7.4.7. Basic failure rate derivation for “non-standard model” items#

When the miscellaneous item under consideration does not match the criteria of the miscellaneous equivalent class (e.g. SSPA – Solid State Power Amplifier- cannot be compared to a TWTA in terms of technology), it is necessary to build a reliability model based on engineering/PoF approach as defined hereafter.

Once the reliability model is established it can be tailored as necessary based on the approach presented in Section 7.4.5.

First, it is necessary, to collect/define all technical information about the miscellaneous item in order to characterize it (refer to Section 7.4.5). Then the basic failure rate \(\lambda_{1}\) = \(\lambda_{B}\) is defined (step 1) as discussed below. Finally, the tailoring of this basic failure rate follows the “standard model” procedure as presented in the subsections of Section 7.4.5) corresponding to step 2 to 7.

7.4.7.1. Design FMEA / Process FMEA#

After the first item characterization, a Design FMEA and a Process FMEA are performed in order to assess the failure modes due to

  1. Design errors and

  2. Manufacturing errors.

It is recommended to use the [BR_MIS_8, BR_MIS_9] which provides the method to apply for a Design FMEA / Process FMEA. Process FMEA is also specified in [NR_MIS_5].

FMEA: Required for every unit

  • Is a tool/method supporting the design of the unit, identifying all the elementary “failure modes” due to “random failure” (part intrinsic reliability) as defined in the failure mode list (refer to [NR_MIS_5]).

  • @unit level: under supplier responsibility

  • @system level: integrated at system level by System RAMS engineer (including HSIA supporting the failure tolerance: observability in-orbit…)

Design FMEA: Required on critical functions only (Risk analysis)

  • Is a tool/method supporting the product quality identifying all the elementary failure modes due to DESIGN errors. It includes, as a minimum, the failure modes identified by FMEA.

  • @unit level: under unit designer / quality Assurance responsibility

  • @system level: reviewed by procurement within the unit acceptance process

Process FMEA: Required on critical functions only (Risk analysis)

  • Is a tool/Method supporting the product quality identifying all the elementary failure modes due to MANUFACTURING errors.

  • @unit level: under unit quality Assurance responsibility

  • @system level: reviewed by procurement within the unit acceptance process

../../../_images/figure4_81.png

Fig. 7.4.7 Differences between FMEA, Design FMEA and Process FMEA.#

Fig. 7.4.7 highlights the differences and the commonalities between FMEA, Design FMEA and Process FMEA.

Fig. 7.4.8 provides an example of Design FMEA and Process FMEA at elementary part level (interconnect between Solar cells). The left-hand side of the table refers to FMEA (blue perimeter in Fig. 7.4.7), whereas the right-hand side refers to Design FMEA data (green perimeter in Fig. 7.4.7).

The main outcome is the ranking of every root cause (Design error, Manufacturing error) in

  • Severity (SEV)

  • Occurrence (OCC)

  • Detectability (DET)

The Risk Priority Number (RPN), which is derived, is an index reflecting the residual technical risk.

../../../_images/figure4_101.png

Fig. 7.4.8 Differences between FMEA, Design FMEA and Process FMEA. The following terms are mentioned in this figure: Solar Array (SA), SA current (I_SA), Telemetry (TM) and Single Point Failure (SPF).#

Table 7.4.6 Occurrence scale with 10 levels#

Rank

Category

Occurence of cause

Explanation

TRL

10

Certain and unpredictable

Failure is unpredictable

New technology / new design no history

1/10

9

Inevitable

Failure is almost inevitable & persistent

Failure is inevitable with new design, new inevitable & application, or change in duty cycle/operating persistent conditions.

TRL1

1/20

8

Almost certain

Failure is almost certain

Failure is likely with new design, new application, orchange in duty cycle/operating conditions.

TRL2

1/50

7

Frequent

Frequent failures

Failure is uncertain with new design, new application, or change in duty cycle/operating conditions.

TRL3

1/100

6

Repeated

Repeated failures

Limited number of failures associated (identical design, simulation and testing). Use of Factor of Safety or Margin of Safety

TRL4

1/500

5

Occasional

Occasional failures

Limited number of failures associated (identical design, simulation and testing). Use of Factor of Safety or Margin of Safety

TRL5

1/2000

4

Infrequent

Limited failures

Limited number of failures associated (identical design, simulation and testing). Use of Factor of Safety or Margin of Safety

TRL6

1/10000

3

Isolated failure

Relatively loaded failures

Limited number of failures associated (identical design, simulation and testing). Use of Factor of Safety or Margin of Safety

TRL7

1/100000

2

Unlikely

Very isolated failures

No observed failures (identical design, simulation and testing). Reduce stress-strength interference

TRL8

1/1000000

1

Remote

Failure is highly unlikely

Failure is eliminated through preventive-type design control Use of proven design guidelines/Standards. Use of field lessons learned

TRL9

\(\epsilon\)

7.4.7.2. Probability assessment#

The occurrence scale of Section 7.4.7.2 is used to derive, based on engineering judgement, either a failure rate or a probability of failure, per failure mode:

  • In the case Detectability (i.e. the capacity to detect on ground the defect) is certain, the probability of failure is set to 0 under the condition that no degradation in time is expected (e.g. if a particle inside a HF/RF passive part which originates from manufacturing is necessarily detected on ground with no possibility to get a particle during the mission, then the associated probability of failure is set to 0).

  • A probability of occurrence is assigned to every failure mode and the sum provides the probability of occurrence of the item.

  • A failure rate could be derived assuming the probability of occurrence on the specified lifetime is equal to the probability assigned to the failure mode, e.g. probability assessed to \(10^{-4}\) as level 4 (Table 7.4.6) leads to a failure rate of \(1,14 10^{-9}\) for 10 years.

This represents the basic failure rate or the basic probability of failure.

Then the general process to adapt this failure rate is described in Section 7.4.5.

It is noteworthy that this probability ranking needs to be clearly justified and documented.