| |
|
|
Design of
Experiments |
|
Project |
Our client, an
automotive supplier, approached SQPS with an assembly
problem. Their customer experienced occasional adhesion
problems on their part when it was mounted on a larger
assembly at the customer's site. The objective of this study
was to identify all failure modes of assembly (both internal
and at the customer's site). Our client was also interested
identifying the best process settings so no failures could
occur. |
|
Approach |
Our approach
can be summarized as follows:
- We
assembled a team of our client's engineers, technicians
and production associates
- All
possible process variables were investigated. This
included controlled and noise variables
- 3
internal factors and one external (customer-controlled)
factor were identified
- An
experimental design was planned. A factorial design was
generated (3 factors at two levels and one factor at 3
levels). The client's customer provided the necessary
samples to complete the experiment under controlled
conditions
- The
experiment was run and analyzed
- A
report was issued with the findings.
|
|
Results |
The results
showed that
- The
"Testing Distance" factor was the most significant. This
meant that most of the failures that occurred in the
experiment were attributed to this factor. Interestingly
enough, this is the factor that was controlled by the
client's customer. Results were confirmed
-
Another factor, "Promoter Applied" was found not to be
significant. This meant that no difference was observed
whether the promoter was applied or not. However, since
the promoter was being applied in the process,
significant savings from not using it could be realized.
- All
factors were set at their best levels
|
|
|
| |
|
|
|
|
|
|
