Compression testing

[hutchman]

I need to run a compression test on my AMC 258 i6 / 4.2l , so today I was searching for acceptable values from a test. What I found was that it should be 120# - AMC 150 # with all cylinders within a 5% - 10% max difference. These numbers seemed good to me and were within the range of what I remembered, but........

I started thinking about just what goes into these readings. Common sense would tell me that since standard atmospheric pressure is 14.7 PSIG, that a 10.0/1 compression motor should have 147 psig if it was a perfect cylinder.

However, in this case common sense would be wrong. I started searching the internet for information on what variables influence the readings of a compression test. And what I found illustrated how much I was forgetting.

To get a true compression reading, you need to have the engine up to operating temperature to ensure that the rings, pistons, and valves have expanded and are where they fall in operation.

All spark plugs need to be removed from the cylinder head with the ignition system grounded to prevent damage to the control module.

The throttle needs to be wide open to allow unfettered air flow to the cylinder.

All of these things need to be in place in order to have a number that compares to the standard.

So now all of these are in place, what variables will effect the compression readings? Remember that I said common sense told me that if we have a compression ratio of 10/1 that we ought to have a 147 psig reading on the gauge? Well, that is just plain wrong.

The gas laws tell us that when you compress a gas it heats up and the pressure it exerts on the containing vessel increases. This has a huge effect on the pressure reading in the cylinder.......I don't remember how much but it seems to me it increased the gauge reading about 2 times what we would expect. But this is not the only thing that effects the indicated pressure.

As the engine cranks, the intake and exhaust valve open and close depending on which stroke the piston is on.....intake, compression, power, and exhaust. But what I was not thinking about was the fact that the intake and exhaust valves are sometimes open at the simultaneously during the 4 stroke process. This fact by itself blows my common sense out of the water. The compression reading will vary greatly based on the camshaft profile in the engine. A stock cam with short overlap will have higher cranking pressure than a long overlap, high rpm, cam will have. So the camshaft grind will effect greatly what is good or bad for a different engine. As it turns out, this effect almost completely reduces the impact of the change in pressure from heating due to the gas laws......almost but not quite.

So where does this leave us? For a brand new and leak free engine with a short overlap cam, we should be around 160#, maybe 170#.......from the best of my memory. So for my motor with 95,000 miles on it, I believe the 120# - AMC 150 # to be an indication of a good motor.

Here is a link to a great technical article on compression testing and how to use the results.......

Puma Race Engines Technical Guide - Compression Tests

Good reading if you are interested in the technical aspect of a compression test. If not then, I believe the numbers above are good.

 

[BajaEdition]

hutch
I have read a lot of your post and thought you had it together,
but here, I am afraid someone has sold you a bill of goods.
one thing you are not doing is a temperature evaluation when taking the the measurements. second you are not even considering that the gauge you are using is designed to read Zero at normal atmospheric pressure at sea level, and that the colder it is the more the pressure is, and at an engine with combustion chambers working at 1500 degrees this pressure changes dramatically. NOT ONLY THAT, but if the gauge reads zero at 14.65 the you double that each time you double the pressure, by 10 to 1 you have 10 times the pressure which brings a curve into the formula, not just a 10x equation.
so if 1 to 1 reads zero what would 2 to 1 read?
there fore, you cannot use atmospheric pressure to determine the cylinder pressure but instead you need to use the amount of air compressed.

 

[hutchman]

hutch
I have read a lot of your post and thought you had it together,
but here, I am afraid someone has sold you a bill of goods.

I haven't bought anything yet, but I am struggling to understand where these compression numbers come from....!

....one thing you are not doing is a temperature evaluation when taking the the measurements.

I don't understand this point entirely......unless you are referring to Standard Temperature and Pressure. But I do not think you have to do a temperature evaluation to have an accurate result. You only need to have the engine warm. As long as this is the way the standard was developed then your results are accurate to the standard.......and should be proportional to STP.......?

....second you are not even considering that the gauge you are using is designed to read Zero at normal atmospheric pressure at sea level, and that the colder it is the more the pressure is,

and at an engine with combustion chambers working at 1500 degrees this pressure changes dramatically.

NOT ONLY THAT, but if the gauge reads zero at 14.65 the you double that each time you double the pressure, by 10 to 1 you have 10 times the pressure which brings a curve into the formula, not just a 10x equation.
so if 1 to 1 reads zero what would 2 to 1 read?
there fore, you cannot use atmospheric pressure to determine the cylinder pressure but instead you need to use the amount of air compressed.

