Safety-Critical Actuator Fails Pressure and Vibration Acceptance
An airborne hydraulic actuator for a military trainer is in acceptance testing. At 210 bar it holds pressure but decays 8 bar/min against a 5 bar/min limit. On the vibration rig, a 2 g sine sweep from 10 to 2,000 Hz shows 4.5 g at the mounting bracket at 850 Hz, above the 3 g limit; the rig accelerometer was last calibrated 13 months ago. The unit shows no external leak, but the test engineer notices a faint seep at the rod seal. The customer's delivery date is five days away and the programme manager asks for a waiver, arguing the decay is within measurement tolerance. The actuator is safety-critical and has no redundancy. The lab has one calibrated accelerometer and a replacement seal kit, but rework and retest will take three days. The specification requires both pressure hold and vibration limits to be met.
1. Should the programme manager's waiver be accepted? Give your decision and reasoning.
Read the worked answer
No. The pressure decay is 8 bar/min against a 5 bar/min limit, 60% over, and that is not automatically measurement tolerance unless the gauge and rig are proven wrong. The vibration response is 4.5 g against a 3 g limit, 50% over, although the 13-month-old accelerometer calibration must be checked before the result is treated as final. The faint rod-seal seep supports a real unit problem: a seal or clearance issue could explain both pressure decay and a change in damping or stiffness. The actuator is safety-critical with no redundancy, so a waiver would carry unquantified risk. Require calibration verification and repeat testing, then rework and retest. A delivery slip is preferable to an unqualified flight-control component; if a bridge is needed, fit a known-good unit.
2. How would you separate test-rig artefacts from genuine unit faults?
Read the worked answer
First verify the accelerometer calibration and mounting. A 13-month-old calibration is a valid reason to doubt the 4.5 g reading. Re-run with a second calibrated accelerometer or a laser vibrometer, check bracket torque, and perform a bare-rig baseline sweep. If the 850 Hz peak persists on the unit but not on the rig, it is a unit resonance. For pressure decay, blank off the test ports, verify gauge calibration and temperature stabilisation, and leak-test a reference actuator on the same rig. The rod-seal seep is a direct unit finding and cannot be explained by rig calibration. Only after eliminating rig artefacts should you rework the unit. Document each check; do not use calibration doubt to waive the pressure limit.
3. Set out the rework and acceptance plan, including what a concession would have to contain.
Read the worked answer
Plan: strip and inspect the rod seal, rod surface and bore; replace the seal kit; check contamination; reassemble; flush; and pressure-test at 210 bar for 10 minutes with decay no more than 5 bar/min and no seep. After accelerometer calibration, repeat the 2 g sweep and require 3 g or less at the bracket with no new peak at 850 Hz. If a concession is requested, it must be time-limited, supported by root-cause analysis, include additional inspections and a reduced service interval, and be approved by the design authority, not the programme manager. No concession on the pressure-hold limit is defensible without proving a gauge error. Three days is realistic; plan the delivery slip and reserve a known-good unit for the immediate fit.