38th   International CAE Conference

2022, November 16-18

NAFEMS symposium

NAFEMS

NAFEMS, Best Practises in Engineering Simulation

Part 1 16 November | 12.00 - 13.00
Part 2 17 November | 16.00 - 17.00

Online symposium

Attendance: free, upon registration

Language: English

Speakers: Miguel Herraez, Anthony Massobrio, Alessandro De Rosis

Review the recorded symposium: part 1!

Review the recorded symposium: part 2!

NAFEMS is the International Association for members of the Engineering Modelling, Analysis and Simulation Community

Nafems’s principal aims are to:

  • Improve the professional status of all persons engaged in the use of engineering simulation
  • Establish best practice in engineering simulation
  • Provide a focal point for the dissemination and exchange of information and knowledge relating to engineering simulation
  • Promote collaboration and communication
  • Act as an advocate for the deployment of simulation
  • Continuously improve the education and training in the use of simulation techniques
  • Be recognised as a valued independent authority that operates with neutrality and integrity

Nafems focuses on the practical application of numerical engineering simulation techniques such as the Finite Element Method for Structural Analysis, Computational Fluid Dynamics, and Multibody Simulation. In addition to end users from all industry sectors, Nafems’s stakeholders include technology providers, researchers and academics.

Agenda part 1

16 November | 12.00 - 13.00

Python: The Key to Generate Complex FE Models
Speaker: Miguel Herraez, TecnoDigital School


Agenda part 2

17 November | 16.00 - 17.00

Evaluation of Designs with AI , the Role of Geometric Deep Learning in Automotive Simulation
Speaker: Anthony Massobrio, Intelligent Simulation Ltd

Multiphysics modelling by Lattice Boltzmann Method
Speaker: Alessandro De Rosis, The University of Manchester

Miguel Herraez
Miguel Herraez

BIO: Miguel Herraez got his BSc in Mechanical Engineering in 2011 after completing an internship in WVU (West Virginia, US).
During his Phd, Miguel collaborated as visiting researcher at NASA Langley (Virginia, US). Upon completion of his Phd on Computational Micromechanics of Composites, he worked as a Postdoc at EPFL (Switzerland) in the development of high-fidelity models of laminated composites.
In 2020 he founded the TecnoDigital School which provides technical training in the field of numerical simulation using finite elements combined with automation and programming capabilities.
His contribution in the field of numerical analysis of composite materials is translated into more than 20 peer-reviewed publications. His doctoral thesis was awarded with the Polytechnic University of Madrid best Phd thesis, the AEMAC best Phd award and the SEMNI award. He is also the developer of the software Viper for the generation of artificial 2D microstructures and FE model generation.

TITLE: Python: The Key to Generate Complex FE Models
Abstract: The development of numerical models for industrial or scientific applications often requires the generation and handling of complex geometries, meshes, etc. Creating these complex numerical models by hand is extremely time consuming in some cases, and not even feasible in many others. One of the most adopted approaches to automate the design and setup of numerical models is the use of programming languages, such as Python. By means of scripts we can automate all the operations required to define our numerical model very fast and efficiently saving lots of time. These automation capabilities are not only limited to the pre-process stage of the model but are also applicable to the postprocessing of the results and the data reduction step. In fact, the possibilities are endless.


Anthony Massobrio
Anthony Massobrio

BIO: For over 30 years present in the world of CAE starting from the Fiat Research Center up to assignments at CD-adapco / Siemens, graduated in physics and specialized in CFD, in 2020 he approached the world of Artificial Intelligence and those developed by the Federal Polytechnic of Lausanne.

TITLE: Evaluation of Designs with AI – the Role of Geometric Deep Learning in Automotive Simulation
Abstract: In the last few years, there has been a growing interest in the role of AI (Artificial Intelligence) in Simulation, especially in its declinations of ML (Machine Learning) and Deep Learning as enablers of data-driven predictions.
The automotive industry is an ideal field of application of Deep Learning because of the maturity of CAx and the richness of data from CAD, CAE and Testing (in terms of Data Science, we have well-consolidated “datasets” of “ground truth”).
There are, however, different approaches depending on the final application.
By circumscribing the application to evaluate designs of different geometrical and arbitrary configurations, it will be possible to highlight in this presentation the role of Geometric Deep Learning in Automotive Simulation with specific use cases from Crashworthiness, Internal and External Flow, and Heat Exchange.
In conclusion: when well understood in its impact and its limitations, we will show how technology can be considered as ready for production.


Alessandro De Rosis
Alessandro De Rosis

BIO: A passionate scientist working in the field of multiphysics modelling by lattice Boltzmann method. Enthusiastic author of more than 40 papers in scientific international peer-reviewed journals. Charmed by challenging physics problems.

TITLE: Multiphysics modelling by Lattice Boltzmann Method


Ansys
BLOM Maritime
Chronos – HPC library
Deeplabs
E4
EnginSoft
ESTECO
FEMFAT
Flownex
HSH
Magma
MathWorks
NAFEMS
Parametric Design
Particleworks Europe
PASS
RBF Morph
RecurDyn
SATE
SDC Verifier
Sigmetrix