definedby thickness and strength. The same master item could be used for different parts,as long as thickness and strength requirements were met, and the part suppliersand forming experts had no concerns. The definition of strength levels as used inULSAB Phase 2 is shown next in the “ULSAB High Strength Steel Definition.”

Material Selection Process

MeetsStatic

Targets

Material / ThicknessSelection,

Design Modification

MeetsStatic

Targets

Create / ModifyPhase 2 Crash Model

Meets Static/Crash

Targets

Parts FeasibleMeets

Static / CrashTargets

Build of FirstTest Unit

Build of FinalDemonstration Hardware

Create / ModifyPhase 2 Shell Model

Modify DesignMaterial / Thickness

AdjustementNo

Yes

Yes YesNo

Yes

Yes

No

No

No

Phase 1 Package /

Concept Design

Phase 2 Package

Refinement

CreateStyling Concept

Modify Package/Styling / Design

Modify Phase 1Shell Model

Steel Supplierand Part Supplier

Input

Start

Chapter 7 - Page 3


MinimumYield Strength Category

140 MPa Mild Steel

210 MPa High Strength Steel




Greater than 550 MPa Ultra High Strength Steel

ULSAB High Strength Steel Definition

The ULSAB program designates steel grades by specified minimum yield strengthin the part. The following steel grades are utilized in the ULSAB design:

This definition was chosen in order to standardize the steel grade definitions for theULSAB Consortium member companies since many countries are involved and thestandards are not the same around the world. This has to be seen together with thegoal that the ULSAB body structure could be built in every region of the world wheresteel is available. This is also the reason that the suppliers of the material for theDHs are kept anonymous within the ULSAB program.

The most suitable material for each part application was chosen with the assistanceof experts from the steel suppliers. This process was especially important for theultra high strength steel because of its more critical forming behavior. Differentmaterials such as dual phase (DP) steels are included in this group of ultra highstrength material parts.

There are several ways to achieve the 280 MPa yield strength level according to theabove definition. This could be done by using microalloyed high strength steel,bake hardening or even dual phase steel. However it is achieved, the minimumyield strength for the finished part has to be 280 MPa in each area of the part.Other material qualities and material types could achieve the same or similarresults; therefore, several factors affected material selection including materialperformance and availability.

Chapter 7 - Page 4


7.1.3. Supplier Selection

Once the “master items” were defined, the material supplier selection was made.This was done in material group meetings attended by all steel supplier experts andthe design and manufacturing team of PES. For every part of the ULSAB, aminimum of two material sources were selected.

The fact that different materials with the same yield strength level were available foreach part (not only from different suppliers, but also in many cases differentmaterial types, such as microalloyed or dual phase) shows that most of the ULSABparts could be made in multiple ways. No specially treated or designed materialwas necessary. Most of the material was taken from normal serial production at thesteel mills.

In order to practice simultaneous engineering most efficiently, the material supplierswere selected by their close proximity to the part supplier’s location (press shop). Ifthe material failed during the first try-outs it was easier to react with corrective stepssuch as circle grid analysis, material tests, or forming simulations.

Similar criteria were used in selecting the welding sources for the tailor weldedblanks. In most cases two different companies could have provided the samewelded sheet, each with slightly different material qualities. This again underscoresthat the ULSAB can be built with widely available material and part manufacturingtechnology.

Chapter 7 - Page 5


7.2. Material Specifications

7.2.1. General Specifications

General specifications for the material used on the ULSAB only concernedthickness tolerances, coating requirements and coating tolerances. Thespecifications are as follows:

• Actual thickness of blanks must measure +0.00 mm/-0.02 mm of thespecified thickness

• Coating may be electro-galvanized (Zn only) or hot dip (Zn or ZnFe)• Coating thickness must be 65 gram/m² maximum (0.009 mm) per side with

coating on both sides

Every delivered material had to be tested at the supplying source before it wasshipped to the part manufacturer. A test report accompanied the material until theparts are finished. This is the basis for the Advanced Quality Planning (AQP) reportthat was performed by the ULSAB Consortium. The test results are also consideredfor welding parameter evaluation at the prototype shop.

7.2.2. Material Classes

7.2.2.1. Mild Steel Definition

Mild steel, which is described in Sec 7.1 Material Selection, is material with a yieldstrength level of 140 MPa. Mild steel can also be defined in terms of “Draw Quality,”“Deep Draw Quality” or “Extra Deep Draw Quality.” The material has no fixedminimum yield strength but does have a minimum elongation. Mild steels are themost common steels used in auto making today. This is because mild steel hasforming and cost advantages compared to high strength steel. On the other hand,the ULSAB clearly shows that the amount of high strength and ultra high strengthsteel can be used up to more than 90% or more without any cost penalty.

Chapter 7 - Page 6


7.2.2.2. High Strength Steel Definition

The steel industry has developed various high strength steel qualities. In theULSAB Phase 2 program the strength levels of 210, 280, 350 and 420 MPa weredefined as high strength steel. The values are related to the strength of the finishedparts as assumed in the FEA model. This includes additional strengthening as aresult of the bake-hardening process also.

High strength steels were used where the design required certain crash andstrength characteristics. Within the range of this material group, differentstrengthening mechanisms can contribute to the final result. The DHs used micro-alloyed steels, phosphor-alloyed steels, bake-hardening steels, isotropic steels,high-strength IF - steels and dual-phase steels, all in the range of the above-mentioned yield strength. This engineering report does not include a detaileddescription of alloying or other metallurgical processes that are used to producethose steel types.

7.2.2.3. Ultra High Strength Steel Definition

Ultra high strength steels are defined as steels with a yield strength of more than550 MPa on the finished part. Parts made from these steels can provide additionalstrength for front and side impact. In the ULSAB structure, all crossmembers of thefloor structure were designed in ultra high and high strength steel.

Today, there are different ways to achieve needed strength levels. This could bedone for automotive sheet panels with dual phase (DP) steels, or with boron-alloyedtypes, which have to be hot formed. Within the ULSAB Phase 2, parts were madefrom DP steels. DP steels were feasible even on parts with a complex shape likethe cross member dash. As of today, those types were also available in anappropriate thickness range, which is interesting for automotive applications, e.g. athickness between 0.7 and 1.5 mm.

Chapter 7 - Page 7


7.2.2.4. Sandwich Material Definition

The use of sandwich material has contributed to considerable mass savings on theULSAB. The sandwich material is made with a thermoplastic (polypropylene) core,which has a thickness of about 0.65 mm. This core is “sandwiched” between twothin outer steel sheets with a thickness of about 0.14 mm each. The polypropylenecore of this sandwich material acts as a spacer between the two outer sheets,keeping the outer surfaces away from the neutral axis when a bending load isapplied (see fig. 7.2.2.4-1). The mentioned material (total thickness about 0.96 mmwhen coated) has a very similar behavior compared to a solid sheet of steel with athickness of about 0.7 mm.

This sandwich material shares many of the same processing attributes with steelsheets, like deep drawing, shear cutting, bonding, etc. But, unfortunately, it cannotbe welded. Even mechanical joining like riveting, clinching or screwing, can be aproblem when the material has to go through the paint-baking oven. The corematerial is softened by the heat and flows away from the area where a pretensionfrom a screw is applied. This may lead to a loss in joining strength.

Therefore, applications used in the ULSAB Phase 2 design were with parts madefrom sandwich material that did not go through the oven. The spare tire tub isdesigned as a prepainted module, preassembled with spare tire and tools. Thismodule will be dropped into place and bonded to the structure during the finalassembly of the vehicle. No additional heat has to be applied. Another applicationof sandwich material is the dash panel insert, which was bolted and bonded into thepanel dash during final vehicle assembly.

Figure 7.2.2.4-1 Sandwich Material

Steel Sheet 0.14 mm

Steel Sheet 0.14 mm

Polypropylene Core 0.65 mm

Chapter 7 - Page 8


F

Figure 7.2.2.4-2 Test Installation

Because there was no application similar to the spare tire tub in the past, anextensive forming simulation was performed on this part. Once the design wasadjusted using the results of the simulation, there were no major concerns about thefeasibility of the spare tire tub. After a small refinement of the best drawable radius,the parts were determined to be manufacturable with no problems.

Furthermore, a physical test with the spare tire tub was performed to check thefatigue behavior of this material for the application. Parts from the describedsandwich material were made and compared to parts made from solid steel sheetsof 0.7 mm thickness. A picture of the test installation is shown below in Fig.7.2.2.4-2.

Chapter 7 - Page 9


The load signal that was applied was taken from Porsche’s proving ground andadjusted to the situation of the ULSAB. The test concluded there are no restrictionsfor the use of the sandwich material for the proposed application when it iscompared to a conventional design using a 0.7 mm solid steel sheet.

The parts that were designed for the ULSAB could be made up to 50% lighter thanthose made of solid steel under similar dimensional and functional conditions. But,higher costs for the sandwich material have to be taken into consideration ascompared to normal coated steel sheets.

7.2.3. Material Documentation

As mentioned earlier, every “Master Item” (material defined by thickness andstrength) was accompanied by a test report, which includes all important strengthproperties, r- and n- values and a coating description. Those tests were performedby the supplying steel mills. All the supplied materials are documented at PES withtheir corresponding values, such as blank size, properties, coatings, material typeetc. The “Master List” was also the base for the documentation of the weldingparameters and the DH build itself.

When the parts were manufactured, the above-mentioned documentation wascompleted with additional information concerning press conditions for parts made atdifferent locations. For those parts where a forming simulation and/or a circle gridanalysis were performed, the documentation was extended with the results fromthese additional steps. These results are included in the earlier mentioned AQPreport.

To ensure proper and comparable documentation, material samples from every part,that goes into the DH were collected by PES and sent to a central testing source.At this neutral location, every collected material was tested in the same way anddocumented again.

Chapter 7 - Page 10


7.3. Tailor Welded Blanks

Introduction

Tailored blanking for vehicle body structures is a well known process with the firstapplications being done for mass production which started in 1985. Below listed arethe main reasons for PES´s decision to use tailor welded blanks in a relatively largenumber compared to vehicles already on the market:

• Mass reduction due to the possibility of placing optimum steel thicknessesand grades where needed

• Elimination of reinforcements with appropriate material gage selection• Simplified logistics due to the reduction of parts• Investment cost reduction of dies, presses etc. due to fewer production

steps• Better corrosion protection by the elimination of overlapped joints• Improved structural rigidity due to the smoother energy flow within the

tailor welded blank parts• Better fatigue and crash behavior compared to a conventional overlapped

spot welded design solution

7.3.1. Selection of Welding Process

Laser welding and mash seam welding are the most common processes for themanufacturing of tailor welded blanks today. Induction and electron beam weldinghave a minor importance and they are still under development. All these processeshave their advantages and disadvantages, related to the process and the machineitself.

Induction welding is a butt welding process. The necessary compressing of the twosheets creates a bulge with the consequence of an increase in thickness in thejoined area. Those blanks could not be used in visible areas without an additionalsurface finishing process. A high accuracy during the movement of the sheets isimportant. The heating of the weld seam by induction / magnetic current over thetotal length leads to a larger heat affected zone when compared to laser weldedblanks.

Chapter 7 - Page 11


The non-vacuum electron beam welding process is similar to laser welding in theresult of the weld seam geometry. This is due to the fact that it is a non-contactprocess as well. The beam is a mass beam and the kinetic energy of this beam isused for heating the material. The beam can be focused by a magnetic spool andthe diameter can be adjusted easily. The advantage of this process compared tolaser is the increased efficiency of about 90% compared to 10% when using laser.But a disadvantage is that the electron beam creates x - rays. This influences themachine design dramatically regarding total investment and material handling.Therefore this process is not used extensively up to now.

Mash seam welding needs a narrow overlapping of the sheets which have to bewelded. The material in this area becomes doughy, not really fluid. During thewelding process the current flows from one electrode to the other one and byresistance heating the sheet material becomes doughy. The electrode force thenmashes the weld area and the sheets are joined together in this way. This lightoverlap and the joining process by force loaded electrodes results in a weld zonebetween 2.5 and 3.0 mm. The coating maybe is affected in this zone negatively.Furthermore, experience has shown that the surface of the weld zone, where littlecaves and pinchers occur due to the mash welding process, may not achieve therequired corrosion resistance.

The laser welding process is used more and more widely. It is a non-contactwelding process, and the heat is brought into the material by a coherent light withhigh energy density. In this way a very narrow weld zone can be achieved. There isalmost no influence on the corrosion resistance when coated material is used. Themain critical point on this process is without any doubt the need for very preciselyprepared edges of the sheet. But this problem could be overcome by today’savailable precise cutting technologies or advanced fixing and clamping devices.One of the biggest advantages is the possibility of a non-linear weld line layout.

Different combinations of laser sources and clamping devices are on the markettoday. In many cases the sheets are moved relative to the fixed laser beam. Thismay lead to a reduction of the cycle time of the whole process.

Chapter 7 - Page 12


Together with the fact that most of the newest installations for welding blanks arelaser equipped devices, and the positive experience of PES, has lead to thedecision to use laser welded tailored blanks on the ULSAB body structureexclusively. The blanks were produced at different locations using differentequipment from the whole range of possible installations. The weld lines werecontrolled during the joining process to maintain the following features:

• width of the remaining gap• mismatching of blank edges• blank position• seam geography (concavity, convexity)• lack of penetration

All of these lead to the high quality of today’s tailor welded blanks.

7.3.2. Weld Line Layout

The weld line layout was mainly driven by the crash calculation results. Formingfeasibility requirements also influenced it. On some of the most critical parts, e.g.the body side outer panel, a forming simulation was performed. Necessary changesfrom this simultaneous engineering process were incorporated in the weld linelayout.

The following parts on the ULSAB body structure were designed as tailor weldedblanks:

• Front Rail Outer• Front Rail Inner• Panel Rocker Inner• Rear Rail Inner• Rear Rail Outer• Panel Body Side Outer• Panel Wheelhouse Outer• Panel Skirt

Chapter 7 - Page 13


The weld line layout is shown in the following pages for each part.

