Turbocharging CFD with Katana · 2020. 10. 12. · Turbocharging CFD with Katana Unleashing cutting-edge Computational Fluid Dynamics from Japan Makoto Shibahara, Jonas Wirgart, Irie

White paper

Turbocharging CFD with KatanaUnleashing cutting-edge Computational Fluid Dynamics from Japan

Makoto Shibahara, Jonas Wirgart, Irie Tomohiro and Keith Hanna

https://www.cradle-cfd.com

Executive summaryIn the natural world fluids act in harmony with structures, fluids with acoustics, fluids with multibody dynamics, fluids with magnetics, fluids with electrics, and all in a myriad of complex, interactive and very complicated ways. Numerical approximations such as Computational Fluid Dynamics (CFD) have to grapple with these phenomena; frequently with simplistic empirical models to approximate boundary conditions, inclusion of chemical and biological reactions, fluid and solid phase changes and multiphase flows, and not least transient phenomena related to turbulence and moving bodies. Nevertheless, the inherent technical benefits of CFD from its predictive insights to product verification & validation, and business benefits such as reducing costs, right-first-time designs, higher manufacturing productivity, elimination of waste, minimizing product recalls, and faster product time-to-market still remain.

The commercial Computer-aided Engineering (CAE) industry is quite mature after nearly 50 years but one of the biggest challenges still to be overcome is to resolve multiphysics simulations when needed for complex industrial applications, especially related to fluid flow and associated transport phenomena. CFD too has come a long way over the last 40 years having migrated from simple models run on mainframe computers to embracing the CAD, PLM and Cloud revolutions. Software Cradle was one of the first commercial CFD vendors to hit the market in 1984, and it is the only one to emerge from Japan.

Hexagon | MSC’s Cradle CFD solutions have been built with the Japanese philosophy of total quality performance in mind and crafted in the spirit of a Samurai’s Katana sword. In other words, we aim to offer the most expertly crafted solutions for CFD calculations that are fast, robust, accurate, multiphysics-focused and all with excellent engineering usability. Cradle’s development processes are user driven and pragmatic; we do not add flashy functions for the sake of ticking a box - the value of our features are defined only by their capability and quality. Software Cradle, therefore, produces CFD software with analyst level accuracy, yet with designer level user experiences. Like Katana swords, Cradle CFD tools are sharp, tough, fast and light when wielded. We take pride that our CFD products are reliable and engineer time efficient ‘out of the box’. Many loyal and long-standing blue-chip customers are drawn to this philosophy and our CFD tools that fit easily into their development processes.

At Software Cradle we offer a unique suite of CFD products that embrace everything from simulating Couette flows to hypersonic shocks, from applications spanning aerospace and automotive to shipbuilding and electronics applications, and everything in between. With a long pedigree of quality CFD tools, we have put in place a strategy to deliver multiphysics-focused CFD solutions to our community of users led by our flagship scFLOW and MSC CoSim software products, the most modern toolsets in the CFD marketplace. Better postprocessing and interrogation of results, design space optimization and virtual reality have all been embraced by Cradle CFD, and the MSC One token system allows for the democratization of CAE with easy co-simulation offerings of best-in-class products.

Backed by MSC Software’s rapidly growing multiphysics software suite, and Hexagon’s long-term manufacturing focus, we strive to solve complicated CFD simulations, even the ones the world thought impossible, through advanced CFD capability and integration with assorted CAE solutions. Hexagon and its Smart Factory concept allows for Cradle CFD to connect its Design ‘Digital Twins’ with Production ‘Digital Twins’ and Lifetime ‘Digital Twins’ to produce end-to-end virtual and real actionable data benefits to all customers in open and customer-centric toolchains.

Table of contentsA brief history of CFD ............................................................................................................................................... 4

Japanese industrial history and the Japanese philosophy of total quality product manufacturing .......... 6

Katana and the art of turbocharging CFD simulation ........................................................................................ 7

The five principles of Katana as applied to CFD .................................................................................................. 7

Katana: Accurate CFD ............................................................................................................................................. 8

Katana: Fast and Robust CFD ................................................................................................................................ 10

Katana: Engineer usability CFD .............................................................................................................................. 12

Katana: Multiphysics-focused CFD ....................................................................................................................... 14

Design & engineering CFD simulation for smarter manufacturing .................................................................. 18

Summary and conclusions ...................................................................................................................................... 22

References ................................................................................................................................................................ 23

2 | hexagon.com | mscsoftware.com

Imperial College CFD Group, London:Parabolic flow codes, vorticity-stream function codes, the SIMPLE algorithm, Upwind differencing,

‘Eddy break-up’ models, ‘Presumed pdf’ combustion models TEACH codeConception of classic k-e turbulence model (Brian Spalding & Brian Lander)

A Brief History of CFD and Software Cradle

1800

1850

1900

19101920

1930 1940 19501960

1965 19701975

19801985

19901995

2000 20052010

2015

2020

ClaudeNavier

(1785-1836)

George Stokes

(1819-1903)

Numerical formulation of equations for fluid flow

Ludwig Prandtl -Boundary layer modelsTheodore von Karman - Vortices Geoffrey Taylor -Turbulence microscales

“Numerical Forecast Factory” for Weather predictions using differential equations

Lewis Fry Richardson(1881-1953)

&

Andrey Kolmogorov - Turbulence length scales work

Brian Spalding(1923-2016)

Francis Harlow

(1928-2016)

Alexander Thom Earliest numerical calculation of ‘Flow Past Circular Cylinders at Low Speeds’, Proc. Royal Society, London,1933

M. Kawaguti in 1953 does 2d simulation in Japan of flow around a cylinder at Re=40

‘The Great Wave’ byKatsushika Hokusai 1830

Wright Brothers First Powered Flight 1903

Henry Ford releases Model T Car in 1908

Yoshida Omnibus released in 1905 in Hiroshima

Sony PlaystationFirst released in 1990

Los Alamos Group develops: Particle-In-Cell, Arbitrary Lagrangian-Eulerian methods, Vorticity-Streamfunctions

Suhas PatankarTextbook, 1980

Software Cradle Founded in 1984 byToshihiko Kano and scSTREAMCFD code released

Flow Science & Adapco founded 1980, Cham, CRADLE & Fluent in 1984

Computational Dynamics founded in 1987, Flomerics in 1988

SCYRU; firstbody-fitted CFD solver released in 1987

HeatDesigner all-in-one electronics thermal CFD package released in 1998

Edwards Deming

(1881-1953) Introduces Total

Quality Management

ideas to Japanese

ManufacturingApple iPhoneReleased in 2005

SC/Tetra for Microsoft Windows - First All-In-One CFD package) on Windows O/S in 1998

AeroacousticsscFLOW releases world’s first coupled capability with Actran, 2019

scFLOWRelease of unique next generation multiphysicsFocused CFD code for Co-simulation2018

MSC Software acquired by Hexagon AB, 2017

Software Cradle acquired by MSC Software, 2016

Toyota PriusFirst commercially successful Hybrid Car released in 1997

Commercial CFD Industry

Human Mannequin Thermoregulation –first 17 section model inside a CFD code, 2007

Laptop / Notebook ComputerFirst patent by Yukio Yokozawaof Seiko, 1980

Toshiba Sord SMP80/08First micro-computer with Intel 8008 chip inside, 1972

2016 20172018

2019

Images courtesy of Wikipedia

DER01 AndroidThe World’s first Android by Intelligent Robotics Lab, 2003

Bullet TrainsWorld’s first high speed passenger

trains built by Kawasaki Heavy

Industries in 1964

Numexa, Blue Ridge Numerics and Exa Corp Founded in 1991-1992

OpenFOAM CFD released in 2004

Osborne Reynolds

Develops fluid flow fundamentals

Probe Landing on an AsteroidJapanese Space Agency, JAXA, lands their probe Hayabusa 2 onto the surface of the Ryuguasteroid in 20198K Ultra HD Television

