I was reading another thread about testing compression and leakdown on a 993TT. There was some discussion that 135 psi was too low. Which got me to thinking, an always dangerous condition. The 993TT has a compression ratio of 8.0:1. Yes, its crazy low, which probably explains our very crappy fuel mileage. But here's where I need help. If normal atmospheric pressure is 14.7 psi, and you compress it 8.0 times, then I would expect a maximum compression pressure of 8.0 x 14.7 = 117.6 psi. And that's assuming ideal conditions.

So where have I gone wrong? No, scratch that. So what's wrong with my math?

 

I thought it was that 35psi was too low and that it should have been at 100psi.

 

Yes, lots of discussion about the leakdown test and the pressures used. I'm not talking about that.

In fact, my arithmetic seems faulty, if you had a compression ratio of 2.0:1, then you would expect to read 14.7 on the compression gauge, not 29.4. So... according to my model, the max pressure for a 993 TT is 7.0 x 14.7 = 102.9 psi. This is clearly incorrect. Why?

 

thats the leakdown input pressure provided by the external air compressor.

 

Compression of air within the cylinder is considered adiabatic - the rise in temperature due to compression causes additional pressure increase beyond that predicted by the CR.

 

Now this sounds right. I'm guessing that there is a formula that WILL predict the max compression pressure based upon the compression ratio plus the added pressure from the heat. And this has probably been made into a simple look-up table for mechanics to use when testing engines. Thanks for this, I'll sleep better tonight.

 

Your equation assumes that temperature doesn't change. But anyone who has put his thumb over the spark plug hole and had the engine cranked knows that the air is bloody hot. When a piston compresses the gas in a cylinder, it is usually assumed to be an isentropic compression meaning no heat is transferred and is reversible (i.e. no change in entropy). The pressure ratio for an isentropic compression is given by:

p2 / p1 = (v1 / v2) ^ (gamma) or p2 = p1* (v1 / v2) ^ (gamma)

v1/v2= compression ratio = 8, and gamma for air is 1.4

p2 = 270 psia = 255.3 psig

much higher than we see. But with a cold engine, the heat transfer assumption is probably not valid. Also, with valve overlap, some of the gas in the cylinder escapes. Both of these factors would yield lower than ideal pressure increase. So the pressure is going to be somewhere between the constant temperature assumption 102.6 psig and the isentropic assumption 255 psig. 135 psig is in that range.

 

OK Vince. Now you're just showing off. Actually, That makes sense, even if I didn't do the math. Thanks for the thermodynamics lesson. Rennlist is pretty amazing.

 

Yeah, I really should stay out of this thread.

 

You just hadda ask, din't you?

As far as the numbers go, there are enough differences between the gauge accuracies, engine temperature, air temperature and just plain technique as not to worry too much about the absolute numbers unless they area grossly out of range. The delta is important though. Same goes for the leak down gauges, I am looking for the difference between the gauge detached and attached to the cylinder. The delta is the leak-down contributed by the engine.

I have a few compression gauges, and one of my older ones always reads a bit low, but it's repeatability is excellent.

I have not seen any calibration devices for these gauges (unlike torque wrenches where you can buy strain gauges now). I sometimes use compressed air set with a gauge and regulator to test my compression gauge as well as comparing two different gauges.

Cheers,

Mike

 

X2

That was strong.

 

Somewhere around here, I have a matrix table from one of the big supercharger rebuilders showing boost pressure on one axis, mechanical CR on the other to display actual CR in the cylinders.

Quite illuminating, to say the very least.

Let me if I can find it over the weekend and if so, I'll post it here.

 

Vince, eve though my eyes are now bleeding I must congratulate you, that was the best answer to this reoccurring post I have ever seen!

 

At least that is how it is explained by Wiki.

I came across these CR and Effective CR (turbo) calculators.

 

A little jealous Alex?