Multiobjective CAE Model Homologation

This presentation was made at the NAFEMS Americas Seminar - Confidence in Engineering Simulation: The Next 10 Years of CAE in Mexico.

What is the future for engineering analysis and simulation in Mexico? Discover innovative engineering simulation processes and tools which are helping companies in Mexico improve production capabilities. Engage with domain experts, industry leaders, and peers in a focused, comprehensive one-day event that covers topics on engineering analysis, simulation, and systems modeling and simulation that every engineer in Mexico should know.

Resource Abstract

Over the past years it has being a common practice and sometimes almost a workflow rule to create a dedicated CAE model for a specific type of analysis, hence it involves the construction of several Finite Element Analysis setups, each one with a different finite element model, therefore, implying over simplifications, many assumptions, under or over constraining, missing interactions, missing contacts between parts, not representing exactly the physical test set up, etc., all the before mentioned circumstances lead to very hard to correlate models, that in most of the cases the simulations performs with a low or a medium/low confidence level, as well as being a highly time consuming process since it implies to obtain each separate analysis result from each type of dedicated model.

This situation remains due to the complexity required to couple the mathematical theories by which the Finite Element Method is build, in which many cases are incompatible.

It is a need in the professional environment to reduce costs, deliverables time and to correctly represent the physical behaviour of the system to achieve good correlations. The present paper introduces a modelling construction technic that allows to greatly improve time and performance of the simulations due to the creation of one complete homologated finite element control model. The difficulties in the homologation processes were passed, by carefully coupling the Explicit and Implicit finite element mathematical solving methodologies, allowing the interaction between every type of entity (shell elements, solid elements, beams, springs, masses, spotwelds, weldlines, etc), the elements formulations interaction, considering all the contact definitions between them, and the precise selection of the material laws including user defined material laws, that can describe the physical behaviour in both type of solving methods.

This modelling technique as before mentioned was applied in one single control model without any simplification nor assumption (every vehicle component including attachments, rubbers, joints, bolts, spotwelds, fixtures, adhesives, glues, etc.) from which eleven different types of analysis were executed at the same time, among them: modal analysis, low speed damageability, step pad, customer loads, isolated thermomechanical, RCAR and thermomechanical plus pressure loading and gravity sag; all simulations conducing to satisfying results reaching medium to high confidence correlation levels, it also allows to perform multitarget or multiobjective analysis to comply between different requirements at the same time.

Some metrics within the results are the following: ‘Step pad’ simulation reached in most cases above 85% correlation confidence level, in some cases the measured displacements differ by tenths of millimetre. ‘Firmness and feel’ simulation increases to medium high and high confidence levels, meaning simulations can predict physical results in a range from 70% to 95% of confidence. ‘Customer load appliance’ simulation reached a medium to high 83% of confidence. ‘Low speed damageability’ in frontal impact reached a high confidence level with 95% of certainty. Homologation in one control model allowed decreasing the payed software licenses from 3 to 1 which significantly reduces error generation and error inherit, concurrently reducing engineering time since it is no need to transform models between software element definitions and fix the loose of contacts, weldings, materials, attachments, etc. Initially 4 engineers were required to perform three different types of analysis, currently only 1 engineer is able to perform eleven different types of analysis running in explicit, implicit or thermomechanical methods according to each type of analysis with the before mentioned homologated control model, and the study definition (model set up) was reduced from taking three days down to four hours.

Document Details

AuthorSinger. R
Date 8th November 2018
OrganisationFord Motor Company


Back to Search Results