The Automotive Industry designs, develops, manufactures, and sells motor vehicles. This sector is known for its complexity, involving a vast supply chain and a high degree of automation.
A manufacturing company fabricates high precision automotive parts received a customer complaint on dimension out of specification. As part of the cost cutting measures, a new type of machining coolant was introduced by engineering department about few weeks ago. The quality engineer suspected new type of coolant may affect the dimension of machined parts.
A two way ANOVA was conducted to study the impact of different coolants on shaft diameters. Furthermore, this experiment aimed to study the consistency of different CNC machines as well. The engineer took five samples from each CNC machine as follow.
Ten raw data were collected from each CNC machines. For each machine, two different types of coolant were used to evaluate the impact on shaft diameter. The following is the raw data in JMP 9 data table:
(Note : the result is generated by making use of Analyze > Fit model)
Before we examine the final results of two-way ANOVA, we suppose to examine the residual by predicted plot as shown below to identify whether there is outlier or obvious trend or pattern which indicates the existence of special causes during data collection. From the chart below, it notices that the distribution of data above and below zero horizontal line was random which indicates that there are no special causes during data collection.
Base on the results of two-way ANOVA table, the p-value for coolant is above 0.05 which indicates that there is no significant difference between coolant A and B. Refer to prediction profiler below, the changes of mean value for shaft diameter is not significant.
The p-value for machine is below 0.05 which indicates that there is significant difference among five CNC machines. Refer to prediction profile below, the performance of each machine is different especially machine 2 and 5 whereby the shaft diameter is on low side.
The p-value for machine*coolant is above 0.05, which indicates that there is no significant interaction between machine and coolant (Refer to the interaction profiles below)
Blue Ocean Data Solutions is a highly experienced company provides turnkey manufacturing solutions. With more than 10 years of experience serving a large customer base in manufacturing and hands-on expertise in programming, software and statistical data analysis combine with six sigma and lean manufacturing methodology, we assist our clients to solve complicated operation problems – yielding profitability increases and driving service and product delivery excellence.
Company
Contact Us
