Framework to plug certification methods to assess margin of safety from the model and result information.

To view the Certification ribbon, click View > Ribbons > Certification.

Restriction: Certification is enabled only in OptiStruct and Nastran user profiles.

The Certification tools are used to asses methods on geometric regions (designpoints) representing a Structural Element. Methods are fed by attributes queried from FE model + FE results as well as external information not held by FE models.

The quick workflow is as follow:
  1. Create some Designpointsets
  2. Create and organize Designpoints in Designpoinsets
    • Assign Structuralproperties to Designpoints
  3. Add Methods to evaluate (from registered methods)
  4. Select loadcases to loop through
  5. Run
  6. Contour or export Table to report

Designpointsets, Designpoints, Structuralproperties and (Designpoint) Methods are HyperMesh entities persistent in the binary file.

In the 2024 release, available configs are:
  • Beam
  • Panel_composite
  • Panel_metallic
  • Rivet
  • Spring
  • Generic
All certification entities are managed in a dedicated Certification Browser, which is invoked by clicking the Certification tool.
Figure 1.

General Concepts

The aim of the Certification tools is to instrument a “region” of an FE-model and evaluate an external math function to assess margin of safety while considering “true” geometry and properties of that region.

In this matter, the so-called region is assigned a “virtual property”, which is a placeholder to provide all required details to better represents physics and geometry.

Such regions are grouped together in sets where functions are assessed under multiple load cases.

Critical margins of safety can then be found to down select critical load cases.

When such an evaluation is done, the aim is to be able to update virtual properties of failed areas to re-assess certification methods.

Certification relies on four new HyperMesh entities:
  1. Designpointset
  2. Designpoint
  3. Structuralproperty
  4. Designpointmethod
Designpointset is the root entity created. It currently offers different configurations, such as:
  • Beam
  • Rivet
  • Spring
  • Panel Metallic
  • Panel Composite
  • Generic
Designpointset is a collection of designpoints of the same config. Functions to evaluate and load cases are defined at this level.

Designpoint is the “region” which is instrumented; designpoint is the FE description of engineering concept denoted as “Structural Element”, that is the geometric zone of interest with details not modeled in the FE world. To fill such a gap, another entity is introduced called “Structural property”. Depending on the designpoint config, Structural property will have a different set of datanames that you can fill. Each designpoint can have its own structural property.

In this documentation, you can either refer to Structural Element or designpoint when it comes to the HyperMesh entity designpoint.

Each designpointset config, except Generic, comes with an algorithm that absorbs Structural Elements from your selection and various options. You will then populate and organize designpointsets with several designpoints to assess methods.

Each designpoint is a collection of several elements which materialize a geometric region on the FE model.

Technically, designpoint is not limited to elements; it could point to any HyperMesh entity. However, the current released configs are either made of elements that form a pattern or freebody sections.

Designpointset is then assigned a series of methods (designpointmethod) to be assessed on all designpoints it holds. The tool relies on a framework (Stresstool) that enables you to plug your math functions in HyperMesh. Such functions can be coded either as:
  • Python functions
  • Compose (OML) functions
  • C++ dynamic libraries (DLL)
  • Tcl procedures

The framework embeds Python/Tcl/Compose interpreters to evaluate such functions. When assigned to a designpointset, registered function becomes a designpointmethod entity. The framework queries all needed results and model information and passes them to methods for evaluation.