Research
AI-assisted creep prediction for plastics
New approach to AI-assisted creep prediction for plastics with fragmentary data.
Project number: 01IF23351N
Project duration
From: 01.07.2024 To: 30.06.2026Description
Material databases offer only fragmentary creep data for plastics. The project aims to close these gaps in terms of time, stress and temperature dependence with the help of artificial intelligence (AI), without additional experimental effort. The ability of AI to recognize complex patterns will be used to complete the data of already existing polymer grades and to transfer these patterns to further polymer grades. The data with which AI models are trained is often key to success. The potential of AI to predict material data remains largely underutilized because material data is usually obtained with high experimental effort and therefore only small amounts of data are available for learning.On the one hand, the scientific-technical challenge is to develop suitable strategies to prepare the data available in databases in the best possible way in order to generate a sufficient training basis. On the other hand, suitable AI models have to be selected from the emerging tools for Small Data and adapted to the problem of material properties.
Contact person:
