The Power of Physics-Based Software for Engineering and Scientific Research

This chapter is from the book

Creating and Using Virtual Prototyping Software: Principles and Practices

Learn More Buy

This chapter is from the book

This chapter is from the book 

Creating and Using Virtual Prototyping Software: Principles and Practices

Learn More Buy

1.7 Virtual Prototyping: A Successful Product Development and Scientific Research Paradigm

As mentioned earlier, virtual prototyping has a long track record of solid achievements. Virtual prototyping for the Manhattan Project was one of the major drivers of the original development of computers during World War II until the mid-1950s (Ford 2015 and Atomic 2014). Manhattan Project scientists developed 1-D virtual prototypes to predict the criticality of nuclear reactors and nuclear explosives and the implosion compression of fissile materials. Following the end of World War II, the development of electronic computers continued at a number of institutions, including the Institute of Advanced Studies at Princeton with John von Neumann and others, where the first “modern” computers were developed (Dyson 2012). The U.S. nuclear weapons programs at the Los Alamos, Livermore, and Sandia National Laboratories have continued the work begun during the Manhattan Project to pioneer the use of supercomputers to successfully design and sustain the U.S. nuclear stockpile. Nuclear tests are expensive and unpopular, so simulating and optimizing nuclear explosions with a computer offers tremendous advantages. Instead of conducting hundreds of real nuclear tests, only a few are needed to calibrate and confirm the computer predictions. These U.S. Department of Energy (DOE) laboratories continue to be among the leading users of high-performance computing hardware and software, especially since the U.S. stopped conducting nuclear tests in September 1992 (Energy 2019).

Motivated by the problems the DoD was having in designing and developing new military platforms, the Office of the Secretary of Defense launched a small experimental program in 2006 to develop and deploy a set of physics-based, high-performance computing software applications. The program goal was to determine whether virtual prototypes could help the department meet the challenges of designing and producing major military systems more rapidly and less expensively than in using physical prototypes. In addition to the main question about the viability of virtual prototypes, other concerns included these:

1. Could government and contractor teams develop the needed software?

2. Would government and industry groups be able to use the software to design and guide the production of successful complex military systems?

The Computational Research and Engineering Acquisition Tools and Environments (CREATE) program proposal was approved in late 2006 to address these questions. It was executed by the DoD High Performance Computing Modernization Program (HPCMP), beginning on October 1, 2007. The HPCMP rapidly organized and sponsored software application development teams to build and deploy 12 software applications for the design and development of naval ships, military aircraft, ground vehicles, and radio frequency antennas, including developing three-dimensional digital product models. The development teams were located at major DoD laboratories and warfare centers. During the following 12 years, the CREATE program was able to provide examples of the high value of virtual prototypes for the development of complex products (Post et al. 2016). The CREATE software tools are now being used by more than 2,000 acquisition engineers (40% government, 50% industry, and 10% other sectors), contributing to more than 180 different DoD programs.

Many other recent commercial examples of the value of virtual prototyping exist. The benefit to Goodyear Tire of adopting the virtual product development paradigm has already been discussed. The process of adoption at Goodyear was evolutionary. During its transition to use of the tire design software, Goodyear continued to build a physical prototype of the final design produced by the physics-based tire design software and continued to test the physical prototype tire before starting to manufacture it. Goodyear found that the virtual tire designs were so successful that it could start manufacturing immediately after the final design and virtual tests were completed. Goodyear then tested the first tires from the initial production run. Design flaws appeared so infrequently that the advantages of the virtual tire design process (faster and cheaper, with greater product innovation and fewer design flaws) outweighed the disadvantages of an occasional flawed design. Goodyear engineers used the lessons learned from those few flaws to improve the tire design software and reduce the rate of flaws in future designs (Council 2009).

