Defect Investigation (DI) After Return to Shop (RTS):
Adour Mk 811 Engines Returned from the Factory Test Bed with a Defect
By Suravarapu Chandrasekhar Aerospace
QA/QC and MRO Engineer
Educational overview. Acceptance limits,
test schedules, rework criteria and retest requirements come only from the
approved Engine Build Manual, Test Schedule, Overhaul Manual and the quality
procedures of the design authority and HAL. No numerical limits are quoted here
on purpose.
Introduction: The Engine That Stopped on the Test Bed
There is a moment on every engine
production line that tests the whole organisation. A brand-new Adour Mk 811,
built from scratch by people who signed for every stage, goes onto the test bed
for its acceptance run. Then something shows up: a leak at a joint, an oil
pressure irregularity, a chip detector finding, a vibration trace that looks
wrong. The run is stopped. The engine is not accepted and comes back to the
shop.
That is RTS, Return to Shop, in the
production context. The engine is not an ageing in-service unit with thousands
of hours behind it. It is a newly built engine, assembled in the factory, run
in the factory's own test facility, and returned to the factory's own shop
floor to find out what went wrong. After the Defect Investigation (DI)
and repair, it goes back to the test bed to prove itself again.
A senior inspector once told me, "A failed test run is the cheapest place to find a build problem. Every stage
after it costs more." I have held that view for decades. This article
explains how DI is carried out after RTS on a production Adour Mk 811, the
engine of the SEPECAT Jaguar built under licence by HAL for the Indian
Air Force, when the engine returns with a defect observed on the test bed:
a visible, specific symptom such as a leak, an abnormal indication,
contamination, vibration or a control fault.
How a New Adour Mk 811 Is Built, Tested and Accepted
To investigate a defect, you must
understand the path the engine took to reach the test bed. The Adour is a
twin-spool, low-bypass turbofan with reheat, built in modules. In simplified
form, the production flow runs like this:
1.
Manufacture and procurement
of parts, with material certification, inspection
and traceability from raw material onwards.
2.
Module build-up and
sub-assembly: compressors, combustion section,
turbines, reheat, gearbox and accessories, each with its own build records and
checks.
3.
Rotor balancing, clearance
and fit checks at module level, per the approved
build manual.
4.
Final assembly: modules joined, pipes, harnesses, controls and accessories fitted,
with locking and torque verified by independent inspection.
5.
Pre-test checks: leak checks, rotation checks, instrumentation fit and
documentation review.
6.
Test bed run: the engine is mounted in the factory test facility and run through
the approved acceptance schedule, covering starting, idle, acceleration, power
settings, reheat and shutdown, with instruments recording temperatures,
pressures, speeds, flows, vibration and oil system behaviour.
7.
Post-test inspection and acceptance, followed by preservation, packing and release.
The test bed is the first place where every
part, every assembly step and every calibration meets real operating
conditions. It is the final examination, and a sound quality system expects
that some engines will not pass first time. It also has a clear route for them:
RTS, DI, repair and retest.
A short history
Early production jet engines were often
accepted on a short run, a visual check and the experience of the test crew. As
engines grew more complex, formal acceptance schedules, instrumented test cells
with recorded data, and structured non-conformance handling became standard.
The Adour's modular construction helped production as well as maintenance,
because a module with a problem could be isolated and corrected without
dismantling the whole engine. Licence production at HAL carried this discipline
into Indian manufacturing, with the quality system and test procedures aligned
to the design authority's requirements.
What a Case A Return Looks Like: Defects Found on the Test
Bed
In Case A, the engine runs, but something
observable is wrong. The defect is usually visible, local and specific.
The investigator starts with a symptom and a location, and the question is: what
went wrong in the build or in the part, and what else could be affected?
