EU Project MERLIN Needs and Demands for the Manufacture …MERLIN – Work Packages WP1 - Project Definition and Specification WP2 - Shape - Topology optimisation, modelling and validation

© 2012 Rolls-Royce plcThe information in this document is the property of Rolls-Royce plc and may not be copied or communicated to a third party, or used for any purpose other than that for which it is supplied without the express written consent of Rolls-Royce plc.This information is given in good faith based upon the latest information available to Rolls-Royce plc, no warranty or representation is given concerning such information, which must not be taken as establishing any contractual or other commitment binding upon Rolls-Royce plc or any of its subsidiary or associated companies.

EU Project MERLINNeeds and Demands for the Manufacture of Next

Generation Jet Engine Components

Jeff Allen

AKL – International Laser Technology Congress

9th May 2012

© 2012 Rolls-Royce plc

MERLIN – Project Summary The concept of the MERLIN project is to:

Reduce the environmental impact of air transport using Additive Manufacturing (AM) techniques in the manufacture of civil aero engines

Develop AM techniques, at the level 1 stage, to allow environmental benefits including:

- near 100% material utilisation- no toxic chemical usage- no tooling costs- impact the manufacture of future aero engine components (current buy to fly

ratios result in massive amounts of waste) These factors will drastically reduce emissions across the life-cycle of

the parts Also be added in-service benefits because of the design freedom in AM Light-weighting and the performance improvement of parts will result in

reduced fuel consumption and reduced emissions MERLIN will seek to develop the state-of-the-art by producing higher

performance additive manufactured parts in a more productive, consistent, measurable, environmentally friendly and cost effective way

2


MERLIN – Areas for Progression

The MERLIN consortia have identified the following areas where a progression of the state-of-the art is needed to take advantage of AM:

Productivity increase Design or Topology optimisation Powder recycling validation In-process NDT development In-process geometrical validation High specification materials process development

3


MERLIN - Beneficiaries Rolls-Royce WSK "PZL - Rzeszow" S.A. Industria de Turbo Propulsores MTU LPW Technologies Ltd Turbomeca Volvo Aero Corporation TWI Limited Fraunhofer - Gesellschaft zur Forderung der angewandten Forschung e.V. ILT Association pour la Recherche et le Developpement des Methodes et Processus

Industries (ARMINES) ASOCIACION CENTRO DE INVESTIGACION EN TECNOLOGÍAS DE UNION LORTEK BCT University West Frederick Research Centre

Total Budget - €7.1m Project Launch – Jan 2011 Project End – Dec 2014

4


MERLIN – Work Packages WP1 - Project Definition and Specification WP2 - Shape - Topology optimisation, modelling and validation WP3 - Process Development - Process development, process

monitoring and control, and thermal management WP4 - In-Process NDT development and integration, geometrical

measurement and control WP5 - Post Processing - Mechanical testing, analysis, recycling

validation and heat treatment WP6 - Technology Demonstration - Environmental and through

lifecycle evaluation WP7 - Management WP8 - Dissemination and Exploitation

5


6

Material Form

Heat Source

Powder Stream

(Various types)

Powder Bed Wire

Laser(Various types)

Electron Beam x

Electric Arc x x SMD

Heat Source/Material Form Combinations Very good surface finish & precision Slow build rate Small parts only Low build rate

Potential for high build rates Material utilisation varies Options for Repair, Hybrid & OEM Potential for component tailoring Electron Beam systems emerging Requirement for vacuum system with EB

Low cost (100% wire utilisation & medium process speed) Good for titanium parts with low/medium complexity Full fusion Good surface finish Large working envelope Cell installed at AMRC (Sheffield)

A wide range of processes exist, with varying attributes in terms of precision, cost, integrity, etc.

As with subtractive manufacture, there is no single “best process”; selection needs to consider application, including material, size, shape type and complexity, access, inspection and validation.

