Leak Down Testing

[steve_wood]

Riding season is pretty much over here in Winnipeg. My engine is a fresh rebuild (less than an hour operating time), so I though it might be interesting to run a leak down test to see how the cylinders were working.

Strange results....

Here's what I did (for both front and rear cylinders):

- select a position to test (i.e. TDC)
- turn the supply pressure up to 50 PSI
- record the pressure
- turn the pressure back to 0
- repeat for several positions starting from TDC going to the bottom of the stroke
(Positions were: TDC, TDC + 30 degrees, TDC+60, TDC+90, TDC+120, TDC+150, BDC)

For the most part, I would get 88% to 96% which is pretty good.

HOWEVER.... a couple of times, when I turned the supply pressure up to 50PSI, the measured pressure would stay at 20PSI for about five seconds, and then suddenly climb back up to as high as 46PSI. (See attached for the data). And actually, there was one measurement where the value stayed at 20PSI for at least 20 seconds. I gave up on any "rebound" at that point.

What could be causing that delayed reaction? Could it be that the rings are not completely worn in and it takes some time for small amounts of oil to settle in to the rings to create the the seal?

Could it be something else is going on?

Anybody got a clue?

[james]

Any leakdown would mean air is escaping.
While conducting your tests is there air coming out of the
intake or carb, the exhaust pipe or the breather?
Valves not completely closed during test?
Valve adjustment slightly tightly?

jim

[steve_wood]

The only leaking is coming from the crankcase which means it's getting past the rings.

[james]

Same for each cylinder or does it differ?
I would run it for 500 miles then conduct your test
again. I think your doing this test too soon.
You could be right about the rings seating.
See if you get any other opinions.

JIm

[Robert Luland]

Steve, plain and simply. You need another motorcycle to keep ya occupied in those cold northern months. Cost wise it beats the shit out of making babies. Bob L

[ncmc47]

In aviation we use 80psi of air pressure at tdc on compression stroke. That way valves are closed. If you listen at the carb and have a leak, the intake is leaking. A leak at the exhaust means a leak through the exhaust valve. A leak through the crankcase indicates the rings are leaking. The test should be run on a warm motor, that way the rings have expanded and are less likely to show a leak. Generally the sprocket shaft needs to be held to keep the motor on top dead center. A loss of less than 80 percent indicates a serviceable motor, so 46psi @ 50 psi is good especially if done on a cold motor. You might try again after a warming of the engine.
Hope that helps.
Steve

[Bosheff]

How did ya keep the flywheels from rollin when ya poured the compressed air to the cylinder/s? This will make for false /inaccurate readings if the piston/s aren't locked at TDC when performing the leak down. In reality the piston/s really don't have to be at TDC, but both valves must be seated for accurate test results....bosheff

[steve_wood]

A bit of amplifying info....

- The pushrods were removed to ensure the valves were always closed
- I used only 50PSI
- I measured at several points in the stroke to look for *differences*. The whole point was to see if the cylinders were sealing better at the top than at the bottom
- The transmission was in 4th with the rear wheel on the bench

Since I was only using 50PSI, and the bike was in gear, I was able to hold the crankshaft in place with a 1/2 drive power bar on the engine sprocket.

The weird leakages that occured were definitely at the rings - you could feel and hear the air coming out of the flywheel inspection hole.

Still think it's weird that the rings would somehow magically "seal". What could possibly cause that?

[rainman]

Steve,

In reading your post, i have a sneaking suspicion your rig is not quite right. A standard diff. comp. rig uses a known orifice between two gages at 80psi inlet pressure. It is that orifice (if I remember back to my tech days, its 0.060" in diameter) that makes the difference between unreliable results and reliable results between the two gages.

I have linked some info from AC43.13, the acceptable methods and practices handbook from the FAA. Its old, outdated, and full of information that relates to our old, outdated hobby. Paragraph 8-14 has some detail on the rig and testing process.

My 0.02 cents is thus (humbly and respectably of course): while your intentions are noble, attempting a test with inputs and variables that are different from standard practices may just lead you down a path where your results cannot be fully validated.
http://rgl.faa.gov/Regulatory_and_Guida ... r%2008.pdf


By the way (off topic, but related to the FAA AC reference book), to all- to put another piece of good info out there about torque adapters and calculations and what not, look in chapter 7 of AC43.13. Specifically Figure7-2.

http://rgl.faa.gov/Regulatory_and_Guida ... r%2007.pdf

....thanks for allowing me to pontificate.....