Whirlpool is another example of a company that successfully adopted the virtual prototyping paradigm. Whirlpool is one of the world’s leading manufacturers and marketers of major home appliances (including the Whirlpool, Maytag, KitchenAid, Jenn-Air, Amana, and Brastemp brands, among others). Today’s home appliances are highly complex products. Their design and manufacture involve trade-offs of cost, safety, reliability, performance, maintainability, efficiency, and convenience. Safety is a major issue; for example, a washing machine must not tip over if a child sits on the door when it is open. The appliance market is international, and the products must be designed to fit into the homes in dozens of countries on almost every continent. A large American refrigerator or stove will not fit into a smaller European or Japanese apartment, for example. The appliance market is also highly competitive. Appliances can be assembled anywhere, with components from a global supply chain. Whirlpool models the appliances as systems of systems, including the packaging for shipment. A refrigerator that arrives at the appliance store scratched or dented during shipment has lost most of its sale value. Whirlpool models the fluid flow (water, refrigerant, and others), airflow, heat flow transport, mechanical strength of the appliance frame and components, electrical layout and switching, motor and pump performance, level of vibration, noise levels, computerized control systems, and mechanical balance. The company has substantially replaced much of its physical prototypes with virtual prototypes. “Testing using virtual prototypes has, for the most part, replaced physical testing—we’re no longer using the old ‘heat and beat’ approach,” says Tom Gielda (Gielda, 2009), Whirlpool’s engineering director for global mechanical structures and systems.

Automobile companies increasingly supplement crash tests with virtual tests, which are faster and less expensive. Virtual tests are easier to diagnose and analyze than real crash tests. According to a 2014 press release, Ford Motor Company increased its computing power by 50% to maximize the speed and number of virtual crash tests it can perform (Ford_Media 2014). The automaker performed more than two million crash test simulations from 2004 to 2014. By comparison, Ford performed its 20,000th full vehicle crash test at its Dearborn, Michigan, testing facility around 2014. The new computing power allows Ford to include up to two million finite elements in its virtual crash test simulations, a significant increase from half a million elements a few years before 2014. Ford’s virtual tests include front impact, side impact, rear impact, roof strength, and safety system checks.

Virtual prototypes with high-performance computers played a pivotal role in the development of Ford’s EcoBoost engine technology that was introduced in late 2010 (Kochhar 2010). Derrick Kuzak, group vice president of global produce development at Ford, stated, “EcoBoost is truly a smart solution for consumers because it provides both improved fuel economy and superior driving performance. The combination of turbocharging and direct injection allows smaller engines to perform like larger ones while still delivering the fuel economy of the smaller powerplant.” Nand Kochhar, the chief engineer at Ford for global materials and standards engineering, said, “A lot of HPC-based computational analysis is involved in simulating the trade-offs between performance, shift quality, and fuel economy. In the case of the engine, we conduct combustion analysis of turbocharging—optimizing a fuel–air mix, for example. To develop overall vehicle fuel efficiency, we use Computational Fluid Dynamics calculations to compute the optimal aerodynamics of the proposed vehicle” (Kochhar 2010).

Procter & Gamble (P&G) has used virtual prototyping in many parts of its product development process (Lange 2009). P&G uses computational tools to design and test its product packaging for plastic bottles intended to hold bleach and other liquids. It modeled the transport of potato chips (for example, Pringles) through baking ovens to maximize the throughput of the chips. The detailed properties of surfactants (the major ingredients in soaps, detergents, lotions, and shampoos) determine how well P&G products meet its customers’ needs. In addition, environmental concerns, both with production and with consumer use and product disposal, have become highly important. Initially, the company’s knowledge and research on surfactants were experimental, but the research demands began to overwhelm what was possible with that approach. P&G turned to modeling the behavior of the atoms and molecules in the surfactants using computational chemistry and molecular dynamics calculations. With this approach, P&G was able to identify ways to produce better products. Kelly Anderson, a senior scientist specializing in molecular dynamics at P&G, said, “Molecular dynamics allows us to make approximations about the interactions—the chemical potentials that are occurring. By investigating the molecular composition of these materials, we are better able to predict what properties a formulation will exhibit—not only its immediate characteristics, but what will happen to the mix 6 months from now. By mixing and matching different molecules containing different configurations of atoms, we can create the most desirable characteristics for our consumer products, such as detergents and shampoos, and at the same time, ensure they are safe and environmentally friendly. That’s really the magic of what we are trying to do” (Lange 2009).