Establishment of Comparability

comparing 2 models is often difficult or needs manual labeling, commodo does not need labeling since the comparison is done on the geometry by similarity features

One essense of Finite Element Analysis is the assessment and comparision of different design variants. Traditionally this comaprision is tedious, when the Finite Element meshes differ significantly.

Functions of the FEM Operations Toolbox allow the comparision of many different abstraction levels.

That way different design variants can be compared easily, even components or design concepts differ significantly or changes in labels, mesh density, position, orientation, or even part geometry hinder comparability.

Changing Models

Component Re-Identification

identify components in your simulation without manual labeling
reidentify components in other simulations
apply it on all shades of similarity (from perfectly equal to vaguely similar)
use this mapping for further comparative analysis (e.g. sensitivity analysis)

Courtesy from Audi

Result Mapping

Based on component re-identification and geometrical interpolation
Project your results onto another model
Compare results locally on a elementar or nodal level
Map many simulations onto a single reference geometry

Result Aggregation

Based on component re-identification
Derive simple feature signals from your results from matching components or regions
Enabler for advanced methods like machine learning on finite element models

results transformation may be done with constraints, e.g. sum over the part elements must be equal

Applied Methods

Automatic part and component identification
Results aggregation at different levels of abstraction
Mapping and geometric normalization

Achievements

Establish comparability of arbitrary design variants and Finite Element models
Improved knowledge discovery from Simulation Data Management