Abstract
The South African foundry industry is faced with a high number of foundry closures. When Castings industries directory was released in 2011,twenty nine foundries were reported to have closed down (Castings Industries Directory, 2011).The main factors leading to these closures has been identified in literature as the inability of South African foundry companies to compete with emerging countries such as China, India and Brazil (National Foundry Technology Network, 2009). A lack of skilled personnel, a high scrap rate, a lack of quality management system, and technology transfer all contribute to South African foundries not being able to compete with these countries (National Foundry Technology Network, 2009). Foundry companies experience a high scrap rate due to a high number of defects occurrence. Casting defects are characteristics that create deficiency or imperfection to quality specifications as a result of design and service requirements. In casting, defects occur due to deviation from the metallurgical processes and poor pattern design. A quality management system is implemented to ensure that the proper process procedures are followed. It also provides a company with an opportunity to document any problems arising during the process. This documentation is used for improvements and future reference. This study investigates the impact of poor utilization of quality management system. The objectives of the study include identifying factors affecting quality in the foundry and identifying defects that contribute to high monetary losses in sand casting. The study, which was conducted at a South African foundry company over a period of one year, followed the mixed research methods to conduct the research. The quantitative and qualitative methods were used respectively to collect numerical data and secondary data. Direct observations and interviews were carried out to explore the whole sand casting process as to examine the procedures and documentation used. The observations began in patternmaking. It was observed that a drawing was not issued, resulting in the patternmaker making his own sketch. The defects of cross jointed and core fault occur due to inaccurate dimensions from the patternmaker’s sketch. The patternmaking department also lacks supervision. In an interview with the human resource manager, it was revealed that patternmakers are a scarce skill in the foundry industry. The major finding in sand casting was the lack of adherence to procedures. The standard temperature set for melting metal is not adhered to by operators. Also, very poor pouring practices are followed. The resulting defects from poor pouring practices are shrinkage and gas porosity. When the production v manager was interviewed, he raised concerns about lack of supervision on the production floor. Ten departmental managers were interviewed. The managers were asked questions related to planning for their departments. Only 20% of managers responded positively when asked if they set goals for their respective departments. Eighty percent said they did not set goals. The crucial finding from the interviews was that some of the managers did not value the importance of the quality management system. Five defects were identified as recurring defects. These defects are cross jointed, core fault, bad mould, shrinkage and gas porosity. The standard multiple regression analysis method was used to analyse the data and answer the research questions. The model was used to explore the relationship between defects (independent variables) and also to find out the correlation between the monetary loss and productivity (dependent variable) and defects (independent variable). The results showed a strong relationship between core fault and cross jointed, with a correlation coefficient of 0.748. This result indicates that as core fault increased so did the cross jointed. If core fault is reduced, cross jointed is also likely to be reduced. There was also a negative correlation of -0.64 between shrinkage and gas porosity. The negative correlation indicates that as shrinkage increases, gas porosity decreases and vice versa. The results of the study demonstrate that the higher the defects the company has, the higher the financial losses. According to the results, core fault had a strong correlation of 0.683 with monetary loss. It might be an indication that the castings affected by core fault were of a high value or their weight was high. The overall model gave the R (0.752) and Adjusted R square showing 0.513, thus predicting 51% of the variance in monetary losses, which indicates a good model. It also presented F = 10.700 showing a huge significance (p<.05). Three models, namely stepwise, forward and backward regression were used to determine the statistical correlation between defects and productivity. All three models predicted significant results. Mallow’s Cp value was used to determine the best model. Backward regression model was the best model and was therefore selected to present the results. The defect making statistical significant unique contribution to the prediction of productivity was core fault with coefficient of 0.353 and a Sig. value of 0.006. The overall variance in productivity explained by the model was 23% indicating a good model.
M.Tech. (Industrial Engineering)