Example 6 - Multi-Phase Planning and Analysis
[Download RGA7 Example File (*.rga7)]
This example presents three features within RGA 7 — the Crow Extended - Continuous Evaluation model, the Growth Planning Folio and the MultiPhase plot — and how they can be combined to conduct a reliability growth analysis.
Background
A manufacturer wants to enter an aggressive reliability growth program to improve the reliability of a system that they plan on releasing to the public. The product is scheduled to be released in about a year. Therefore, nine months have been allocated for testing. To accumulate as many hours of testing as possible within the time allotted, the manufacturer has decided to test five of these expensive systems. The testing will be divided among two phases. Phase 1 will last for 12,500 hours and phase 2 will last for 14,500 hours. The total testing time is 27,000 hours. The total accumulated number of hours across all systems is recorded when a failure occurs. All systems maintain the same configuration throughout the testing. The goal is to estimate an MTBF of 1000 hours at the end of phase 2.
Analysis and Results
Growth Planning
To begin implementation of the reliability growth program, a reliability growth plan was developed to outline the expected progress so that the goal can be achieved. The average fix delay to incorporate fixes is estimated to be 1500 hours. The average fix delay is the amount of test time from when the failure mode is discovered until the fix will be implemented in the units under test. The Growth Planning Folio representing this scenario is shown next.
The inputs for and results of the planning calculations are shown next.
Note the following about the inputs:
- The GP Design Margin is a "safety factor" that can be adjusted to make sure that the growth potential will be reached. The higher the GP Design Margin, the less the risk in the program.
- The average effectiveness factor for the corrective actions (Avg EF) is determined based on engineering expertise, specific product complexity, prior history, etc.
- The Management Strategy determines what percentage of the discovered failure modes will be addressed (i.e. fixed). In case an MTBF goal cannot be reached, the management strategy can be changed to address additional failure modes.
- A Discovery Beta of less than 1 indicates that the inter-arrival times between unique B modes are getting larger. A value of 0.65 for the discovery beta was chosen based on knowledge of a similar system.
The Growth Planning plot associated with these results is shown next.
This is the overall reliability growth program plan for the new product. Using the Crow Extended-Continuous Evaluation model, the test data at each phase will be compared to the overall goals.
Developmental Testing
The Multi-Phase Failure Times data type is used for this data set. This data type is analyzed using the Crow Extended - Continuous Evaluation model. It is recommended to calculate the unbiased beta whenever performing Crow Extended calculations. To apply this setting, select the Calculate Unbiased Beta option on the Calculations page of the User Setup.
The Data Sheet for the testing data will look like the one shown next.
Note that only a small portion of the data set is shown here. The full data set is available in the Multi-Phase Data Folio in the wbex6.rga7 file.
The effectiveness factors for the BD failure mode fixes have been specified as follows:
The Implemented at End of Phase column indicates when a delayed fix will be implemented for each BD mode. These fixes are not implemented during testing. For example, BD 101 is planned to be fixed at the end of phase 1, before the start of phase 2. Corrective actions are not planned for every BD mode that has been found.
The following results were calculated using the unbiased beta calculation option, indicated as (UnB):
The demonstrated MTBF at the end of phase 2 is estimated to be 929.1971. This is obviously below the requirement of 1000 hours. However, this does not take into account the delayed fixes that are planned at the end of phase 2. The projected MTBF based on these corrective actions can be viewed using the Growth Potential MTBF plot.
Based on the planned corrective actions, the projected MTBF is estimated to be 1006.6053. This meets the specified goal. While it was not specified for this example, a more detailed goal can be provided if a confidence level is also specified along with the MTBF goal.
MultiPhase Plot
During the testing, we can continuously compare the extended planning model with the test data via the MultiPhase Plot. A MultiPhase Plot that includes the planning information from the Growth Planning Folio and the testing information from the Multi-Phase Failure Times Data Sheet is shown next.
In the MultiPhase Plot, the demonstrated, projected and growth potential MTBF at each analysis point and phase are compared to the planned growth at each phase. For the second phase you can see that the demonstrated MTBF coincides very closely with the planned value for that phase. The projected MTBF at the end of phase is equal to 1006.6053, which is just above the goal of 1000 hours.
