Automated Computational Mechanics Laboratory: News
ACML News
2022
Posted: January 08, 2022
Congratulations to Mohamad Mohamadsalehi for successfully defending his PhD thesis last month. Mohamad's thesis title is "Expansion of Conforming to Interface Structured Adaptive Mesh Refinement Algorithm to Higher Order Elements and Crack Propagation". He joined ACML in August 2017 and has authored/co-authored 2 journal articles (both first-authored). After earning his PhD, Mohamad is joining Transdigm as a senior software engineer. We at ACML wish Mohamad all the best in this new career path.
2021
Posted: November 11, 2021
ACML, together with a research group at North Carolina State University led by Prof. Jason Patrick, received a 3-year $680K project from the Department of Defense (DoD) on the manufacturing and computational design of self-healing 2D woven composites. The NCSU team will lead the manufacturing and experimental aspects of the project, while ACML will lead the failure analysis and computational design of the composite microstructure.
Posted: September 03, 2021
Prof. Soghrati received a one-year research award from OSU's Center for Clinical and Translational Science entitled "AI-enhanced approach to minimize the risk of vertebral compression fracture in patients with spinal metastasis undergoing stereotactic radiotherapy". This project, which will support a PhD student at ACML, is part of a larger, interdisciplinary research effort focused on cancer engineering in collaboration with multiple faculties from colleges of medicine and engineering.
Posted: June 9, 2021
Congratulations to Ming Yang for successfully defending his PhD thesis this summer. Ming's thesis title is "Advanced Algorithms for Virtual Reconstruction and Finite Element Modeling of Materials with Complex Microstructures". He joined ACML in September 2016 and has authored/co-authored 6 journal articles (4 first author) and presented 2 conference papers/posters during his PhD studies. Ming will join Cadence Cadence Design Systems as Lead Software Engineer and we wish him all the best in this new journey.
Posted: August 12, 2021
Prof. Soghrati gave a keynote talk entitled "An AI-enhanced computational framework for modeling materials with complex microstructures" at 2021 U.S. Congress on Computational Mechanics (USNCCM). This presentation summarized some of the recent research activities at ACML.
Posted: June 19, 2021
Prof. Soghrati was awarded a 3-year grant from Air Force Office of Scientific Research (AFOSR) to develop a set of novel techniques combing advanced computational algorithms and deep learning for simulating the failure response of materials with complex microstructures. The project will support two PhD students at ACML.
2020
Posted: December 1, 2020
Prof. Soghrati received funding from the Technical Data Analysis (TDA) company to study the failure response of multi-layered carbon fiber reinforced composites at the mesoscale. This one-year project will support a PhD student at ACML
Posted: October 30, 2020
Prof. Soghrati and Prof. Jason Patrick (North Carolina State University) received a $690K grant from Strategic Environmental Research and Development Program (SERDP), which is the DoD’s environmental science and technology program executed in partnership with DOE and EPA. This three-year research program entitled "Optimization of Self-healing Fiber Reinforced Polymer Composites via Convolutional Neural Networks", takes a collaborative, interdisciplinary approach by combining polymer mechanics/chemistry, emergent manufacturing, advanced computing and deep learning to accelerate the development of self-repairing structural composites. This synergistic experimental-computational project relies on: (i) a newly realized in situ self-healing platform in epoxy-matrix fiber-composites via thermal remending of an environmentally inert 3D printed thermoplastic interlayer; and (ii) novel microstructural material optimization using automated finite element simulations and deep learning via convolutional neural networks.
Posted: May 16, 2020
ACML PhD student, Ming Yang, was selected as one of this year’s recipients of the Ohio State Presidential Fellowship.
The fellowship is given to students who “embody the highest standards of scholarship” in the graduate programs at the university going into the last stages of their dissertation research or terminal degree project. Recipients are given a monthly stipend for living expenses so they can focus solely on completing their research, as well as help with travel expenses to present at national conferences.
Yang’s research (supported by AFOSR's Computational Mathematics Program) focuses on creating an efficient numerical framework, including microstructure reconstruction, mesh generation, finite element simulation, and deep learning algorithms, for the computational modeling of novel materials with complex microstructures. It will significantly reduce the time and labor cost associated with the modeling process. The framework has the capability to go beyond some of the limits in computational material design, and it can impact all engineering disciplines by facilitating the numerical investigation of novel material behaviors.
2019
Posted: October 15, 2019
ACML received a 3-year funding from Honda R&D to study fatigue failure response of structural adhesives. A PhD student (Mingshi Ji) will be supported through this project to develop realistic microstructural models and novel micromechanical fatigue models to simulate the high-fatigue failure response of these heterogeneous structural adhesives. ACML will also closely collaborate with Honda R&D researchers to acquire experimental data needed for the calibration and validation of these fatigue simulations.
