Reasoning for Assembly

Classification of parts during the DFA/Manufacturing analysis

 

Geometry

Support for Reasoning for Assembly

 

Reasoning for Assembly Overview
There is much work in the field of assembly sequence planning which addresses the determination of valid component insertion trajectories. Insertion operations are complicated by the existence of almost- and partial-symmetries. Ideally, a component should be exactly symmetric, parallel and perpendicular to the direction of insertion, or be distinctly asymmetric. However, determination of insertion trajectories also requires knowledge of mating faces and degrees of freedom. Hence, methods for determination of these characteristics form another significant aspect of the Sandpit geometric reasoning requirements, although previous research, already described, offers some potential solutions.

However, in order to reduce the complexity of the problem user input may be utilised by development of the CAD interface such that relevant information is easily accessible. Many commercial CAD packages require the engineer to define the position and orientation of a component in the assembly in relation to its adjacent parts, using a set of constraints. These constraints consist of commands such as ‘Mate’ or ‘Align’ and therefore provide additional information about component interactions. Usually two constraints are sufficient to fully define the location of a component but obviously, there may be further unspecified contacts between parts. Geometric reasoning functions may then be invoked to identify all remaining contacts and the degrees of freedom may then be inferred from the accumulated data. There is much work in the field of constraint-based modelling which may provide a useful foundation from which to develop suitable algorithms.