ICME Helps Designers Living in a Materials World

The promise and challenges of a systems-level approach to lightweighting.

Lightweighting relies on a combination of both design and manufacturing innovation, as well as the use of new, lighter materials. Current approaches, however, are often limited by focusing on parts of a design, rather than the whole.

A team of researchers led by ORNL’s Amit Shyam is using high-performance computing to speed the development of new high-temperature aluminum alloys for automotive cylinder heads. Image courtesy of ORNL. A team of researchers led by ORNL’s Amit Shyam is using high-performance computing to speed the development of new high-temperature aluminum alloys for automotive cylinder heads. Image courtesy of ORNL.

Lightweighting can be achieved by switching a component to a lighter material that can meet structural requirements, changing the structural shape of a component, or altering the configuration of materials to improve efficiency. Still, these efforts are often pursued on a component-by-component basis. The design, materials and production methods used to produce each component are not often considered in an integrated fashion.

A systems-level approach to lightweighting that takes all of these factors into consideration, and relies on collaboration among materials engineers, designers and other stakeholders can result in new design innovations.

“In a systems-level approach, it is possible to predict the specific in-service conditions of each component in the overall system being modeled,” says Dr. Will Marsden, director of industry relations at Granta Design. “If you understand those specific conditions, you can understand the interplay of the system as a whole, and then identify the key areas (e.g., materials selection), which dictate the performance of the overall system rather than designing components in isolation.”

Approaching Integration

One such systems approach is Integrated Computational Materials Engineering (ICME), which takes into account design, material and processing methods to optimize the use of materials for a given component. ICME requires collaboration across various specialties to succeed.

“It’s not just the tools that have to talk to each other, the people have to talk to each other, too,” says George Spanos, technician director at The Minerals, Metals & Materials Society (TMS). “That’s really key for practical implementation because it involves engineers that are used to working on the product development cycle, as opposed to the fundamental science framework.”

TMS has produced a lengthy report (ICME: Implementing ICME in the Aerospace, Automotive, and Maritime Industries) that includes detailed steps for implementing ICME in those verticals.

ICME involves the integration of personnel, models, computational tools, experiments, tests, analyses, design and manufacturing processes across the entire product development cycle. It allows designers to explore a much larger design space more quickly than using traditional experimental approaches.

“When you integrate this optimization scheme, you approach design without the restraint of the exact materials being fixed up front,” Spanos says. “The designer can have more latitude. It doesn’t tie your hands to a limited type of alloy. Topology optimization, which is done in the mechanical world, can help free up the parameters of the materials a bit.”

For example, engineers could vary the properties of the chosen material across a component to maximize the utilization of “local” properties to meet the performance requirements of the system, Marsden says.

“This can unlock the full potential of the selected material/process combination,” he says. “This assumes precise control of the processing parameters needed to create the required structures within the materials (grain structure for metallics or fiber distributions for composites) that exhibit the desired properties from which the desired system performance is derived.”

It is an iterative process. “You want to identify the basic geometric envelope of a part, and to optimize the geometry assuming homogenous materials properties,” Marsden says. “You then identify the areas in which the stresses are greatest and see if the introduction of locally-focused material properties would be advantageous.”

The processing of the materials in each location across a part can be modeled using the appropriate techniques (e.g., calculation of phase diagrams or finite element analysis) so that the relevant heat/deformation history (in the case of metals) can be derived. The local processing history can then be used to derive the local properties to be predicted. The models can be tested in order to find an approach that produces the desired results.

“Another angle is that there may be multiple steps involved in manufacturing a part,” Marsden says. “If you can model each step, and then combine the results of those different modeling approaches sensibly, you will be closer to modeling reality.”

By using computational models at different length scales in the ICME environment, companies can model ways to improve material development and design optimization, model new assembly processes and predict finished product performance. It can also reduce the need for prototype production and testing, while accelerating time to production and improving weight reduction.

“It cuts down on the matrix of experimental testing that you need to do,” Spanos says. “You don’t eliminate experiments, because you have to validate and certify the materials, but you don’t have to go into the granularity of doing hundreds of tests.”

The Materials Data Challenge

An integrated approach to lightweighting requires a significant increase in the availability of materials information, as well as a new family of material models that can predict the properties of the materials based on a whole range of relevant processing parameters.

