Constraint solver for a fixture design: results of a student case study

Abstract

For teaching students the skills of programming and usage of knowledge-based engineering systems, we conduct student projects in a lecture in which they independently represent a configuration solution space and resolve it using a constraint solver. For this purpose, the lecture is conducted in a flipped classroom concept to not only teach the students the theoretical basics but to enable them to independently formulate and integrate design problems, which can be abstracted as configuration problems, so that they develop a sustainable competence through learning-by-doing. The configuration problem of the student case study represented here is the positioning of a cast part for manufacturing, where the positioning is done via three subassemblies consisting of parts from a fixture toolbox. For this purpose, a development environment written in the Python programming language was set up, which uses an external Excel database as a knowledge base to provide the sizes of the fixture elements. Through a graphical user interface, the designer can specify how the fixture should be used so that the constraint solver can find a solution. If there are several possible solutions, an optimization loop is executed so that the designer can be given a clear recommendation. An interface to the CAD program Autodesk Inventor offers the possibility to build the fixture assembly of the selected solution from parameterized CAD models of the components by linking their custom coordinate systems. To reduce computing time, a case base is also provided for configurations that have already been created, so that existing subassemblies can be used if the same or similar configuration problem arises.