ULSAB 008 - Rail Front Outer

2.0 (350 MPa)1.5 (350 MPa)

1.6(350 MPa)

ULSAB 010 - Rail Front Inner

1.6 (350 MPa) 1.8 (350 MPa)1.5 (350 MPa)

ULSAB 042 - Panel Rocker Inner

1.7 (350 MPa) 1.3 (350 MPa)

Chapter 7 - Page 14


ULSAB 046 - Rail Rear Inner

1.6 (350 MPa) 1.3 (350 MPa) 1.0 (350 MPa)

ULSAB 048 - Rail Rear Outer

1.6 (350 MPa) 1.3 (350 MPa) 1.0 (350 MPa)

ULSAB 060 - Panel Body Side Outer

1.5(350 MPa)

0.9 (280 MPa)

1.3 (280 MPa)

1.7 (350 MPa)

0.7 (210 MPa)

Chapter 7 - Page 15


ULSAB 070 - Panel Wheelhouse Outer

0.8 (210 MPa)

0.65 (140 MPa)

ULSAB 096 - Panel Skirt

2.0 (140 MPa)

1.6 (140 MPa)

Chapter 7 - Page 16


7.3.3. Production Blank Layout

7.4. Hydroforming


Today, tubular hydroforming is a well-established process in automotivemanufacturing. When ULSAB Phase 1 began several years ago and hydroformingwas chosen as the manufacturing process for the side roof rail, the technology wasbeing used mainly for exhaust pipes and some front cradles. These had a muchsmaller diameter-to-thickness ratio compared to the ULSAB side roof rail. But withthe focus on mass savings, it was assumed that hydroforming could reduce thenumber of parts while helping to optimize available package space.

Figure 7.3.3.-1 For the Economic Analysis cost calculation purposes, the production blanklayout for the tailor welded blank parts was developed.

Chapter 7 - Page 17


The hyroforming process is described very simply as: “put a tube between a lowerand an upper die, close the die, fill the tube with water and increase the internalpressure in order to force the tube to expand into the shape of the die.” However,several things must be taken into consideration within this process technology. Thismethod will work only for straight tubes. In all other cases the tube has to be pre-bent or preformed depending on the final shape. The various steps necessary forthe manufacturing of the ULSAB side roof rail will be explained in the next section.


As explained in the Phase 1 report, the use of hydroformed parts instead ofconventionally formed and spot-welded structures have certain apparentadvantages. Because of the absence of flanges, available space could be utilizedwith higher efficiency (bigger cross sections were achievable). The homogeneoushydroformed parts also provide an improved load flow in comparison to otherstructural members made of several parts joined by spot welding. The side roof railrepresents a significant structural member in the ULSAB structure and provides anoptimal load distribution from the A-pillar along the roof into the B and C-pillar. Thisis true for the static as well as for the dynamic behavior of the body structure. Alsothe side impact and the rear crash support is affected positively. The interior of thevehicle is well protected by the “roll bar” design of these two structural membersintegrated into the body structure.

The hydroformed parts described in ULSAB Phase 1 already have led to similarapplications in vehicles that are on the road today. There is a high potential forfurther steel applications on comparable parts that are loaded with high forces.Other opportunities for hydroformed steel structures will be in the area of protectionsystems for convertibles.

Chapter 7 - Page 18


7.4.3. Forming Simulation (Review)

First, a feasibility check was made using the predicted bending line along withanalyzing the material distribution over the circumference in different cross sections.Next, the design of the side roof rail was analyzed and optimized for feasibility byconducting a forming simulation. Simultaneous engineering was used by the teamconsisting of PES and the part manufacturer; a similar approach was used for thedevelopment of the conventional stamped parts.

Conducting a forming simulation for parts like the side roof rail is much morecomplex than for stamped parts. This is because material properties that areaffected by a combination of processes such as prebending, preforming andhydroforming are very difficult to calculate. The first forming simulation has shownthat wrinkles will occur during a very early stage of the forming process in the areawhere the tube was first prebent. The next step is to preform in a different directionto make it fit into the hydroforming tool. A picture of this area taken from theforming simulation program is shown in Figure 7.4.3-1.

As a result of this analysis the design of the side roof rail was modified so thatsome bending radii were softened. Also some other areas were slightly changed inorder to prevent excessive material thinning or cracking during the forming process.The forming simulation also led to the decision of using a separate preforming tool(described in Sec. 7.4.5).

Figure 7.4.3-1 Forming Simulation

Chapter 7 - Page 19


7.4.4. Tube Manufacturing

Certain material qualities have to be defined. Standard tubes, beside the fact thatthe required diameters with the needed thin wall were not available commercially,have no high demand concerning transversal elongation. But this is one of the mainfactors during the hydroforming process when the tubes are expanded. Even if thedifference in diameter on different cross sections of the tube is relatively low, certainareas of the ULSAB hydroformed side roof rail required a high degree of elongation.During the design process, differentiation must be made between local elongation(between two points of the circumference) and the overall elongation (totaldifference in circumference in a cross section). These two factors must also betaken into consideration for the longitudinal shape of the part. Transitions betweenshape changes of the cross sections should be as smooth as possible and highelongation is needed.

The above mentioned facts led to the decision to manufacture tubes for the ULSABside roof rail from material different to what is used for conventional tubes. Tubeswere made, therefore, from high strength steel sheets to meet yield strengthrequirements and to have uniform elongation in both directions. High workhardening, which should be achievable by this material, is an important factor aswell.

Tubes can be made in several different ways. One way is to manufacture them witha continuous roll forming and high frequency welding. This has to be done withextremely high accuracy of the weld geometry especially on such thin walled largediameter tubes. Because the burr (which is unavoidable in this process) has to beremoved in an additional planing operation (scarfing), not all of the welds are able tomeet the tube specifications. Another approach is to use non-contact laser weldingfor the joining process. This eliminates the burr and therefore no additionaloperations are needed; it also creates a much-narrowed heat-affected and de-zinced zone. For these reasons the tubes for the ULSAB structure were laserwelded.

Chapter 7 - Page 20


For the prebending process, which requires a tube with small tolerances and afinished part with high strength, the following tube specifications were created:

QualityFeature: Precision steel tube according to the following tolerancesMaterial: Zinc coated on both sides details see belowYield Strength: > 260 N/mm² (> 280 N/mm² on finished parts)Total Elongation: > 32% (longitudinal and transverse)Uniform Elongation: > 20%r - Value: > 1.80

Dimensions and TolerancesOutside Diameter: 96 mm +0.1 / 0Wall Thickness: 1.0 mm; tolerances according to ULSAB specificationTotal Tube Length: 2700 mm +/- 1Cutting of Tube Ends: Free of Burr

No ovalization or cave-inNo chamfersRectangular to longitudinal axis +/- 0.5°

Appearance of TubesSurface: Free of mechanical damage, splatters, etc.

No collapsed areas (no indents, bulges, etc.)Free of impurities (swarf, weld chips etc.)

Welding RequirementsWelding Process: Laser- or high-frequency weldingWeld Seam Area: Outside of tube: Undercut 0.0 mm, no expansion

Inside of Tube: Undercut < 0.2 mm, no expansionNo mismatch of edgesFree of any porosityStrength similar to base material

Chapter 7 - Page 21


7.4.5. Process Steps for Rail Side Roof

Because the side roof rail has several 2-dimensional bendings with different radiiover its length and two 3-dimensional curves in the rear portion, the straight tubehas to be prebent. At the beginning of the design phase, bending tubes with such ahigh diameter (96 mm) -to-wall-thickness (1.0 mm) ratio resulted in very poor bendquality. At first, the tubes were bent by using a conventional mandrel-bendingmachine modified in such a way that the mandrel was replaced by internal fluidpressure. This inside pressure is working as a substitute for a mandrel. Thepurpose of this was to maintain stricter tolerances which are directly related to theaccuracy of the bending tools, the diameter of the mandrel used, and the tubediameter and wall thickness. In this way, the tubes could be bent into the neededshape without any wrinkles. However, because the pressure was applied inside thewhole tube, the tube diameter increased to a point that the tube would not fit intothe next die. Therefore, Porsche went back to using the solid mandrel. By holdingto stricter tolerances and taking certain other steps, wrinkle-free tubes could beformed. With this process, the clamping force needed to avoid wrinkles or damageto the tube has to be kept within a tight tolerance.

Once the tube is prebent, preforming is the next step. This is done in a three-piecetool under low internal pressure to avoid collapsing. The tube is then flattened andbent again in order to fit into the final hydroforming die. The basic layout of thepreforming tool and the tool itself is shown in Figure 7.4.5-1, 2 & 3.

Figure 7.4.5-1 Preforming Tool Concept

Outer tool part

Tube

Inner tool part

Moving direction ofouter tool part

Section A - A

Upper tool partnot shown

Chapter 7 - Page 22


Tube filled with waterunder low pressure

Outer tool part movedto inner pert

Upper tool part closed Pressure released and die opened

Figure 7.4.5-2 Sec. A-A of Preforming Tool Concept

Figure 7.4.5-3 Preforming Tool

Upper tool part

Inner tool part

Outer tool part

Chapter 7 - Page 23


Figure 7.4.5-4 Hydroforming Tool

The final step is the hydroforming process itself. During the down movement of theupper half of the die there is another area preformed again (under low internalpressure) on the tube. This must be done because the hydroforming process isvery sensitive to die locking. Once the die is finally closed, the internal pressure isincreased and the side roof rail tube is calibrated into its final shape. The pressurehas to be raised to 900 bar for the side roof rail in order to set the final shape of thepart. This required a closing force of about 3200 tons. This internal calibrationpressure was higher than predicted by calculation and forming simulation. A pictureof the hydroforming tool is shown in Fig. 7.4.5-4.

Chapter 7 - Page 24


7.4.6. Results

Hydroforming has never been used previously to form a high strength steel tubewith such a high diameter-to-wall-thickness ratio. Nevertheless the goal tomanufacture the side roof rails was achieved. There is still room for improvement,but the main problems related to the bending and preforming operations wereresolved. Hydroforming will be only a calibration operation if all-important stepsbefore this were optimized. With the experience gained from the ULSAB Phase 2,producing similar hydroformed applications should be easier in the future.

Chapter 7 - Page 25


Figure 7.5.1-1 Active Hydro-Mec Process Step: Loading / Unloading

7.5. Hydromechanical Sheet Forming


Hoods, roofs and door panels (large body outer panels) produced by conventionalforming methods often lack sufficient stiffness against buckling in the center area ofthe part. Due to the low degree of deformation in the center, there is only a little workhardening effect that could be achieved. Therefore, material thickness has to beincreased to meet the dent resistance requirements on those parts. This of courseleads to heavier parts and creates extra costs. The “active hydromechanical sheetmetal forming process” is a forming technology that uses an active fluid medium.The die consists of three main components: a drawing ring, which is designed as a“water box,” the blankholder (binder) and the drawing punch itself. At the beginning,the die is open and the blank is loaded on the ring (see figure 7.5.1-1).

Blankholder Cylinder

Slide

Blankholder

Moving Balster

Slide Cylinder

Chapter 7 - Page 26


Figure 7.5.1-2 Active Hydro-Mec Process Step: Pre-forming

In the second stage, the die is closed and the blankholder clamps the blank. Thedie punch has a defined, part specific regress against the clamped blank, as infigure 7.5.1-2. A pressure intensifier is used to introduce the water emulsion intothe water box, where a pre-set pressure is generated. The blank is inflated in acontrolled manner and stretched over the complete area until it is pressed againstthe punch. This is the reason why the process is called “active hydromechanicalsheet metal forming.” Forming with fluids (or flexible rubber layers) is well knownalready, but previously there was no forming in the “opposite” direction within thoseprocesses. The plastic elongation produces a work-hardening effect, especially inthe center of the part. This effect significantly improves the dent resistance of theformed part.

Chapter 7 - Page 27


Figure 7.5.1-3 Active Hydro-Mec Process Step: Forming Completed

Once the first plastic elongation process is done, the draw punch is moveddownward, as in figure 7.5.1-3. At the same time, the emulsion is evacuated fromthe water box and the pressure of the fluid is lowered in a controlled process. Aftercompletion of the drawing operation, pressure is increased once more in order tocalibrate the part into the final shape. The later visible surface of the part (outerside) is turned towards the active fluid medium. There is no contact to metal on thissurface and an excellent surface quality of the part was achieved.

Source: SMG Engineering Germany

Chapter 7 - Page 28


Figure 7.5.1-4 Roof Panel


The active hydromechanical sheet metal forming process is characterized byimproved component quality and potential mass and cost reduction. The essentialfeatures of this new technology are: higher dent resistance achieved by anincreased work-hardening effect during the first “counter” forming operation, andsuperior visible surface quality achieved by using water instead of a metal die forthe final forming operation. This leads to a reduced component mass due toincreased stability. Sheet thickness could be reduced to 0.7 mm and reinforcementelements could be saved, while all other requirements were still fulfilled. Inaddition, the cost of dies can be reduced by about 40% because only one polishedhalf of the die is required. In addition, the average lifetime of the dies will lastlonger, under mass production conditions, than usual because there is little wearingoff when forming with a fluid medium.

In order to get the most benefit out of this process a forming simulation should beperformed. This simulation may help to predict the maximal prestretching amountachievable without damaging the sheet. The absence of friction between the blank

A picture of the formed roof panel is shown below in figure 7.5.1-4.

Chapter 7 - Page 29


and the conventionally used second half of the die makes the result of thesimulation very reliable. Furthermore, the process parameters, (e.g., preformingpressure, etc.) could be easily adjusted.

7.5.3. Process Limitations

Depending on the grade of prestretching, which is related to the preformingpressure, the size of the forming press (locking force) has to be chosen. This isalso influenced by the overall projected area of the part (e.g., for the ULSAB roofpanel, a press with a locking force of 4,000 was chosen.) A double (or triple) actionhydraulic press must be used to make the process reliable.

This press can be used for conventional forming, and with the use of someadditional equipment, for the tubular hydroforming process.

The filling time for the fluid medium pressure bed has to be taken into account aswell. This leads to a calculated cycle time for the ULSAB roof panel of about30 - 40 seconds. Depending on the design of the part, this has to be compared to atwo-step conventional forming operation.

Due to potential die locking, it appears that an undercut on the hydroformed parts isnot feasible in this process without using a separate tool. This is also relevant forthe cutting of flanges. This has to be done separately using laser or conventionaltrimming operations.

Chapter 7 - Page 30


7.5.4. Results

Roof panels for the ULSAB could be manufactured by using the activehydromechanical sheet metal forming process. Different material qualities, likeisotropic, IF and bake-hardening types, were formed successfully. Due to the work-hardening effect, which was applied through the above-described process, the sheetthickness of the roof panel could be lowered to 0.7 mm, while the dent resistancerequirements were still met.

In order to limit the needed locking force of the press, the flange radii should bedesigned not too small. The radii are directly related to the needed pressure duringthe final forming operation, and if too small lead to an uneconomic high-lockingforce/press size. The surface quality on the visible side of the ULSAB roof panel,which was not in contact with any metal tool, was very high compared toconventional formed (prototype) parts.