It also requires information to be transferred efficiently between different methods, and for the links between related items of information to be captured and available for interrogation.

“For example, if you can understand the link between the processing parameter, the structures within the material they create, and the resulting property displayed by that structure, then the performance of the part can be predicted—and optimized,” Marsden says. He adds that Granta provides a materials information “backbone” for the ICME process—a single system in which all of the material and process data from simulation and experimentation can be captured, together with its inter-relationships and technologies to get the information into and out of simulation codes, analysis packages and test programs.

However, there is currently a lack of standards for constructing and maintaining database structures so that materials information and tools can be easily accessed and exchanged. In addition, intellectual property concerns may impede these sharing efforts, as well as cost concerns about the effort required to share the data in the first place. In most cases, there is also imperfect knowledge of the precise conditions at each location of a part within every step in each process.

There are efforts underway to improve knowledge sharing. The White House Materials Genome Initiative was launched to help accelerate the development and deployment of new materials, for example. TMS is currently conducting a study on storing and sharing materials data in a way that companies can pass the information back and forth reliably and securely. “In this systems framework, it’s really important to determine how we share data, and that goes across the product development cycle,” Spanos says.

A promising approach is using a federation of databases in different communities that can speak to each other. “The government is going to play a big role in how to share that data,” Spanos says. “Once industry gets involved, there are proprietary considerations, of course. The idea is you define that precompetitive place, and then you do as much as you can within those definitions.”

New Levels of Collaboration Required

The systems-level approach requires collaboration among designers, materials specialists, mechanical engineers and manufacturing to create an integrated product development team.

“It involves design engineers interacting more closely with other engineers. There have to be champions within management to break down certain stovepipes and change the culture internally,” Spanos says.

This requires internal education and training, as well as the need for ICME-experienced staff.

Verification of models and simulations will also be critical, along with managing and mitigating uncertainty quantification and risk in the modeling results.

“There is a lot of great modeling and simulation, but not enough experimental validation,” Spanos says. “It’s important to validate models and simulations, and that’s a key activity that needs to be promoted.”

There are additional challenges. In some vertical markets, there is a lack of good quantitative modeling tools, as well as data on structure-property relationships for materials. It can also be difficult to effectively integrate different codes and models.

Marrying advanced materials modeling and simulation to ICME principles could help simulate the performance of a completed part (like an aircraft or automotive component) without an extended build and test period.

Tailoring New Materials

The use of ICME could also help speed the development and use of new materials. Engineers often don’t understand the properties or failure tendencies of new materials. Because it’s more difficult (and often more expensive) to test and understand them, designers often fall back on traditional materials.

In the aerospace industry, this has been a particular problem as qualification time and costs have made it difficult to qualify new materials for use in aircraft. In addition, the materials have to be understood in the context of different types of production methods to predict their performance. ICME can help accomplish that.

ICME can also generate information could speed development of new materials, a process that can take decades.

DARPA’s Materials Development for Platforms (MDP) program, for example, uses ICME to help compress the development cycle by as much as 75%. The program focuses on rapid materials development with specific capabilities and intended missions. Researchers will start with an application, and then work their way back to creating an appropriate material.

Oak Ridge National Laboratory, Nemak of Mexico and automaker Fiat/Chrysler are also using ICME to help accelerate the development of new materials to help achieve new fuel efficiency targets. The project is part of the DOE’s Vehicle Technologies Office initiative that Ford, GM and Fiat/Chrysler are spearheading to develop a high-strength cast aluminum alloy that can be used to produce lighter powertrain components. The project is initially targeting the development of aluminum cylinder heads. The ICME approach (in combination with DOE’s Titan supercomputer) is allowing the researchers to customize new alloys at the atomic level to help reach the desired material properties. They can model new alloys, then use the results to narrow their choices for additional experimentation.

As companies continue to struggle to drive weight out of their products, a systems-level approach can help speed those efforts while providing insight into new materials and production methods that can be leveraged in additional applications.

More Info

Share This Article

Subscribe to our FREE magazine, FREE email newsletters or both!

Join over 90,000 engineering professionals who get fresh engineering news as soon as it is published.

About the Author

Brian Albright's avatar
Brian Albright

Brian Albright is the editorial director of Digital Engineering. Contact him at [email protected].

Follow DE