How Should Leakage, Proof-Pressure and Dimensional Results Be Correlated for a Cast Pump Housing?
Correlate the results by testing and inspecting the same identified part state. Each dimensional, leakage and proof-pressure record should carry part serial or traceable sample identity, drawing revision, alloy and lot, cavity or tool, machining program and fixture revision, finish state, test fixture, sealing method, medium, pressure, temperature, duration and date. Separate reports without these shared identifiers cannot establish cause or release.
Dimensions show whether passages, ports, walls and seal faces are where the design expects. Leakage testing evaluates escape under a defined condition. Proof-pressure testing evaluates the housing response at a specified load and duration. Passing one does not replace the other, and a failed result should be diagnosed with all three streams before the supplier changes casting, machining or sealing.
Freeze the Part and Test State Before Comparison
Define the test sequence. A housing tested before machining does not represent opened surfaces, threads and seal faces. A test before coating may not represent masking or coating damage, while a test after impregnation or sealant introduces another process state. The control plan should state exactly when each test occurs and whether the part can proceed, be retested or be reworked after failure.
Leakage and proof pressure need complete method definitions. State test medium, fill and vent procedure, stabilization time, pressure ramp, test pressure, duration, temperature, external sealing, detector or measurement method and acceptance limit. Air, helium and liquid methods do not produce interchangeable numbers. A bubble observation, pressure decay limit and measured flow rate answer different questions.
Dimensional evidence must cover features connected to the fluid boundary. Include seal-face flatness and roughness, port position, thread or bore condition, minimum relevant wall, passage intersection and any distortion after pressure exposure. A report containing only overall length and width has little diagnostic value when leakage appears near a port.
Diagnose Dimensional and Functional Results Together
Suppose a hypothetical pump housing leaks near a threaded outlet. If port position is shifted toward the cored passage, the remaining wall may be too small or machining may break into a local discontinuity. If position and wall are stable but leakage clusters by material lot or cavity, casting process or local porosity becomes more likely. If leakage changes with fixture or sealing plug, the test setup may be the cause.
A proof-pressure deformation or crack should be reviewed against section, material identity, machining marks, thread depth, local radius and fixture load. Do not classify every functional failure as porosity. A poor gasket, damaged O-ring, warped seal face, over-deep thread, blocked vent or incorrectly supported test fixture can produce similar symptoms.
Evidence Stream | Shared Identifier | Failure Interpretation | Release or Escalation Action |
|---|---|---|---|
Material and casting record | Alloy, heat or melt lot, tool and cavity | Shows whether failures cluster with material or casting state | Hold affected population and compare chemistry/process records |
Dimensional report | Part identity, drawing, fixture and program revision | Tests port, wall, passage and seal relationships | Correct datum, stock or machining cause before functional retest |
Leakage test | Part state, medium, pressure, fixture and sealing method | Locates or quantifies escape under the stated condition | Confirm setup, identify leak location and follow disposition rule |
Proof-pressure test | Part state, ramp, pressure, time, temperature and support | Evaluates gross structural response under the specified proof condition | Inspect deformation or damage and escalate per drawing owner |
Finish or rework record | Coating, impregnation, repair and retest status | Separates original performance from altered part state | Release only if the authorized process and retest rule are met |
Set Release, Sampling and Change-Escalation Rules
Define whether leakage and proof-pressure checks apply to development samples, each part, each lot or a risk-based sample. The decision depends on consequence, process evidence and customer or regulatory requirements. A test that destroys or permanently loads the component needs a sample rule and a clear relationship to production parts. Do not infer a sampling plan from convenience alone.
The testing-equipment overview can help buyers discuss available inspection categories, but the drawing and control plan must name the method that fits the failure mode. Instrument ownership does not define resolution, calibration, fixture suitability or acceptance criteria for a particular housing.
For repeat production release, trend dimensions and functional outcomes by lot and cavity instead of storing only pass/fail reports. A gradual port-position shift, seal-face change or leakage increase can identify process drift before complete failure. Retain enough identifiers to connect a trend to tool maintenance, core change, machining fixture or material source.
Define changes that trigger revalidation: alloy or source, casting tool repair near the pressure boundary, core or passage change, machining datum or fixture, thread operation, sealant or impregnation, coating at the interface, test fixture and acceptance method. A finish or post-process change matters if it alters sealing, corrosion or dimensions.
The release package should contain one traceable story: the part matches the controlled geometry, the fluid boundary meets the specified leakage criterion, the housing survives the defined proof condition, and no unauthorized process altered that evidence. Correlation makes the records useful for diagnosis and repeat orders; three disconnected certificates do not.