Chapter 8 - Page 1



8.1. Supplier Selection

The main criterion for supplier selection was quality. Although the process used“soft” tools and lasers, the contract required production representative parts.Therefore, it was decided to identify companies that specialize in one or more of thefollowing system groups:

• Front End Structure• Floor Panels and Body Side Inner• Body Side Outer• Rear Structure• Roof and Roof Side Rails

Extensive discussions took place with approximately 30 suppliers on a worldwidebasis to identify the sources for the ULSAB program. The criteria used torationalize the final selections were:

• Supplier must have major OEM quality rating or ISO 9000• Must be a system supplier to a major OEM• Must be prepared to enter simultaneous engineering prior to contract

release• CAD/CAM systems compatible with CATIA• Program management system established• Experience in match metal checks• Cost competitive

Chapter 8 - Page 2


Based on the foregoing, the following companies were selected:

• Front End Structure – Stickel GMBH, leading supplier to Porsche AG• Floor Panels and Body Side Inner – Peregrine FormingTechnologies,

supplier to GM, Chrysler and Ford• Body Side Outer – AutoDie International, leading Body Side supplier to

Chrysler, also supplying Ford and GM• Rear Structure – Fab All Manufacturing, commodity supplier to Ford• Roof and Roof Side Rails – Schaefer Hydroforming

Company Name Address

Major products

Other Divisions Customers Major Equipment

Autodie International 700+44 Coldbrook, Grand Rapids, Michigan, USA

Tools, Dies and Molds, Prototypes & ProductionAutomated SystemsTransfer EquipmentWelding FixturesRobotic Vision Systems

Progressive ToolWISNE DesignWISNE Design - Die TechnologyWISNE AutomationEagle EngineeringFreeland Manufacuturing+ Others

FordChryslerTowerSpartanburgNavistarCambridge

Presses up to 3000 tBed Size to 200 x 1004 CMM5 Axis Control Laser1 Lamoine Machine SystemCNC MillsPDGS CGS CATIA

GMJaguarBMWKarmaxHaworth

Number of Employees

Chapter 8 - Page 3


Company Name Address Number of Employees

Major products


Peregrine Forming Technologies 16026269 Groesbeck, Warren, Michigan, USA

Prototype ToolingStampings and AssembliesDoors Inner / OuterCowls, Fenders, Deck LidsRoof Panels and Floor Panels

APG - Technical ServicesBattle Creek StampingWarren StampingWarren Assembly

FordGMDanaTowerOgiharaHondaSpartanburg

Presses up to 1500 tBed size to 192 x 793 CMM5 Axis Control LaserFoundry3 CNC MillsPDGS CGS CATIA


Major products


Fab All Manufacturers 95645 Executive Drive, Troy, Michigan, USA

Prototype ToolsStampings and AssembliesSpecializing in Underbody, Front Structuresand Inner Structures

Hubert GroupSharp Mold EngineM & T Design ServicesModels & Tools

GMFordChrylserAG SimpsonVeltriNarmco

Presses up to 1700 tBed size to 144 x 1322 CMM6 Axis LaserNC MachiningCATIA PDGSCGS Unigraphics

Number of Employees

Chapter 8 - Page 4



Major products


Stickel GmbH 40Porschestrasse 2, D - 74369 Loechgau

Prototype BuildPrototype Tooling, Prototype StampingsLow Volume Production Stampings and Subassemblies

None AudiBMWMannesmannMercedes BenzOpel AGPorsche AG

Presses up to 800 tBed sizes up to 2m x 3m3D LaserCMM EquipmentCATIA CGS

Number of Employees


Major products


Schäfer Hydroforming, Schuler 135Auf der Landerskrone 2, D - 57234 Wilhelmsdorf

Hydroforming Presses (Development, Fabricating)Prototype and Production PartsTechnology Development (Active Hydro Mec)

Tool ShopFEM Forming SimulationHydroforming Componenets

AudiAerosmithGMBentelerPorsche

Hydroforming presses to 3000t10.000 t under ConstructionHigh Speed MilingPrebending Equipment

Number of Employees

Chapter 8 - Page 5


8.2 Simultaneous Engineering

In order to achieve the optimal design from a manufacturing and assemblystandpoint, reviews were held with the suppliers and the assembly facility toevaluate all designs six months prior to design release.

Each supplier was represented by specialists in CAD/CAM, tool making andmanufacturing. Every detail was reviewed for formability, spring back issues,aesthetic consideration, tolerance control and assembly issues. In addition to thepart suppliers, steel companies also attended these sessions in order to discussand resolve any material issues.

These reviews continued after design release, primarily in the suppliers’ facilities,but in addition to the design for manufacture and design for assembly, the reviewsalso included the supplier maintaining quality and timing plans.

8.3. Part Manufacturing Feasibility

Introduction

At the request of the ULSAB Steel Consortium and PES, Phoenix Consulting Inc.has assisted in the investigation and documentation of the manufacturing feasibilityof the ULSAB components. The study includes the following objectives.

• Demonstrate that the processes used to fabricate the ULSAB componentsmeet the following conditions:, Used design intent materials., Can repeatedly produce parts that meet dimensional requirements., Can repeatedly produce parts that meet formability requirements.

• Demonstrate that through continuous improvement, these processes canbe evolved to production capable processes., Mechanisms are in place and are being followed to address

manufacturing feasibility concerns., Action plans have been developed to address remaining barriers to

production capability.

Chapter 8 - Page 6


• Demonstrate that state of the art methods and technologies have beenused to develop the demonstration hardware processes, such as:, Forming Simulation., Early Steel Involvement., Dies and fixtures developed from CAD, CNC Machining and CMM

Inspection.

Overall Assessment

Although the components of the ULSAB body structure certainly present asignificantly greater challenge to production capability than a conventional design,we are convinced that these components can be fabricated with production capableprocesses under the following conditions:

1.The process of continuous improvement that has been undertaken by Porscheis continued, including additional soft die tryout and minor product revision.

2.With the use of the more sophisticated press equipment that can be madeavailable in hard tool construction: Multiple Nitrogen Cushions, TogglePresses and with the superior surfaces encountered in hard tooling.

3.With the implementation of further enhancements in materials, blankdevelopment and binder development.

The team assembled to fabricate these components has made excellent progressalong the learning curve of fabricating with high strength steel and laser weldedblanks, advancing the state of the art. The prototype processes have undergonesignificant continuous improvement toward production capability

Documentation Overview

The components on the ULSAB body have been classified into three levels ofdifficulty or criticality. Level C being the most critical, level B the next most criticaland all other parts are level A. The extent of documentation provided for a givencomponent has been determined accordingly. The purpose of these documents isto validate the objectives outlined in the introduction. These documents have beenassembled into a notebook that can be provided through the ULSAB Consortium.

Chapter 8 - Page 7


These documents are described below, followed by a list of B and C level parts. Inthe pages that follow is an example of the detailed summaries for each individual Band C level part that can found in the notebook.

Level A - Non Critical

• Material Characterization. This validates that the parts are made ofmaterial that meets structural requirements and that these materials canbe worked into the forms of the respective parts.

Level B - Moderately Critical. All Level-A requirements plus the following:

• Strain Analysis (Circle Grid and or Thickness Strain): Demonstrates that aformability safety margin exists and that parts are not merely split free.The goal and conventional buy off requirement is a 10% safety margin.These Strain Analyses are the responsibility of the Steel Vendors as partof the Early Involvement Program. They should include material propertiesof metal used to form the evaluated panel and the associated pressconditions. This information is documented in AQP Parts format.

• Process Set Up: After extensive tryout, die shops have arrived at, anddocumented, optimum press conditions that will repeatedly yield qualitypanels. These Press Conditions along with other details of die set up aredocumented on Set Up Sheets. These Set Up Sheets can serve asbaseline for further continuous improvement to develop production capableprocesses.

• Part submission warrants: These certify that prototype parts meetdimensional requirements.

Chapter 8 - Page 8


Level C - Most Critical: All level A and B requirements, plus the following.

• CMM Reports: Computerized measurement of dimensional integrity.• Development Logs: Show that state of the art methods and technologies

were used to develop prototype processes and that these processes areundergoing a continuous improvement of evolution toward productioncapable processes.

• Proposed Production Process: This is the capstone of the above efforts. Itis the culmination of lessons learned in prototype tryout and ademonstration of Porsche’s confidence that the next step of setting upproduction processes can be taken.

• Forming Simulation: Finite Element Analysis based on CAD data wasused to identify formability concerns before the construction of tools.

B and C Level Parts

Part Name Part Number Die Shop Level

Pan Front Floor 040 Peregrine C

Panel Rocker Inner 042 / 043 Peregrine C

Panel B-Pillar Inner 064 / 065 Peregrine C

Rail Rear Inner 046 / 047 Fab All C

Rail Rear Outer 048 / 049 Fab All B

Panel Wheelhouse Outer 070 / 071 Fab All B

Panel Body Side Outer 060 / 061 Autodie C

Member Dash Front 026 Stickel C

Panel Skirt (& Shock Tower) 096 / 097 Stickel C

Rail Front Inner 010 / 011 Stickel B

Rail Front Extension 012 / 013 Stickel B

Panel Dash 021 Stickel B

Member Kick Up 091 Stickel B

Rail Side Roof 072 / 073 Schaefer C

Panel Roof 085 Schaefer B

Spare Tire Tub 050 Stickel B

Chapter 8 - Page 9


Summaries of individual B and C level parts.

On the following pages you will find an example of the documented data. Includedwill be:

1.Summary page, including observations and recommendations.2.Part diagram.3.Documentation checklist, listing and/or summarizing required documentation.4.Material characterization sheet.5.Forming limit diagram (part of strain analysis).

NOTE: Complete documentation for all A, B & C level parts is contained

In a separate report obtainable through the ULSAB Consortium.

Documentation Responsible Format Parts Forming Simulation Steel Co. Steel Co. Report Select Parts

Strain Analysis

(Circle Grid, Thickness Strain) Steel Co. AQP B & CSteel Co.

Material Characterization and Phoenix AQP A, B & C

Process Set Up Steel Co, Die Shops Phoenix Summary & (Set UP Sheets) and Phoenix Die Shop Set Up Sheet B & C

Proposed Production Process Porsche & Phoenix Process Sheet C

Certification of Dimensional Die Shops Die Shop Form B & C Integrity (Warrant)

Die Shops CMM or Checking C

Inspection Report Fixture Report Development Log. Demonstrates state of the art procedures used to develop capable prototype processes & action plans for Die Shops Die Shop Log C making processes production capable. Observations and

Recommendations Phoenix Phoenix Summary B & C

Chapter 8 - Page 10


Pan Front Floor - 040Part Manufacturing Feasibility Summary

The process involves first forming the front of the panel down, then the middle ofpanel the down and finally the rear of the panel up. This had to be done in separateoperations for several reasons. One was press bed size. Another was the fact thatall these areas are on separate levels and proper control of metal cannot beobtained without a more elaborate process involving nitro cushions and dydro units.

The availability of these resources for production will enable a reduction in thenumber of operations, which will be necessary to reduce the total number ofoperations once trim and flange dies are added. Trimming and flanging is currentlyperformed by laser and hammer form and will require cams in production due to theorientation of some of the trim and flange lines.

Marginal strains detected in tryout and GD&T (geometric dimensioning &tolerancing) issues would have to be reassessed after implementing therecommendations below.

Recommendations Based on Documentation ChecklistInvestigating grade change to a dent resistant steel that meets yield strengthrequirements but has a higher n-value. A dry film lube trial is also recommended.

Consider use of a wider blank. This will allow for better control of metal outside ofthe kickup area by adding a more gradual transition in the addendum and binder.This may also enable the use of patches of higher formability metal where they areneeded the most. This exercise would be well worth the effort, considering theportion of overall weight represented by the floor pan, and the challenging formingcharacteristics associated with it.

Consider ways of forming embossed areas as late as possible in the process, eitherby using restrike die or by delayed action in draw dies, to avoid metal locking onand/or skidding over embossed area when it is required for feeding deep formations.

Forming Simulation of first draw predicted wrinkling in tunnel near kickup. This isone of the areas where wrinkling was encountered in tryout.

Chapter 8 - Page 11


First Form

Second Form

Third Form

Increase blank widthand implement smoothtransition & drawbar.

Embossments impede metalflow; result in double draw lines. Implement laser weld

for wider blank.

Marginal Forming Strainsat locations #2 and #15.

#2

#15

Chapter 8 - Page 12


ULSAB Part Manufacturing Feasibility StudyDocumentation Checklist

Level

Part#

Part Name Supplier SpcThk

YieldStrength

Coating Blank

C 040 Pan Frt Fl oor Peregrine 0.7 mm 210 MPa 60G60GU Rectangle

Document Format Status / SummaryFormingSimulation

Steel Co LS-Dyna3D simulation of 1st draw predicted significant wrinkling in the steparea of part near the tunnel. This is one of the areas where wrinkling wasencountered in tryout. The other areas occurred mainly during subsequentoperations.

StrainAnalysis

Material TestPressConditions

AQP Reports 40_D1.TXF (First Form) & 40_D3.TXF (Third Form)Safety Margin = 3%. Dry film lube trial suggested. Marginal Strains (#2,#15) need to be re-assessed after implementing blank config, binder anddie process improvements.Included in AQP. Also see Process Set Up below.

Material TestFinal / Conam

AQP Samples shipped to Conam on 12/11/97

Process SetUp

Peregrine Peregrine Set Up Sheet summary:Blank Size = 1829mm x 2057mm1) PreDraw = Three piece stretch forms tunnel and kickup2) Draw = Single Action with Upr Binder on Nitro forms deep pocket at rearof kickup 3) Three piece stretch forms shape at rear of panel 4) Flange.Flange at kickup is hand formed. Would have to be Cam Flanged inproduction. All trimming is by laser.Form #1 Ram = 1000 ton Binder = 160 ton (40 cyl @ 1600 psi)Lube = Quaker PrelubeForm #2 Ram = 400 ton Binder = 100 tonLube = Super DrawForm #3 Ram = 400 ton Binder = 200 ton (toggle press)Lube = Super Draw

ProposedProductionProcess

1) Draw 2) 1st Trim 3) Re-strike 4) Form/Cam Form5) Final Trim/Cam Trim

DimensionalCheck

Warrant Included

DimensionalCheck

CMM Report CMM detected points that deviated from nominal by more than +/- 0.5 mm,however all were vertical and attributable to part length and flexibility, orhammer formed flanges. No difficulty experienced in assembly.

DevelopmentLog

Simultaneous Engineering procedures were used to develop the process,and continuous improvement was implemented to evolve the processtoward production capability. Supplier concerns were fed back to Porscheand product revisions were subsequently implemented. Summary ofdevelopment history and log of product changes is included. Also includedis sketch of part showing significant manufacturing related changes.