Typical triggers:
•
Fuel, oil, hydraulic or bleed
air leaks at joints, unions or housings
•
Oil pressure or scavenge
irregularities
•
Chip detector or filter
contamination
•
Abnormal vibration or noise at
a particular speed
•
Instrumentation or indication
discrepancies
•
Slow or hesitant starts
•
Control or linkage sticking or
hunting
•
Damage found at post-run
inspection
Defects and the areas they usually point to
These are general patterns for production
engines of this class. They are not diagnostic rules, so always follow the
manual.
|
Symptom
on the test bed |
Areas
to investigate |
Common
build-related causes |
|
Oil or fuel leak at a joint |
Sealing faces, seals, unions, torque and
locking |
Damaged or omitted seal, scratched face,
incorrect torque, misalignment |
|
Oil pressure fluctuation |
Pump, filters, scavenge lines, sensors |
Debris or blockage, loose or misrouted
line, sensor fault |
|
Chip detector or filter contamination |
Bearings, gears, accessory gearbox, build
debris |
Cleanliness failure during build, fit or
lubrication issue, early wear |
|
Vibration or noise at a speed |
Rotor balance, fits, bearings, rubs,
mounting |
Imbalance, incorrect clearance, loose
item, misassembly |
|
Indication discrepancy |
Sensor, probe, harness, connector |
Wrong connector mating, chafed harness,
incorrect sensor fit |
|
Hesitant start or control hunting |
Fuel control, linkage, ignition, settings |
Rigging or setting error, linkage
binding, defective accessory |
|
Damage found at post-run inspection |
Area of damage and everything downstream |
Loose hardware, debris left in the
engine, fit problem |
Do not forget the test facility
Before blaming the engine, ask whether the test
facility is behaving correctly. A faulty instrument, a wrong connection to
the engine or a facility-side contamination source can mimic an engine defect.
Checking this early, with documented evidence, is quick and protects the team
from chasing a fault that is not in the engine.
The DI Process After RTS, Step by Step
Your approved procedures set the exact
sequence. What follows is the logic behind them.
Step 1: Stop, secure and record
•
Shut down safely per the test
schedule and record the exact point of the event: speed, power setting, elapsed
time and what the instruments showed.
•
Secure the engine, tag it as under
investigation and prevent any unauthorised work.
•
Retain oil, fuel and filter
samples. Do not clean or wipe away leak trails, scorch marks or deposits until
they are photographed.
•
Preserve the test data,
including raw traces, not only the summary printout.
Step 2: Review test data and build records
This step separates a guided investigation
from a guessing game.
•
Test data: compare the run against the acceptance schedule and against healthy
engines tested earlier. Note when the symptom began and whether it repeats.
•
Build records: module build sheets, torque and clearance records, balancing data,
part and serial numbers, batch numbers, inspection stamps, concessions and
non-conformances raised during build.
•
Change history: any parts replaced during build, any rework, any approved
deviation.
•
Similar events: other engines from the same batch, supplier lot or build team
showing similar symptoms.
A build record is a map. Many failures are
found by asking, what was unusual about how this particular engine was
assembled?
Step 3: Eliminate the test facility and instrumentation.
•
Check calibration status and
recent history of the test cell instrumentation and data system.
•
Verify the engine's own
sensors, probes, and harness connections.
•
Compare against the previous
engine run in the same cell.
Step 4: Non-disruptive investigation
Gather as much information as possible
without dismantling.
•
Visual inspection of the engine
externally: leaks, chafing, loose items and heat marks.
•
Borescope inspection of the
compressor, combustor and turbine sections where relevant.
•
Functional checks of fuel, oil
and control systems using approved equipment.
•
Filter and chip detector
examination.
•
A repeat run only where the
approved procedure allows it and where running will not destroy evidence or
risk damage.
Step 5: Fault isolation
1.
List possible causes from the
symptom.
2.
Rank them by likelihood,
evidence and ease of check.
3.
Change or test one thing at
a time, so cause and effect can be separated.
4.
Record negative results as
carefully as positive ones.
5.
Narrow to an assembly or
component.