Laser heat source

Argon chamber / vacuum for EB

Mirrors (focussing /deflection coils for EB

Powder bed

Filters

Focussing Lens

Lens series

Overview of Technologies & Selection


7Overview Powder Bed

Small Parts

Blown Powder Repair Hybrids

Advantages V.good precision of deposition Good surface finish Good material properties Good utilisation of powder Uses established basic technology Automatic operation Better at overhung surfaces Can make shapes which were

previously impossible

Limitations Relatively low build rates Not straightforward to

achieve sub 10ppm oxygen levels

Need to build on a flat base

Relatively low build volume

Limited range of materials at present

Advantages System development Easy Inert atmosphere operation High Deposition rate capability Usable for additive manufacture Usable for repair Relatively wide process window Low heat input to substrate

Limitations Low Powder usage efficiency No so good at overhung

surfaces Tend to have rough surface

finish Need complex manipulation

systems for 3D parts Can get scatter on material

properties


8Business Opportunities

Reduced time to marketSimple/minimal tooling,

Reduced material costReduced waste material (high buy-to-fly ratio)

Reduced ‘removal’ operationsNo roughing, Minimal finishing, Less WIP

Repair capabilities to support aftermarket opportunitiesLower total life cycle cost

Enhanced product opportunitiesDesign for NNS

Additive Bosses

Built up Leading / Trailing EdgeAdditive Flanges


9

Graded & Tailored Materials Can change the alloy composition throughout the shape of the component, avoiding abrupt

changes in material properties Add desired properties to a part only in the areas they are needed without abrupt changes in

alloy compositioneg: Hard facing for integral bearing tracks on shafts

Organic Shapes Can be used to produce components previously not capable of being manufactured, minimum

weight and maximum strength structures akin to ‘natural’ shapes such as: butterfly wings (eg: isogrids) graded spheroidal void structures such as wood and bone

Geodesic structures Skeletal or polygon faceted structures with high specific stiffness, such as the Beijing National

Stadium. Material efficient designs

Design Opportunities

Image courtesy of Optomec Inc.


10Powder Bed - Modes

Rapid Prototyping ModeDevelopment partsShort Lead timeAbility to change design until the last minuteExpect to pay a premium on price

Powder Bed - Modes

Rapid Prototyping ModeDevelopment partsSmall numbersShort Lead timeAbility to change design until the last minuteExpect to pay a premium on price

Powder Bed - Modes

Rapid Prototyping Mode Development parts Short Lead time Ability to change design

until the last minute Expect to pay a

premium on price

OEM* Parts Manufacture Mode Have to compete with

established processes High cost of

Qualification Need Powder Supply

Chain Need Parts Supply

Chain

* Original Equipment Manufacture


11Issues – Specific Cost of Deposited Material

Assumes a self checking, self controlling, serialised, productionsystem

Specific cost of deposited material is primarily a function ofbuild rate and powder cost

- Raw Material Cost- Material Usage Efficiency- Technical support time (programming and development)- Process Labour Cost- Capital Cost - depreciation- Machine Utilisation- Power Cost and usage- Power Conversion Efficiency (wall:delivery)- Build Rate (incorporating feature and surface control)- Consumable form- Set-up time- Planned maintenance & consumables- Associated costs of ancillary processing,

- e.g. machining, heat treatment, inspection


Issues – Multiple Spots In very simple terms – to a first approximation!

Material Props = fn Crystal Structure Crystal Structure = fn Cooling rate Cooling Rate = fn Molten Pool Size Molten Pool size = fn Spot size

Specific Cost of deposited material = fn deposition rate Deposition rate = fn Spot size

Surface Finish = fn Spot Size (larger spots give coarser surface finish)

Ie: Practical systems reach a limit on max spot size This limits the deposition rate & surface finish and defines the material

properties

Points towards the use of Multiple Spots

12


13

Issues – Material Properties UTS

Between Cast and Forged Fatigue

Dependent on level of micro cracking Creep

Fine grain structure tends to lower creep performance Anisotropy

Deposition Parameters can be optimised to minimise anisotropy Fracture Toughness (Hot Creep Rupture)

Can be problematic if the resulting grain structure exhibits epitaxial grain growth

Surface Finish External Surfaces Internal Surfaces Effect on fatigue life


14

Issues – Maximum Component Size

Limited by the tank size of the powder bed system This is presently typically 250 x 250 x 250 Development systems are planned for 400 x 400 x 400 Then 600 x 600 x 600


15Issues – Powder Source

Limited number of sources approved for powder supply into Aerospace

Expensive and long term to Approve new suppliers Development usage is small cf Production Difficult to predict future requirements Ti 6-4 supply particularly difficult Gas Atomised

Plasma Rotating Electrode Powder (PREP) Specialist Nickel Superalloys difficult to supply to

specification Particle size distribution Chemistry Morphology Trace elements eg Lanthanum & Yttrium are difficult to

control at low ppm levels


16

Issues – Residual Stress/Distortion

Material structure is defined when the deposit is hot Thermal contraction must take place during cooling to ambient Results in a combination of:

residual stress distortion in the part Cracking

Aim to accommodate thermal contraction effects in the form of residual stresses which are below the material UTS