Sharp releases first 85" 8K LCD TV at CES in 2012

HeatPathViewerUnique market leading 1d thermal flow viewer inside HeatDesignerreleased, 2012

Industry 2.0Industry 3.0 Industry 4.0

Autono

my

Moon Landing 1969

Infographic of the History of CFD and Software Cradle

Manufacturing Intelligence hexagonmi.com | mscsoftware.com4

A brief history of CFD

Fluids are all around us in our atmosphere, in our oceans and on our landscapes; they are inside us, and they even extend out into outer space and other planets. They have fascinated humanity since the dawn of recorded history, and probably well before that as the magic of water flowing or ever-changing weather patterns have impacted the progress of humanity as a species on earth. The earliest civilization to study fluid effects was probably the Ancient Egyptians with their irrigation systems, ship hulls and sail designs. Certainly, by the time of the Ancient Greeks we hear of philosophers like the pre-Socratic Heraclitus in the 6th century BC - who was based in Ephesus - noting that “Everything Flows” without fully understanding what that meant in a modern scientific sense. The Roman Empire, in particular, went on to harness the power of water and fluids more generally and had a basic understanding of water engineering as is witnessed by their magnificent public baths, city sewer pipes and massive aqueducts stretching over hundreds of kilometers. And, of course, ancient Chinese and Indian engineers had also mastered many of these disciplines separately. It wasn’t really until the European Medieval period, however, after Islamic scholars had preserved and extended ancient manuscripts, that the ancient knowledge seeped back into Europe and Renaissance Italy where fluids and the study of fluid flow came back to the fore again. The work of the great artist and engineer, Leonardo da Vinci, a polymath genius who was fascinated by all sorts of topics included the study of flows of water, air and blood (1,2)

A truly modern mathematical understanding of the fundamental scientific equations of fluid flow, heat and mass transfer was not really outlined until the early 19th Century when the French mathematician, Claude Navier and, in parallel, but slightly later, the Irish researcher, George Stokes, formulated the fundamental underlying equations that describe pressure, flow and thermal effects in nature – see Figure 1. These complex, non-linear coupled differential equations that we now call the Navier-Stokes equations describe the flow of all fluids, but historically they proved to be very hard to solve numerically for well over 100 years. It was only with the development of modern digital computers over the last 50 years or so that we have been able to solve these equations numerically effectively. This advance has paved the way for the modern scientific discipline of what we call Computational Fluid Dynamics (CFD).

After the theoretical work of Claude Navier and George Stokes, the English engineer Osborne Reynolds added considerable experimental measurements to the body of science investigating fluid flow, and insightful deductions as to the nature of laminar and turbulent flows in the 1880s. Deeper understanding of turbulence theory by Theodore Von Karman and Geoffrey Taylor in the early 20th Century and subsequent research by Ludwig Prandtl in Germany helped to create further mathematical equations to describe these phenomena. The English mathematician, Lewis Fry Richardson postulated in the 1920s that weather could be forecast if atmospheric space could be broken up

into zonal areas, and equations solved by binary numerical calculations in a stadium full of people who either raised or did not raise flags in their hands when orchestrated by a “conductor” in the stadium; in effect he was describing a prototypical CPU. He spent nearly 6 months doing one weather forecast for Europe with a simple mechanical calculator machine and got the wrong answer (3). Nevertheless, he was on to something, and Alexander Thom in 1933 published the first calculation of flow past a circular cylinder. After the Second World War, Mitutosi Kawaguti in Japan completed calculations on a mesh and published the first ever 2D CFD simulation of laminar flow at a Reynold Number of 40 past a cylinder (1).

During the 1960s a research group at Los Alamos in New Mexico, USA, led by Francis Harlow started to put together some of the elements of CFD numerics that have since

https://www.hexagonmi.com

https://www.mscsoftware.com

2 | hexagon.com | mscsoftware.com

Imperial College CFD Group, London:Parabolic flow codes, vorticity-stream function codes, the SIMPLE algorithm, Upwind differencing,

‘Eddy break-up’ models, ‘Presumed pdf’ combustion models TEACH codeConception of classic k-e turbulence model (Brian Spalding & Brian Lander)

A Brief History of CFD and Software Cradle

1800

1850

1900

19101920

1930 1940 19501960

1965 19701975

19801985

19901995

2000 20052010

2015

2020

ClaudeNavier

(1785-1836)

George Stokes

(1819-1903)

Numerical formulation of equations for fluid flow

Ludwig Prandtl -Boundary layer modelsTheodore von Karman - Vortices Geoffrey Taylor -Turbulence microscales

“Numerical Forecast Factory” for Weather predictions using differential equations

Lewis Fry Richardson(1881-1953)

&

Andrey Kolmogorov - Turbulence length scales work

Brian Spalding(1923-2016)

Francis Harlow

(1928-2016)

Alexander Thom Earliest numerical calculation of ‘Flow Past Circular Cylinders at Low Speeds’, Proc. Royal Society, London,1933

M. Kawaguti in 1953 does 2d simulation in Japan of flow around a cylinder at Re=40

‘The Great Wave’ byKatsushika Hokusai 1830

Wright Brothers First Powered Flight 1903

Henry Ford releases Model T Car in 1908

Yoshida Omnibus released in 1905 in Hiroshima

Sony PlaystationFirst released in 1990

Los Alamos Group develops: Particle-In-Cell, Arbitrary Lagrangian-Eulerian methods, Vorticity-Streamfunctions

Suhas PatankarTextbook, 1980

Software Cradle Founded in 1984 byToshihiko Kano and scSTREAMCFD code released

Flow Science & Adapco founded 1980, Cham, CRADLE & Fluent in 1984

Computational Dynamics founded in 1987, Flomerics in 1988

SCYRU; firstbody-fitted CFD solver released in 1987

HeatDesigner all-in-one electronics thermal CFD package released in 1998

Edwards Deming

(1881-1953) Introduces Total

Quality Management

ideas to Japanese

ManufacturingApple iPhoneReleased in 2005

SC/Tetra for Microsoft Windows - First All-In-One CFD package) on Windows O/S in 1998

AeroacousticsscFLOW releases world’s first coupled capability with Actran, 2019

scFLOWRelease of unique next generation multiphysicsFocused CFD code for Co-simulation2018

MSC Software acquired by Hexagon AB, 2017

Software Cradle acquired by MSC Software, 2016

Toyota PriusFirst commercially successful Hybrid Car released in 1997

Commercial CFD Industry

Human Mannequin Thermoregulation –first 17 section model inside a CFD code, 2007

Laptop / Notebook ComputerFirst patent by Yukio Yokozawaof Seiko, 1980

Toshiba Sord SMP80/08First micro-computer with Intel 8008 chip inside, 1972

2016 20172018

2019

Images courtesy of Wikipedia

DER01 AndroidThe World’s first Android by Intelligent Robotics Lab, 2003

Bullet TrainsWorld’s first high speed passenger

trains built by Kawasaki Heavy

Industries in 1964

Numexa, Blue Ridge Numerics and Exa Corp Founded in 1991-1992

OpenFOAM CFD released in 2004

Osborne Reynolds

Develops fluid flow fundamentals

Probe Landing on an AsteroidJapanese Space Agency, JAXA, lands their probe Hayabusa 2 onto the surface of the Ryuguasteroid in 20198K Ultra HD Television

Sharp releases first 85" 8K LCD TV at CES in 2012

HeatPathViewerUnique market leading 1d thermal flow viewer inside HeatDesignerreleased, 2012

