Bachelor thesis or Master thesis
Topic of the work
Fiber-reinforced plastics are suitable as lightweight construction materials as they combine high strength with low weight. They are used in many industries such as aerospace and the automotive industry, where complex geometries and load-specific designs are required. Manufacturing processes such as injection molding enable the production of such structures while meeting the necessary performance requirements. This performance is directly influenced by the microstructure of the material, which is created during the manufacturing process and depends on the geometry to be produced. Currently available simulation methods are able to capture this microstructure, but they are too computationally intensive for the calculation of entire structural components. This is where artificial intelligence (AI) methods come into play: they offer a faster method for predicting these properties and thus change the approach to optimizing the lightweight structure.
© IKVObjective
During your work, you will develop an AI-based model to predict the mechanical properties of fiber-reinforced plastics. As part of your final thesis, you will generate experimental data by producing test specimens and testing them. In addition, you will carry out numerical simulations to generate virtual material data for the training data set. These serve as an input data set for the AI-based model and also validate it. The end result? A validated and optimised AI model that delivers accurate and reliable predictions – faster than traditional methods.
Your task
| You will work on the following tasks for a Bachelor's thesis: | For a Master's thesis you will work on the following tasks: |
| Experimental investigation: Production and testing of samples to determine the material properties. | Extended experimental investigations on microstructure and mechanical properties. |
| Carrying out numerical simulations to obtain additional training data. | Combination of experimental and simulative data to optimise the training data set. |
| Development of a simple ANN (Artificial Neural Network) model to predict mechanical properties. | Development and optimisation of a deep neural network. |
| Validation of the AI model by comparison with experimental data. | Comparison of AI predictions with numerical simulation results and experimental data. |
Your profile
- Degree in mechanical engineering, computational engineering science, materials science or similar
- Basic knowledge of numerical simulation (FEM)
- Basic knowledge of programming (Python)
- Interest in artificial intelligence and machine learning. Knowledge of AI/ML is an advantage
- Motivated, structured and independent way of working