amtech / 1d-abaqus-optimizer-tool

This tool is provided to Abqus users in welding applications in order to automate the pre processing, the processing and the post processing in a loop until the key parameter is identified using a gradient descent method. This work is a preparation for the master thesis .

1D-Abaqus-optimizer-tool

This tool is provided to Abaqus users in welding applications in order to automate the pre-processing, the processing and the post-processing in a loop until the key parameter is identified using a gradient descent method. This work is a preparation for the master thesis .

Road-map:

Parameters control of a heat source model in Abaqus welding simulations

Heat source model

Parameters of control

The heat source model is based on the Goldak model and has 5 key parameter:

• Q: the heat source maximum intensity,
• Af: Front length of the molten zone
• Ar: Rear length of the molten zone
• B:Half of the width of the bead
• C: Penetration of the bead

In order to familiarize with the topic, only Q is considered as an optimization parameter within this work. The other parameters are fixed.

Target function

In order to find the optimal parameter Q, one should define an objective function based on one of the following inputs:

• 2D or 3D Images
• Measurements over the welding simulated model

Abaqus welding simulations

Pre-processing scripts

The purpose is running automatically a number of simulations given predefined parameters. The parameters are found in the input file and are modified based on a python script. The material model file is also required to run the main script launching the job on the computation server.

Post-processing scripts

Once the simulation is done, a post-processing script has the task to extract the predefined measures of temperature according to the target function variable.

References

This work is done in Fraunhofer IWM at the Structural Integrity and Fracture Mechanics group under the supervision of Mr. Florian Dittmann and Mr. Igor Varfolomeev.