In the design of mechanical assembly, the dimension-chain tools take into account the manufacturing dispersion of the parts and assembly defects. This ensures the interchangeability of the different components and guarantees that an assembly can carry out different service functions, as it is modeled in infinitely rigid solids. However, this approach does not take thermo-mechanical effects and deformation due to inertia effects like gravity, angular velocity etc., into account. Most materials change length as they change temperature. As a result of this change, the dimensions and tolerances of a product become at variance with the design values. Hence, thermal effects must be taken into account when designing a product that will undergo temperature cycling and yet, the different operating regimes of an assembly make it indispensable that the effects caused by the thermodynamic cycle should be integrated. In this regard, a finite element model of a machine assembly is created in order to determine the deformation due to change in temperature and inertia effects. The aim of this article is to include the deformation determined by Finite element analysis in the dimension chain thereby controlling clearances in the mechanical assembly. The approach first generates a Cost-tolerance model using neural network where the inputs are parameters and tolerance levels. Then, Finite element analysis of the machine assembly is carried out. The deformation obtained by FEA is then included in the dimension chain. Finally, optimization is done using Non-dominated sorting genetic algorithm II (NSGA II). The results provide designers with optimal component parameters and tolerance values, and the critical components and the manufacturing cost. The approach can also guarantee that the parameter and tolerance values found remain within tolerance for the temperature variation. Then, the product can function as intended under a wide range of temperature conditions for the duration of its life.
Key words: Dimension chain, thermo mechanical tool, finite element analysis, neural network and NSGA II.
Copyright © 2022 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0