Can stress relieve in some cases May result in enforced overbuild to allow the finished shape to be machined

from the distorted geometry EB processes preheat the powder bed – results in reduced residual stress Dev work to use auxiliary lasers to preheat laser powder bed processes at

ILT Can use Preheated base plates to minimise distortion & cracking on cooling

Need 900° - 1000°C for high performance Ni superalloys


17Issues - Process Window Note that for superalloys there are definite practical limits to rate:

The effect of distortion and cracking – though these are not insurmountable Fine melt pool size allows development of optimised parameters for reduced

segregation Rate is limited by consumable feed process control and thermal management

DilutionPorosity

No fusion

Pow

er d

ensi

ty (i

ncre

asin

g as

a

func

tion

ofin

crea

sing

spo

t rad

ius .

Consumable feed rate(hypothesised)Superalloy Process Window

Bead profile aspect ratio

Energy per unit length –related to velocity

(After Steen)

General (laser powder) deposition processing windowFor high temperature alloys, solidification effects and the thermo-mechanical effects of deposition and heat treatment / post processing must be considered. This is due to crack sensitivity / strain aging and end microstructure requirements.

Pow

er d

ensi

ty (i

ncre

asin

g as

a

func

tion

ofin

crea

sing

spo

t rad

ius.

Lin

ked

to

cons

umab

le fe

ed ra

te a

nd lo

cal

dyna

mic

hea

t sin

k.

Travel speed, linked to location

Liquation

Excess segregationSolidification

cracking

Lack of inter-run fusion

Adapted (After Reed)


18

Process Capability

Need a process suitable for routine use on the shop floor Stable Repeatable Predictable

Eg: Powder Bed Intra bed variance Inter-build variance Inter machine variance Inter Operator variance

Depends on the maturity of the system

Measurement and control of Key Process Variables


19Issues - Integrity of Deposited MaterialThere are many claims with regard to density and forged properties,What do they mean?Usually that the porosity is usually fine…

BossBoss

PadPad

The maximum flaw-size is significant for component life

Flaws of various sizes will randomly intersect test bars.

Scatter in performance and tight acceptance thresholds show the importance of control

But aero-applications can experience arduous loading regimes, including:• Stress rupture• Fatigue• Dynamic loading

Schematic (not to scale) of flaws dispersed through a mechanical test block .

…does that always mean that the material is homogeneous?

Interface encompassing fusion zone and HAZ


20Properties of Thin Walls

Outer surfaces of parts are subjected to abrasive blast

Removes adhered powder particles Normalises surface finish

Abrasive blast results in a 100µ thick surface layer

contains cracks (fatigue initiation sites) Material properties would not be expected to be

the same as the bulk material Particular problem when performing stress

analyses of components with thin (c.0.5mm) walls

Unknown material properties for FE Model FE modellers tend to err on the cautious

resulting in designs which are perceived to be overweight

Erodes the business case for WXB Noise attenuators

Need to establish the properties of thin walls for input in to FE Stress models

20

100µ 100µ

500µ

300µ

100µ thick layer from abrasive blast

- Contains cracks

- Unknown material Properties

Usable thickness of base material reduced to 300µ

Unknown effect on base material from defects in the surface layer


21Surface finish External Surface Finish Surface finish is generally a

function of: Orientation of surface in build tank Deposition Rate Powder Size distribution

Laser Processes Smoother Surface Finish Lower Deposition rate

Electron Beam processes Rough surface Finish Higher deposition rates

General approach is to have different sets of parameters for ‘skin’ and ‘fill’ sections of the build

Need to optimise this approach to maximise surface finish quality whilst maximising the overall deposition rate

Need to look further at improving surface finish on overhung & low angle surfaces

21

Internal Surface Finish Need a method of post processing

internal surfaces to achieve an acceptable surface finish

Chemical? Electro Polish? Abrasive media?

Need to use designs which are optimised to give best internal surface quality


22Summary

Need to Develop: Systems with economically competitive

deposition rates and capacities Deposition Parameters and Techniques for High

Temperature Nickel Superalloys Powder sources with suitable Quality and Cost Capable NDE techniques Methods of Topological Optimisation Post processing methods to give acceptable

Surface Finish & Materials Properties

Documents

EU Project MERLIN Needs and Demands for the Manufacture …MERLIN – Work Packages WP1 - Project Definition and Specification WP2 - Shape - Topology optimisation, modelling and validation