بهینه سازی عددی چند پارامتری عناصر دستگاه نازک دیواری انتخاب شده با استفاده از یک الگوریتم بهینه سازی stigmergic

This paper presents a method of multi-parameter optimization of shell structures of constant thickness from linearly elastic homogeneous material. Our goal is to present how a relatively complex shape of shell can be optimized in terms of its stiffness only by changing its geometry and at the same time preserve the primary volume. Loading cases are not known exactly and so we use the criterion function which on the basis of arbitrary chosen eigenvalue frequencies, gives a stiffness of the shell structure. The optimization algorithm is based on stigmergy, which is becoming increasingly popular in combinatorial optimization and lately also in multi-parameter optimization. The method was verified on a real industrial problem on an electric motor casing.

In the design of various thin-shell structures we often encounter the problem of modeling a shell-structure as light as possible and, at the same time satisfying the requirements of high stiffness. To achieve this, one possibility is that at the proper places on the shell material is added, thus reinforcing the shell, structure. However, in this way we increase the time of manufacturing, amount of material and difficulty of processing. Another possibility is trying to make the shell stronger only through variation of its shape. Our starting condition in this optimization study was that the volume of the newly shaped structure has to remain the same as in the existing one, that the consumption of material is not higher and that the manufacturing process and time are not affected by the change in the design. This kind of approach is especially important for cases of low-cost components which are mass-produced by deep drawing. In the continuation of this paper, an example of such optimization will be presented in detail on the casing of the electric motor produced by Domel Ltd., Železniki, Slovenia.

The most important achievement of this optimization is that in future the wall of the casing of the electric motor will have a thickness 0.8 mm instead of 1 mm as is the case in the existing casing. So the result of geometry optimization will be smaller consumption of material and energy for forming which means a lower cost of manufacturing process and the product itself. In the chosen criterion function in the optimization process the aerodynamic and other mechanical loadings were not considered but are present during the running of dry vacuum cleaner motor. For further optimization the dry vacuum cleaner motor should be considered as a casing-impeller loading collective which means that aerodynamic and acoustic effects should be added to the criterion function. This leads to multi-criterion function with more parameters added and will be the subject of further research.