Providing a Model to Determine of Powder Factor using Principal Component Analysis Technique

Objectives: Powder factor is one of the main technical and economic parameters in the design of drilling models and tunnels blasting. Therefore, the prediction and optimization of powder is so important. Methods/Statistical Analysis: The value of powder factor is dependent upon several parameters such as geological conditions, mechanical properties of the rock and geometric design parameters. In this study, based on geotechnical properties of the rock mass in construction operations for water delivery tunnel of Seimare Dam, a suitable model has been presented to determine the powder factor using statistical methods. Findings: In this regard, PCA analysis was used to eliminate the effect of co-linearity between input variables in prediction models and coefficient of determination (R2) and mean square error (MSE) were used to assess and compare the constructed models. Comparison of models shows that the elimination of co-linearity between input variables using PCA has given better prediction results. Application: In conclusion, a model proposed to determine the Powder Factor effects on cement by using principal component analysis technique, which is valuable in civil industry.