For a production engine, always consider
the build-related cause families:
•
Assembly errors: wrong or missing parts, incorrect torque or locking, misrouted
lines, damaged or omitted seals, contamination left inside
•
Fit and clearance problems: rubbing, incorrect running clearances, misalignment, imbalance
•
Part non-conformance: dimensional or material deviations that escaped earlier inspection
•
Accessory and control
faults: pumps, control units or sensors defective
or out of calibration
•
Setting errors: rigging or control settings not matched to the engine
•
Cleanliness issues: foreign material or fluid contamination from the build
Step 6: Disassembly where required
•
Strip only as far as the
evidence demands, following the approved sequence.
•
Keep every part identified,
positioned and traceable. Photograph before and after removal.
•
Retain suspect parts untouched
for examination.
•
Record findings in sequence in
a strip report.
Step 7: Examination and root cause
•
Visual and dimensional
inspection to approved criteria.
•
NDT such as penetrant, magnetic
particle and eddy current, where applicable.
•
Metrology on clearances, fits,
distortion and runout.
•
Laboratory examination if a
part has failed or is suspected of a material or manufacturing defect.
Then ask three questions. What failed
first? What is secondary damage? Why did it happen: design, part,
process, assembly, setting, facility or handling? A fix without a root cause
sends the engine back to the test bed to fail again, and it can send the same
flaw to the next engine on the line.
A Shop-Floor Scenario: The Leak That Was Not the Seal
This is a composite scenario
representing the kind of case engineers meet.
A new engine was stopped during acceptance
because an oil leak appeared at a joint. The simplest response would have been
to replace the seal and retest. The investigator instead reviewed the build
record and noticed that the same joint had been disturbed during a late rework.
Examination showed the sealing face had picked up a minor scratch during that
rework, so a new seal would have leaked too.
Further review found that the rework
instruction made no mention of protecting the face. The fix had three parts:
repair the face under an approved scheme, correct the work instruction, and
brief the shop. The engine went back to the test bed and passed. The line
stopped producing the same scratch.
The lesson: a leak is a symptom. The
question is why the joint failed to seal, and how the problem got past the
earlier inspections.
Rectification, Rebuild and Retest on the Test Bed
Finding the cause is half the job. The
other half is correcting it properly and proving it.
Rectification
•
Rectify only per approved
data: build manual, approved rework or repair schemes, or a concession
authorised by the design authority. A fix not covered by approved data goes to
engineering, not to the next shift.
•
Use correct parts with full
traceability. Replaced parts are tagged, quarantined and examined, not
discarded.
•
Raise a non-conformance
report (NCR) for every defect, with the disposition recorded: use as is
(only where approved), rework, repair, return to supplier or scrap.
•
Mutilate and record scrapped
parts so they cannot re-enter the line.
•
Where a module or assembly is
replaced, follow the associated installation, setting and check procedures.
Rebuild and pre-test checks
•
Reassemble to the approved
standard with independent inspection of critical tasks: locking, rotating
components, control connections, seals and pipework.
•
Verify cleanliness before
closing up, because foreign material introduced during repair is a classic
cause of a second failure.
•
Complete leak and rotation
checks as for a new build.
•
Update the engine's build and
repair record so its history is complete and truthful.
Retest
The engine returns to the test bed. The
scope of the retest depends on what was found and repaired, and is set by the
approved procedure and engineering authority.
•
After a localised repair,
the retest may focus on the affected system and a defined run to prove the fix.
If the repair could influence overall engine behaviour, the engineering
authority may require a wider or full acceptance run.
•
Compare the new run against the
failed run, against the acceptance criteria and against healthy engines.
Improvement alone is not acceptance. The engine must meet the criteria.
•
Carry out a post-test
inspection, including filters, chip detectors and borescope checks as
required.
If it fails again
A second failure is a signal. Either the
root cause was not found, or a second fault was hidden behind the first. The
discipline is to escalate, widen the investigation and resist the temptation to
keep swapping parts. On a production line, repeated RTS of the same engine is a
process problem, and it needs a process answer.
