Font Size: a A A

Studies On The Application Of Multivariate Statistical Process Control In Casting Production

Posted on:2005-03-20Degree:DoctorType:Dissertation
Country:ChinaCandidate:B F LiFull Text:PDF
GTID:1101360152967982Subject:Materials Science and Engineering
Abstract/Summary:PDF Full Text Request
Multivariate Statistical Process Control (MSPC) is a method for monitoring, analyzing and controlling process operating performance. As an efficient technology for quality control and improvement, MSPC has been successfully applied in both continuous processes and batch processes. With the development of foundry technology, improving casting quality and reliability has been considered as the most important target of foundry industry. In order to assure casting quality, it is necessary for an efficient method to monitor the performance of production process, which is a very complex one with many influencing factors. Therefore, to apply MSPC in casting production is an important way in this respect.According to the characteristics of casting production, a plan for foundry application of MSPC has been developed. The kernel technologies of the plan have been studied.Univariate Statistical Process Control system has been established in foundry. The key procedures of casting production have been studied with univariate Statistical Process Control methods. The studies demonstrate the requirement and feasibility of the application of MSPC in foundry.Some research for Principal Component Analysis (PCA) modeling of molding sand preparation process has been developed. Based on the analysis of molding sand preparation process, a method of collecting the data for MSPC has been designed. To avoid the influence of a small quantity of samples to the result of PCA modeling, an improved PCA modeling method has been proposed.The rules of the variation Squared Prediction Error (SPE) statistics and T2 statistics with variance of process variables and variance increasing of process variables have been proved by theoretical derivation and simulation. The MSPC charts for molding sand preparation process have been determined.The PCA model of molding sand preparation process has been established in a leading foundry. SPE chart and T2 chart have been used to monitor the real process. Some faults have been detected by the MSPC charts. When the special cause of the faults has been confirmed and some measures have been taken, the casting quality has been improved significantly. A true case of the application of MSPC in foundry has been shown.
Keywords/Search Tags:casting production, quality control, multivariate statistical process control
PDF Full Text Request
Related items