Chapter 8 - Page 13


Chapter 8 - Page 14


Chapter 8 - Page 15


Chapter 8 - Page 16


8.4. Quality Criteria

The quality assurance system utilized on the ULSAB project followed the samestandards as normal automotive practices. The key elements of control were:

• Material• Engineering levels• Process control• Dimensional accuracy• Parts submission

Material: All material received was checked for dimensional accuracy by the partsuppliers, the steel suppliers provided the material characterization data which wasverified by an independent laboratory. Additionally, Porsche checked the materialfor weldability.

Engineering Levels: A strict engineering change control system was implementedfor this program. At each weekly review meeting all product levels were checkedagainst the design status to insure compatibility. Suppliers were not allowed toimplement any change without the authorization of PES.

Process Control: As previously stated, the components were produced toproduction intent standards. Therefore, to insure this occurred, regular audits of theprocess were undertaken.

Dimensional Accuracy: For each component, automotive standard checking fixtureswere produced. These fixtures were used throughout the development process toprovide verification of dimensional accuracy. Additionally for all major parts, thecontract with the suppliers called for two fully CMM checked samples. As furtherassurance, where possible, match checks were undertaken to insure fit and functionfor the assembly process.

Parts Submission: The approval process was based on PPAP (Production PartApproval Process) as outlined in QS 9000 guidelines. Before any part was shipped,the supplier had to provide documentation that showed all material, engineering,process and dimensional controls had been completed and met with thespecifications set within the program.


9. DH Build

Chapter 9 - Page 1


9. DH Build

9.1. Introduction

After ULSAB Phase 1 was successfully completed, the ULSAB Consortium decidedto proceed with the ULSAB program into Phase 2. This involved proceeding from aconceptual study to the real world hardware, whereby the predicted mass savingsand improved performance could be proven by actual product.

Due to the experience in laser welding, Porsche’s R & D Center in Weissach,Germany was chosen for the execution of the 13 DH builds.

Figure 9.1-1 Prototype Shop

Chapter 9 - Page 2


Roof• Audi • BMW • Ford

• GM • Mercedes • Opel• Renault • Volvo

• Volkswagen

Roof• Audi • BMW • Ford

• GM • Mercedes • Opel• Renault • Volvo

• Volkswagen

B/C Pillars• Audi • Mercedes

B/C Pillars• Audi • Mercedes

Decklid / Tail gate• BMW • Daihatsu

• Honda • Opel • Suzuki• Volkswagen

Decklid / Tail gate• BMW • Daihatsu

• Honda • Opel • Suzuki• Volkswagen

Front Structures• BMW • Mercedes

Front Structures• BMW • Mercedes

Hood• Opel • Volvo

Hood• Opel • Volvo

Doors• Honda • Porsche

Doors • Honda • Porsche

Laser weldin g applications on production auto-bodies

9.2. Joining Technologies

9.2.1. Laser Welding

For more than 10 years the laser has shown its production capability. The first autobody application was the blank welding of the floor panel for the Audi 100. Laserwelding in the assembly process was first brought into a production plant by BMWfor the roof welding of its former touring model 3 series and Volvo for the roofwelding of the 850 model.

Since then, especially during the last three years, an increasing number of automanufacturers have installed laser welding equipment within their production lines.

Today laser welding applications in production plants are utilized all over the autobody, such as the front end, under body, closure panels and roof panel.

Fig. 9.2.1-1 Laser Welding in Assembly

Chapter 9 - Page 3


1314

7

(12)

17

18

12

12

3

20

9

(14)

4

(3)10

5

8

6

19

1516

11

The major reasons for using laser welding is the predominantly high static anddynamic strength of the joints, one side weld access for the welding equipment,small thermic impact zone and good aesthetic look at the joint area. The totallength of the laser welding seams for the assembly on the demonstration hardwareis 18.28 meters.

Figure 9.2.1-2 Laser Welding on ULSAB Demonstration Hardware

1. Rail Front Outer to Rail Front Inner 2. Rail Fender Support Inner to Rail Fender Support Outer 3. Panel Body Side Outer to Panel A-Pillar Inner Lower 4. Rail Fender Support Outer to Panel Body Side Outer 5. Panel B-Pillar Inner to Rail Side Roof 6. Bracket Member Pass Through Lower to Member Pass Through 7. Panel Wheelhouse Inner to Rail Side Roof 8. Panel Back to Rail Rear Inner and Rail Rear Outer 9. Panel Dash to Rail Front Extension10. Panel Cowl Upper to Panel A-Pillar Inner Lower

11. Panel B-Pillar Inner to Panel Rocker Inner12. Panel Roof to Panel Body Side Outer13. Rail Side Roof to Panel A-Pillar Inner Upper14. Panel Body Side Outer to Rail Side Roof15. Panel Package Tray Upper to Support Package Tray16. Support Panel Rear Header to Rail Side Roof17. Panel Roof to Rail Side Roof18. Member Pass Through to Brkt Member Pass Through Upr Frt & Rear19. Rail Rear Outer to Rail Rear Inner20. Panel Package Tray Upper to Panel Gutter Decklid

Chapter 9 - Page 4


current measurement

voltage measurement

controlunit

powerunit

transformer

9.2.2. Spot Welding

Spot welding is for all OEMs a well-experienced, reliable, affordable joiningtechnique for steel auto bodies, even with zinc-coated steel materials. Porsche, forexample, has been producing cars since 1977 with 100% zinc coated steel sheetmetal and was the first company in the world practicing this. Now, more and moreOEMs are switching to 100% zinc coated materials to improve corrosion protectionand to give a long time anti-corrosion guarantee. Also for ULSAB, 100% of thematerial is double side zinc coated.

Porsche’s R & D Center Body Assembly Facility utilizes computer controlledmedium frequency (1000 Hz) welding equipment. This system uses calibration toensure that the welding current is maintained at a constant level. Thereby providinga good weld without disturbances and achieving optimum settings for welding time,welding current and electrode force. Having established the optimum setting, thedata is stored in the computer enabling the use of the ‘control mode’ to ensure allsubsequent welding operations achieve the same optimum integrity.

Figure 9.2.2-1 Configuration of a Welding System

Chapter 9 - Page 5


wel

d cu

rren

t

AC welding operation (50 Hz)

wel

d cu

rren

t

medium frequency inverter welding operation

(1000 Hz)

Comparison of the control response of thyristors and inverter controllers

These control processes inevitably necessitate fast welding current sources. Thisrequirement is fulfilled by medium frequency inverters with a response time of onemillisecond at an inverter frequency of 1000 Hz and by the substantially fastertransistor DC technology.

The system is sensitive to:

• main voltage fluctuations• shunts• electrode wear (automatic stepper function)• electrode force fluctuations• small edge distances• welding splashes• changes from two sheet to multiple sheet welds

Figure 9.2.2-2

Chapter 9 - Page 6


The control process compensates the various influencing factors by increasing orreducing the current strength and extending the welding time. Extension of thewelding time can be limited.

Welding splashes are monitored via output of an error message, with optionalshutdown of the welding current.

Optimum adaptation to each weld spot guarantees that the required strength forweld joints is maintained throughout broad ranges.

Spot welding is used on ULSAB in all areas with suitable weld access and normalstructural loads.

The assembly of the demonstration hardware uses 2,126 spot welds.

Figure 9.2.2-3 Medium Frequency Spot Welding Equipment

Chapter 9 - Page 7


9.2.3. Active Gas Metal Arc Welding (MAG)

Active Gas Metal Arc Welding, or similar joining techniques, is used at all OEMs inlocations with no weld access for spot welding or in areas with high stresses due toits strong structural behavior in comparison to spot welding.

The disadvantages of this process, like slow welding speed, big heat impact zone,and pollution by weld fumes, especially with zinc coated materials, forced manyOEMs to reduce it to a minimal amount.

The targets for ULSAB were established to minimize the MAG welding seams.MAG welding is only used on the ULSAB body structure at locations without weldaccess for spot and laser welding.

In total, there are 1.5 meters of MAG welding on the DH structure.

Figure 9.2.3-1 MAG Welding on ULSAB Demonstration Hardware

5

6 7

1

2

34

1. Panel A-Piller Inner Lower to Panel Cowl Upper2. Door Hinges to Panel Body Side Outer3. Door Hinges to Panel B-Pillar Inner4. Door Hinges to Panel A-Pillar Inner5. Support Package Tray to Rail Side Roof6. Bracket Roof Rail Mount to Rail Side Roof7. Bracket Member Pass Through Lower to Rail Side Roof

Chapter 9 - Page 8


9.2.4. Adhesive Bonding

The ULSAB steel sandwich material cannot resist the high temperatures during thepainting process for body structures. Therefore this material is only suitable forparts which are assembled to the body after the painting procedure. Another factoris the non-weldability of the ULSAB sandwich material.

So for the two parts on ULSAB made of steel sandwich adhesive bonding is thechosen joining technology.

It has not only a structural function, it also provides sealing. The two panels madefrom steel sandwich material are the Panel Dash Insert (Part No. 022) and thePanel Spare Tire Tub (Part No. 050).

Figure 9.2.4-1 Bonding at Panel Dash Insert

Chapter 9 - Page 9


In the production line, the panel dash insert will be assembled to the painted bodystructure as part of the instrument panel module. This includes the instrumentpanel, steering column, air conditioning system and pedal system. The panel dashinsert is adhesive bonded and additionally bolted to dash panel. The bolting isnecessary to keep the part in position until the bonding material is hardened.

The panel spare tire tub will be assembled to the painted body structure as amodule including the spare tire and the repair tools. The module is bonded to thestructure. The operation does not require additional fixturing.

The bonding material is a two component, non-conductive, high modulus, highviscous, chemically-curing polyurethane adhesive/sealant that cures almostindependently of temperature and moisture. It is Betaseal X 2500 produced byGurit Essex.

Figure 9.2.4-2 Bonding at Panel Spare Tire Tub

Chapter 9 - Page 10


Technical Data Basis Polyurethane prepolymer

Color black

Solids content >98%

(GM 042.0)

Flash Point >100° C

Processing temperature ideal 10° C - 35° C

Working time approx. 10 min. at 23° C/50% r.h.

(Processing time)

Sagging behavior good, non-sagging

Ultimate tensile strength > 5.5 MPa

(DIN 53 504)

Percentage elongation > 200%

(DIN 53 504)

Combined tension (GM 021) > 4.5 MPa

and shear resistance

G-Modulus > 2.5 MPa

Specific electrical > 10 cm

(volume resistivity)

Abrasion resistance Extremely high

Recovery (DIN 52 458) approx. 99%

Temperature stability - 40° C at 100° C (for short periods up to 140° C)

Resistance to chemicals Highly resistant to aqueous chemicals, petrol

(in cured conditions) alcohol and oils.

Conditionally resistant to esters, aromatics and

and chlorinated hydrocarbons.

Preparation of bonding surface All bonding surfaces must be free of dirt, dust,

water, oil and grease. In general, surfaces

should be primed.

W

Chapter 9 - Page 11


9.3. Flexible Modular Assembly Fixture System

The body shop in Porsche’s R & D Center used a highly flexible modular fixturesystem for the DH assembly. It is based on standardized units, which areadjustable in all directions.

There are many advantages of this fixture system. 95% of the elements in a fixtureare from the standardized module system and can be used also for other carprograms.

Figure 9.3-1 Assembly Fixture Module

Chapter 9 - Page 12


Figure 9.3-2b Assembly Fixture Module Detail

Figure 9.3-2a Assembly Fixture Module Detail

Chapter 9 - Page 13


The fixture design performed in CATIA was very efficient, because all models wereaccessible from the CAD data bank. Therefore, the construction time for assemblyfixtures was reduced and modifications or corrections of existing assembly fixturescould be implemented rapidly.

Figure 9.3-3 Assembly Fixture - Bodyside Inner Subassembly

Chapter 9 - Page 14


Porsche is using the flexible modular system in two ways.

The first is the so-called shuttle system, which is related to the set-up pallets. Theshuttles for different assemblies are stored in a shuttle magazine. During theassembly operation the shuttle is fixed on a set-up pallet. The changeover ofvarious assembly shuttles on a set-up pallet is a very fast process. Theseassembly shuttles are mobile and can be used at different locations.

Figure 9.3-4 Assembly Fixture Shuttle on Setup Pallet

Chapter 9 - Page 15


The second method is the utilization of a rolling device that supports the modularassembly fixtures independent from set-up pallets. These assembly fixtures work atany location.

Figure 9.3-5 Mobile Assembly Fixture - Shock Tower Front SubAssembly RH/LH

Chapter 9 - Page 16


9.4. Design of Assembly Fixtures

All fixtures are developed with a CAD system (CATIA) based on the existing designdata. The CAD data models of the fixture system modules are available from a databank.

Figure 9.4-1 Fixture Development on CAD System

Figure 9.4-2 CAD Data Modules of Fixture System

Chapter 9 - Page 17


The DH assembly sequence is exactly the same as it is foreseen in the productionplant. Due to the fact that in prototype productions no cycle time limit is given onefixture can be used for more joining operations than in a production line. Thisresults in a drastically reduced number of assembly fixtures in relation to aproduction line.

For the ULSAB assembly, the Porsche body shop used the following fixtures:

• Assembly Shock Tower Front• Assembly Front End• Assembly Floor Complete• Assembly Under Body Complete• Assembly Body Side Inner• Assembly Body Complete

An example of a fixture design is shown in Figure 9.4-3.

Figure 9.4-3 Fixture Shock Tower Front

Chapter 9 - Page 18


9.5. DH Build

9.5.1. Assembly Team

The Porsche BIW assembly team consists of the following personnel:

• 1 foreman• 1 expert/deputy foreman• 23 workers which include 5 with foreman’s / technician’s degree and 5

workers trained for CATIA

Figure 9.5.1-1 Body Shop

Chapter 9 - Page 19


In a workshop space of 1200 m2, the following equipment is installed:

• 12 setup pallets (6x3m) with surface measuring device• 4 mobile welding machines, 1000 Hz with control equipment• 5 mobile welding machines, 50 Hz with constant-voltage regulation system• 5 overhead spot-welding devices with 3 secondary guns each and a 50 Hz

Bosch control system• 1 Rofin Sinar Laser device, 2.5 kW

Two applications with special interest for ULSAB will be described in more detail.

All spot welds on ULSAB were manufactured with a mobile Duering welding cartand a Matuschek medium-frequency inverter device with master control system.

Figure 9.5.1-2

Chapter 9 - Page 20


The welding gun changeover system allows a rapid change between different typesof welding guns, whereby a special gun coding provides the correct weldparameters from an automatic program selection.