The QA/QC Engineer's Perspective
On a production line, the RTS investigation
is where the quality system proves it works. Every defect that reaches the test
bed is a defect that escaped the earlier inspection stages, and that deserves
as much attention as the repair does.
Escape analysis
For every RTS, ask: at which earlier
stage could this have been caught? A leak caused by a damaged seal should
have been caught at assembly inspection. A dimensional deviation should have
been caught at part inspection. The answer shapes corrective action: sharper
inspection criteria, better tooling, clearer instructions or retraining.
Non-conformance and corrective action
•
Every defect has a recorded NCR
with its disposition.
•
Repeated or similar defects
trigger root cause analysis and corrective and preventive action (CAPA).
•
Supplier-related findings are
fed back through supplier quality channels.
•
Findings are reviewed for batch
implications, including engines already delivered.
Traceability
A good RTS record lets anyone reconstruct
the story: which parts, serial numbers, batches, operators, tools and
inspection stamps. If a particular supplier lot or process step is implicated,
you must be able to identify every engine that used it.
Test cell and tool calibration
The test bed, its instruments, torque
tools, gauges and borescopes must be within calibration and traceable. An
uncalibrated instrument can make a good engine look bad, or a bad engine look
good. I have seen both, and the second is the more dangerous.
Independent inspection and authorisation
•
Critical assembly steps need a
second, independent inspection.
•
Only authorised, trained and
current personnel may inspect, certify and release.
•
Concessions and deviations must
come from the proper authority and be recorded.
Documentation expected in an RTS file
•
Test run report and raw data
for the stopped run
•
Build records and inspection
stamps
•
NCRs and dispositions
•
Strip report, NDT, metrology
and laboratory reports, where applicable
•
Rectification record with parts
used and traceability
•
Retest data and acceptance
sign-off
•
Root cause report and
corrective action record
Audit considerations
Expect auditors to ask:
•
Is every RTS investigated to a
depth proportionate to the defect?
•
Are repeated defects identified
and closed with real corrective action?
•
Are scrapped parts controlled
so they cannot return to the line?
•
Was the retest sufficient to
prove the repair?
•
Is test cell calibration
current, and are changes to test procedures controlled?
•
Do records show who did what,
with what, under what authority?
Why it all matters
A defect found at part inspection costs
little. One found on the test bed costs more. One found at the customer's
squadron costs far more. One found in flight can cost everything. A production
quality system exists to move discovery earlier, and a disciplined RTS process
is its safety net.
The Production Engineer's Perspective, Safety and Future
Trends
Working between the line and the test cell
The production floor and the test bed have
different pressures. The line wants engines moving. The test cell wants engines
ready. When an engine comes back, both sides feel it, and the temptation is to
find someone to blame. The better approach is to treat the defect as shared
information. The test crew's observations (what was heard, felt and seen) often
hold the first clue, and the build team's knowledge of how the engine was
assembled holds the second.
Handling schedule pressure
Delivery commitments are real. But an
engine pushed through investigation too quickly may fail again, or carry a
hidden problem to the customer. A senior engineer protects the depth of the
investigation from schedule pressure and explains the reason clearly to
management.
Human factors on the shop floor
•
Confirmation bias: deciding on the cause early and seeing only what supports it.
•
Anchoring on the first fix: replacing the obvious part, seeing a pass, and closing the case
without a root cause.
•
Blame culture: if people fear punishment, they hide small errors, and that is
where the next failure begins.
•
Fatigue and handover loss: half-finished investigations passed between shifts with incomplete
notes.
•
Normalisation of deviance: accepting a marginal reading because the last few engines also had
it.
•
Repetitive assembly: attention drops when a task is done many times. Independent
inspection and clear checklists protect against that.
Why this matters for flight safety
An engine that stops on the test bed is a
success for the quality system, because the defect was found before it reached
an aircraft. The danger lies in the engine that passes with a hidden problem. A
disciplined RTS process reduces that risk by forcing root cause thinking,
feeding corrective action back to the line, and keeping the acceptance standard
from drifting.