Posted: September 13, 2019
Congratulations to Hossein Ahmadian for successfully defending his PhD thesis earlier this summer. Hossein's thesis title is "Integrated Multi-Scale Modeling Framework for Simulating Failure Response of Fiber Reinforced Composites". He joined ACML in January 2014 and has authored/co-authored 5 journal articles (3 first author) and presented 2 conference papers/posters during his PhD studies.
After earning his PhD, Hossein is still one of the effective members of ACML as a postdoctoral research scholar. In this new effort, in collaboration with OSU’s Spine Research Institute, Hossein is modeling and analyzing microstructural damage mechanisms in the vertebrae of cancer patients.
Posted: January 12, 2019
ACML received a new two-year research grant from Ford on a project entitled "Automated finite element modeling of fiber reinforced composites with high carbon fibers volume fractions". In this project, we develop a novel computational framework for modleing chopped fiber composites, including microstructure reconstruction and mesh generation.
2018
Posted: November 30, 2018
ACML received funding to start a new project in collaboration with Prof. William Marras from OSU's Spine Research Institute (SRI) to develop a new predictive model for the assessment of mechanical behavior of vertebrae in cancer patients. Cancer patients with vertebral body metastasis treated with radiation therapy are at risk of vertebral fractures. There is a general lack of knowledge on the systemic effects of radiation in these patients that often receive Immuno-Oncology (IO) therapy as part of their standard of care and what histology or patient is most susceptible to fracture risk in IO therapy patients. Assessing who may need prophylactic treatment using vertebroplasty to minimize these risks will help in clinical decision making for both local control and quality of life.
The goal of this project is to review the effects of stereotactic radiotherapy in the setting of immune therapy and develop a computerized predictive algorithm to predict vertebral strength and fracture risk for cancer patient’s treatment with radiation therapy. ACML will further develop and implement a suit of existing algorithms for the automated microstructure reconstruction and mesh generation in this project to analyze the impact of variations in the bone microstructure on its fracture response.
Posted: November 15, 2018
Congratulations to ACML PhD candidate, Anand Nagarajan, for receiving the presidential fellowship award from OSU's Graduate School. The Presidential Fellowship is the most prestigious award given by the Graduate School. Recipients of this award embody the highest standards of scholarship in the full range of Ohio State's graduate programs. The Presidential Fellowship gives fellows one year of full-time financial support so they can complete their dissertations or terminal degree projects unimpeded by other duties.
Posted: November 14, 2018
Congratulations to Ehsan Taghipour for successfully defending his PhD thesis on 11/14/2018. Ehsan's thesis title is "Development of Reduced-Order Computational Models for Digital Manufacturing of Flexible Wire Harnesses". This challenging project was sponsored by Honda. Ehsan joined ACML in August 2014 and has authored/co-authored 3 journal articles (2 first author) during his PhD studies.
We congratulate Ehsan for his excellent academic performance and have best wishes for him in his future career.
Posted: November 14, 2018
Congratulations to Bowen Liang for successfully defending his PhD thesis on 11/08/2018. Bowen's thesis title is "Integrated Multi-Scale Modeling Framework for Simulating Failure Response of Materials with Complex Microstructures". He joined ACML in August 2013 and has authored/co-authored 13 journal articles (5 first author) and presented 3 conference papers/posters during his PhD studies.
After leaving ACML, Bowen will join ANSYS as a software engineer. We congratulate him for his excellent academic performance and have best wishes for him in his future career.
