Confidential Human Factors
Incident Reporting Programme
The Charity
Aviation
Maritime
GA1375
Initial Report
I had recently carried out a 75-hour maintenance cycle on my Jabiru 2200 engine. Whilst doing so, I noticed that the fuel pump was leaking oil from around it, the pump itself is driven by a pushrod from the engine, lubricated by the sump oil. I sourced some 1mm gasket paper, 2x pieces being required, also RTV (Room Temperature Vulcanising) gasket sealer.
The pump was removed, very simple, and new gasket paper cut out to match the shape of the old gaskets which were in poor condition. The new gaskets were cut out and made using a wad-cutter tool to make perfect holes, gaskets fitted, and screw/bolts torqued up. Having waited 24 hours for the gasket sealer to cure, the engine was run up, no problems were encountered, no oil leaks were found.
The following day all normal checks were run prior to take off, power checks etc to 2000rpm, including a long taxi over grass to the in-use runway. Having called taking-off and applied power, I was passing through 200ft when the engine started running rough. I put the nose down and pulled the power back, intending to land back on, which I did. However, I was surprised when having put the nose down the engine ran smoothly again and was fine all the way back to the hanger. Whilst taxying, I saw that the fuel pressure gauge was reading ZERO, which was surprising as the engine was running fine.
I did note however that the last thing I had touched was the fuel pump. I realised that my fuel is overhead in wing tanks, and that some pressure is gained from gravity. Further inspection of the fuel pump revealed that the push rod only moves 1-2 mm, and that, with thick gasket paper, I suspect that the pushrod was not making contact with the fuel pump, therefore no pressure. New thin gasket paper was used, 1/2 mm, and fuel pressure was restored in the gauge, and of course now no rough-running engine. Looking at the fuel pressure was not part of my take-off checks, (it is now) with the gauge being on the far-right passenger side.
I do not believe I could have foreseen this issue, perhaps someone will say different? There was no indication of a fuel-flow problem from sound and power checks. I have, as stated, now included a visual check of the fuel pressure gauge in my checklist but perhaps someone out there without a pressure gauge may well change the gasket on their engine and not be aware of the pitfall. I have now received some manufacturer’s gaskets, which look to the human eye to be a bit thinner. I have not fitted them as 1/2 mm gasket paper seems fine.
CHIRP Comment
Making such things to pattern should only be done as a last resort (such practice is completely prohibited in the commercial world) because of the uncertainty in ensuring correct tolerances. Even if the reporter had chosen to use a micrometer to establish gasket thickness, the old gasket will have been crushed on initial installation and so any measurement would have given an inaccurate result (below thickness in this case).
Fundamentally, the lesson is to use manufacturer-approved materials whenever possible, but there ought to be some guidance as to what thickness gasket is required for those who have the skills to manufacture such materials themselves because it’s quite common for people to do so. We understand that Maintenance Manuals for GA aircraft can sometimes be rather lightweight in detail, and often allow the use of standard engineering practices, but this particular item should be in the manual due to the obvious close tolerances.
Whilst we understand that the self-manufactured thinner gasket seems to work fine, we’d recommend that the reporter replaces the gasket with the manufacturer’s official versions; you just don’t know how close you might be to the tolerance limit with the self-made version. More importantly, it would also be apposite to check the fuel pump push-rod at the same time to ensure it is not worn or deformed because this could also be a problem rather than just the thickness of the gasket paper alone; the pump itself may well be at the limit of tolerances. We recommended that the reporter contact the LAA/BMAA as applicable to their aircraft type for advice, and it may well be worth them engaging an engineer to check the fuel pump for such wear, or just get a new pump if there’s any doubt.
On a handling note, well done to the reporter for reacting appropriately when encountering the rough-running engine, and for not subsequently assuming all was ok and trying to resume the climb once the rough-running had stopped in the descent. Reduced-power emergencies on climb-out can be real killers because, instead of making a decision to land and enacting that decision, people sometimes try to nurse the aircraft along and can end up in situations that they might not be able to recover. The reporter’s comments about checking fuel pressure before take-off speak for themselves; Ts & Ps don’t just apply to oil pressure; if you have a gauge, check the fuel pressure too.
Key Issues relating to this report
Dirty Dozen Human Factors
The following ‘Dirty Dozen’ Human Factors elements were a key part of the CHIRP discussions about this report and are intended to provide food for thought when considering aspects that might be pertinent in similar circumstances.
- Resources – self-manufactured gasket vs OEM version.
- Knowledge – did not check what the required gasket thickness was.
- Complacency – did not check fuel pressure before take-off.