APP-RMP-01 — Pump-P4 accelerated seal evidence

This constructed application connects RMP.1–RMP.8. It is not evidence about an actual pump, seal, failure mechanism, test Method, instrument or authority.

Pump-P4’s reliability council asks whether a high-temperature accelerated seal test supports changing the maintenance interval. The maintenance authority retains that choice. Domain engineering and assurance own the qualified failure relation, test Method and criteria; instrument, laboratory, safety, access, and material owners supply their own results.

First fork — routine qualification, supplied answer, or research question

RMP.1 recovers the receiving use and current frame before treating the laboratory request as research.

The brief states the receiving use, the proposed change to the mechanism-to-evidence claim, the materials and operating conditions covered, current sources and rivals, required criticism, constraints and results supplied by other practitioners. It distinguishes the answer condition from conditions for finishing work with uncertainty. The maintenance authority retains the interval decision.

Second fork — answer the current research-allocation question

With the brief fixed, RMP.2 uses an adequate current answer or compares the real research relations with no new study. An experimental mechanism comparison, field observation or model challenge is serious only when its criticism and obtaining conditions bear on this question.

If a missing value prevents a particular selection still needed, a hold is the useful return. A settled no-new-study allocation instead retains the unknown as a limitation without inventing an awaited investigation. On the selected-design branch, the decision remains the source of the comparison and its limits; RMP.3 supplies the operational work it does not itself perform.

From the selected design to operational choices and actual trace

Suppose the mechanism marker, measurement qualification, specimen plan, safe conditions and access are supplied. RMP.3 connects the marker and comparison to the question of whether acceleration changes the mechanism. The protocol identifies each seal and material batch, its assigned condition, measurement timing and procedure, and the records needed to criticize selection and missingness. Three readings from one seal remain repeated observations, not three independent seals.

Now suppose the domain owner reports that the marker cannot distinguish thermal degradation from the competing field mechanism under the selected conditions. Return that result and its consequences to RMP.2. Replacing the marker with a convenient wear score is not harmless protocol completion; the new observation needs a justified relation to the original question.

On the valid-marker branch, specimen P4-07 is damaged in handling before a planned observation. RMP.4 retains its identity and allocation, the supported handling history, absent observation, prior exclusion intention and actual decision. It connects the remaining comparison data to their specimen and measurement records. A later critic can examine the selection and missingness; the trace does not decide whether the remaining data support the scientific inference.

The operational result is an executable specification for its supported scope and a trace whose examinations and limits are clear. RMP.5 uses the actual specimen, measurement and exclusion history to analyze the bounded claim under its qualified Method. A missing observation does not by itself establish the direction or importance of any resulting bias; preserve the analysis and criticism that bear on that consequence.

From analysis to a relevant credibility examination

Suppose the proposed field conclusion rests instead on a calibrated simulation whose identified data and code reproduce the reported curve. RMP.5 retains the computational result and exposes the unsupported mechanism-to-field step. Repeating the same computation cannot establish the field claim.

RMP.6 can finish with that limited result when no relevant worthwhile examination is obtainable. If independent qualified field material later makes a discriminating mechanism comparison feasible, assess its contribution and select the appropriate validation or comparison. The selection supplies no performed result. A later returned comparison may retain, narrow or defeat the proposed field use; the maintenance authority retains the interval decision.

From a corpus answer to the exact model return

Suppose four supplied reports include two accounts of one accelerated study, an independent compatible test and a service-use account with different conditions. RMP.7 retains the supplementary report’s useful detail while recovering only two independent test contributions. It can combine the compatible laboratory contrast under its justified statistical model; the service account remains evidence for its different question. RMP.7’s numerical case shows the calculation and its assumptions.

RMP.8 compares the synthesis with the model’s particular claims. The laboratory relation can remain supported while the claimed transfer to service remains unestablished and a dependent replacement-time prediction loses its warrant. The returned correction identifies that missing relation and preserves the engineering owner’s supplied criterion. RMP.8’s model case shows those distinct changes.

The reliability council receives that bounded result. A current source answer, limited synthesis or exact correction can finish the research question; a future field study earns a separate selection, and the maintenance authority retains its decision.