a. The Starting Point
In the absence of historical data, engineering assessment of the design and of early development test results may be used to estimate the reliability starting point. An arbitrary selection of a starting point for RDT is not recommended, although traditionally values from 10% to 30% of the predicted MTBF have been used. Techniques for determining a more realistic value are given In MIL-HDBK 189 and are based on design analysis and experience on similar systems.
b. The Growth Rate
The growth rate is a function of the amount of control, rigor, and efficiency by which failures are corrected. An arbitrary selection of a growth rate for RDT is not recommended. Analytical techniques exist for predicting more realistic rates as a function of equipment attributes and development program characteristics.
When Duane or AMSAA models are used, rates in the vicinity of 0.3 are common, 0.5 is achievable if resources are applied intensively, and values as low as 0.1 (or less) and as high as 0.7 have been reported. Early tests can lead to relatively fast growth as the most glaring design errors become evident. Later tests usually are less productive.
A range of acceptable growth rates may be established (e.g., growth rates from .45 desired, to .35 minimum) as a means of tracking contractor performance, with the minimum rate being used to trigger more aggressive actions needed to improve unsatisfactory growth.
c. Hardware to be Tested
To determine the optimum number of equipment/systems to be allocated for the RDT, trade-off analyses must be made. Of primary concern is the availability of systems, subsystems, or even functional modules for the extended test periods involved. Ideally, two or three systems would be dedicated for the required test duration. This is rarely possible, particularly for large or complex systems, because resources are not allocated for a sufficient number of FSED systems to accomplish all test objectives as well as other development tasks. The logical solution, in order to optimize the use of equipment and test time, is to carefully perform pre-RDT analyses and select high risk equipments for emphasis in the RDT. As mentioned before, this may involve management decisions to build and test more of the high risk equipment but this is necessary to assure the system will have the desired reliability. Similarly, Hems such as test chamber availability, performance monitoring capabilities, personnel availability, other test program requirements, and the overall program schedule must be considered in the trade-off analyses.
Because of unavoidable program constraints, you may need to test systems that are incomplete or not configured with realistic operational relationships among system elements. In such cases, meaningful reliability measurement is not possible unless surrogate performance measurements at interfaces have been defined and implemented.
Use of early hardware has drawbacks such as:
- Hardware with tolerance or performance problems may be switched with RDT hardware to allow performance tests to proceed. Holding onto assets is difficult and should be a major consideration in planning any RDT program.
- Early hardware will contain early software; RDT hardware may not be completely representative as software changes may be made which are not functionally compatible or easily installed in the older hardware. Test set and spare parts compatibility may also cause delays. If early hardware is used, plans should call for the cycling in of up-to-date hardware.
d. Test Time
Reliability growth takes place when weaknesses are corrected by changes in designs or processes. Such changes must be preceded by identification of weaknesses, development of corrective actions, implementation of these actions, and verification. The rate at which this process can progress in calendar time is constrained by the rate at which weaknesses can be identified (affected by test time per unit and by the starting reliability) and by the speed of development and implementation of corrective actions (affected by engineering capability and management policy). Growth therefore is non-uniform and discontinuous. Variations are discussed in considerable detail in MIL-HDBK-189.
The overall test hours needed for a RDT may be estimated by using the reliability growth curve planned starting point, growth rate, and required final MTBF. The calendar time needed to complete the RDT will be dependent on the number of units under test and the overall test efficiency, sometimes as low as 50% or less. The following are some of the many factors which must be considered when estimating test efficiency:
- Downtime for maintenance and repair of test chambers or debugging of new test chambers and test setups.
- Availability of spare and repair parts.
- Availability of resources for troubleshooting and failure analysis.
The estimation of total test time for RDT becomes
significant because of the resources involved. Contractual requirements, as
spelled out in Appendix
. require that the proposal submitted in response to
a RFP details all tasks proposed to perform a satisfactory TAAF. Included
would be specifics on quantity and type of equipment, test environments, total
test time, starting point, projected growth rate, growth model used, and
FRACAS. These details, including any government/contractor improvements, are
then incorporated in the contract.