Figure 9.5.1-3 Weld Gun Station

Chapter 9 - Page 21


The laser welding and laser cutting cabin is equipped with a KUKA KR 125 robot.The maximal load is 125 kg and the working range of 2410 mm.

Figure 9.5.1-4 Laser Cabin

Chapter 9 - Page 22


The laser source is a Rofin Sinar CW 025 Nd:YAG Laser. The maximum output of2500 W is transferred through a switching device with two outlets via two 15-mglass fibre cable of 0.6 mm diameter to the laser optic.

Besides a laser cutting head three different types of laser welding heads areavailable.

Figure 9.5.1-5 Laser

Figure 9.5.1-6 Laser Picker

Chapter 9 - Page 23


Figure 9.5.1-7 Single Roller

Figure 9.5.1-8 Double Roller

Chapter 9 - Page 24


9.5.2. Build of the Test Unit

The construction of the test unit, internally called “workhorse,” started on May 26,1997, and began testing on June 27, 1997.

The following series of photographs shows steps of the assembly sequence of thetest unit.

Due to the extensive preparations, the construction worked out excellent, but therewas still room for small improvements.

Figure 9.5.2-1 Rear Floor Subassembly

Chapter 9 - Page 25


Figure 9.5.2-2 Subassembly Front End

Figure 9.5.2-3 Subassembly Underbody Complete

Chapter 9 - Page 26


Figure 9.5.2-5 Assembly Body Side Inner to Underbody

Figure 9.5.2-4 Subassembly Body Side Inner

Chapter 9 - Page 27


Figure 9.5.2-6 Subassembly Body Side Inner with Underbody

Figure 9.5.2-7 Sub-Assembly Body Side Outer, with Body Side Inner and Underbody

Chapter 9 - Page 28


The build of DH #2 started on December 1, 1997. The assembly sequence for DH#2 to DH #13 remained the same as test unit.

9.5.3. Build of DH #2 to DH #13

After build and testing of the test unit, a design review meeting in Porsche’s R & DCenter was held with the experts in the fields of body design, safety, CAEcalculations, parts manufacturing and body assembly. Ideas for improvements inrespect to performance, parts feasibility, weld access and appearance weregenerated in this meeting.

The next step was a redesign of the ULSAB body structure reflecting the ideas ofthe design review meeting. The CAE calculations of the changed FE model provednearly the same performance. Now new parts were manufactured incorporatingthese changes in the construction of DH #2 to DH #13.

Figure 9.5.3-1 Demonstration Hardware #2 in Body Shop

Chapter 9 - Page 29


9.6. Quality

9.6.1. Body Quality Control Team

The Porsche Body Quality Control Team includes the following personnel:

• 1 engineer• 2 technicians• 5 foremen• 2 specialist workers

In a working area of 300 m2 the following equipment is used for body quality controlmeasurement:

• 1 Stiefelmeyer double-column coordinate measuring machine (CMM)• 1 Stiefelmeyer single-column manual measuring machine• 1 Zeiss double-column CMM

Figure 9.6.1-1 DH #2 during Measuring Procedure

Chapter 9 - Page 30


The general range of services includes:

• Part acceptance at supplier’s premises• Model acceptance at supplier’s premises• Body measurement• Digitalization of data for design• Trouble-shooting• Prototype quality statistics

9.6.2. Quality Control Measurements of DHs

The basis for part and assembly quality was the early involvement of all relevantparticipants in the design and engineering process. Regularly simultaneousengineering meetings were established with designers, engineers, materialsuppliers, tool and part manufacturers and body shop personnel.

The expert group defined locator holes, tooling holes and fixing points. To ensureexcellent quality, these defined points were used for the complete process chainfrom parts manufacturing over subassemblies to final assembly.All manufactured parts were inspected by the supplier’s quality control personneland approved by Porsche specialists.

The first proof of feasibility and design for manufacture was the successfulconstruction of the test unit. This demonstration hardware was fully inspected byPorsche’s quality control team. In total, about 200 different points on the ULSABbody structure were measured and compared to the original CAD data.

The measured dimensions were, especially for a first time assembled bodystructure, in a close range to the nominal values.

Chapter 9 - Page 31


Nevertheless, the results of the test unit were used to develop modifications of thetools for part manufacturing and of the assembly fixtures for improved quality,meaning smaller tolerances for the following DHs. Each DH is or will be inspectedto evaluate a quality statistic for the ULSAB program.

Figure 9.6.2-1 Measuring protocoll

Chapter 9 - Page 32


9.7. Conclusion

The assembled demonstration hardware proved to be a successful execution of thebody structure construction. The measured tolerances are in a comparable range inrelation to average car programs.

The challenges of laser welding in assembly, assembly of hydroformed parts, 90%high strength steel, and steel sandwich material, were mastered. The principlecondition for success was the simultaneous engineering process. All projectpartners contributed to the realization of Phase 2 of the ULSAB program.

Through early involvement in the project, all parties involved incorporated all of theirexpertise into the realization of the demonstration hardware.

Figure 9.7-1



Chapter 10 - Page 1



Fig 10.1-1 Aerial View

10.1. Scope of Work

To prove the structural integrity of the ULSAB demonstration hardware, the followingtest procedures were executed as part of the ULSAB program in Phase 2.

• Static rigidity• Static torsion• Static bending

• Modal analysis• 1st Torsion mode• 1st Bending mode• 1st Front end lateral mode

• Mass• DH mass in test configuration

All testing work was performed at Porsche’s R & D Center in Weissach.

Chapter 10 - Page 2


10.2. Targets

The main factors affecting the ride and handling of the vehicle are Noise, Vibrationand Harshness, known as NVH behavior. To achieve the desired levels of comfortfor the occupants, the vehicle body must have high static and dynamic rigidity. Inother words, the auto body should have high stiffness. This is required because theincreased rigidity improves the vehicle resistance to excitement caused by the drivetrain, the engine or by road conditions such as bumps and potholes. When excited,the car body vibrates at particular frequencies, called its natural frequencies, andalso in a particular manner called its mode shape. The mode shapes are forinstance on: global torsion mode, global bending mode and front end lateral mode.

Another result of good rigidity would be minimal deviations in the dimensions of thebody structure openings such as the hood, front door, rear door and deck lid underload conditions. These movements between the body structure and the closurepanels often create sounds.

Furthermore, it should be proven that the received numbers from the analysis byFE-calculations are in correlation with the results gathered by the testing procedure.

Based on the current average of selected, benchmarked vehicles in Phase 1, thefollowing targets for the ULSAB structure were established:

Performance Targets

Mass 200 kg

Static torsional rigidity 13,000 Nm/deg

Static bending rigidity 12,200 N/mm

First body structure mode 40 Hz

NOTE: Structural performance with windshield and backlight; masswithout windshield and backlight.

[

m

m

m

Chapter 10 - Page 3


10.3. Static Rigidity

10.3.1. Test Setup

10.3.1.1. General

The DH in full test configuration consists of the following parts:

• Welded Body Structure• Bonded Windshield and Back Light• Bonded and bolted Panel Dash Insert (Part-No. 022)• Bonded Panel Spare Tire Tub (Part-No. 050)• Bolted Reinforcement Panel Dash Brake Booster (Part-No. 115)• Bolted Braces Radiator (Part-No. 188)• Bolted Reinforcement Radiator Rail Closeout RH/LH (Part -No. 094/095)• Bolted Reinforcement Radiator Support Upper (Part-No. 001)• Bolted Tunnel Bridge Lower/Upper• Bolted Brace Cowl to Shock Tower Assembly

Figure 10.3.1.1-1 DH with Bonded / Bolted Parts

Chapter 10 - Page 4


The unpainted body structure was measured without front and rear suspensionsystem. The body structure was held at four points: the front; at Panel Skirt RH/LH(Part-No. 096/097) and the rear; at Plate Rear Spring Upper (Part-No. 110).

Along the front rails, the rockers, and the rear rails 12 stadia rods were attached.Twenty-four electronic feelers measured the movements of these rods.

Aluminum panels with glass thickness were used to simulate the bonded windshieldand backlight. Due to the fact that the related material property for rigidity andstiffness, the Youngs modulus, shows a close similarity for glass and aluminum.This can be done without compromising the test results, but taking advantages intiming and cost.


The DH was mounted to the test rig with rigid tubes. Two rear locations at the platespring rear upper were constrained, while the load was applied to panel skirt RH/LHby a scale beam.

Figure 10.3.1.2-1 Test Configuration for Static Torsion

Chapter 10 - Page 5


The measurements were taken with four different loads from Mt=1000Nm to

Mtmax=4000Nm.

Before starting the measuring procedure, the maximum load was applied to the DHto eliminate the sag rate.


The DH was mounted to the test rig by rigid tubes. The four fixing points of the DHwere constrained.

The loads were applied to the center of the front seats and to the center of the twoouter rear seats.

Figure 10.3.1.3-1 Test Configuration for Static Bending

The measurements were taken with four different loads from Fb = 1000 N

(4 x 250 N) to Fbmax = 4000 N (4 x 1000 N).

Before starting the measuring procedure, the maximum load was applied to the DHto eliminate the sag rate.

Chapter 10 - Page 6


10.3.2. Results


Figure 10.3.2.1-1 DH on Test Rig for Static Torsion

The torsional rigidity for the test unit in the configuration described in section10.3.1.1 is:

With glass 21,620 Nm/degWithout glass 15,790 Nm/deg

Chapter 10 - Page 7


In general, the graph plot is running harmonic. There is only a jump in rigiditybetween x = 3800 to x = 4200. This is related to the positive impact of the MemberPass Through (Part-No. 090) to the torsional stiffness.

Figure 10.3.2.1-2 Torsion Lines 4 Load Cases with Glass

Figure 10.3.2.1-3 Gradient of Torsion Line with Glass

The above graph shows the gradient of the torsion line. The disharmonies of thetorsion line can be seen in a higher resolution.

Test Unit Gradient Torsion

500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Longitudinal Axis X [mm]

-0.2

-0.1

0

0.1

0.2

0.3

0.4

Gra

dien

t [°/

m

Rear AxleFront Axle

Test Unit Displacement Torsion

500 1000 1500 2000 2500 3000 3500 4000 4500 5000


-5

0

5

10

15

20

Ang

le o

f Tw

ist [

min

]

4000 Nm3000 Nm2000 Nm1000 Nm

Rear AxleFront Axle

Chapter 10 - Page 8


The torsional rigidity for DH #2 in the configuration described in section 10.3.1.1 is:

With glass 20,800 Nm/degWithout glass 15,830 Nm/deg

Figure 10.3.2.1-4 Torsion Lines 4 Load Cases with Glass

As expected, the results are very close to the test unit.

This assumption is based on the test results without glass, because these arenearly identical (15,790 Nm/deg vs. 15,830 Nm/deg).

DH #2 Displacement Torsion

500 1000 1500 2000 2500 3000 3500 4000 4500 5000


-5

0

5

10

15

20

Ang

le o

f Tw

ist [

min

]

4000 Nm

3000 Nm

2000 Nm

1000 Nm

Rear AxleFront Axle

Chapter 10 - Page 9


The above graph shows the gradient of the torsion line. The disharmonies of thetorsion line can be seen in a higher resolution.

Figure 10.3.2.1-5 Gradient of Torsion Line with Glass

DH #2 Gradient Torsion

500 1000 1500 2000 2500 3000 3500 4000 4500 5000


-0.2

-0.1

0

0.1

0.2

0.3

0.4

Gra

dien

t [°/

m]

Rear AxleFront Axle

Chapter 10 - Page 10


As the numbers show, only the bolted brace cowl to shock tower assembly has asignificant impact on the torsional rigidity of 6.3%.

80

90

100

110

100.098.3 98.3

92.0 92.0

Test Configuration

Tors

ion

Rig

idity

[%

]

1 2 3 4 5

Torsion Rigidity

Figure 10.3.2.1-6 Torsion Rigidity Five Test Configurations

To investigate the impact of several bonded and/or bolted parts, additionalmeasurements in various test configurations were undertaken with the test unit.

Test Configurations:

1. Full configuration as described in Section 10.3.1.12. As 1, but without braces radiator (Part-No. 188)3. As 2, but without radiator support upper (Part-No. 001/094/095)4. As 3, but without bolted brace cowl to shock tower assembly5. As 4, but without tunnel bridge




Figure 10.3.2.2-1 DH on Test Rig for Static Bending

The bending rigidity of the test unit in the configuration described in Section10.3.1.1 is:

With glass 20,460 N/mmWithout glass 17,150 N/mm



Figure 10.3.2.2-3 Deviation from the Average Bending Line with Glass

Figure 10.3.2.2-2 Bending Lines 4 Load Cases with Glass

The graph is running harmonic. There is only a local increase in bending rigiditybetween x = 3500 and x = 4200. This indicates a stiff joint between rocker and rearrails. Furthermore, Porsche relates this to the design of the side roof rail.

The above graph shows the deviation from the average value of the bending line.The disharmonies can be seen in a better resolution.

Test Unit Average Deviation Bending

500 1000 1500 2000 2500 3000 3500 4000 4500 5000


-30

-20

-10

0

10

20

30

40

50

De

via

tion

fro

m t

he

ave

rag

e [

%] Front Axle Rear Axle

Test Unit Displacement Bending

500 1000 1500 2000 2500 3000 3500 4000 4500 5000


-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

Ver

tical

Dis

plac

emen

t [m

m]

4000 N3000 N2000 N1000 N

Rear AxleFront Axle



The bending lines show the same characteristics as for the test unit, but theabsolute value decreased by 11%. The local increase between x=3500 and x=4200is not so evident as it was on the test unit. This could be created by localmodifications of the side roof rail and the rear rails for improved manufacturing.Furthermore, the material gage of the panel roof changed from 0.77mm to 0.70mmdue to material availability problems for the test unit; this was also a factor for thedecrease of the absolute value.

Additionally Porsche has experienced that static rigidities of body structures differby plus/minus five percent (5%) even under series production conditions.

Figure 10.3.2.2-4 Bending Lines 4 Load Cases with Glass

The bending rigidity for DH #2 in the configuration described in Section 10.3.1.1 is:

With glass 18,100 N/mmWithout glass 15,950 N/mm

500 1000 1500 2000 2500 3000 3500 4000 4500 5000


-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

Ver

tical

Dis

plac

emen

t [m

m]

4000 N3000 N2000 N1000 N

Rear Axle

DH #2 Displacement Bending

Front Axle



Figure 10.3.2.2-5 Deviation from the Average Bending Line with Glass

The above graph shows the deviation from the average value of the bending line.The disharmonies can be seen in a better resolution.

