CIVILICA We Respect the Science
(ناشر تخصصی کنفرانسهای کشور / شماره مجوز انتشارات از وزارت فرهنگ و ارشاد اسلامی: ۸۹۷۱)

Variation Source Diagnosis in Compliant Sheet Metal Assemblies Using Neural Networks

عنوان مقاله: Variation Source Diagnosis in Compliant Sheet Metal Assemblies Using Neural Networks
شناسه ملی مقاله: ICME07_205
منتشر شده در اولین کنفرانس بین المللی و هفتمین کنفرانس ملی مهندسی ساخت و تولید در سال 1384
مشخصات نویسندگان مقاله:

M Shariat-Panahi - Assistant Professor Department ofMechanical Engineering,University of Tehran, Tehran, Iran
S.H Sadat - Graduate Students, Dept. of Mechanical Engineering, Sharif University of Technology, Tehran, Iran
S Bagheri - Graduate Students, Dept. of Mechanical Engineering, Sharif University of Technology, Tehran, Iran
M.H Mokhtari - Senior Design Engineer, Iran Khodro Company, Tehran, Iran

خلاصه مقاله:
When a structure is made up of deformable sheet metal panels joined together by means of spot welds, rivets, adhesives or the like, it is commonly referred to as a compliant sheet metal assembly (CSMA). The dimensional quality of a CSMA is primarily determined by the variations of its constituent parts and fixture(s). While conventional variation analysis methods such as Worst Case, Root Sum Square and Monte Carlo may be able to simulate the contribution of rigid part variations to the dimensional variation of the final product, they definitely do not produce realistic results for the case of compliant sheet metal assemblies. In recent years, a number of CSMA-specific models have been proposed that take into account the deformability of the constituent parts. These models can fairly accurately predict the variations of the final assembly, given the variations of the individual parts and of the fixture(s). However, they are more of a prediction tool rather than a diagnosis tool which the designers really need. In other words, the designer often wishes to identify the source(s) of a particular variation in the final product (assembly) as a function of the variations of the components and the fixtures; whereas these models only predict the effect of a component-level variation in the assembly.

کلمات کلیدی:
Sheet Metal Assemblies, Variation Analysis, Fault Diagnosis, Neural Networks

صفحه اختصاصی مقاله و دریافت فایل کامل: https://civilica.com/doc/82664/