Thesis
Validation of the material behavior of thermoplastic composites using covariance analysis
Motivation
With the help of numerical models, processes can be virtually investigated and optimized without having to rely exclusively on complex test series. Many different material models are available for this purpose, but their suitability is not always clear.
In practice, the quality of a model is often judged by the fact that simulation and experiment match as well as possible.
match as closely as possible. However, this approach can be deceptive, as unsuitable models can also deliver seemingly good results by adjusting model parameters without correctly describing the actual material behavior.
behavior.
An alternative approach considers not only individual measured values, but the entire dynamic behavior of the material. For this purpose, a covariance-based validation approach is used, which was originally developed in a different field of application.
was developed. Statistical correlations between input and output variables can be used to check how well a model reflects the essential properties of the system.
The aim of this bachelor thesis is to transfer this method to thermoplastic composites and to test various
material models on the basis of experimental characterization data.
Work content:
- Familiarization with existing simulation models and software implementations
- Application of the covariance-based validation approach to experimental characterization data
- Systematic evaluation and comparison of different material models with regard to their ability to reproduce the experimental results
results - Documentation of the methodology and results in a written paper
Requirements:
- Initiative and independent working style
- Strong analytical skills
- Interest in simulation and numerics
- Programming experience in Python (advantageous)
Subject area: mechanical engineering / computational engineering
Type of work: simulative/numerical/experimental
Start: immediately
Contact: M.Sc. Johannes Mitsch
Phone: +49 721 608-45390
E-Mail: Johannes.Mitsch∂kit.edu