Abstract

In order to develop mechanical products faster and reduce development costs, industries rely more on and more on simulations, following the Model-Based Systems Engineering (MBSE) strategy. But to build more confidence in more refined simulation models, these models have to be backed up by a lot more real-world data, in order to avoid large design margins. Simulation validation today is often a long and energy-consuming part of CAE engineers’ job. Whether or not they follow closely ASME’s guide on verification and validation for solid mechanics, teams have to confront larger simulation models to larger amounts of test data. The goal is clear: proving higher-level simulation models have credibility will allow further removing tests from the development process, which in turn can radically reduce development time and costs. Optical measurement techniques like Digital Image Correlation (DIC) are currently used in test labs to increase the amount of test data and are viewed as a possibility to feed higher-level models. They indeed create a massive amount of test data, which is actually a new kind of problem when the current tools and methods are not fit to process these very large datasets. The problem is that current Finite Element (FE) simulation validation processes involve a substantial amount of pointwise simulation-to-test comparison, using the likes of Excel sheets or Python/Matlab scripts, for lack of capable solutions. In the context of DIC, this leads to engineers being forced to process virtual gauges instead of actually comparing displacements and strain fields and taking advantage of the large dataset. Optical methods, when used in conjunction with traditional sensors, are a way to drastically increase confidence in simulations. But they alone will not bridge the gap between simulation and testing. The bigger challenge is to connect all test data to simulation and SDM (Simulation Data Management) tools, in order to provide full access to all data in one place for fast decision-making. This paper’s goal is to propose an analysis of the situation CAE engineers currently meet, as well as a path to integrating all test data in the design thread, in the light of recent advances of optical measurement techniques. Recent development projects will be used as examples of how simulation validation and model calibration can benefit from a more systematic and extensive test data integration to the Finite Element models.