Welcome to MMLDE-CSET 2023 in El Paso, Texas.

ANNOUNCEMENT:  
Plenary Speaker Dr. Chris Hill, Ansys will speak on TUESDAY from 7:30 - 8:15am.
Plenary Speaker Dr. Joel Quintana, UTEP, will speak on WEDNESDAY from 7:30 - 8:15am

You will find here information on Plenary Speakers and the Technical Program as well as other information.  We encourage you to check back frequently as there may be changes in the sessions due to cancellations, etc.  

Tuesday, September 26, 7:00am – 10:00am
Tomas Rivera Conference Center in the Union Building, 500 W. University Ave.

Conference Banquet
Tuesday, September 26, 5:30 - 8:00pm

Technical Sessions will take place at the University of Texas El Paso Union Building. 
Tomas Rivera Conference Center in the Union Building, 500 W. University Ave.

2nd IACM Mechanistic Machine Learning and Digital Engineering for Computational Science Engineering and Technology 

International Association of Computational Mechanics

On the Epistemology, History and Future of Data Driven Digital Twins in Materials Science

An overview of the development, impact and future potential of data-driven digital twins (D3T) will be presented from the perspective and context of material science. First the delineation between a model and D3T will be introduced in the context of epistemology and systems theory. Then, an existential and ontological framework of D3Ts will be presented in terms of the fundamental questions and needs that make their utility practical. Several historical examples of D3Ts along with custom-developed multi-degree of freedom robotic testing systems for generating data that were developed since the 1960s at the U.S.-Naval Research Laboratory involving composite materials will be presented. Finally, an overview of the drivers that push the extension of D3Ts beyond their present state, along with potential examples of future D3Ts in the context of material and structural systems will close the presentation.



Plenary Speaker - Dr. John G. Michopoulos, U.S. Naval Research Laboratory

Machine Learning Applications in Industrial CFD

Three separate areas in which machine-learning is employed in industrial CFD will be presented. The first of these is fusion modeling. A low fidelity model for a system can be combined with some high-fidelity data to arrive at a fusion model. Such models can provide high-quality results at a fraction of the computational cost of a full-fidelity model. An example of using this approach in industrial CFD will be shown. Secondly, the use of machine-learning in the tuning of two-equation turbulence models will be discussed. An adjoint flow solver is used to support the training of a modification to the GEKO turbulence model that is available in Ansys Fluent. This modification helps to match high-fidelity scale-resolving computations better. Some results will be shown, and challenges discussed. Finally, an approach to using machine-learning to solve CFD problems is presented. This approach uses autoencoders and a local training approach to provide a generalizable strategy for identifying high-quality flow solutions efficiently. Some results from such a solver will be shown.

Dr. Chris Hill has a B.S.c in Natural Philosophy and Mathematics from the University of Glasgow, Scotland. He holds a Ph.D. from the University of Cambridge Engineering Department, Cambridge, England, in acoustics and fluid mechanics. His research interests include the interaction between hydrodynamics and acoustics, boundary layer transition and receptivity, flow control, adjoint methods, optimization, lattice Boltzmann methods, finite volume methods, and machine learning for CFD. Dr. Hill is the original author of the Ansys Fluent adjoint solver. He has spent time at NASA Ames Research Center as a Post Doc. and Stanford Center for Turbulence Research as a Senior Fellow. For the last 25 years he has worked at Fluent and Ansys where he is currently a Technical Fellow and Chief Technologist for Fluids.

Plenary Speaker - Dr. Chris Hill, Ansys, Inc.

A Digital Twin Framework for Materials-Aware Manufacturing Process Design

Concurrent materials, geometry and manufacturing process optimization involves many computational challenges such as high-dimensionality associated with location dependency, material heterogeneity, multi-modal information, and nonlinear material behaviors such as large deformations and plasticity. The recent growth of using physics-based machine learning creates opportunities for incorporating data-driven methodologies with physical models into design. Furthermore, digital twin is an emerging technology in the era of Industry 4.0 that holds promises for real time optimization of manufacturing processes and quality control. We will present in this talk a digital twin framework that integrates materials modeling, part-scale manufacturing simulation and mechanics analysis, as well as state-of-the-art data science techniques to achieve off-line surrogate modeling and model uncertainty quantification and on-line predictive model control. To support materials-aware product and process design, techniques of microstructure characterization and reconstruction, machine learning of materials laws, differentiable simulation based optimization, and deep-learning based computer vision metrology will be introduced.



Plenary Speaker - Dr. Wei Chen, Northwestern University