Industry 2.0Industry 3.0 Industry 4.0

Autono

my

Moon Landing 1969

Infographic of the History of CFD and Software Cradle

mscsoftware.com | hexagonmi.com Manufacturing Intelligence 5

become ubiquitous; for instance, particle-in-cell, arbitrary Lagrangian-Eulerian methods, and vorticity-stream functions. But then the real ‘Father’ of Commercial CFD came to the fore, Brian Spalding at Imperial College, London, England. In over a decade of pioneering research, he, his students and collaborators like Brian Launder, David Tatchell, David Gosman, Akshai Runchal and Suhas Pantankar defined the finite volume based CFD approach with innovations like parabolic flow codes, vorticity-stream functions, the SIMPLE algorithm, upwind differencing, ‘Eddy break-up’ models, ‘presumed pdf’ combustion models, the TEACH code, and various forms of the classic two-equation k-ɛ turbulence model that is still the workhorse of the commercial CFD industry nearly 50 years later. Indeed, the famous textbook written by Patankar, “Numerical Heat Transfer and Fluid Flow” that was published in 1980 (4) spawned an amazing number of CFD codes written in Fortran throughout the 1980s (5). You could write your own cartesian CFD code back then based on the equations within its text. And so it was that an engineer called Toshhiko Kano at a small consulting company called Software Cradle Co. Ltd in Osaka, Japan, created a finite volume CFD code that he called STREAM (now scSTREAM) in 1984 to provide a commercial CFD solution fashioned by Japanese engineers for the needs of the Japanese market.

Software Cradle had the distinction to be one of the first CFD vendors in the market along with companies like CHAM, Flow Science, Fluent, and Adapco in the early 1980s – see Figure 1. It was managed by Mr. Komada and Mr. Nakanishi for many years and its history mirrored that of the other commercial CFD companies with a body-fitted CFD code - as opposed to cartesian stair-stepped codes - appearing in the late 1980s called SCRYU (now SC/Tetra), and a specialist



Figure 2: Typical wide range of simulations across multiple industries today using Cradle CFD Products


electronics cooling CFD package called HeatDesigner in 1998. Software Cradle developed specialist knowledge for itself in the HVAC and built environment industry that was cemented by the release in 2007 of the world’s first Human Mannequin Thermoregulation Model inside a CFD code with 17 sectional parts to represent each human body part in order to capture comfort indices correctly. A patented HeatPathView capability inside the HeatDesigner release of 2012 reinforced Cradle’s reputation for easy-to-use and intuitive CFD software in the electronics cooling space and has been well received by the CFD industry and users ever since.

In 2016, Software Cradle was acquired by the world’s oldest Computer-Aided Engineering (CAE) company, MSC Software Inc. based in the United States and within 6 months MSC Software was in turn acquired by Hexagon AB, a world leader in sensors, hardware, software and autonomous solutions (Figure 1). This coincided with the release of scFLOW in 2016, one of the most modern, most advanced, very accurate, and yet easy to use commercial CFD codes in the world today, which is extremely focused on co-simulation and coupled multiphysics. The acquisition of Cradle by MSC Software allowed for excellent couplings with world class structural, acoustic and multibody dynamics codes. Indeed, in 2019 the world’s first embedded coupled aeroacoustics capability was released when the FEA (Finite Element Analysis) solver Actran (from FFT in Belgium) was located seamlessly underneath Cradle’s Finite Volume (FV) scFLOW interface thus jointly solving in one user experience for complex engineering grade aeroacoustics related CFD phenomena.

日本製MADE IN JAPAN

Japanese industrial history and the Japanese philosophy of total quality product manufacturing

Until the second half of the 19th Century, Japan kept itself separate from the rest of the world. However, once it opened up to global commerce, it was quick to industrialize in the early 20th Century (Figure 1) with the gasoline powered Yoshiba Omnibus being released a good three years before Henry Ford’s Model T Automobile Factory was set up and selling cars from Detroit in America. In addition to the auto industry, Japan ramped up its heavy manufacturing industries after the first World War, but on entering the second World War it suffered a devastating defeat and its industrial base was effectively crushed; manufacturing machines were broken, and engineers had to work with very few materials on production lines that were prone to shortages in the post-war economic depression. However, an American academic, Professor Edwards Deming came to Japan, and he encouraged Japanese manufacturing industry to adopt the wholesale embracing of his ‘Total Quality’ methods in their factories – something he could not get traction for in America. As a consequence, slowly over two decades Japan pulled its manufacturing base up to be one of the best in the world and its industries like automotive, manufacturing, shipbuilding, steelmaking, and




Figure 3: The benefits of front-loading CFD in any manufacturing process

consumer electronics became the envy of the world in the 1960s and 1970s. Japanese manufacturing has produced several world firsts over the years such as bullet trains, assorted electronics innovations, successful hybrid electric cars, cutting edge robots and HD televisions (Figure 1). This transformation of Japanese products over the last 75 years has been done with characteristically high Japanese quality standards attached to them so that Japan produces consumer products that the world loves to buy.

Japanese industry very much invented the term ‘Just in Time Manufacturing’ that in turn caused their factories to become world class in the 1960s. Indeed, the Japanese word kaizen, 改善 (かいぜん), represents total quality manufacturing – and the word itself is synonymous with quality control at Japanese car assembly lines. Kaizen comes from the Japanese words “kai-” which means “change” and “-zen” which means “good.” It literally means incremental performance improvement within existing manufacturing processes and it soon led to better products in Japanese Factories such that by the 1960s their productivity was twice that of equivalent Western European and North American factories, with minimal waste and high-quality products. The kaizen concept has been widely adopted not only in manufacturing but also in many service industries in Japan. This includes retail, transportation and even education. It is hard to ‘understand’ the cultural connotations of kaizen; in Japan, an entity or product without kaizen is like a bowl of rice without miso soup – one simply flourishes by enhancing the other (6). When one considers Japan’s auto manufacturers like Toyota, Mitsubishi and Nissan, and electronic giants such as Sony and Hitachi, it shows how Japan has successfully incorporated the concept of kaizen into the economy’s corporate sector. The popular meaning of kaizen from Toyota is “continuous improvement” or “small incremental improvements” of all areas of a company, not just manufacturing. Kaizen usually means all personnel are expected to stop their work when they encounter any abnormality and, along with their supervisor, suggest an improvement to resolve the abnormality.

The concept of ‘Just in Time Manufacturing’ in Japan was rechristened as ‘Lean Manufacturing’ in the 1980s by western academics as it came back into Europe & America and was widely adopted by industries. Manufacturing has since moved towards automation in the late 20th Century (so-called Industry 3.0) are we are now seeing a shift towards autonomy in the 21st Century (with the advent of Industry 4.0)

Katana and the art of turbocharging CFD simulation

Since its nascent use over five decades ago, Computational Fluid Dynamics has because of its very predictive nature been at the forefront of making more efficient and higher quality products and product development processes. We see this repeatedly across the world because CFD engages with research & development design processes so that better and more innovative products can be designed with fewer (and sometimes no) prototypes being

built during their manufacturing process. This means less material waste, low scrappage rates, lower energy usage and less recalls and warranty issues. In short, CFD helps to design products ‘right-first-time’ and ‘fit-for-purpose’ in each and every industry it is applied to today (Figure 2). CFD is usually ‘front-loaded’ in every product design and manufacturing process – and concurrent throughout that process – across the world. This means it can be used to design out undesirable performance features and to avoid negative environmental impacts at the early conceptual design stage when ROI (Return on Investment) is highest for engineering simulation tools (Figure 3). And CFD can conceptualize and virtually test as-yet-not-prototyped products and manufacturing processes. CFD is, therefore, a very cost effective and safe way to test ‘what if..’ scenarios without hazardous consequences to both people or the environment. CFD also allows for high-tech engineering products to be conceptualized and designed in PLM (Product Lifecycle Management) software driven by pure CFD and multiphysics-focused CFD simulations.

