Several months ago I was combining the delivering of a DH Beaver (aka The Aerial 1 Ton Moving Truck) to have her floats installed with commuting to work. The early evening light was gorgeous as I started cranking the faithful Pratt, and I was truly looking forward to the short flight across the Gulf Islands. As the engine fired on her first few cylinders, I glanced down at the oil pressure gauge, waiting for the 50 psi to push the propeller pitch control to fine pitch. Ten seconds or so later, with all cylinders firing and now at a smooth 400 RPM idle, there was no oil pressure. None, Nada, Zilch. I quickly shut down the engine with the mags, not really believing what I was seeing. After a few minutes of careful thought, I convinced myself that it had to be a gauge problem. To confirm, I set the propeller pitch control to fine pitch and started the engine again, this time with two buddies watching for the blades to change pitch as the engine oil reached the propeller – if the governor was getting oil, then so too would the rest of the engine and I could get the gauge replaced when the floats were installed. After another 5 to 10 seconds, they were both giving me the thumbs down and I again hurriedly shut the engine down. Now I knew I had a real problem. And I was wondering what kind of an idiot tests a three hundred dollar gauge with a forty thousand dollar engine.
The most usual cause of a lack of oil pressure in these engines is a piece of crud stuck under the pressure relief valve so the valve was pulled , inspected, and cleaned but nothing untoward was found. Next I connected an electric pressure pump and oil supply to a port in the engine oil system. The gauge now read 50 psi and this not only confirmed that the gauge was working, but also pre-oiled the engines bearings again. At this point I also knew that the problem lay with the oil supply or in the pump itself. Now due to the oil tank being located in the cockpit, the Beaver has an emergency fuel and oil shut off valve at the firewall. The first thing was to check that this valve was open and free, and that nothing was obstructing the line supplying the engine driven pump. The valve was open and the supply line clear, but there was little to no oil in the line. This was the first clue. Time to exercise the speed dial. Now I have used engines built by Tulsa, Covington, and Aero-Recip in Canada and I have a high opinion of all three shops. So just to be sure I covered all the bases, I called all of them describing my tale of woe. Rex at Tulsa, Blaine at Covington, and Wayne at Aero- Recip all independently came up with the same possible diagnosis. It was interesting enough as well a highly relevant to many of the 450 Stearmans, to cause me to write this article.
The oil pump in the R-985 is really three pumps in one stacked on top of each other and running on common shafts (Photo 1) – the lower portion is the pressure pump (Photo 2), the middle section a small scavenge pump for the rocker oil system (Photo 3), and the upper section is a large scavenge pump for the crankcase. Each section comprises of two gears, one driven by a keyway on the drive shaft, and one free turning on an idler shaft. The oil is pumped by virtue of the very close fit of the gears with the housing, and not between the gears themselves. In other words, like most hydraulic pumps, everything is a “kiss me” fit. The first thought was that something had become lodged between the gears and locked them up, causing the keyway on the drive shaft to shear. Removal and disassembly showed this not to be the case and the pump turned over like a fine Swiss watch. The lesson here is that if something falls into the oil tank, however small, it must be removed before flight. Additionally, even if you have a spin on oil filter, leave the engine oil screen in place as this will hopefully protect the oil pump from a debris ingestion induced failure. But we still didn’t know why the engine had no oil pressure. Speed dial again twice to Oklahoma and once to Winnipeg. Rex, Blaine, and Wayne were initially all puzzled until I mentioned that the engine was installed on a DH Beaver. “O’h, it has lost its’ prime. That sometimes happens, especially after an oil change,” was the universal reply.
It turns out that with the DH Beaver the oil level in the tank is about 5” lower than the pump inlet. And gear pumps are great at pumping fluid, but terrible at sucking it. Any air lock in the line (Remember there was little to no oil in it when I removed it for inspection?) will prevent the pump from picking up. It was recommended that the line be loosened at the pump, the tank vent sealed, and a few pounds of air applied to the tank to force oil to the pump. Once oil was evident at the pump, the engine could be turned over by hand to fill the pressure section of the pump. I would have been happy with this had we just changed oil and introduced air into the system, but this had happened out of the blue while the airplane sat parked for a month. Something else was at play here. Blaine Abbott from Covington supplied the missing piece of the puzzle.
It turns out that the pressure and scavenge sections of the pump are separated by oil seals around the shafts. They are nothing more than split leather packing rings that appear to have been manufactured by the local town saddlemaker or cobbler . If these wear, they can allow air in the crankcase to leak past them when parked, and that then in turn allows the oil in the pressure system to drain back into the tank creating an airlock. Obviously, if the oil tank is located above the engine this isn’t an issue, but in the Beaver, and even more so with a 450 Stearman equipped with an oil tank in the baggage compartment, this is the gotcha. Blaine also mentioned that they have seen enough problems with these seals that Covington has developed a modern replacement seal that should hopefully avoid this issue in the future. I am pleased to report that after fitting an overhauled oil pump the engine made instant oil pressure on startup and all is well. So what are the takeaways if you have a radial engine with an oil tank located below the oil pump:
- Checking for engine oil pressure at startup is good airmanship. If the installation has the oil tank located below the oil pump, it is critical airmanship.
- When changing oil, it is highly advisable to bleed the oil system. After this debacle, I have installed a Curtiss drain valve on my installation for just this purpose (Photo 5). It was also mentioned by Rex at Tulsa that it is best to pour the first gallon or so of oil in slowly to avoid creating an air bubble in the line.
- The Pratt & Whitney engine operator’s manual allows for 30 seconds of operation at less than 500 RPM without oil pressure. All three overhaul shops were of the opinion that these parameters were reasonable, but anything past that was risking damage to the master rod bearing. I mention this because after changing oil some local operators are revving the engine on startup to “suck” the airlock through the pump. But at what cost to the rest of the engine?
- The fit of the keyways on the oil pump shafts, the end clearance of the oil pump gears, and the fit of the metal seal ring in the oil pumps are very closely toleranced. They are all supplied in various standard and oversize dimensions to accommodate in service wear. For this reason metal parts between different oil pumps should not be interchanged.
- If this oil pressure issue happens unrelated to an engine oil change as it did to us, it may be worth changing the oil pump while safely and conveniently at home base. (Photo 6)
- Lastly I would like to express my sincere thanks to Rex Thompson at Tulsa Engines, Blaine Abbott at Covington, and Wayne Cathers at Aero-Recip. All three shops build great engines and all three gentlemen are incredibly generous with their time and expertise when you have a problem, whether or not you bought the engine from them. Without their help I would still be lying under a Beaver, covered in black oil, and scratching my head!