Description
The importance of data is increasing in all areas of life. This effect can also be observed in product development. The combination of virtual product development with the continuous and holistic use of data is referred to as «digital engineering». The implementation of digital engineering results in a transformation process and is accompanied by a change in the previous roles of people involved and the tools used. The aim is to use as much data as possible and to process this data with machine learning algorithms. In product development there is numerous geometric data (e.g. CAD models or measurement data) or data which is linked to geometry (e.g. numerical simulations and their results). In this dissertation, the method of spherical detector surfaces was developed, which makes it possible to transform arbitrary geometries into a uniform numerical matrix. The developed method can also be used to convert information associated with the geometry into further uniform matrices and thus provide this information for machine learning algorithms. The developed methodology is implemented in three different application examples and all necessary substeps are described in detail. This also includes the transformation of finite element simulations to the so called «DNA of an FE simulation».
Reviews
There are no reviews yet.