If one thinks of CFD as the ability to deliver the same aerodynamic predictions as you would get from wind tunnel testing, but you can do it inside your computer rather than in a huge wind tunnel test facility, then you can understand the power of CFD simulation software. Indeed, CFD can produce much more diagnostic data than wind tunnels ever can; as well as predict forces of drag, lift, pitch etc. at a much cheaper price point than usually expensive and time-constrained testing facilities. CFD today is as fast as the computer one is using, and with the rise of cloud computing, many CFD ‘experiments’ can be executed online giving ‘results’ that are much faster than ever before, and in many instances simulation predictions can be produced within usable engineering timeframes and increasingly in near real time.

The five principles of Katana as applied to CFD

There are many commercial CFD codes in the world today, but what really sets Software Cradle’s approach apart is that the development team decided to base its software philosophy for delivering quality CFD products on the Japanese concept of katana 刀 (かたな). Katana is the Japanese word for ‘sword’ and it literally refers




to a particular type of sword carried by pre-modern warriors called Samurai in medieval Japan - see Figure 4. These swords were typically light, strong, fast and very deadly. Software Cradle’s development philosophy is to want to engage with customers all round the world in a collaborative contract to deliver the software features, functions and enhancements that they want to make them more productive, accurate, fast, and application focused in their usage of CFD. This is viewed as a social contract with high levels of software service delivered in return for the customer’s favour (Figure 5). It is with this Japanese software philosophy ethos in mind that Software Cradle has developed CFD products for the last 36 years, and, in a CFD sense, it has four major distinguishing characteristics – Accurate, Fast & Robust, Engineer Usability, and Multiphysics-focused. The approach has been successful because of the level of renewals of the software each year that is in the high 90s percent and its fast growth rate now within the CFD market worldwide.

Katana: Accurate CFD

Accuracy matters in all computer-aided engineering (CAE), and never more so than in computational fluid dynamics

where it is as easy to get the wrong answer in a CFD prediction as it is to get the right one dependent on the mesh you use, the empirical models employed, and the quality of boundary conditions and material properties utilized. This is multiplied even more when you are solving an application that has strong multiphysics elements added to the mix. To counter the old maxim, ‘garbage in, garbage out’ it is essential to have a strong quality assurance process and a rigorous set of CFD benchmark tests that are integral to software development sprints to produce a trusted general purpose CFD code.

Japanese engineers are renowned the world over for obsession to detail and extreme levels of accuracy in their

products and solutions. That is the standard that Cradle CFD sets itself to achieve each software release. Today, all supported versions and development versions of Cradle CFD products are built on Windows and Linux platforms where nightly regression tests are executed automatically if there are any changes to the code. Cradle CFD products have to go through over 300 test cases as part of the QA process. Those test cases include basic validation cases

Figure 5 – Social contract between development and customer built on the concept of 奉公 “Service” and 御恩 “Favour”.

Figure 4: A Katana is a traditional Japanese sword: light, practical, strong and very sharp

Figure 6 – Examples of Motor Sport CFD customer applications of Cradle CFD software over the years.




(130), industrial grade benchmarks (over 40), example cases in our CFD User’s Guides (134), and bugfix confirmation cases. These cases help to ensure that Cradle CFD products are correct ‘out of the box’ and ready for all blind CFD applications. Indeed, we have many long-time motorsports customers like Honda and Yamaha (Figure 5) who demand extreme CFD accuracy within incredibly tight competitive motorsport design timescales.

Cradle CFD products also engage in a large number of CFD benchmarks from around the world including for JSAE (Figure 6), AIAA (Figure 7), JAXA (Figure 8) and JoRes (Figure 9). These help to maintain our high accuracy fluid dynamics predictions. A good example of Cradle CFD’s prowess was the 2013 Society of Automotive Engineers of Japan (JSAE) blind benchmark for commercial CFD software to demonstrate CFD accuracy against test wind tunnel validation data for a new ¼-scale ‘Ahmed’ car body shape (7). The external

aerodynamic test model consisted of the classical Ahmed vehicle body with and without an ‘additional part’ at the rear of the vehicle. All CFD simulation codes had to provide results for drag, lift, and pitching moment coefficients as well as pressure coefficient at various sections of the vehicle body. Seven organizations provided submissions to the JSAE blind benchmark spanning most of the main commercial CFD codes available in the market at that time. They all had three months to submit their simulation results and technical information on their CFD computation approaches, physical models, and resolution scales. Figure 6 shows that Cradle’s SC/Tetra code produced similar numerical results compared to experiment to that of the other traditional commercial vendors in the two cases studied. But since the calculations were conducted in transient mode, which usually results in very high CPU time for the calculation with a high number of computer cores, it was able to do it much faster than alternative codes by up

Figure 6: 2013 JSAE CFD Benchmark of External Aerodynamics around a modified Ahmed car body

Figure 7: 2016 AIAA CFD Benchmark of external aerodynamics around a modified Ahmed car body




to a factor of 7. This elapsed time to useful CFD results is an important factor in the practical usage of commercial CFD software in the world today.

In the aerospace sector, several classical American Institute of Aeronautics and Astronautics (AIAA), Japan Aerospace Exploration Agency (JAXA) and ONERA benchmarks have been carried out in Cradle CFD software ranging from low speed compressible flows to hypersonic bodies – see for instance Figures 7 and 8. Similarly, in the shipping industry, hull drag predictions and the effects of propeller-hull-water interactions are critical for CFD codes to simulate correctly. Figure 9 shows a Japanese Bulk Carrier (JBC) industrial benchmark that Cradle CFD’s software performed very well in (8).

Katana: Fast and Robust CFD

The ideal in any CFD simulation is to have “real time” results (1), but in reality, there is usually is a trade-off between accuracy and time-to-solution with the computational resources you have to hand. And the biggest time bottleneck inside CFD is usually geometry clean-up and mesh generation, followed by solver speed-to-solution; which can be long in elapsed time especially for large simulations or complex transient calculations, along with the time for final postprocessing and visualization of results. Solver speed-to-solution is

Figure 9: Hydrodynamic ship hull drag predictions using scFLOW for the JBC, Japan bulk carrier, benchmark

linked to HPC (High Performance Computing) and how the CFD code scales on multicore machines, networks, supercomputers and ultimately on massive data centres of computers in the Cloud. Speed to solution and robustness in solution go hand in hand within the CFD software world. Robustness revolves around several factors. First and foremost comes speed of pre-processing including CAD Cleanup. Secondly, the availability of different types of mesh elements for different CFD applications (hex-, tet- polyhedral, voxelated...). And third is solvers that don’t fall over even with confronted by the most complex physics and multiphysics problems to address. We have implemented a host of technologies inside Cradle CFD software - some unique to Software Cradle and some are patented - to allow for robust CFD simulation predictions whatever is thrown at our codes. Users value robust solvers they can trust to go to accurate results quickly.