500 1000 1500 2000 2500 3000 3500 4000 4500 5000


-50

-40

-30

-20

-10

0

10

20

30

40

50

Dev

iatio

n fr

om th

e av

erag

e [%

]

DH #2 Average Deviation Bending

Front Axle Rear Axle



80

90

100

110

100.0 100.0 99.0 98.8 100.0

Test Configuration

Ben

ding

Rig

idity

[%

]

1 2 3 4 5

Bending Rigidity

To investigate the impact of several bonded and/or bolted parts, additionalmeasurements were undertaken:

Test Configurations:

1. Full configuration as described in Chapter 10.3.1.12. As 1, but without braces radiator (Part-No. 188)3. As 2, but without radiator support upper (Part-No. 001/094/095)4. As 3, but without bolted brace cowl to shock tower assembly5. As 4, but without tunnel bridge

As the numbers show, none of these parts display a significant impact on bendingrigidity.

The increase from test configuration four (4) to test configuration five (5) is causedby local effects of the tunnel bridge to the displacement of the rocker. This behaviorwas also noticed in other body structures.

Figure 10.3.2.2-6 Bending Rigidity Five Test Configurations



Figure 10.4.1-1 Test Configuration for Modal Analysis

10.4. Modal Analysis

10.4.1. Test Setup

A modal analysis describes the vibration behavior of a structure. Results of a modalanalysis are the resonance frequencies of the specific structure and thecorresponding mode shapes (how the structure vibrates).

The ULSAB structure was suspended on a test rack held by rubber straps todecouple the test unit from the supporting structure of the test rack.

In order to find the mode shapes and the resonance frequencies, energy is appliedto the structure. The response of the structure (in general the acceleration atdifferent points) is measured in relation to the input forces. From the contribution ofeach input force to each response value, the dynamic behavior of the structure iscalculated.

In the case of the ULSAB, the body structure is excited by means of fourelectrodynamic shakers that are coupled to the corner points of the structure.



The simultaneous excitation with four shakers is necessary to provide good energydistribution into the structure and to minimize the influence of possible nonlinearitiesto the quality on the results. In addition, the torsion and bending modes of the bodycan be excited definitely. Torsion and bending are the most important global modesof a body structure.

Each of the four shakers is driven by a computer-generated, statistical independentband limited (0 to 100 Hz) Gaussian random noise spectrum. The response of thestructure is determined by measuring vibration transfer functions between theacceleration at each measurement point in three orthogonal directions and eachdriving force.

Figure 10.4.1-2 Set-Up for Modal Analysis

The global parameters of the structure, frequency and damping are determinedthereafter by a Least Squares Complex Exponential (LSCE) fitting.

HP 9000/700LMS CADA-X

DAC Interface

ADC Interface

Memory

Power Amplifier

Charge Amplifier

Aliasing Filterand Amplifier

Accelerometer

ElectrodynamicShakers



Figure 10.4.2-1 DH on Test Rig for Modal Analysis

The modal displacement is calculated subsequently by fitting a Multiple Degree ofFreedom (MDOF) model to the transfer functions in the time domain.

The test configuration of the test unit was exactly the same as the testing of staticrigidities described in section 10.3.1.1.

10.4.2. Results



The dynamic rigidity of the ULSAB structure is remarkably good, as it was alreadyindicated by the static test results. Windshield and backlight have a significantimpact on the first torsion mode. The difference is in the same range, as knownfrom other sedan body structures.

The effect on first bending and first front-end lateral mode is relatively small. Forthe test configuration with glass, the first torsion mode and the first front-end lateralmode are coupled at 60.6 Hz.

Figure 10.4.2-2 Modal Analysis Results - Test Unit

The global modes of the test unit in the described test configuration can be seen inthe following chart:

40

50

60

70

49.1

60.8

64.3

60.662.4

60.6

Firs

t M

odes

[H

z]

Torsion Bending Front End Lateral

without glass with glass

Test Unit Modal Analysis



Figure 10.4.2-3 Frequency Response Functions - Test Unit

The graph plot above shows the frequency response functions, measured at thefour driving points. Second bending mode at 63.5 Hz occurs mainly in the rear;whereas the first bending mode occurs in the front and rear of the structure.

2

Test Unit Modal Analysis with ScreensFrequency Response Functions, measured at the body corner pointsPower input by means of electrodynamic shakers at the body corner points

Test Unit Modal Analysis with Screens

Freq

uenc

y R

espon

se F

unct

ion

Am

plit

ude

[(m

/s2)/

N]

Frequency [Hz]

50 52 54 56 58 60 62 64 66 68 700

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8First Torsion 60.6 Hz Bending 63.5 Hz

First Bending 62.4 Hz

Front Left

Front Right

Rear Left

Rear Right

Corner Points



The dynamic rigidity of DH #2 is in the same range as the values of the test unit.The front-end lateral mode changed remarkably. This is created by the change ofthe material gauge of the rail fender support inner from 0.9mm to 1.2mm.

The torsion mode and bending mode without glass decreased slightly, but withglass, the loss of dynamic rigidity is compensated.

Figure 10.4.2-4 Modal Analysis Results - DH #2

The global modes for DH #2 in the described test configuration can be seen in thefollowing chart:

40

50

60

70

47

57.2

66.5

60.1

63.9 64.9

Firs

t M

odes

[H

z]

Torsion Bending Front End Lateral

w ithout glass w ith glass

D H #2 M oda l Ana lys is



The graph plot above shows the frequency response function, measured at the fourdriving points. The amplitude of the first bending increased in relation to the testunit. This is in correlation with the decrease of the static bending rigidity.

Additional modal analysis was conducted on the ULSAB structure, to investigate theinfluence of several bolted and/or bonded parts.

Test configurations:

1. Full test configuration as described in chapter 10.3.1.1.2. As 1, but without bolted brace cowl to shock tower assembly3. As 2, but without braces radiator (Part-No.188)4. As 3, but without tunnel bridge5. As 4, but without radiator support upper (Part-No. 001/094/095)

Figure 10.4.2-5 Frequency Response Functions - DH #2

DH #2 Modal Analysis with ScreensFrequency Response Functions, measured at the body corner pointsPower input by means of electrodynamic shakers at the body center points

Front Left

Front Right

Rear Left

Rear Right

50 706052 54 56 58 62 64 66 68

0

4

1

2

3

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

2.2

2.4

2.6

2.8

3.2

3.4

3.6

3.8

18-12-97

ULSAB_DH2_mS

ulsabdh2

ULSAB DH2Body Structure with Screens

Project:

Test:

Date:

Vehicle:

Driving Points:

Body Corner Points

Measurement Points:

Body Corner Points

Frequency Hz

Fre

quen

cy R

espo

nse

Fun

ctio

n A

mpl

itude

[(m

/s2)

/N] First Bending 63.9 Hz

First Torsion 60.1 Hz



40

50

60

70

60.6

47.0 47.3 47.2

53.4

60.6 61.0 61.0 60.8 60.3

62.4 62.4 62.4 62.3 62.3

Test Configuration

Firs

t M

odes

[H

z]

1 2 3 4 5

Front End Lateral Torsion Bending

Modal Analysis

Figure 10.4.2-6 Modal Analysis Five Test Configurations

The influence of the bolted brace cowl to shock tower assembly on the front-endlateral mode of 13.6 Hz is evident.

Test configuration 5 shows an improvement in the front-end lateral mode, but this ismainly caused by the influence of the mass of assembly radiator support.

The other modifications have no evident impact on dynamic rigidity.



The measured mass in full test configuration included the mass of the bolted bracecowl to shock tower assembly and tunnel bridge, which were installed for testingonly (see 10.3.1.1 Test Configurations). The mass of Windshield and backlightwere not included. The mass in this test configuration was the following:

Test Unit 197.3 kg

DH #2 198.5 kg*This mass includes 2.86 kg for the bolted brace cowl to shock tower assembly and tunnel bridge

The calculated mass for non-constructed reinforcements and brackets has to beadded (see Chapter 5 on Design and Engineering).

10.5. Masses in Test Configuration

A crane with a scaled load cell balanced the DH.

Figure 10.5-1 DH #2 on Crane



ULSAB Testing Results Overview vs. CAE Results

Testing CAETest Final Test Benchmark

Testing DH #2 Unit Version Unit Average Targets

Static Rigidity

Torsion (Nm/deg) 20,800 21,620 20,350 19,020 11,531 ≥ 13,000

Bending (N/mm) 18,100 20,460 20,540 20,410 11,902 ≥ 12,200

Modal Analysis

Torsion (Hz) 60.1 60.6 61.4 61.1 38* ≥ 40

Bending (Hz) 63.9 62.4 61.8 64.1 38* ≥ 40

Front End Lateral (Hz) 64.9 60.6 60.3 58.5 38* ≥ 40

10.6. Summary

All test results proved excellent performance and coordination between test resultsand CAE results for structural performance values.

This is caused by the fact that the approach from former times, to define thestructural body parts by these requirements, is superseded. Nowadays, these bodyparts are mainly specified by safety requirements.

The results gained by CAE calculations are in good, if not excellent, correlation withthe test results.

* 1st mode shape varied for each vehicle benchmarked



Chapter 11 - Page 1



11.1. Introduction

The objective of this program was to establish a credible cost estimation of theULSAB body structure by using automotive practices of manufacturing engineering,process engineering and cost estimating.

To undertake this program, Porsche Engineering Services, Inc. (PES) organized aninteractive process between product designers, stamping process engineers,assembly line designers and cost analysts. The team was comprised of thefollowing organizations:

Porsche Engineering Services .... Program Management Knight Engineering .... Stamping Process Engineering Schaefer GmbH .... Hydroform Process Engineering Classic Design .... Assembly Process Design Porsche AG .... Process Validation Camanoe Assoc. / IBIS Assoc .... Cost Analysis

Because end users would want to analyze “what if” scenarios and compare existingor potential body structures to ULSAB, the entire program used a technical costmodel program developed by Camanoe Associates (a group of MIT researchers)and IBIS Associates.

The technical cost model is programmed to allow the user to change any of thegeneral inputs to suit their specific environment or to change specific inputs foralternative processes.

In addition, because the costs shown on the ULSAB cost model reflect only factorycosts and are relative to the level of product development as of today, a user maywish to enter additional cost categories for both ULSAB and a comparative bodystructure. The cost model has been arranged to accommodate this.

Chapter 11 - Page 2


Some of the areas not included in the ULSAB Cost Analysis are:

• SQA (Supplier Quality Assurance), quality testing, auditing• Impact on body structure through other system developments, i.e., electrical,

trim, powertrain, etc.• Changes as a result of physical body structure testing• Start up and production launch costs• Marketing campaigns• Transportation costs• Departmental costs, marketing, finance, purchasing, human resources, etc.• Preparation for paint

11.2. The Process of Cost Estimation

11.2.1. Overview

The Economic Analysis of ULSAB began with the establishment of the basicassumptions regarding general inputs. This was achieved through a series ofmeetings between the Economic Analysis Committee of ULSAB and the EconomicAnalysis Team.

The program then commenced to establish the estimated production costs against anextremely well defined design. Having a process design meant that costs could beanalyzed based on exact definitions concerning fabrication and assemblyrequirements.

On the parts fabrication side, each stamping and hydroformed component wasstudied to determine the process. This step was undertaken by Knight (Stampings)and Schaefer (Hydroforming) who provided the initial inputs on operationrequirements, equipment requirements, tooling costs, manpower requirements, etc.On all major components Porsche, Germany confirmed the data.

Chapter 11 - Page 3


Assembly Requirements(number and type of welds)

Fabrication Process Parameters(line run rate, tool cost, press cost,number of hits)

Assembly Process Parameters(total equipment cost, number of workers, etc.......)

Complete Porsche Design

Consensus amon g:- Camanoe / IBIS- Kni ght En gineerin g- Porsche En gineerin g- Porsche AG

Part Definitions(mass, area, etc........)

Assembl y Line desi gned explicitl y for ULSABby Classic En gineerin g

General Inputs - ULSABEconomic Analysis Committee

Cost Model Cost Model Algorithmby Camanoe / IBIS

Figure 11.2.1-1 Mechanism for Determination of All Part Inputs

This data was then compared to the mass industry data bank at the MassachusettsInstitute of Technology (MIT) to ensure reasonableness before being used for costestimating.

For the assembly line design and processing, PES provided Classic Design with adetailed bill of materials (BOM) and parts sequencing. From this, each area andstation was developed in a macro view, which established the equipment, tooling,building and manpower required to fulfill the production requirements. Followingvalidation by Porsche, Germany this data was then forwarded to Camanoe for finalcost estimation.

Chapter 11 - Page 4


11.2.2. Cost Model Algorithm Development

In this section the methodology for development of the technical cost models isdescribed. The cost models can be used not only for determining manufacturingcosts for the ULSAB design, but also for costs associated with alternative designs.The models allow the capability to track the major cost contributors and todetermine opportunities for target areas for reduction.

The principal objective for this project includes development of a cost estimationtool to aid automotive designers specifically interested in costs associated with theULSAB design. The cost model permits any user to easily adapt various inputparameters, allowing cost investigations for alternative designs on a consistentbasis.

The cost model must account for various processes used in the manufacture of thebody structure, including stamping, hydroforming and assembly. Based onnumerous input parameters, both economic and technical, the model tracks costcontributions to the stamping process from blanking, welding (for tailor weldedblanks) and stamping for all parts. Similarly, hydroformed part costs are brokendown into contributions from bending, pre-forming and the final hydroformingfabrication. The assembly process costs include cost contributions from spotwelding, active gas metal arc welding (MAG), laser welding and adhesive bonding.

Technical cost modeling is a technique developed and used by Camanoe and IBISfor simulating manufacturing costs. The technique is an extension of conventionalprocess modeling, with particular emphasis on capturing the cost implications ofmaterial and process variables and various economic scenarios.

The focus of the technical cost models developed for ULSAB are limited to directmanufacturing cost, although the models could be expanded to include indirectcosts and aspects of the entire product life-cycle. Direct manufacturing costsinvolve specific processes: fabrication and assembly of the body structure. Indirectmanufacturing costs, including executive salaries, marketing and sales, shippingand purchasing, research and development, and profits are not considered.

Chapter 11 - Page 5


Cost is assigned to each unit operation from a process flow diagram. For each ofthese unit operations, total cost is broken down into separately calculated individualelements.

• Variable cost elements: Materials, labor, and energy• Fixed cost elements: Equipment, tooling, building, maintenance,

overhead labor and cost of capital

Developed to breakdown and track contributions from variable and fixed costs, themodels identify the major cost contributors to manufacturing. After the directmanufacturing costs are established based on an initial set of input parameters,sensitivity analysis can be performed to indicate the cost impact of changes to keyparameters. Technical cost models provide an understanding not only of currentcosts, but also of how these costs might differ in the face of future technological oreconomic developments. Typical parameters investigated via sensitivity analysesinclude: annual production volume, throughput (cycle time or production rate), rawmaterial prices and tooling costs.

