Compression testing

phantomphixer · Post by **phantomphixer** » Thu May 06, 2004 10:41 pm

Need someone to explain in layman's terms how exactly a compression tester actually works? As in why the 80psi indication remains steady and does not move while the moving the prop. And saying that a miracle occurs does not cut it.

I've done numerous leak downs but I can't explain the working of the tester. Any rocket scientists out there?

Thanx

Post by **GAHorn** » Fri May 07, 2004 5:12 pm

The compression tester is made up of two pressure gauges connected by a calibrated orifice. One gauge measures regulated "input" or supply air pressure, while the other measures the non-regulated pressure downstream of the orifice as it is contained/released by the piston/cylinder assembly.

phantomphixer · Post by **phantomphixer** » Fri May 07, 2004 6:16 pm

George, Is there a check valve between the two gauges or just the orifice(sp)? It seems almost too simple.

55B N3585C s/n 26659

Post by **GAHorn** » Tue May 11, 2004 1:51 am

Why would there need to be a check valve? The stationary cylinder will never produce more pressure or volume than the 80 psi supply side.

All the downstream gauge does is measure the rate of leak-down from the supplied 80 psi. This is not rocket science, it's merely an estimation of relative health of a cylinder, rings, valves, etc. It's not even an "absolute" measurement. It's merely a relative one. If we didn't have the supply-side gauge there to accurately measure the input air (80 psi) then we'd simply have even less comparison to make. (But we'd still have the indicators of failure such as hissing at an exhaust valve, or crankcase (rings) etc. The problem is that without a regulated supply pressure (the 80 psi again) how would you be assured that you'd not just simply overpowered the valve springs and blown healthy valves to an open position, or overpowered the rings and blown them off the pistons? After all, if we simply used a compressed bottle of nitrogen at 2,000 psi I imagine we could develop a leak somewhere! No?)

lcranton · Post by **lcranton** » Thu May 13, 2004 4:41 am

George, Maybe I missed somthing in this discussion, but how would increasing the pressure in the cylinder blow open the valves? I would think it would seat them tighter. I am not sure but would not the cylinder pressure during combustion be quite a bit higher than 1000 to 2000 PSI. Would you happen to know what the typical cylinder pressure is in the O300 at full rated RPM and max MP?

Thanks

Post by **GAHorn** » Thu May 13, 2004 10:19 pm

You didn't miss anything especially. It's a little known fact that all valves will leak (if only microscopically/indetectably) and if sufficient pressure is applied over a period of time enough will slip between a valve and it's seat to break the spring tension and lift the valve. Will it happen in the course of ordinary combustion? No. I'm talking about the relatively long time-frame of a compression test not the relatively instantaneous time frame of a running engine.
Believe it or not, I actually do have a graph showing the relative combustion pressures at the moment of peak combustion. I't somewhere on this hard drive and it's under an article about leaning. Since it's only a supportive piece of information I cannot readily locate it but I'll keep it in mind in case I run across it. If memory serves peak combustion pressures are in the order (depending upon a lot of factors such as engine size, speed, compresison ratios, etc.) of 1700-2800 psi.

zero.one.victor · Post by **zero.one.victor** » Fri May 14, 2004 5:34 am

gahorn wrote:You didn't miss anything especially. It's a little known fact that all valves will leak (if only microscopically/indetectably) and if sufficient pressure is applied over a period of time enough will slip between a valve and it's seat to break the spring tension and lift the valve................

Say what? I have to disagree with this one,George. Remember,the however-many-psi pressure inside the combustion chamber is pushing against the face of the valve. I don't think that a miniscule leak past the valve could result in enough pressure behind the valve to open it against that pressure. Don't forget,at the same time the valve springs are also doing their best to keep the valve from opening.

Eric

Post by **GAHorn** » Fri May 14, 2004 4:04 pm

Well, Eric, I realize that it's a rare thing that you and I might ever differ on things

But,...it's true. I've witnessed it on two engines that were deliberately tested for it.
In normal operations, the surface area of the valve is so much greater than the bearing surface of the valve against it's seat ...that the phenomenon (a technical term -

) is not observed. But imagine this: If sufficient air pressure is applied to the system, a miniscule leak will occur at the valve/seat interface. When that pressure is increased it progresses eventually to overcome the relatively static pressure against the valve head area and the ability of the valve springs to prevent lifting of the valve. As the air pressure increases it causes a widening rift all along the valve seat and the valve will then suddenly pop open. It's more commonly observed on engines with broken/weak valve springs, and rarely with loose valve seats.
The pressure required to do this on a healthy engine is enormously beyond the 80 psi we use to perform compression tests. But it's a real thing. (In fact, if one wishes to take the illustration to it's ultimate conclusion,...if surface area of the valve head and the air pressure against it is a guarantor of valve seating, then in theory there'd be no need for valve springs except to initiate closing of the valve. But, in the extreme, proof will be found in the fact that an already closed, unsprung valve, with pressure applied to the cylinder, will not long remain closed. As pressure is applied to the cylinder, the valve will indeed lift at some point. Perhaps with only 3 or 4 lbs of pressure, but it will lift without the benefit of valve springs. The presence of valve springs only increase the air pressure required to lift the valve. Carrying the experiment further with ever-increasing spring tensions will in each case continue to increase the air pressure required to lift the valve.)
Admittedly, this is an aberration and not within the purview of ordinary compression testing. It will not ordinarily ever be observed during compression testing, and perhaps I shouldn't have complicated the discussion with it. I was just using it as an extreme example of one reason we use a calibrated input air pressure during compression testing.
Better reasons are: To have a comparison by which we can easily derive a percentage value. 80 psi is a level at which a human can reasonably still hold a prop in position during the test. And it's become a de facto standard in the industry and allows ready comparison to other engine situations.
Sorry if I fogged up the glass with this one.

Post by **Bruce Fenstermacher** » Fri May 14, 2004 9:07 pm

eric I'm with George onthis one. I've rapped on the valve spring with a soft mallet will it is under pressure in order to unstick the valve keepers and had the valve blow open rather easily. I could see it happening on its own with enough pressure.