Engine Testing Problems
Why you shouldn’t get out of bed some days
There are days when testing goes really well and the time spent doing the actual test is minimal. Then there are those days where almost nothing goes right. Here’s an example of the latter. It was the end of an extensive testing program as contracted by an aerospace company on its new two-stroke diesel engine intended for a small unmanned aerial vehicle (UAV). The manufacturer wanted just one more test of a new tuned expansion-chamber muffler that it hoped would boost the engine’s horsepower a bit. It was a simple test, really: Just start the engine, adjust it to maximum rpm, and record the rpm with a specific propeller. What followed was almost comical.
I allowed the engine to warm up after a normal startup. As I adjusted the engine’s compression and leaned in the primary needle valve, it backfired, throwing the drive nut, washer, propeller, and shaft extension. Oh, yes, the engine continued to run without a load. A “shaft run” is always a stressful occurrence, but fortunately, I was able to pinch off the fuel line and stop the uncontrolled increase in rpm before something catastrophic happened to the engine’s components. These things happen from time to time—that’s why I always wear safety glasses and stand behind the engine when making adjustments. As I began picking up all the scattered parts, I found everything but the specially fabricated aluminum shaft extension. This is no small piece. It’s 1 1/8 inches in diameter and 1 inch long, with a 3/8-inch-diameter hole through the middle. After unsuccessfully looking for it for about 10 minutes, I enlisted the help of my wife, Carolyn. We looked everywhere: the grass on both sides of the test area, inside the shop (the door had been open), behind all the dusty machines and benches. We retrieved a stepladder and looked in the roof gutters behind the test area, with no luck. I even resorted to using a lawn rake in all the grassy areas, while Carolyn carefully sifted through her flower garden next to the shop. After more than an hour, we still hadn’t found the thing, and I was starting to contemplate the possibility that I was going to have to make another piece. As I was walking the stepladder back to its storage place, I caught a reflection from something shiny in the grass. There it was, more than 50 feet away from the test stand, in the opposite direction that I had thought it had traveled. After a bit of touchup with a fine-tooth file, the entire prop assembly was again fastened to the engine for another run.
Everything started off well, but suddenly the engine stopped.
An inspection revealed that the throttle barrel had exited the end of its carburetor housing.
The screw that travels in the helical groove of the barrel had apparently been loosened by the shaking shaft run and was now missing. After spending 20 minutes looking through my supply of metric machine screws for the correct thread size (to fabricate a new barrel screw), I decided to tap the carb housing for an S.A.E. thread (10-32 worked well). After turning the appropriate follower end on the new screw, the barrel assembly and carb housing were finally reunited, no worse for wear.
There’s a special O-ring on the bottom of the carburetor housing that mates with the engine’s crankcase. You know the one—it’s used to ensure that air leaks won’t occur and alter the mixture during engine operation.
Yes, it’s the same O-ring that slipped from my fingers as I was attempting to reattach the carb to the engine. No problem, though: I saw exactly where it fell into the grass at the edge of the test stand—except that I still hadn’t found the thing after 10 minutes! Rather than spend any more time looking, I retrieved a similar piece from my stash of O-rings in my basement shop. Total time for the O-ring extravaganza: 35 minutes.
After restarting the engine, the tuning procedure was well underway when the engine again suddenly quit. This time, it was out of fuel. Damn! Is there anything else that could go wrong? Yes! A new batch of fuel needed to be mixed. Trying to save time, I got a mouthful of kerosene while priming the siphon tube; that’s something I normally avoid by incorporating a simple hand pump—which today I couldn’t find! By wasting the better part of a glass of lemonade, I was able to flush out most of the bad taste.
The next run, although successful, proved that the new pipe didn’t offer a power gain. So goes the testing process. Total time: 3 1/2 hours.