At Software Cradle we have been working on solver speed for many years on all of our CFD codes. CFD Software speed-to-solution is related to how the codes are architected as well as the type and quality of mesh being solved on, and errors in both the geometry used and boundary conditions solved. Moreover, sometimes the real-world physics being solved by CFD can be inherently unstable or so complex it is hard to pick it up numerically. Software Cradle’s CFD products – scFLOW which employs polyhedral meshes, scSTREAM which uses Cartesian meshes, and SC/Tetra

Figure 8: Separation of an orbiter from a two-stage-to-orbit (TSTO) Launch vehicle booster flying at hypersonic speed in scFLOW using overset meshing



Figure 10: Cradle CFD meshing technologies including tet, hex, poly, voxel-based and overset meshing

Figure 11: Speed to a meshing and overall solution for a hypersonic missile benchmark example

Figure 12: Scaling of scFLOW on up to 512 cores


take. Our flagship modern polyhedral mesh CFD solver scFLOW that was released in 2018 has very impressive HPC scaling as shown by these benchmarks for up to 512 cores in a CFD simulation below – see Figure 12. You can see that for certain applications the code is super-linear in places where solver speed increases as more and more cores are added to the simulation. We aim for this level of scaling for all of our CFD product suite. Linear HPC scaling effectively

which uses tetrahedral meshes – have a long history of usage by large OEMs (Original Equipment Manufacturers) around the world including many big-name market leading Japanese companies, and come with strong technical pedigrees (see Figure 10). We have built robust, stable, reliable CFD solvers that are fast to solution for years. In some application benchmarks like that for liquid sloshing tanks we can be up to 500 times faster – from taking a week with competitor software to getting a solution in 20 minutes that an engineer to interpret. Figure 11 shows typical time-to-solution benefits for scFLOW polyhedral meshing approach, yet it has no compromise in solution accuracy for hypersonic flows over missiles.

If a CFD solution crashes after a week’s simulation that can be a disaster, and that is something we try to avoid. Indeed, we retain different meshing element type solvers as ‘horses for courses’ because cartesian meshes like those used by scSTREAM are still recognized to be some of the fastest CFD approaches in the world today - even after 30 years - to get to usable engineering CFD simulation results. Our tetrahedral meshers are some of the fastest in the CFD industry as witnessed by SC/Tetra’s ongoing popularity, especially with good boundary layer mesh Y+ capabilities. Our customers consider us to be excellent in meshing, automation and moving bodies as we can complete hard CFD applications faster and with fewer manhours spent than our competitors typically



Figure 13: Photorealistic postprocessing with scPost related to Cradle CFD and multiphysics simulations

Figure 14: Cradle CFD viewer for human thermofluid mannequins and augmented reality postprocessing capabilities and CFD viewing handheld device apps


means that our customers can get results quicker and quicker as their computational resources increase. We have also made our solvers available on the Rescale Cloud platform (www.rescale.com) where users can solve for CFD simulations on an infinite number of cores and port their CFD solver licenses online.

Katana: Engineer usability CFD

Customers usually say they typically spend 60% of their precious simulation engineers’ time resources into cleaning geometries and meshing prior to doing their CFD simulations. And, similarly, enormous amounts of time can be spent in postprocessing CFD predictions for company reports and in CFD presentations to senior management with advanced animations and visualizations. As reflected in our company name, we aim to deliver to our customers CFD technology that has been nurtured, cultivated and refined by user feedback. Our user-centric software development policies stretch back to our birth in 1984 and the legacy of our consulting roots that has been core to our DNA ever since. We aim to provide the best CFD simulation experience possible for our users and especially for non-expert users. We have done this with industry leading scripting automation and ‘drag-and-drop’ capabilities all within an open ecosystem so that

our customers can plug into their preferred toolchains of best-in-class software for their particular industry applications. Our CFD software APIs and plug-ins are well respected in the industry and highly valued by customers; especially for the multiphysics-focused CFD simulations we excel at. Openness in our CFD solutions and platforms is also important so that our customers can choose ‘best-in-class’ products to fit their product research, development and production simulation workflows. We let our customers choose best-in-class products, and we deliver plug-ins and connections to work with their preferred toolchains.



Figure 15: HeatPathView capability inside HeatDesigner for LED light electronics thermal paths


Classically, CFD simulations have always been dominated by the disciplines of pre- and post-processing, where pre-processing includes geometry cleanup and meshing, and postprocessing encompasses visualizing your CFD results either graphically or in contour or vector plots on your screen. Postprocessing of CFD results can also be quite time consuming as you have to set up viewing angles, surface lighting scenarios, data visualization points & data reporting lines, and fluid pathlines releases to be viewed etc. Indeed, increasingly, photorealistic postprocessing (Figure 13) and virtual and augmented reality visualization (Figure 14) has come into the PLM/CAE/CFD space, and it can also be time consuming to produce a picture or animation as an output from your CFD simulations especially for very large domain models.

Figure 16: JOS Thermoregulation human mannequin model inside Cradle CFD tools

We released our multiphysics-focuses CFD post-processor, scPost, a few years ago to deliver the sort of customer user experience required for the modern engineering simulation world. It has the visualization quality of open-source Blender software yet offers professional multiphysics-focused CFD postprocessing outputs for a non-expert visualization user.

At Software Cradle we pride ourselves in our ability to listen to our customers in order to produce software products that yield high-end analyst level accuracy yet designer level usability and user experiences. This manifests itself in our desire to simplify CFD software user interface buttons and workflows that mean an engineer gets to what they want to achieve intuitively and quickly. Indeed, we have released capabilities like HeatPathView for electronics

thermal applications (Figure 15) that has patented algorithms and visualization abilities embedded within it so that the user can quickly see where heat is dissipating through very complex PCB systems and assemblies to reveal heat dissipation paths which are not always intuitive to spot.

At Software Cradle we also strive to pursue true CFD usability with orders of magnitude meshing speedups and significantly faster solver speeds than our competitors while offering a rich set of CFD functionalities and multi-physics connections. Our laser focus is to be a trusted CFD solution provider/partner for our customers to make them successful with CFD. This is where our CFD domain expertise resides rather than in being another purveyor of an unwieldy PLM software suite. Some of our customers have done CFD solver




Figure 17: ODYSSEE LUNAR design of experiment of a diaphragm pump using scFLOW and MSC Marc

Figure 18: FSI simulation of a flag flapping in the wind in scFLOW with its unique DEM coupling capability

simulations with over half a billion meshing elements coupled with complex fluid flow, heat transfer and moving meshes rather than simple billion cell aerodynamics calculations, for instance. These are some of the most demanding and complicated CFD simulations in the history of commercial CFD.

We devised and released one of the commercial CFD industry’s most sophisticated thermoregulation human mannequin models to make it easy for CFD users in the transportation and built environment industries to assess the perception of heat and thermal effects in a simulation environment scenario where all exposed and hidden parts of the human body can be reported upon (Figure 16). Cradle CFD products are also geared up to work with all PIDO (Process Integration and Design Optimization)

software in use across the CAE space especially for simulation design space exploration and not just DoE (Design of Experiment). We have also produced a close coupling with the cutting-edge capabilities for AI/ML/ROM from the ODYSSEE suite of software and its LUNAR platform in particular (from www.CADLM.com) as witnessed by the results shown in Figure 17 for a fluid-structure interaction calculation of a moving membrane diaphragm pump coupling between MSC Marc and scFLOW from Cradle CFD.

Katana: Multiphysics-focused CFD

Over the last 20 years one of the ultimate ‘holy grails’ of computer-aided

engineering (CAE) has been ‘multiphysics’ simulations, i.e. co-simulations between different physics simulation types (1). Multiphysics, or multi-disciplinary physics simulation, even though it is ill-defined within CAE, intuitively feels right to engineers from the perspective of real-world engineering simulations being inherently interconnected. Fluids don’t usually exist in isolation from structural effects, or with acoustics, or with multibody dynamics, or with magnetics, or with electrics in the real world we all inhabit, especially for complex systems, yet for many years engineers have simulated these subdomains of physics as isolated point simulation solutions inside CAE. Realistically, there is ultimately a need for vertical co-simulation applications that are democratized, i.e. a single user interface for enough multiphysics to satisfy the user’s application needs



Functional Mock-Up Interface

EngineModels

ThermalSystems

Chemical Systems

MotionSystems


- be they an analyst or a designer – e.g. fluid-structure interaction between CFD and FEA, or aeroacoustics with CFD coupled to acoustic FEA rather than jumping between two separate software packages (9).

