Appendix F - Actuator Certification Test Plan Rev A.pdf

CORRELATION TEST PLAN

for

PRODUCTION TESTING FOR ACTUATORS

on

ACTUATOR TEST STANDS

BUILDING 3902

10 JULY 2020

Main Engine Control Certification Test Plan For Production Testing on new MEC Test Stands

Introduction This document describes the certification test plan to qualify the new MEC test stands (hereinafter referred to as New Stands), to be located in Building 3902, to perform and meet the production testing of requirements of all Main Engine Controls per applicable T.O.(s).

Correlation Process The New Stands will be correlated to prove that performance meets or exceeds the existing production test stands and applicable T.O. requirements. The New Stands will be approved for production testing of the End Items listed in Table 1 if the Master Stand data and the New Stands data correlate as follows:

1. The final acceptance test will be run five times, per T.O., on the Master Stand to obtain correlation data. Data will be checked for outliers utilizing the Weisberg t- Test. See the section below titled Weisberg t-Test for Outliers for a brief description of testing for outliers. If an outlier is found, results from an additional run will be utilized to ensure that five samples are available for each test point. If an outlier is still indicated, it will be considered normal data scatter and all six runs will be used in the 95% confidence interval analysis.

2. The final acceptance tests will be run five times, on the New Stands to obtain correlation data. Preset tests will be checked to ensure functionality only. No data will be submitted for Preset tests. Data will be checked for outliers utilizing the Weisberg t-Test. See the section below titled Weisberg t-Test for Outliers for a brief description of testing for outliers. If an outlier is found, results from an additional run will be performed to ensure that five samples are available for each test point. If an outlier is still indicated, it will be considered normal data scatter and all six runs will be used in the 95% confidence interval analysis.

3. Correlation between the two data sets will be analyzed utilizing the MOSI Method with confidence intervals of 95%. See section below titled Mosi Method for brief description of analysis method. See Attachment 4 for detailed description of analysis method.

4. The results from the correlation methods will be documented in an Excel spreadsheet.

See Attachment 1 for an example Certification Test Worksheet.

5. After the correlation data runs are completed a preliminary test report will be submitted to the System Program Office Cognizant Engineering Authority for review.

Following approval, a certification test report will be submitted, via AFTO 202, for formal review and approval of the New Stand. See Attachment 2 for an example Certification Test Report.

5 Run Correlation Method The methods to be used for correlation of test stand data are as follows:

• Weisberg t-Test for Outliers

• Mosi Method for comparing two means with a 95% Confidence interval

Weisberg t-Test for Outliers All correlation data will be tested for Outliers at an Alpha Error Rate of 5%. All data will be checked for outliers at the time of data generation from the test stand and at the time of data correlation. See Attachment 3 for additional information.

Mosi Method The Mosi Method (See Attachment 4 for additional information) for comparing two means with a 95% Confidence interval involves the following steps:

• Average the five samples from each stand

• Calculate the Standard Deviation of the five samples from each stand

Main Engine Control Certification Test Plan For Production Testing on new MEC Test Stands

• Calculate the 95% confidence intervals for each mean of stands

The data from each stand is statistically comparable if the intervals overlap such that each interval overlaps with the mean value of the other interval as shown in the diagram below. In our case there is a potential bias that must be taken into account to properly apply the MOSI technique. This bias is the measurement uncertainty of the test stand instrument.

For example:

Test stand A measures an average flow (5 readings) of 960 pph and Test stand B measures an average flow (5 readings) of 1040 pph. The measurement uncertainty for the flow measurement is ± 50 pph. For test stand A we can state the true average flow measurement is between 910 and 1010 pph and for test stand B we can state the true average flow measurement is between 990 and 1090 pph. Due to this bias we must include the measurement uncertainty as a range of acceptable mean values to properly apply the MOSI technique. This is graphically displayed on the following page.

Main Engine Control Certification Test Plan For Production Testing on new MEC Test Stands

Example of two stands that are statistically comparable

Example of two stands that are not statistically comparable

Actuator Certification Test Plan For Production Testing on new MEC Test Stands

Table 1: End Items Tested Noun

PN to be Certified

T.O. Test Type

F110 IGVA 3820071-114 9H2-7-42-3 5 – Run Correlation F101 IGVA 3810051+-103 9H2-7-37-3 5 – Run Correlation F110 VSVA 1211335-009 9H2-7-44-3 5 – Run Correlation