The meeting is free of charge but Registration is mandatory (Participation subject to approval)
8:30 – 9:30 : Registration and Coffee
9:30 – 9:40 Welcome Remarks. Shmulik KEIDAR, CEO, ADCOM
9:40 – 10:30 SIMULIA Brand Update and Highlights, MONTI Indro, EUROMED SIMULIA Industry Process Consult – Mgt Director, SIMULIA Industry Consultants, Dassault Systèmes.
10:30 – 11:00 Model Based Multiphysics Approach for Airborne Radome Design, Davide TALLINI, EMAG CoE Euromed Senior Technical Manager, Dassault Systèmes.
11:00 – 11:40 – Break & Networking Time
The session will present the major technological enhancements introduced in SIMULIA Abaqus, TOSCA and XFLOW version 2022.
Specifically, an overview of new features will be presented on: Solvers, Performance, Contact, Element Technology, nonlinear mechanics and more …
Biomedical Simulations of Healthy and Pathological Aortic Valves in an Electro-Mechanical Full Heart Model: Fluid-Structure Interaction using Lattice Boltzmann Method
Adi Morany (1), Karin Lavon (1), Ricardo Gomez Bardon (2), Danny Bluestein (3), Ashraf Hamdan (4), Rami Haj-Ali (1)
(1) School of Mechanical Engineering, Tel Aviv University, Tel Aviv, Israel
(2) Dassault Systemes España, SIMULIA XFlow, Madrid, Spain
(3) Department of Biomedical Engineering, Stony Brook University, Stony Brook, NY, USA
(4) Department of Cardiology, Rabin Medical Center, Petach Tikva, Israel
The aortic valve (AV) is located between the left ventricle and the aorta and responsible for maintaining an outward unidirectional flow. Many AV hemodynamic and structural aspects of tha AV have been extensively studied, however, more sophisticated models are needed to better understand the AV biomechanical behavior.
In the first part, this study deals with integrating a new parametric AV structural model with the electro-mechanical Living Heart Human Model® (LHHM). The LHHM is a finite element model simulating human heart capable of realistic electro-mechanical simulations. Different geometric metrics of AV have been examined and clinically validated. The proposed model within the LHHM better predict local stresses during the cardiac cycle due to the realistic boundary condition.
In the second part, this study examines a new hemodynamic-structural co-modeling approach using Lattice-Boltzmann (LBM) and Finite-Element (FE) methods. The proposed new fluid-structure interaction (FSI) approach is used to investigate healthy and pathological AV. Towards that goal, FE-LBM FSI models of tri and bi cuspid AVs under physiological pressure are studied. Different parameters have been examined, such as effective orifice area, hemodynamic metrics and stress magnitudes. We found that LBM-FE FSI models can produce good predictions for the flow and structural behaviors of AVs.
The final goal is to introduce a new FSI LBM-FE model of integrated aortic valve within a full heart model examining the effect of electro-mechanical boundary conditions on AVs structural and hemodynamic responses.
Prediction of material properties and geometry in deep drawing process, by using Machine learning methodology.
By: Avner Shmuel (Rafael), Prof. Shmuel Osovski (Technion)
To demonstrate confidence in the use of ML tools for predicting plastic properties, work will be presented that predicts the geometric dimensions of a product created in a deep drawing process based on the anisotropic material properties and vice versa, the material properties prediction based on the final product geometry.
Two predictors were presented in each direction. The first predictor is GPR , the second predictor is the Extra Trees Regressor from the Ensemble family. Both predictors gave very good results, on the order of 90% or more.
The demonstration showed that with a small number of samples (81) it is possible to obtain good predictions for complex plastic processes
The work was done by constructing finite element (FE) simulations, which simulate different loading modes for different material parameters. The results from the simulations serve as a mechanism for producing synthetic information for the Machine Learning [ML] model that attempts to train a deep drawing process, based on the relationship between the load, the anisotropic material properties, and the final geometry.B
13:20 – 14:20 – Lunch