‘Multiphysics’ was coined as a term in the 1990s and in many ways is a consequence of a failure of the CAE industry to solve the fundamental underlying physics equations in a combined way. Partly, this is due to practical problems that we find because of different mathematical techniques. For instance, Finite Volume Methods (best for fluids), Finite Element Methods (best for structures, multibody and acoustics) meant that for more efficient solver convergence in real world engineering problems, one or the other is chosen as the best methodology. And, partly, software vendors have struggled to grapple with the multiphysics challenge to deliver usable engineering simulation tools because the mathematical approaches don’t tend to gel well together when combined. A classically difficult example for multiphysics simulation that has proved to be hard for commercial CAE vendors to do is a flag flapping transiently in a cross wind – see Figure 18. In a modern CFD code, like scFLOW this simulation of 8 seconds of transient real time took about 20 elapsed days to calculate accurately

Co-Simulation Multiphysics area Products Involved Industries Applications

Fluid-Structure Interaction (FSI) Nastran / Marc + scFLOW/scSTREAM

All Aeroflutter, Valve opening, MEMs, VIV, Suspension Loads, Thermo-Mechanical Stress…

Structural & Aeroacoustics Nastran / Marc +scFLOW + Actran

All Cabin Noise, Door Rattle, Noise & Vibration…

Multi-Body Dynamics & Fluids Adams + scFLOW All Large Particle Movement, Vehicle Side WindEvents, Vehicle Running Over a Puddle…

Virtual Drive & Vehicle Dynamics Adams + Vires VTD Automotive Autonomous Vehicles, ADAS Validation, Real Time Vehicle Driving Simulator…

Particulates & MBD & CFD Adams + EDEM, EDEM + scFLOW

Auto, Aero, Chem & Proc,

Car Stability on a Surface, Filtration, Bulk Material Handling…

MBD & Nonlinear FEA Adams + Marc Automotive Door Sag & Closing, Vehicle Extreme Load Cases (eg. hitting a kerb), Running Over an Obstacle, Battery Pack Deformation…

1d Systems & MBD & Controls Adams + Easy5 / MatlabSimulink / Maplesoft / GT Suite etc.

All Robot Arms, Machinery, Landing Gear system, Vehicle ABS, ESC, Traction Control…

using 144 CPU cores in parallel with scFLOW coupled to our inhouse Discrete Element Modeling (DEM) capability.

When Software Cradle was acquired by MSC Software in 2016 a development effort was kickstarted to couple fluids simulation into easy-to-use co-simulations with long established and well respected FEA codes like MSC Nastran, Actran and Marc. The multibody dynamics code Adams has been connected to scFLOW and combinations of all of these codes are connected in unique CAE toolchains and couplings. Everything from acoustics to multi body dynamics (MBD), and CFD to structural analysis, plus

API

User Defined Functions

Pre-Processing Meshing Solv ing Post-

Processing

ExamplesCAD; Linear Structural; Non-Linear Structural; Aero-Elastics; Detailed electro-magnetic; Accoustics; Quenching…

Co-Sim

Process Automation

Figure 19: Software Cradle’s open coupling CFD platform (left) and FMI interfaces (right)

Figure 20: MSC’s Cradle CFD and Actran Acoustics co-simulation of an automotive exhaust and muffler

Table 1: A cross section of co-simulation CAE applications connected to the MSC Software Solution Suite



+ +


Figure 22: MSC’s Adams – Marc- scFLOW multibody-fluids-structure prediction of a laden drum washing machine motion

Figure 23: MSC’s Adams - scFLOW predictions of a dynamic vehicle suspension movement through a puddle and a vehicle in a crossflow wind

Figure 21: MSC’s Nastran - scFLOW - Marc co-simulation of a flap bending and twisting in a crossflow

Figure 24: MSC’s Nastran - scFLOW prediction of propeller fluid-structure interaction with free surface waves and cavitation creation

Figure 25: Adams - scFLOW dynamic prediction of wing flap deployment aerodynamics (top) and a vibrating supersonic plate shock structure co-simulation in MSC Nastran – scFLOW CFD (bottom)

explicit crash dynamics can be connected together inside Hexagon | MSC Software. A multiplicity of multiphysics co-simulation applications can, therefore, be done today, both in two-product couplings, as well as in toolchains of three or four product combinations that were mere pipe dreams a few years ago (see Table 1).

Software Cradle offers remarkable open couplings for its scFLOW product via FMI, Functional Mock-up Interfaces, as shown in Figure 19. Cradle CFD’s technology is recognized by many inside the CFD market to be leading edge in extremely fast, accurate and robust solution approaches for free surface and overset meshing techniques. Physical quantities can easily be passed between third party and MSC’s own CAE software and Cradle’s CFD tools using the FMI that supports general physical quantity settings, user-defined functions, and various script languages (Figure 19). The release of scFLOW 2020 in 2019 allowed for unique aero-acoustic prediction couplings where MSC’s Actran, a recognized market leader in acoustics simulation with a twenty-year history, has been embedded under the hood of the scFLOW CFD code operating seamlessly to the users in

order to predict aero-mechanical load predictions for noise and vibration assessments (see Figure 20). Very complex and challenging dynamic co-simulations, as shown in Figure 21 are now possible where a non-linear bending and twisting metallic flap in a cross flow was simulated in an scFLOW - Marc - MSC Nastran toolchain.

With the Adams multibody dynamics solver in particular, a Cradle CFD co-simulation approach can simulate large particulate flows with free surfaces such as can be found in vane pumps, washing machine drum flows and vibration (Figure 22), and multibody dynamics fuel tank sloshing for instance. A really exciting multibody application is that of a vehicle driving through a large puddle with suspension effects taken into account via Adams or the same vehicle’s dynamics in a cross wind (Figure 23). Co-simulation with Marc means that Cradle CFD couplings can do aircraft fuel




MSCOne offers designers and engineers a smarter way to access a huge array of world-class CAE simulation software all from one flexible, subscription-based platform.

tank baffles, liquid quenching, flexible plates / membranes / valve seals, and Sirocco fans. When coupled to MSC Nastran, scFLOW can also do fluid-structure interaction of ships’ propellers and include cavitation effects for example which is a very hard CFD simulation to do (see Figure 24).

Software Cradle’s scFLOW when coupled with Adams multibody dynamics uniquely allows users to be able to simulate dynamic aircraft wing and flap deployment aerodynamic predictions.

Another exciting new aerodynamic co-simulation application is that between scFLOW and MSC Nastran / Marc for non-linear panel flutter of a supersonic plate involving a highly unusual coupled undulating vibration of the flat plate’s surface at high speeds resulting in moving shocks on the surface (see Figure 25). The CFD domain deformation prediction was done with sliding mesh capabilities and then the FEA was performed by MSC Nastran and Marc. The CFD elements used could be tetrahedrals, hexahedrals and polyhedral elements.Another challenging multiphysics area is thermo-mechanical stress predictions for electronic components on chips and PCBs (Printed Circuit Boards). In Figure 26 an electronic chip resistor has been simulated that undergoes repeated heat generation by its own on-off actions. This will eventually cause the breakage of its solder joints over the lifetime of the component and ultimately product failure. Being able to predict it in advance and to locate high stress areas is very important. Temperature distribution predictions from scSTREAM were mapped onto the mesh of a structural analysis solver (eg MSC Nastran or ANSYS Mechanical). The stress on the solder connection can then be accurately predicted. This is a very powerful way of predicting failure mechanisms and their likely locations in the consumer electronics industry and is relatively easy to do in scSTREAM as a co-simulation.