Models can be implemented in either a descriptive or predictive manner. In eithercase, direct inputs are specified for the product material, geometry andmanufacturing scenario. With descriptive models, the user directly inputs theintermediate parameters such as production rate, equipment cost and tooling cost.In the predictive approach, the model as a function of the product material andgeometry calculates the intermediate parameters. These predictive functions arederived from analyzing a continually expanding range of case studies, and areupdated routinely. It is this predictive nature of technical cost models that separatesthem from other cost estimating tools.

Chapter 11 - Page 6


11.2.3. General Inputs

As stated previously, the Economic Analysis began with the establishment of thegeneral inputs. An example of these inputs is as follows:

11.2.4. Fabrication Input

For each part in the ULSAB design, a press line time requirement was calculated.The machine clean running rate, the line downtimes, the part reject rates and thetotal annual production volume are used to determine the total time needed on theline for the given year. This information, combined with the technical requirementsfor stamping each part is used to calculate the total number of each press line typeneeded to produce the ULSAB body structure. For ULSAB, it was determined thata total of 15 press lines and five blanking lines were needed to produce all thenecessary parts and blanks.

Input

Production Volume 60 jobs per hour

Working Days per Year 240

Production Location Mid-West USA

Wage including Benefits $44.00 per hour

Interest Rate 12%

Equipment Life 20 years

Production Life 5 years

Building Life 25 years

Chapter 11 - Page 7


1 2 3 4 5 6 7 8 9 10 11

12

13

14

15

B1 B2 B3 B4 B5

A

A

A

A

A

A

B

B

B

B

C

C

C

C

C

C

C

C

C

C

C

C

D

D

D

D

D

D

D

D

D

D

D

D

E

E

E

E

4500tons

3600tons

400

t

400

t

400

t

600

t

1000

t

The accompanying press shop layout shows the distribution of these 15 press linesand five blanking lines among the various equipment types shown in the previousslide. The layout also shows the number of presses required on each line. Forexample, there is only one line using “Press Group A” and it contains six presses;there is one line using “Press Group B” containing four presses; three lines using“Press Group C” containing four presses each; and four lines using “Press Group D”containing three presses each. In addition, one of each large transfer press typesand four smaller transfer presses suitable for the progressive die parts were alsoused. Finally, one large, one medium and three small blanking lines were required.

Figure 11.2.4-1 ULSAB Press Shop Layout

Chapter 11 - Page 8


The press line descriptions are as follows:

Press Capacity Size

Press Group A: 1600 ton DA/1000 ton SA 4572 mm x 3048 mm

Press Group B: 1000 ton DA/800 ton SA 3048 mm x 2032 mm

Press Group C: 800 ton DA/500 ton SA 2743 mm x 1524 mm

Press Group D: 500 ton DA/350 ton SA 2438 mm x 1220 mm

Press Group E: 350 ton SA 2134 mm x 1220 mm

(Progressive Dies)

Transfer Presses: 4500 ton & 3600 ton

Blanking Lines: 400 ton 2438 mm x 1220 mm

600 ton 2743 mm x 1524 mm

1000 ton 3048 mm x 2032 mmDA = Double ActionSA = Single Action

11.2.5. Assembly Input

The assembly line was designed explicitly for ULSAB by Classic Engineering whichincludes equipment and tooling investment, assembly plant area and labor force.Cost enhancements concerning material, energy, overhead labor and maintenancewere performed by Camanoe and IBIS.

It is very important to remember that the assembly line was designed for a net linerate of 60 jobs per hour. Because of the various line downtimes, this requires arunning rate of 72 body structures per hour, which in turn implies that there are only48 seconds per station to perform assembly operations and transport the body tothe next station. In practice, increasing (or decreasing) the line running ratechanges the time available at each station to perform the assembly operations andthus changes the line configuration, resulting in different levels of requiredinvestment. Because the line was actually designed for one line speed (net rate of60 body structures per hour), the model is unable to adjust the investment based onthe different line rates. Consequently, the user MUST change the assemblyinvestment inputs in order to have an accurate estimate of the assembly cost atother production volumes. Additionally, ULSAB is costed against specific spotwelds and laser welds, any alteration to this situation would require a re-evaluationof the equipment and manpower needed.

Chapter 11 - Page 9


Overall Costs

Cost Summary

Cost Breakdown

Investments

General Inputs

Part Inputs

Calculations

Machine Rents

Information

11.3. Cost Model Description

The following chapter describes the salient information and input parameters withinthe ULSAB Technical Cost Model. With the enormous quantity of input parametersrequired for cost calculations, validation and consensus among all participants arecritical for appropriate ULSAB cost determination. A description of the process forgenerating consensus on all of the input parameters for the ULSAB design isdiscussed.

Figure 11.3-1 Technical Cost Model Layout

The ULSAB technical cost model consists of the following nine major sections orsheets, in order of appearance: Overall Costs, Cost Summary, Cost Breakdown,Investments, General Inputs, Part Inputs, Calculations, Machine Rents andInformation.



The Overall Costs sheet, appearing first, reports the total cost for body structurefabrication. This sheet provides the user with a brief synopsis of the model outputs,which include cost contributors for stamping and assembly of a body structure. Asmentioned in the introduction, the user will be able to input additional costs asrequired. The second sheet, Cost Summary, provides more detail by listing costcontributors for each part ID number or assembly area. The next sheet, CostBreakdown, gives further detail on the contributors to part cost. Cost contributorsfor each part ID are broken down by process step, and the information in this sheetis organized slightly differently than in the Cost Summary sheet. No information onassembly is contained on the Cost Breakdown sheet, only costs related to partproduction. The 2 input sheets (General Inputs and Parts Inputs) contain all of thepertinent input parameters for cost calculation. The Calculations sheet listsintermediate cost output calculations that may be of interest.

The model includes a sheet that can be used to test the effect of various sets ofinput parameters on the machine rents. Finally, the Information sheet givesinformation concerning the size and the gages of the blank sizes to be used forULSAB.

Organizational Format of Model Sheets

Stamped Parts:General Output Costs Process Specific Information

Tubular and Purchased Parts:General Output Costs Cost Breakdown by Element

Assembly:General Output Costs Cost Breakdown by Element

Figure 11.3-2 Organizational Format of Model Sheets



Most of the eight sheets are organized in a similar manner, as shown schematicallyin the figure above. This organization is consistent for cost sheets and calculationsheets. By paging down each sheet, three sections become apparent: StampedParts, Tubular and Purchased Parts and Assembly.

By paging across the sheet within each of these sections, the costs for specificparts or assembly processes (listed by ID) are identified, and sorted into twocategories: General Output Costs and Cost Breakdown by Element.

Within the General Output Costs regions, the total cost for fabricating parts is listedfor each part, identified by part ID and name. Hence part cost information for eachstamped, tubular and purchased part is readily available. The total cost forfabrication is summed at the bottom of each column and section.

Paging across to the Cost Breakdown by Element region, the total cost for each partis broken down into nine cost categories, including material, energy, labor,equipment, tooling, overhead labor, building, maintenance and working capitalcosts. Addition of all cost elements in a given row sums to the total part cost. Eachof the nine cost elements is also totaled at the bottom of each column for all parts toprovide a total cost breakout by element in the Stamping, Tubular and PurchasedParts and Assembly sections.

11.4. ULSAB Cost Results

11.4.1. Overall Cost Results

The cost analysis for the ULSAB design is presented, including a breakdown ofcosts by processes, factor elements, and investments. The costs associated withnew technologies are focused upon, specifically for all the tailor welded blankstamped parts and for the hydroformed side roof rail. Sensitivity analyses areincluded for changes in input parameters, which may affect the cost of TWBprocessing.

The manufacturing costs for the ULSAB body structure at 203.2 kg with 158 partsresult in an overall value of $947 per body structure.



The body structure cost can be broken down into $666, from parts fabrication and$281 from assembly. Of the 158 parts in the ULSAB design, the 94 major stampedparts make up the majority of the mass (184 kg) and represent the largest costelement at $584. Tubular parts, such as the two hydroformed side roof rails and themember pass through beams, as well as a large number of small brackets andhinges (normally out-sourced by the auto maker), make up only a small portion ofboth the overall mass and cost.

Figure 11.4.1-1 ULSAB Overall Cost Results

The breakdown of the variable costs (and the remaining fixed cost total), both forparts fabrication and assembly, shows the importance of the material and fixedcosts. Material (steel) is the single largest cost driver, accounting for 37% of thetotal body structure cost. Total fixed costs (for parts fabrication and assemblyoperations), which primarily derive from the investments in plant equipment andoverhead, also lead to 44% of the body structure cost. The labor and energycontributions are relatively small at a combined total of only 10% for the entireassembled body structure.

Number Mass ofCost of Parts Parts (kg)

Stamped Parts $584 94 184.3Tubular & Purchased Parts $82 64 18.9

Assembly $281 --- ---Total Body Structure $947 158 203.2



ULSAB % of Total

Material $353 37%Labor 36 4%

Energy 6 1%Fixed Costs 189 20%

Stamping Parts Fabrication $584 62%

Hydroforming $41 4%

Purchased $41 4%Material $0 0%

Labor 45 5%Energy 10 1%

Fixed Costs 226 24%

Assembly $281 30%

Total Body Structure Cost $947 100%

ULSAB % of Total

Stamping $584 62%

Hydroforming $41 4%

Purchased $41 4%

Assembly $281 30%

Total Body Structure Cost $947 100%

Total Number of Parts 158

Total Mass 203.2 kg

Figure 11.4.1-3 Cost Breakdown by Factor

Figure 11.4.1-2 Cost Breakdown by Process Step



Investments ULSAB(Millions)

% of Total

Blanking Tooling $4.4 1.4%

Blanking Lines $10.1 3.2%

Blanking Building $1.2 0.4%

Welding Line $37.2 11.9%

Welding Building $5.9 1.9%

Stamping Tooling $37.1 11.8%

Stamping Lines $102.9 32.8%

Stamping Building $6.1 1.9%

Hydroform Tooling $1.3 0.4%

Hydroform Lines $16.3 5.2%

Hydroform Building $0.5 0.2%

Assembly Tooling $19.0 6.0%

Assembly Equipment $40.4 12.9%

Assembly Building $31.3 10.0%

Total Investments $313.7 100%

Investments for each process step show that the assembly line and related toolingand building expenses account for less than one-third of the total. The press shopis the major source of investment. Press lines account for over 30% of theinvestment total. Welding lines for producing tailored blanks are also significant,despite the fact that there are only 16 tailor welded blank parts used in the bodystructure

11.4.2. Cost Breakdown for Fabrication

The parts fabrication total can be further broken down into $584 for major stampedcomponents (including the Panel Roof which is produced with the Active Hydro-MecProcess), $41 for the two hydroformed side roof rails and $41 for the remainingsmall purchased parts (including ordinary tubes such as the pass-through beamsand a number of small brackets and hinges).

Figure 11.4.1-4 Distribution of Investment Costs



Breakdown for Stamped PartsCost perVehicle

Material Cost $353

Labor Cost $36

Energy Cost $6

Total Variable Costs $395

Equipment Cost $88Tooling Cost $51

Overhead Labor Cost $27

Building Cost $7

Maintenance Cost $15

Working Capital Cost $1

Total Fixed Costs $189

TOTAL COST OF STAMPED PARTS $584

The primary driver for the major stamped parts is material. Due to the stage ofprogram development, a very cautious approach was taken in determining blanksizes; therefore the level of engineered scrap results in a relatively high materialcost.

Figure 11.4.2-1 Overall Cost Breakdown for Stamping

As is typically the case, the other main cost components for the stamped parts arethe equipment (press lines) and the tooling.



11.4.3. Cost Breakdown for Assembly

Body structure assembly contributes less than one-third of the overall bodystructure cost. The main cost elements are the labor (mostly the indirect oroverhead labor) and the assembly line equipment. Notable is the relatively lowequipment cost which results from the reduced assembly effort required as a resultof the parts consolidation.

Breakdown for AssemblyCost perVehicle

Material Cost $0

Labor Cost $45

Energy Cost $10

Total Variable Costs $55

Equipment Cost $50Tooling Cost $23

Overhead Labor Cost $125

Building Cost $18

Maintenance Cost $9

Working Capital Cost $1

Total Fixed Costs $226

TOTAL COST OF ASSEMBLY $281

Figure 11.4.3-1 Overall Cost Breakdown for Assembly



Part #Material

CostBlanking

CostWelding

CostStamping

CostTotal Cost

008/009 $11.96 $0.75 $2.75 $3.97 $19.43

010/011 $18.25 $0.99 $3.02 $4.16 $26.42

042/043 $25.39 $1.07 $2.20 $4.63 $33.29

046/047 $19.08 $1.10 $3.30 $4.94 $28.42

048/049 $9.27 $0.74 $1.95 $4.75 $16.71

060 $39.44 $1.90 $9.53 $11.06 $61.93

061 $39.43 $1.90 $9.53 $11.06 $61.92

070/071 $9.13 $0.49 $4.40 $3.91 $17.93

096/097 $6.78 $0.49 $1.64 $3.61 $12.52

$178.73 $9.43 $38.32 $52.09 $278.57TOTAL 64% 3% 14% 19% 100%

11.4.4. Cost Analysis for New Technologies and Materials

While there are only 16 parts (eight left/right part pairs) that use tailor weldedblanks, they make up a considerable fraction of the mass of the body structure.These 16 parts weigh 88.38 kg, which is 45% of the total body structure mass. Notsurprising, they also represent a significant portion of the total body structure cost.These parts cost $279 to produce, which is 42% of the cost of all parts fabrication.This, of course, means that these parts cost more per kilogram than the rest of thebody structure. This result is not unexpected because the additional welding step isrequired. However, this relatively small cost increase is compensated for by thereduced part count and thus reduced tooling and assembly costs. Further, the tailorwelded parts offer the mass savings, which is the main objective of the ULSABdesign.

Figure 11.4.4-1 Tailor Welded Blank Part Cost Breakdown



The costs of tailor welded parts are still primarily driven by the material costs, whichmakes up 63% of the total. This is also true for the body sides (parts 060 & 061)where the blanking process was especially productionized to decrease the scrapassociated with the large cutouts for the door openings. Processing costs dividefairly evenly between the welding and stamping operations, with the blanking stepcontributing only a small percentage.

