Author: Daniel Ackerman | MIT News Office<\/p>\n
Isaac Newton may have met his match.<\/p>\n<\/p>\n
For centuries, engineers have relied on physical laws \u2014 developed by Newton and others \u2014 to understand the stresses and strains on the materials they work with. But solving those equations can be a computational slog, especially for complex materials.<\/p>\n<\/p>\n
MIT researchers have developed a technique to quickly determine certain properties of a material, like stress and strain, based on an image of the material showing its internal structure. The approach could one day eliminate the need for arduous physics-based calculations, instead relying on computer vision and machine learning to generate estimates in real time.<\/p>\n<\/p>\n
The researchers say the advance could enable faster design prototyping and material inspections. \u201cIt’s a brand new approach,\u201d says Zhenze Yang, adding that the algorithm \u201ccompletes the whole process without any domain knowledge of physics.\u201d<\/p>\n<\/p>\n
The research appears today in the journal Science Advances<\/em>. Yang is the paper\u2019s lead author and a PhD student in the Department of Materials Science and Engineering. Co-authors include former MIT postdoc Chi-Hua Yu and Markus Buehler, the McAfee Professor of Engineering and the director of the Laboratory for Atomistic and Molecular Mechanics.<\/p>\n<\/p>\n Engineers spend lots of time solving equations. They help reveal a material\u2019s internal forces, like stress and strain, which can cause that material to deform or break. Such calculations might suggest how a proposed bridge would hold up amid heavy traffic loads or high winds. Unlike Sir Isaac, engineers today don\u2019t need pen and paper for the task. \u201cMany generations of mathematicians and engineers have written down these equations and then figured out how to solve them on computers,\u201d says Buehler. \u201cBut it\u2019s still a tough problem. It\u2019s very expensive \u2014 it can take days, weeks, or even months to run some simulations. So, we thought: Let\u2019s teach an AI to do this problem for you.\u201d<\/p>\n<\/p>\n The researchers turned to a machine learning technique called a Generative Adversarial Neural Network. They trained the network with thousands of paired images \u2014 one depicting a material\u2019s internal microstructure subject to mechanical forces, \u00a0and the other depicting that same material\u2019s color-coded stress and strain values. With these examples, the network uses principles of game theory to iteratively figure out the relationships between the geometry of a material and its resulting stresses.<\/p>\n<\/p>\n \u201cSo, from a picture, the computer is able to predict all those forces: the deformations, the stresses, and so forth,\u201d Buehler says. \u201cThat\u2019s really the breakthrough \u2014 in the conventional way, you would need to code the equations and ask the computer to solve partial differential equations. We just go picture to picture.\u201d<\/p>\n<\/p>\n That image-based approach is especially advantageous for complex, composite materials. Forces on a material may operate differently at the atomic scale than at the macroscopic scale. \u201cIf you look at an airplane, you might have glue, a metal, and a polymer in between. So, you have all these different faces and different scales that determine the solution,\u201d say Buehler. \u201cIf you go the hard way \u2014 the Newton way \u2014 you have to walk a huge detour to get to the answer.\u201d<\/p>\n<\/p>\n But the researcher\u2019s network is adept at dealing with multiple scales. It processes information through a series of \u201cconvolutions,\u201d which analyze the images at progressively larger scales. \u201cThat\u2019s why these neural networks are a great fit for describing material properties,\u201d says Buehler.<\/p>\n<\/p>\n The fully trained network performed well in tests, successfully rendering stress and strain values given a series of close-up images of the microstructure of various soft composite materials. The network was even able to capture \u201csingularities,\u201d like cracks developing in a material. In these instances, forces and fields change rapidly across tiny distances. \u201cAs a material scientist, you would want to know if the model can recreate those singularities,\u201d says Buehler. \u201cAnd the answer is yes.\u201d<\/p>\n<\/p>\n The advance could \u201csignificantly reduce the iterations needed to design products,\u201d according to Suvranu De, a mechanical engineer at Rensselaer Polytechnic Institute who was not involved in the research. \u201cThe end-to-end approach proposed in this paper will have a significant impact on a variety of engineering applications \u2014 from composites used in the automotive and aircraft industries to natural and engineered biomaterials. It will also have significant applications in the realm of pure scientific inquiry, as force plays a critical role in a surprisingly wide range of applications from micro\/nanoelectronics to the migration and differentiation of cells.\u201d<\/p>\n<\/p>\n In addition to saving engineers time and money, the new technique could give nonexperts access to state-of-the-art materials calculations. Architects or product designers, for example, could test the viability of their ideas before passing the project along to an engineering team. \u201cThey can just draw their proposal and find out,\u201d says Buehler. \u201cThat\u2019s a big deal.\u201d<\/p>\n<\/p>\n Once trained, the network runs almost instantaneously on consumer-grade computer processors. That could enable mechanics and inspectors to diagnose potential problems with machinery simply by taking a picture.<\/p>\n<\/p>\n In the new paper, the researchers worked primarily with composite materials that included both soft and brittle components in a variety of random geometrical arrangements. In future work, the team plans to use a wider range of material types. \u201cI really think this method is going to have a huge impact,\u201d says Buehler. \u201cEmpowering engineers with AI is really what we\u2019re trying to do here.\u201d<\/p>\n<\/p>\n Funding for this research was provided, in part, by the Army Research Office and the Office of Naval Research.<\/p>\n<\/div>\n Go to Source<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"