Simulations of components subjected to mechanical and thermal loads

By simulating the component, taking into account its shape, material and the mechanical and thermal loads present, it is possible to determine the state of the material. It is possible to determine the state of stress, plasticisation zone, temperature distribution, etc. The simulation result gives an indication of the areas most susceptible to damage or wear. It can also help to select a material suitable for the manufacture of a part.



Microstructure evolution - high plastic deformation

The microstructure of plasticised materials can change in a very significant way. Simulations of changes in modulus of elasticity, crystallographic texture or distribution of misorientation angles can be used to determine the extent to which the properties of a material are altered by a particular plastic treatment process. Examples are rolling, drawing, extrusion and so-called intensive plastic deformation processes.

Determination of material parameters of non-standard models

The determination of parameters for advanced constitutive models describing anisotropic materials, composites, strongly textured materials (e.g. rolled sheets or extruded rods) from typical experiments can be a complex task. We propose to use optimisation methods (e.g. genetic algorithms) to determine the parameters of constitutive models on the basis of available experimental results.



Customised software solutions for individual data sets

For many production facilities, large amounts of data are collected from various types of sensors. The influence of process parameters on the final result is often made possible by the intuition of experienced workers with many years of experience. We propose to use state-of-the-art artificial intelligence and machine learning techniques to develop software that will automatically tell the operator how to set the process parameters to achieve the desired result.

Artificial intelligence-based image analysis software tailored to users' needs

The automation of industrial image analysis saves a significant amount of time and money. An example is the detection and classification of defects on the surface of steel strips. The detection itself can be carried out using relatively simple methods, e.g. by comparing grey levels. However, to determine the type of defect, the use of machine learning, such as deep convolutional neural networks, is usually required.


Ask for more details:

Karol Frydrych – specjialist
phone number: +48 22 273 25 15

Stefanos Papanikolaou – head of department
phone number: +48 22 273 25 07