Figure 11.4.4-2 Effect of Welding Parameters on TWB Total Costs

A key question regarding the use of a relatively new technology (i.e. tailor welding ofblanks) is the certainty of the process variables and the effect of changes in theseparameters on the part cost. Three major input parameters were considered for thissensitivity: the weld speed, the line unplanned downtime and the line cost. Thebaseline values used in the cost analysis were 100 mm/sec, 30% (four hrs/daydowntime) and $3.8 million respectively. These factors were allowed to vary withina range of reasonable values. The graph shows that the cost of the parts is most

$250

$275

$300

$325

$350

$375

Weld Speed Downtime Equipment Total

Tot

al C

ost o

f TW

B P

arts

Min: 50 mm/sBase: 100 mm/sMax: 150 mm/s

Max: 40%Base: 30%Min: 15%

Max: $5.3 MillionBase: $3.8 MillionMin: $3 Million

Best of All Inputs

Worst of All Inputs

Baseline

$313

$286

$268$272

$288

$274

$348

$260

$279



Breakdown for One Side Roof RailCost per

RailMaterial Cost $11.08

Labor Cost 1.53

Energy Cost 0.11

Total Variable Costs $12.72Equipment Cost $4.87

Tooling Cost 0.82

Overhead Labor Cost 1.23

Building Cost 0.15

Maintenance Cost 0.58

Working Capital Cost 0.05 Total Fixed Costs $7.70

TOTAL COST PER RAIL $20.42

sensitive to assumptions regarding the weld speed. A weld speed reduction to only50 mm/sec would raise the cost by approximately $35. The downtime and lineequipment costs have much smaller effects that might result in increases (orsavings) of less than $10 each. Even under the worst case scenario of low weldspeeds and high downtimes and equipment costs, the total part cost would only riseby about $50, or about 18%.

Figure 11.4.4-3 Cost Breakdown: Hydroformed Side Roof Rail

Hydroforming is the other new parts fabrication technology used in the ULSABdesign. While there are only two hydroformed parts, the left and right side roofrails, these components enable design changes in numerous other parts in the bodystructure. Because this process produces only two parts the cost significance isrelatively low. Each side roof rail is estimated to cost $20, of which the majority ofthe non-material related costs result from the hydroforming equipment.



The draw operation of the panel roof is planned in hydro-mech technology using a10,000 ton hydraulic press. The investment cost of this press is $84 million,excluding installation and auxiliary equipment, the resulting operation cost includingmaterial is $18.41. The subsequent operations (trimming and flanging) are done inconventional presses. As the draw operation needs a far longer cycle time than theother operations (100 per hour vs. 400 per hour), the production sequencing hasbeen separated.

Laser welding has been incorporated into four areas of the assembly system. Thetotal number of laser welders used is 13 at an average cost of $1.2 M each.

High strength steels range in cost from $0.85 kg to $1.16 kg compared to mild steel,which costs $0.77.

Laminate materials used on the spare tire tub and dash insert is at $3.60 kg. Thisresults in relatively high prices for these parts.



Additionally, Tailored Welded Blanks, Hydroforming and Laser Welding are relativelynew processes. As the utilization of these technologies increases so shouldefficiency and this would result in cost reductions.

11.4.5. Sensitivity Analysis

A key element of the Economic Analysis is to determine the potential costmovements as a result of sensitivity analysis and other scenarios that could impactcost.

Areas investigated are labor wage, unit energy costs, equipment life, building unitcost, production life and material costs:

$875

$900

$925

$950

$975

$1000

LaborWage

Ove

rall

BIW

Cos

t

+ 20%$44 p/hour-20%

$994

$950

$900

$943

$955

$944

$1013

$912

+ 20%0.10 $/kWh-20%

15 years20 years25 years

+ 20%$1500 p/m2

-20%

3 years5 years8 years

+ 10%$352-10%

$952

$942

$909

$982

$947

EquipmentLife

BuildingUnitCost

ProductionLife

MaterialCosts

StampingParts

UnitEnergyCost



11.5. Body Structure – Comparative Study

11.5.1. Overview

Due to the fact that ULSAB’s holistic design approach uses new technologies suchas hydroforming, laser welding, etc., a comparative study using conventionalprocesses was created in order to analyze the overall competitiveness of ULSAB. Abrief description of the models follows:

••••• Year 2000 Reference Model – Base (A)Year 2000 Reference Model is based upon a generic four door passengercar body structure. The general body structure definition consists of abroadly described parts list made of groupings based on their size andcomplexity, and grouping of assembly operations based on their level ofautomation and size. Costs are generated via existing data, automotiveindustry inputs, predictive processes and general assumptions establishedby the Economic Analysis Group. The manufacturing processes used inthis study were conventional stampings, spot welding and limited MAGwelding.

••••• Year 2000 Reference Model – PES Internal Study (B)To further analyze ULSAB’s competitiveness, alternative refinements weremade to the Year 2000 Reference Model (A) in order to establish thepotential range of costs for “classical” structures. To establish this,engineering judgment was used to integrate the general manufacturingassumptions of the Year 2000 Reference Model (A) with the designconcept of ULSAB. Allowances for additional parts and gage increasesdue to the lesser use of high strength steel were made in an effort tosimulate the performance characteristics of ULSAB. The result of thisexercise was Year 2000 Reference Model (B).

As the above described comparative study does not utilize the specific design ordetailed manufacturing cost estimates contained within ULSAB, detail or technicalcomparisons with ULSAB cannot be made.



For the purpose of direct comparisons, a specific detailed cost model of ULSAB inspreadsheet format is available and will be provided by the ULSAB Consortium toautomotive manufacturers. This will allow the automotive OEMs to directly comparein detail, their current or future planned models with ULSAB.



Cost Model InputsYear 2000

ULSAB (A) (B)* Body Structure Mass

Stampings (kg) 184 230 248Hydroformings (kg) 10 0 0

Purchased Parts (kg) 9 20 10

Total Mass (kg) 203 250 258 Material Utilization (Stampings) 49% 55% 50%

Parts Fabrication Direct Labor (Manpower) 59 79 40

Indirect Labor (Manpower) 47 36 28

Total Parts Count 158 200 171

Large Stamped Parts 11 6 12Medium Stamped Parts 39 79 54

Small Stamped Parts 44 50 40

Hydroformed Rails 2 0 0Purchased Parts 62 65 65

Total Number of Die Sets 61 109 65Transfer 14 20 14

Tandem 27 59 33

Progressive 18 30 18Hydroform 2 0 0

Hits per Die Set

Transfer 4.1 3.8 4.1Tandem 3.6 3.2 3.6

Hits per Part

Transfer/Tandem Combined 2.5 2.5 2.3

Assembly Direct Labor (Manpower) 64 80 74

Indirect Labor (Manpower) 178 210 202 Number of Spot Welds 2,206** 3,250 3,060

Length (mm) of Laser Welds 18,286 0 0

Number of Robots 136 200 154 Number of Laser Welders 13 0 0

Number of Assembly Stations 114 130 128

Assembly Building Area (m²) 20,865 30,000 28,156

11.5.2. Assumptions

** Includes 80 spot welds for brackets and reinforcements * PES Internal study



11.5.3. Overall Results

Year 2000

ULSAB (A) (B)*

Stamping $584 $609 $592

Hydroforming 41 0 0

Purchased 41 41 41

Assembly 281 329 308

Total Cost $947 $979 $941

Total Mass (kg) 203 250 258 * PES Internal Study



11.6. Conclusion

The ULSAB design is aimed at achieving two significant goals:

• Major mass savings• Improved performance

These goals have been met by implementing appropriate materials andtechnologies in to a holistic design approach. Individually some of the processes,such as, high strength steels, tailored welded blanks, hydroforming and laserwelding are considered expensive, but when used in conjunction with a good designconcept, gage reduction, part consolidation and efficient manufacturing methods, itresults in an extremely cost competitive product.

The results show that the Year 2000 Reference Model iterations are within 3.5% ofthe ULSAB cost but carry a major weight penalty.

As this cost difference is smaller than the recognized level of variance generallyconsidered for a calculated cost estimate, it is accepted that all models would costapproximately the same.

Therefore, in conclusion, when coupled with good design, the technologies of highstrength steel, tailor welded blanking, hydroforming and laser welding can be usedto achieve mass reduction and performance improvements at no cost penalty.


12. Summary of Phase 2Results

Chapter 12 - Page 1


Phase 2 Benchmark Difference

Performance Results Average Difference (%) Mass (kg) 203 271 - 68 - 25%

Static Torsional Rigidity (Nm/deg) 20800 11531 + 9269 + 80%

Static Bending Rigidity (N/mm) 18100 11902 +6198 + 52%

First Body Structure Mode (Hz) 60 38 + 22 + 58%

12. Summary of Phase 2 Results

The Phase 2 of the ULSAB program has come to its conclusion with the build of thedemonstration hardware.

The test results of the demonstration hardware are remarkable.

Relative to the benchmark average vehicle mass of 271 kg, the mass reductionachieved is 68 kg (25%).

The static torsional rigidity exceeds the target. The efficiency (rigidity / mass) hasincreased, in relation to Phase 1, to 102.5 [(Nm/deg)/kg] (Fig. 12-2). The Phase 2structural performance results are shown in the graphs as a tolerance field ratherthan a fixed point. To indicate that the mass and the performances can vary fromone demonstration hardware structures to another, as it would also do in real massproduction. The static bending rigidity as well as the first body structure mode havealso been increased in comparison to the Phase 1 results (Fig. 12-3 and 12-4).

These high levels of static and dynamic rigidity provide an excellent basis for acomplete vehicle development in respect to its NVH behavior.

Figure 12-1 Structural Performance Summary

Chapter 12 - Page 2


50

60

70

Cb

(x10

00)

[N/m

m]

Cb/m

m [kg]All data adjusted to target vehicleCb with Glass, m without Glass

Reference Vehicles: Acura Legend, BMW 5 series, Chevrolet Lumina, Ford Taurus, Honda Accord, Lexus LS 400,Mazda 929, Mercedes Benz 190 E, Toyota Cressida

Bending Rigidity vs. Mass

8090100110

180 200 220 240 260 280 300 320 3404

6

8

10

12

14

16

18

20

22

24

20

30

40ULSABTarget

FuturePerformanceReference

CurrentAverage

ULSABPhase I

ULSABPhase II

18.1

203

Figure. 12-2 ULSAB Phase 2 Torsional Efficiency

Figure. 12-3 ULSAB Phase 2 Bending Efficiency

180 200 220 240 260 280 300 320 3404

6

8

10

12

14

16

18

20

22

24

20

30

40

50

60

70


CurrentAverageC

b(x

1000

) [N

m/d

eg]

Ct/m

m [kg]All data adjusted to target vehicleCb with Glass, m without Glass

Reference Vehicles: Acura Legend, BMW 5 series, Chevrolet Lumina, Ford Taurus, Honda Accord, Lexus LS 400,Mazda 929, Mercedes Benz 190 E, Toyota Cressida

Torsional Rigidity vs. Mass

8090100110

ULSABTarget

ULSABPhase I20.8

203

ULSABPhase II

Chapter 12 - Page 3


180 200 220 240 260 280 300 320 34020

25

30

35

40

45

50

55

60

65

70

ULSABTarget

CurrentAverage

f [H

z]First Body Structure Frequency vs. Mass

m [kg]Lowest global frequencyf with Glass, m without Glass

Reference Vehicles: Acura Legend, Chevrolet Lumina, Ford Taurus, Honda Accord, Lexus LS 400,Mazda 929, Toyota Cressida


ULSABPhase I

ULSABPhase II

60.1

203

Figure. 12-4 ULSAB Phase 2 Frequency Efficiency

The results of the crash analysis confirmed the integrity and safety of the ULSABstructure. The AMS Offset Crash is considered one of the most severe crash testsof today. In recently performed comparison crash tests of AMS, with the samevehicle towards a deformable barrier with 40% offset at 64 km/h versus the AMSOffset Crash barrier with 50% offset at 55 km/n, the results were nearly equal. Thisconfirms that the decision to analyze the ULSAB structure for its offset crashbehavior using the AMS test configuration, determined at the beginning of Phase 2in 1995, was the right choice.

The NCAP 100% Frontal Crash was run at 35 mph, 5 mph above the federalrequirement of FMVSS 208, meaning 36% more energy had to be absorbed.

In both the 50% Offset and 100% Front Crash low footwell intrusion and structuralintegrity proved the safety of the structure.

Chapter 12 - Page 4


The rear impact crash analysis, also run at 5 mph above the required speed of30mph and showed fuel system integrity, passenger compartment integrity, residualvolume and door opening after the analysis.

The side impact crash analysis showed good results for criteria, such as passengercompartment intrusion, B-Pillar displacements and overall shape of deformation.

The roof crash analysis proves that the roof meets the federal standardrequirements and is stable and predictable.

The crash analysis was run with a vehicle crash mass of 1612 kg, meaningsecondary weight savings of other components such as engine; suspension, etc.were not considered, to achieve a conservative approach.

Apart from the design of the structure and its optimized smooth load flow from frontand rear rails into the rocker and the side roof rail concept; the use of high strengthsteels in 90% relative to the ULSAB structure mass was the key to achieve thiscrash performances at low mass.

This need to use high strength steel to achieve this crash performance with thegiven target for mass was a challenge for the part design and our suppliers.

Together with steel suppliers, part manufacturers, designers and engineers, theright materials were selected and the design was modified until it was feasible.

Significant mass reduction was also achieved with the use of tailor welded blanks incombination with high strength steel. The elimination of reinforcements and jointsbetween parts reduced mass and enhanced crash and structural performance.Furthermore, the total number of parts and assembly steps was reduced. With theuse of the tubular hydroforming manufacturing process for the side roof rail andsheet metal hydroforming for the roof panel, parts could be manufactured,contributing to performance and weight reduction. The hydroformed side roof railmade from a tube with a relatively large diameter of 96mm and a wall thickness of1mm from high strength steel was made feasible in Phase 2.

Chapter 12 - Page 5


The assembly sequence of the ULSAB structure with the body side innersubassembly, first assembled to the underbody structure and the body side outer inthe following step, gives better weld access, especially in the rear of the structure.With this assembly sequence, weld access holes can be avoided and structuralperformance can be maintained.

Laser welding in assembly is successfully applied to weld the body side outer paneland the roof to the side roof rail. In addition, it was used to join the fender supportrails and the front rails to enhance the performance.

In terms of the cost analysis, following extensive work in detail processing ofcomponents and assemblies, it was established that ULSAB would cost $947 tomanufacture. The competitiveness of this cost is due to the design concept, whichconsolidated parts and eliminated many reinforcements, therefore saving stampingand welding operations.

These savings were partially offset by the cost of high strength steel and the newtechnologies such as laser welding and hydroforming, but the final conclusion of theanalysis is that ULSAB can be produced without cost penalty.