Computer Solutions
remove burrs, create a smooth, shiny surface, and protect the device against corrosion, it also must minimise metal removal so that the dimensions of the device remain within specifications.
To reduce costs and improve quality control, Smith & Nephew decided to move their offshore electropolishing operations to their own facility. To make this move successful, they had to demonstrate that the in-house process met critical performance requirements. The project team identified four key factors that affected the electropolishing process:
* Specific gravity of solution; * Voltage; * Cycle time;
understand the effects of the three process variables, while also considering the noise variable, ambient temperature. They also needed to assess the interactions between the factors.
Solving the problem Using Minitab’s Design of Experiment (DOE) tools, Mr Gopal quickly designed an efficient experiment to evaluate the electropolishing process and get answers to the team’s questions.
First, he used Minitab to create a design for their experiment based on the number of factors and the number of runs that they could feasibly perform, given their available resources.
The noise variable ambient temperature, which was controlled during the experiment, needed to be treated as a blocking factor. So he selected a full factorial design with three factors, two blocks for the low and high ambient temperature settings, and two replicates to increase the experiment’s statistical power. He also added centre points to the design to detect any curvature if it existed.
Fig. 2. The Pareto chart allows users to clearly see main effects and interactions that cross over the red line and thus are statistically significant. This chart above shows that both voltage (B) and the interaction between voltage and time (BC) have significant p-values.
* Ambient temperature (noise variable). They conducted preliminary tests to estimate the range of settings for each factor that would yield acceptable corrosion resistance and appearance.
Now the team needed to design an experiment that allowed them to fully
The end result? A lean, efficient experiment that required only 20 runs, accounted for variation due to temperature, and allowed all interactions between the factors to be
evaluated independently. Results
Based on Minitab’s DOE analysis results, the team found that ambient temperature, a potential source of unwanted variation that was difficult to control, had no statistically significant effect on
www.engineerlive.com 19
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39