Grinding is a machining process that is used for the fine and finish machining of components. In addition, grinding is classified as a cutting process with a geometrically undefined cutting edge, as the shape, distribution and number of cutting grains in engagement are fundamentally unknown. The cutting grains, consisting of natural or synthetic material and bound in a tool, remove material from the workpiece at high relative speed, whereby the cutting grains and the workpiece are not permanently in contact with each other. This process changes both the shape and the dimensional accuracy of the workpiece and improves the surface quality.

The grinding process is characterized by a high degree of complexity and dynamics, so that a detailed process prediction is only possible with a valid simulation model. With such a physical force model, the effort for experimental parameterization of empirical force descriptions can be significantly reduced. In addition, such a physical force model can be a useful tool for virtual process planning.

To realize such a physical force model, single and multi-grain scratch tests are first simulated at the CPE using an FEM. Using scratch tests, i.e. real experiments, force measurements with different test parameters such as scratch speed, workpiece materials and scratch grain geometries will be performed and used to validate the physical force model. In the further course of the project, this force model will be used for statistical evaluations and will allow scaling from individual cutting grains to a complete grinding wheel and its simulation.

Since industrial grinding processes usually take place under the influence of cooling lubricants, the basic force model is to be adapted or extended so that the influence of any cooling lubricant can be taken into account. For this purpose, real experiments will be performed in order to first identify the general influence of a cooling lubricant on the resulting force of the scratch tests and then to implement this behavior into the existing model using a suitable approach.