Problems it helps keep out of service
include:
•
Assembly errors leading to
leaks, fire hazard or loss of lubrication
•
Incorrect clearances causing
rubs, vibration or early wear
•
Defective or out-of-tolerance
parts entering the fleet
•
Contamination leading to
premature bearing or accessory damage
•
Control and fuel system faults
causing thrust irregularities
Future trends
•
Data analytics on test
results: comparing each engine's run against the
population to spot subtle drift early.
•
Digital build records and
digital twins: linking assembly measurements to
expected behaviour, so a problem is flagged before the engine reaches the test
bed.
•
Automated inspection and
image analysis as decision support, with qualified
people retaining responsibility.
•
Stronger safety management
and human factors practice in production and test
environments.
•
Closer links between
production quality and in-service reliability data,
so a field pattern can be traced back to the build process behind it.
None of this replaces a skilled person at
the engine asking why.
Advice for Young Engineers
1.
Treat the stopped engine as
evidence from the moment it comes off the run.
2.
Read the build record before
you pick up a tool. The story often starts there.
3.
Rule out the test facility with documented checks before blaming the engine.
4.
Change one thing at a time, so you know what worked.
5.
Do not stop at the first
fix. Ask whether it explains the whole defect.
6.
Ask how it escaped. Every RTS tells you something about the process upstream.
7.
Report your own errors at
once. A reported mistake can be fixed. A hidden one
travels.
8.
Prove the repair on the test
bed and do not assume it.
Frequently Asked Questions (FAQ)
1. What does RTS mean for a newly built
engine? RTS means Return to Shop. A newly built
engine that is stopped or not accepted on the factory test bed is returned to
the shop for investigation and repair, then retested.
2. What does DI mean here? DI means Defect Investigation: the structured process of finding
the root cause of a defect, correcting it with approved data and proving the
correction.
3. Is a defect on the test bed unusual? Not for a complex engine, and a sound quality system expects it.
What matters is that every defect is investigated properly and that the lessons
reach the production process.
4. Why check the test facility first? Because instrumentation, connections and facility contamination can
mimic an engine defect. Eliminating them early avoids unnecessary teardown.
5. Why are build records so important in
a DI? They show how the engine was actually
assembled: parts, settings, clearances and any rework. Many production defects
trace back to something unusual recorded there.
6. Does every returned engine need a
full strip? No. The depth depends on the evidence
and the approved procedure. A localised defect often allows targeted work,
extended if findings require it.
7. How much retest is needed after
repair? The approved procedure and engineering
authority decide. A localised repair may need a focused run, while a repair
that could affect overall behaviour may need a full acceptance run.
8. What happens to replaced parts? They are tagged, quarantined and examined, then dispositioned under
the non-conformance process. Scrap parts are mutilated and recorded.
9. What is an escape analysis? It is the review of which earlier inspection stage should have
caught the defect, so that the process can be improved.
10. Can this article be used as an
official procedure? No. It is educational. Official
procedures, limits and criteria come from approved manuals and quality
procedures.
Conclusion: The Test Bed Tells the Truth
The test bed is honest. It does not care
about schedules, signatures or assumptions. It shows what the engine really is.
When a new Adour Mk 811 stops there with a defect, the factory has been given
something valuable: a chance to find a problem while it is still cheap, still
controlled and still invisible to the pilot.
A Defect Investigation after RTS turns that
chance into learning. The work is to find the true cause behind a visible
symptom, to check what else it touched, to correct it under approved data, to
prove it on the bed and to feed the lesson back to the line.
Technology will bring better data and
sharper tools. But the foundation remains a disciplined, honest engineer who
will not sign for what has not been proved.
Respect the engine, respect the evidence,
and respect the person who will fly it.
Disclaimer: This article is for
educational and informational purposes only. It does not replace the approved
Engine Build Manual, Test Schedule, Overhaul Manual, airworthiness requirements
or the quality procedures of the design authority and manufacturer.
Comments
Post a Comment