Posted: August 16, 2018
Prof. Soghrati gave a keynote presentation entitled "Integrated Computational Framework for Simulating the Failure Response of Materials with Complex Microstructures" in the 13th World Congress on Computational Mechanics in New York City. Four PhD student in ACML (Bowen Liang, Anand Nagarajan, Ming Yang, and Hossein Ahmadian) were co-authors on this paper. The abstract is provided below:
We present an integrated computational framework relying on a new microstructure reconstruction algorithm and a non-iterative mesh generation technique, named Conforming to Interface Structured Adaptive Mesh Refinement (CISAMR), for creating high fidelity FE models of composite materials. A NURBS-based reconstruction algorithm is implemented to synthesize the material microstructure by packing arbitrary shaped particles, morphologies of which are extracted from digital data such as micro-computed tomography images. A genetic algorithm (GA) based optimization framework is also employed to simulate the target statistical microstructural descriptors such as the size distribution, volume fraction, and spatial arrangement of particles. CISAMR is then employed to create a FE model of the material by transforming a structured mesh into a high quality conforming mesh with low element aspect ratios and a negligible discretization error. This non-iterative transformation is carried out by combining customized versions of four algorithms: h-adaptivity, r-adaptivity, face-swap, and sub-tetrahedralization. Compared to enriched methods such as extended FEM, CISAMR obviates the additional computational burden associated with evaluating enrichment functions and provides a higher accuracy for recovering the gradient field. Further, unlike conventional mesh generation algorithms such as the Delaunay triangulation and octree-based methods, CISAMR can easily handle problems with highly complex geometries without the use of iterative smoothing or relaxation algorithms to improve the elements quality. In this work, we show the application of this integrated reconstruction-meshing framework for simulating the failure response a variety of heterogeneous materials, including particulate and fiber-reinforced composites, as well as non-woven entangled materials such as fiberglass insulation packs. Multiple sources of material and geometrical nonlinearity, including the damage, contact, and cohesive debonding are considered in each simulation.
Posted: June 4, 2018
Mr. Bowen Liang received the best paper award in 2018 EMI conference (Boston, MA) in the Modeling Inelasticity & Multiscale Behavior Student Competition. The title of his talk and the abstract are below:
Title: Image-based multi-scale modeling of effects of microstructural features on the failure response of adhesive-bonded joints
Authors: Bowen Liang, Anand Nagarajan, and Soheil Soghrati
Abstract: Adhesive-bonded joints have been widely used in the automotive and aerospace industry as a key component for the assembly and manufacturing of lightweight structures. In such applications, adhesive bonding has provided several advantages over other conventional joining techniques such as welding and riveting, including reduced stress concentrations, high corrosion resistance, and the ability to join thin and dissimilar materials. Therefore, a reliable 3D multi-scale computational framework to predict the failure response of adhesive joints subject to different loading conditions is highly demanded and will be presented in this work. The modeling process begins with processing high-resolution micro-computed tomography (micro-CT) images of a heterogeneous adhesive to determine the realistic shapes, volume fraction, size distribution, and spatial arrangement of embedded particles. A new reconstruction algorithm is then employed to synthesize the adhesive’s microstructure, which combines the random sequential adsorption (RSA) and genetic algorithm (GA) to replicate statistical microstructural descriptors extracted from the imaging data. A non-iterative mesh generation algorithm named Conforming to Interface Structured Adaptive Mesh Refinement (CISAMR) is then utilized to transform the virtual microstructures into high-quality finite element (FE) models. Finally, appropriate constitutive models are employed to simulate two major damage mechanisms in the adhesive layer: a surface-based cohesive model for the delamination phenomenon along particle-matrix interfaces and a ductile damage model for the elastoplastic behavior and damage response of the epoxy matrix. After determining the appropriate size of the representative volume element (RVE), this multi-scale cohesive model will be used to study the effects of microstructural features such as heterogeneities, pre-existing voids, particle-matrix adhesion strength, and the adherends surface roughness on the failure response of the joint. These computationally intensive simulations rely on a parallel computing scheme to partition the microstructure , where each processor discretizes the background mesh of its own domain, generates the conforming FE model, and solves the resulting equations at the micro-scale.
Posted: March 31, 2018
Congratulations to Weijie Mai, the first PhD candidate in ACML who successfully defended his PhD thesis on 03/30/2018. Weijie's thesis is entitled "Numerical Modeling and Simulation of Electrochemical Phenomena", with focus on developing meshfree and phase field methods for simulating localized corrosion phenomena. He has also studied high-fidelity finite element modeling of the electro-chemo-mechanical performance of cathode electrodes used in lithium ion battery in the last year of his PhD. The outcome of his great research efforts have been 6 published journal articles (4 first authored) and one under-preparation article in high-impact journals of numerical modeling and electrochemistry, including the International Journal of Numerical Methods, Journal of Computational Physics, Corrosion Science, and Electrochimica Acta. Weijie has also given 4 presentations in international conferences such as ASME IMECE and NACE throughout his PhD.
After leaving ACML, Weijie will join the National Renewable Energy Laboratory as a postdoctoral researcher. We congratulate him for his excellent academic performance and have best wishes for him in his future career.
Posted: March 14, 2018
The College of Engineering will recognize the research contributions and productivity over the last five years of faculty and research scientists by presenting the Lumley Research Award at the annual College of Engineering Faculty Awards Banquet. Funds for these awards are made possible by an endowment established by John H. and Mildred C. Lumley. Awardees will be selected on the basis of their research accomplishments and productivity made during the five-year period prior to the nomination, at The Ohio State University.