Finally, one of the advantages of being part of MSC Software has meant that Cradle’s CFD software is available

from within the unique MSC One CAE token system that allows for easy multiphysics-focused CFD simulation. This means that designers and engineers can flexibly access all the CAE software they might need with a low barrier to entry and without having to raise lots of purchase orders all from one flexible, subscription-based platform. This can accelerate innovation in your product development by taking advantage of co-simulation solutions to yield productivity and quality benefits.

MSC One is a clever way of democratizing multiphysics-focused CFD and is being used extensively across the world (9) as shown by the Panasonic domestic roof fan example in Figure 27.

Panasonic Ecology Systems have been working on ceiling fans to improve indoor air quality. Ceiling fans may be damaged or can crash due to blade vibration. Panasonic needs to perform a fluid-structure interaction simulation to validate safety. Cradle scFLOW and MSC Nastran were co-simulated to calculate the aerodynamic pressure on the blade surface and structural deformation. The MSC Co-Sim engine was used to manage the co-simulation task and enables fast data transfer between the two solvers. With

Figure 26: scSTREAM thermo-mechanical stress prediction of electronic component solder stresses

Figure 27: Fluid-structure interaction of a Panasonic Roof Fan using scFLOW and MSC Nastran




co-simulation predictions for a fan model in its real-life working environment, engineers were able to observe the complicated multiphysics behavior of the fan blades and achieve safety validated products as well as performance improvements in their designs.

Design & engineering CFD simulation for smarter manufacturing

Today, CFD impacts product design & engineering across all the main industries in the world from automotive through aerospace, electronics, chemical processing, energy, water, food, and even health sciences. It is well known that

CFD boosts efficiency, productivity and quality outcomes in all of these sectors because it deals with improvements in fluid flow, heat transfer and materials (mass) transfer. Figures 28-31 show examples of CFD - and multiphysics-focused CFD in particular - across four of these industries: automotive, aerospace, electronics and shipbuilding by way of example. Design and engineering predictive simulation software have very much been in the vanguard of new technologies to meet customer innovation, product requirements, and meet ever more stringent legislative compliance demands. In short, CFD simulation has resulted in leaner operations and reduced energy usage footprints by right-first-time conceptual design.

Figure 28: Multiphysics-focused CFD applications for the automotive & ground transportation industry

Figure 29: Multiphysics-focused CFD applications for the aerospace & defense industry



Figure 31: Multiphysics-focused CFD applications for the shipbuilding & marine industry


The automotive industry is going through unprecedented changes due to tighter emission regulations and two emerging massive dual disruptors of a switch to electrification and a rising demand for autonomous mobility. In order to achieve emission reduction targets the electrification of powertrains for all OEM (Original Equipment Manufacturer) vehicles is being required around the world. Indeed, the advent of electric powertrains inside cars will remove considerable complexity in manufacturing. Typically, a conventional internal combustion engine in a car has 1,400 separate parts yet an electric vehicle powertrain will only have 200, so in principle manufacturing will be easier. And, of course, the major disruption that autonomous vehicles will bring to the ground transportation industry in the next 5 to 10 years will also inevitably lead to sustainability benefits such as near zero emissions, better acoustics and longer vehicle lifetimes and more usage because most

self-driving cars will be electric. In terms of automotive, extensive exhaust and drive-by environmental noise CAE predictions and coupled fluid-acoustic simulations of exhaust silencers are being simulated around the world as well as particulate emissions modeling. Simulation can quickly and accurately simulate modern composites and additive manufacturing components dealing with NVH (Noise, Vibration and Hardness) simulations. With over 100 CPUs in every modern sedan car and 5 kms of cables, the modern car is a consumer electronics box on wheels and all of these changes will require more multiphysics-focused CFD in future.

Modern Airbus and Boeing aircraft are some of the most technologically advanced machines in the world and they have been extensively designed by CAE taking into account aeroelasticity and aero-acoustic improvements.

Figure 30: Multiphysics-focused CFD applications for the electronics industry




Their engines can generate thrusts that are a quarter of the thrust of the NASA Space Shuttle and their range now means they are able to travel from London to Perth Australia in one flight. In addition, optimal fan performance and other efficiency improvements conceptualized by CAE simulations mean that they will use 25% less fuel than last generation aircraft. They are therefore eco-friendlier than their predecessors. Moreover, engine acoustics are 40% lower than aircraft engines from the past. Boeing and Airbus aircraft airframes are not made out of aluminium exclusively these days; rather, exotic new titanium composite materials are also being employed that are stronger and lighter. Lightweighting requires even more multiphysics CAE simulations especially in the age of integrated computational materials engineering (ICME).

The Integrated Chip (IC) transistor and the Printed Circuit Board (PCB) are inside multifarious electronics applications and they have become pervasive in our modern world with computer processors and microchips embedded in almost every gadget and product we buy these days. The arrival of the internet in the early 1990s also caused an explosion of data and information being shared around the world on the back of massive data centers that have been built to

maintain the ecosystem around the world. Moore’s Law for CPUs has meant that computer chips have doubled in size and power every 18 months or so for the last 50 years. What was impossible to simulate 10 years ago is commonplace today. And whole industries like social media have emerged on the back of the internet over the last 15 years.It is a little-known fact that nearly 90% of all the goods in the world are transported across the planet on container ships, and CO2 emissions from international shipping make up ~2% of the world’s total greenhouse gas emissions today.

Work is going on in the world today to reduce the carbon footprint of all ships by 30% by making them more energy (fuel) efficient via better ship hull and propeller design, more bio friendly engines, and reduced underwater noise to preserve damage to marine mammals.

Digital transformation is a mega trend emerging across all manufacturing industries today and when coupled with the rise of both ‘Digital Twins’ that are virtual representations of actual products and assets, and ‘Digital Threads’ of data across the virtual and real worlds of manufacturing, it can often be found at the center of many modernization discussions (10).

The use of virtual predictive data and measured real data generated during manufacturing operations is, therefore, key to enabling unprecedented levels of manufacturing productivity and product quality in terms of delivering Smart Manufacturing. The move towards the so-called ‘Industry 4.0’ associated with digital transformations has compelled many manufacturers to start using ‘actionable’ information more effectively. This actionable data can be used intelligently to cut manufacturing lifecycle times, enhance products, and ultimately drive industrial sustainability. Modern CAE/CFD design and engineering simulation is, therefore, a big part of this transformation as part of the virtual Digital Twin within the Product Lifecycle Management (see Figure 32).

Figure 32: Hexagon Manufacturing Intelligence’s unique portfolio including CAE/CFD for innovation in automotive product lifecycle across the full digital twin and digital thread




‘Siloing’ is common within different departments of manufacturing companies especially between research & development, design, production and manufacturing

departments and they are a major productivity block in many organizations where CAE design and engineering, production software, and quality assurance of the final product are not communicating as effectively as they could or should be between each other. Data produced by these different functions can become siloed causing missed opportunities to drive greater process efficiency, improved products, and reduce resource wastage during manufacturing. Connecting data throughout a product’s lifecycle via the Industrial Internet of Things (IIoT) and cloud-based data handling will enable real-time insights into different departmental operations, thus driving greater collaboration and quick-response decision-making. The emergence of big data analytics also ensures that manufacturers are quickly identifying patterns in the relevant data and enabling them to capitalize on the right opportunities and quickly identify the most crucial areas for improvement. This connecting up of departments with virtual and real data as a ‘digital thread’ between them throughout a product’s full lifecycle is called the Smart Factory and it is happening more and more.

