Simulation expertise at CTR
Computer-aided simulation techniques depict reality and accelerate development processes dramatically. The simulation competence found at CTR is based on an interdisciplinary team of experts from various branches of the sciences. This professional know-how is employed both in research and to develop solutions for application-oriented objectives. In particular, this comprehensive simulation expertise complements research into system integration with the objective of enabling individual components to be integrated into a higher-level overall system, both virtually and in reality. This can lead to more functionality and improved performance. With a wealth of simulation expertise and corresponding software tools, feasibility studies and innovation concepts can be tested and optimized in advance.
CTR’s areas of competence are focused on the following fields:
- optical simulation (ray tracing, wave optics, etc.),
- electromagnetic field simulation (1D, 3D & transient),
- CFD simulation of fluids and multi-phase systems, and
- FEM simulation of structural or thermal behaviour, plus
- their combination in multi-physical / multi-domain models, from the micro scale up to large-scale devices.
Our team of experts makes use of specialized software to do so, such as Ansys, Comsol, Zemax OpticStudio, LightTrans VirtualLab, CFS++, Simulink, NI Multisim, Octave, Phyton etc., usually in project-specific combinations.
In focus: Static & dynamic expertise
An interview mit Johannes Schicker, simulation expert for mechanical stresses, oscillation behaviour and deformation.
What is meant by the term FEM simulation?
Schicker: We use the Finite Element Method to calculate the reactions of solid bodies to external influences, such as applied forces or temperature changes. Put simply, we replace the geometry with a similar geometry composed of a number of elements with simple shapes such as small cubes which can be calculated easily, and then replace the integration over all (infinitely small = infinitesimal) material points with a summation over our (finitely small = finite) elements. This provides us with the deformations, internal forces and/or the temperature distribution as a reaction to external loads. The reactions of liquids and gases, by the way, cannot be calculated this way; to do this, we use a similar method known as Computational Fluid Dynamics or CFD. We can combine the two methods with each other, however.
Where can FEM simulation be used?
Schicker: Principally speaking, wherever solid materials are subjected to thermal or mechanical stresses. This spans the range from simple stress-deformation calculations for components, entire structures, aircraft engines or aircraft, to natural frequency analysis of oscillation-prone components such as bridges and high-rise buildings as well as temperature development in microchips, and on to failure analysis of complex components such as laminated or sandwich materials or even reinforced concrete structures in earthquakes. There are limitations, however, when it comes to areas in which the “normal” principles of mechanics no longer apply, meaning especially when the dimensions become too small and we need to look at the atomic level.
What advantages does FEM simulation offer in terms of development and optimization?
Schicker: Summarized briefly, it saves costs and time. In contrast to analytical methods, we are usually able to arrive at a result fairly quickly and reliably, even, or especially, when the geometry is somewhat more complex or several components or materials interact with each other. If we have reliable raw data, i.e. data for the behaviour of the material and the boundary conditions, we can often predict how a component will behave in a certain situation without having to first build a prototype. In this case, we use the CAD data directly for the simulation. This allows comparisons of design variations to be made while still in the development phase. Often, we are also simply able to analyse a test scenario and point out what factors need to be taken into consideration.
FEM simulation is a computer-aided tool – you put the data in and obtain a perfect result. Does it really work that easily?
Schicker: The software industry would like us to believe that, especially since the introduction of more sophisticated graphical user interfaces. For simple elastic stress analysis, that may in fact work fine, at least when the user has a good working knowledge of technical mechanics. The question of correct interpretation of the results still remains, however. A challenge which recurs time and again in practice is the selection of the data. It is also possible to make fundamental errors in the selection of the model parameters. With a seemingly simple deflection test for a thin plate, for example, we were able to calculate differences in the deflection of several hundred percent, namely from 16 to 133% of the plate thickness. So what does the correct result look like? Because the deflection tends to be smaller working from the assumption of large deformation, the value of 16% would actually be the better result!