Software Designed for Genetics Shows How Materials Behave Under Extreme Loading

Civil engineers by tradition are concerned with the big picture, but some are refocusing their vision, zooming in to solve minute problems we can't see with the naked eye, like tiny fractures in polymers, silicon or the molecular structure of proteins.

This work involves understanding the mechanics of a material--its ability to withstand pulling, twisting and heavy loads--at the atomic level. But the classroom technology for teaching this in a short timeframe doesn't exist--until now, that is.

An educational experiment during IAP demonstrated that students can learn to apply sophisticated atomistic modeling techniques to traditional materials research in just a few classes, an advance that could dramatically change the way civil engineers learn to model the mechanical properties of materials and provide enormous benefit to industry.

"Taking an atomistic approach to the study of materials' design and analysis offers opportunities for making significant improvements in materials' strength, reliability and sustainability," said Markus Buehler, an assistant professor in the Department of Civil and Environmental Engineering who collaborated with Ivica Ceraj, a software developer in MIT's Office of Educational Information Technology (OEIT), to prepare the new simulation techniques.

"While scientists often rely on quantum mechanics in their study of materials, engineers tend to use a more traditional continuum approach that relies on empirical parameters to model processes such as a crack forming, without considering the mechanisms at the atomic scale that give rise to these phenomena," said Buehler.

However, a fracture in a concrete bridge doesn't begin as a long, jagged scar; it starts off as a vibration at the atomic scale and progresses.

Engineering students usually study typical weight-bearing problems during their first years in college, but they aren't taught how a material's response to forces at larger scales relates to its structure and mechanisms at the atomic level. The problem Buehler faced was finding a way to teach students to model the material's atomic response without getting too caught up in the complexities of a computer program.

This is where Ceraj and the OEIT initiative came in. "We are looking to reduce the operational fog and help students focus on the subject they are learning without stumbling over new software tools," said Ceraj.

Buehler and Ceraj employed a web interface called GenePattern, an award-winning software program developed in 2004 by a team at the Broad Institute of MIT and Harvard to help scientists perform gene expression analysis. Ceraj created an interface between GenePattern and the software code Buehler uses in his own research. The interface, a derivative application that Ceraj calls StarGP, provides a simple-to-use, yet very accurate tool for modeling the behavior of materials under extreme loading.

"We are expanding StarGP's use in different fields: civil and environmental engineering, materials science and biology," said Ceraj. "Each discipline has different challenges, but it provides us with the opportunity to bring the latest research tools to undergraduate and graduate students." Ceraj collaborates with Jill Mesirov and Michael Reich at the Broad Institute in his work with StarGP. Mesirov and Reich are part of the original GenePattern development team.

In Course 1.978 (From Nano to Macro: Introduction to Atomistic Modeling Techniques), Ceraj and Buehler introduced students to the new atomistic simulation program with great success. They found that when using the web interface method, students learned the basics of atomistic modeling quickly, then applied the technique to predict the mechanical properties of silicon, copper nanowire and a structural protein called vimentin that plays a crucial role in stabilizing eukaryotic cells under deformation.

Previously, such simulations required students to learn technical details of a Linux workstation before they could get to the heart of the numerical method. With the web interface, students need enter only a few pieces of key information about how and where a material will be pulled, pushed or twisted, and the program will prepare an accurate video simulation. For instance, one video shows a fracture in one side of a small piece of silicon zig-zagging until it cleaves the material. Another demonstrates a vimentin protein being pulled at both ends until it unwinds from a tight tangle into a long string.

As a result of the class, students not only learned the atomistic simulation quickly, some have already adopted it for their own applications.

Michelle Hyers, who is working toward a Ph.D. in mechanical engineering, took the class to find a more accurate way of modeling at smaller scales. "My research involves modeling self-assembly at the micro and nanoscales, but we currently use macroscale theory with various assumptions to describe the system," said Hyers. "As a result of this class, I formed a collaboration with Professor Buehler to work on a more accurate model for our system than our current approximate methods."

"It is the best class I've had so far at MIT in terms of engaging content, as well as excellent teaching," said Hyers.

Buehler plans to use this method next spring to teach a section of Course 1.021J (Introduction to Modeling and Simulation), an undergraduate class that provides an overview of simulation techniques. This subject will expose undergraduates to state-of-the-art atomistic modeling methods to teach the next generation of engineers how to make a big impact by thinking small.

http://www.mit.edu

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.