CAE design & engineering simulation (including Cradle CFD software) generates petabytes of data, and as CFD becomes ever more pervasive even more data will be generated. The single biggest technology challenge businesses and manufacturing faces today is putting all this data generated to work because data creation is outpacing our ability to use it. Think about the 25 billion devices connected to the internet today and the expectation that this number will more than double by 2025 – it means the gap between data creation and usage is just going to get bigger and bigger. This ‘Data Leverage Gap’ is what Hexagon excels at -

increasing data visibility in your manufacturing supply chain.Today, one of the biggest challenges that manufacturers face is to capture factory data and evaluate it during the testing process. Product developers want to keep record of the information of pre-testing, actual testing and post-testing processes easily and systematically and then create complex relational information between these data. To meet this need, an engineering lifecycle management (ELM) tool within a data and process management system that focuses on capturing and managing the data all the way from the test data to the consumption of this data by the CAE products and various end users is needed. ELM provides multiple methods to capture your data and is usually a web-based interface that allows you search-and-compare-validate data and apply multiple levels of security access to make sure users only see data that they are authorized to see.

Using MaterialCenter or SimManager from MSC Software as a process engine allows for the ability to automate virtual testing, do report generation, and many other activities straight out of a browser. Leveraging Hexagon’s unique approach to CAD/CAM and CAE Software Simulation & Testing solutions via precision metrology measurements today will be key to generating the autonomous Smart Factories of the next twenty years. Hence, better predictive simulation from CAE/CFD/CAM Design Twins produces virtual data, sensors and manufacturing machines produces real data in Manufactured Twins, and their mutual ability to communicate to one another with everything connected to the internet will allow for a better understanding of any end-to-end manufacturing process value when combined with the as-installed Lifetime Twin.

These uses of end-to-end manufacturing data can and will promote direct business benefits in terms of lower manufacturing costs, the ability to get products faster

Figure 33: CAE simulation data is becoming part of the Digital Twin of a modern Smart Factory




to market, and the ability to personalize products. Cradle CFD is, therefore, leveraging Hexagon’s unique approach to smarter manufacturing today to produce fast R&D and manufacturing. This will ultimately deliver better productivity, lower wastage levels, improved energy consumptions savings, and ultimately higher product quality in 21st Century manufacturing facilities.

Summary and conclusions

The Computational Fluid Dynamics (CFD) industry is reaching maturity after nearly five decades, but there is an increasing need for easy-to-use multiphysics-focused CFD tools. This is because the world we all inhabit naturally includes coupled multiphysics in many applications and being able to accurately and effectively simulate such phenomena from a CFD perspective is important across all industry sectors. Software Cradle, Hexagon and MSC Software have grappled with this co-simulation conundrum and come up with many usable multiphysics solutions and tool chains for real world engineering applications; all democratized by the MSC One licensing token system. Cradle CFD tools are designed to be practical for all users because they deliver simulation analyst accuracy yet have built-in designer level usability and user experiences with our many meshing enhancements, automation scripts, design-centric models, APIs and open ecosystem, and post-processing features.

The rapid industrialization of the 20th Century that Japan went through led to some amazing inventions, not least in consumer electronics, and improvements in manufacturing processes with global implications. Cradle CFD software has imbibed the Japanese total quality manufacturing revolution to produce unique and innovative CFD tools that are robust, fast, accurate, multiphysics-focused and with good user experience and usability embedded within them, that are capable of opening new frontiers for CFD.

We call the philosophy behind our approach to developing CFD software ‘Katana’ after the sharp and powerful Samurai swords of Japanese folklore. Multiphysics-focused CFD simulation may sound easy in principle, yet many commercial CAE vendors have had point physics simulation solutions for several decades but have failed to implement usable coupled solutions for industrial grade engineering applications - either loosely or closely coupled CAE. The classical ‘multiphysics’ challenges that CFD users typically face include fluids and structures (FSI, VIV), fluids and acoustics, fluids with structures and dynamics, and fluids with multibody dynamics. All of these have been addressed by Cradle CFD’s solutions.

Many manufacturing industries in the world today are experiencing a drive towards accelerating the usage of CFD to make it more accessible from conceptual design front-loading all the way through the development, manufacturing and deployed product’s lifecycle. In fact, CFD can be used all the way through these stages to a saleable product from the first design idea, through prototyping, tooling, manufacturing, utilization and the product’s ultimate retirement or its recycling. Advances in computer hardware and software have made this possible with CFD now being able to virtually design, “test” and optimize products before the first physical prototype is even constructed, thus saving companies time and money and unnecessary prototyping, testing and tooling costs as well as minimal waste which usually means no recalls, no scrappage and no rework. Hexagon uniquely offers the complete manufacturing ‘Digital Twin’ with its attendant ‘Digital Thread’ data backbone of actionable data from Design Twins to Production Twins and Operational Lifetime Twins. This approach will become ever more pervasive as CFD is democratized and multiphysics-focused CFD become ubiquitous in manufacturing.




References

1. “Back to the Future; Trends in Commercial CFD” by J. Parry & K. Hanna, NAFEMS World Congress, Boston, USA, 2011

2. “The Democratization of CFD”, Hanna R.K., Weinhold, I., Mentor Graphics White Paper, 2017:

3. “Lewis Fry Richardson: scientist, visionary and pacifist” by A. Vulpiani, Lett Mat Int, 2014, 2:121–128

4. “50 Years of CFD in Engineering Sciences: A Commemorative Volume in Memory of D. Brian Spalding, Akshai Runchal, Springer Nature, Book, ISBN 978-981-15-2670-1, 2020

5. “Numerical Heat Transfer and Fluid Flow”, Patankar, S., Taylor & Francis, ISBN 9781315275130 1980

6. “Understanding the Japanese Word, Kaizen” Website accessed July 2020.

7. “JSAE Benchmark of Automotive Aerodynamic Test Measurements”; Website accessed July 2020

8. “Tokyo 2015 Workshop on CFD in Ship Hydrodynamics”; Website accessed July 2020

9. “Co-simulation - Breaking the Back of Multiphysics CAE Simulation” by K. Hanna, Engineering Reality Magazine, MSC Software, Volume VIII - Winter 2018, 73-79

10. “Digital Twin: Values, Challenges and Enablers”, by A. Rasheed, O San & T. Kvamsdal, IEEE, DOI: 10.1109/ACCESS.2020.2970143, 2020

Acknowledgements

This white paper has emerged from various discussions over the years and has been very much inspired by our colleagues Romain Baudson and Diego d’Udekem (FFT, Belgium), Keith Perrin (MSC Software, UK) and Keita Fujiyama (Software Cradle, Japan). We are indebted to the founders and leaders of Software Cradle over the last 36 years for the long and illustrious history of the company and the distinctive pedigree and DNA woven into all its CFD products that we have inherited.



© 2020 Hexagon AB and/or its subsidiaries and affiliates. All rights reserved.

Hexagon is a global leader in sensor, software and autonomous solutions. We are putting data to work to boost efficiency, productivity, and quality across industrial, manufacturing, infrastructure, safety, and mobility applications.

Our technologies are shaping urban and production ecosystems to become increasingly connected and autonomous – ensuring a scalable, sustainable future.

Software Cradle, part of Hexagon’s Manufacturing Intelligence division, provides highly reliable, multiphysics-focused computational fluid dynamics (CFD), thermal dynamics software and integrated simulation tools that enhance customers’ product quality and creativity. Learn more at cradle-cfd.com. Hexagon’s Manufacturing Intelligence division provides solutions that utilise data from design and engineering, production and metrology to make manufacturing smarter.

Learn more about Hexagon (Nasdaq Stockholm: HEXA B) at hexagon.com and follow us @HexagonAB.

https://www.cradle-cfd.com/

https://hexagon.com

https://twitter.com/HexagonAB

https://www.cradle-cfd.com

Documents

Turbocharging CFD with Katana · 2020. 10. 12. · Turbocharging CFD with Katana Unleashing cutting-edge Computational Fluid Dynamics from Japan Makoto Shibahara, Jonas Wirgart, Irie