I did consider the difference between gauge pressure and absolute pressure and that it should be calibrated to be accurate at 70 degress F and 14.7 PSIA. So when you screw it into the head and measure the chamber pressure, you are getting inaccuracies due to the heat of the chamber. But this will not be 1500 degrees, but somewhere above atmospheric temp + the heat addition from compression + the intrinsic heat added from the block, head, and piston. But again, as long as the testing conditions are the same as the standard then the results are accurate to the original standard and proportional to STP.........I think.

As to the issue of calibration at gauge or absolute pressure........

Very clearly, a complete vacuum in the cylinder would result in 0 psia or 0# absolute. But we know that with the the intake and exhaust valve open during overlap, we have atmospheric pressure in the cylinder or 14.7 psia and 0 psig or 0# gauge pressure. So a reading of AMC 150 psig on the compression gauge is equal to 164.7 absolute. So it doesn't matter whether we use gauge pressure or absolute as long as the gauge we are using is the same as the standard we compare our results to. In this case all compression gauges read in psig and not psia, so we are OK.

I think though that I have the same thoughts/doubts about the conversion from atmospheric pressure to compression pressure. And this is the real core of my question......where does this compression value come from.

We know from the gas laws that for an ideal gas at a constant temperature that the product of the pressure and volume is always constant.......

P1(V1) = P2(V2)

From good ole' Boyle we can see that if we have 14.7 psia with the piston at BDC and 10/1 compression (10 volumes), we would have 147 psia at TDC with the volume of the area above the piston now only 10% of what it was or 1 volume, which gives the following:

14.7(10) = 147(1)

So assuming no change in temp, we would get 147 absolute at TDC. And we also know that an instrument calibrated in gauge pressure would indicate 132.3 as it's zero point is 14.7 psia or absolute. With a pressure instrument calibrated in gauge pressure your indication is = psia - 14.7 psia.

Back to real world......We do not have an ideal gas at constant temp. The temperature of the gas increases as it is compressed and it picks up temperature from the block, head, and piston. So the resulting pressure is much higher than just the pressure from the change in compression volume. According to Puma, we would expect about 300# from the change in volume and heat added.

This was where the valve overlap affects the compression. Intake and exhaust overlap reduce the effective volume compression effect and temperature effect. Puma states that you can expect a compression pressure value of 17 - 20 times the value of the compression ratio after factoring in all of the factors involved.....

So for a brand new Jeep AMC 258 i6 / 4.2l , we should expect a compression pressure of 9.2 X 17 for a pressure of 155 - 165. That's why for my AMC 258 i6 / 4.2l , 120 - AMC 150 would be good with no more than 5% difference between them.

Does that make it more clear or not?

 

[BajaEdition]

Hutch. You are trying to put words in my post I did not say
Leave it like this, atmospheric pressure is not the base for reading pressure on a. Pressure gauge. It is volume of air compressed that is significant
The ratio is calculated by the following formula:
, where

= cylinder bore (diameter)

= piston stroke length = clearance volume. It is the volume of the combustion chamber (including head gasket). This is the minimum volume of the space at the end of the compression stroke, i.e. when the piston reaches top dead center (TDC). now to put a fly in the ointment, this formula has not taken into account Valve Overlap, in other words the fact that the intake valve and exhaust valves are usually open slightly at the same time, and is the reason no automobile engine since the model T has had what is called positive compression, or in laymans terms, a point when both valves are closed and are compressing air with no loss through exhaust valve.
not having positive compression, and the fact of the difficulty to calculate the internal space of a super complex variable angle inside a head, that mathmatics are not used to predict inch pounds of pressure in an automobile engine but instead the figures are gathered by using a tester.

 

[hutchman]

I agree completely with what you say about compression ratio........however, the compression tester or gauge is measuring compression pressure and not the ratio. Under ideal conditions, it is this ratio that does determine pressure, but an engine is not ideal. There are temperature changes both from the compression of the gas and the intrinsic energy added by the surroundings.

The other thing we have to remember is this device is driven by the laws of physics, so starting from an absolute pressure of 14.7 psia is a logical place to start.

Good discussion BTW.......

One of the advantages of these discussions is that we do usually all learn something. We don't always agree, but we learn....and that is what makes it worthwhile to me.

 

[BajaEdition]

If such were true you would get significant variations according to the weather.remember atmospberic preasure varies hourly.

 

[hutchman]

Glad to see you're up early this morning......!

I have thought about this some more and believe the methodology of my thought process is correct. In fact I think this would be a good question for a physics class studying the gas laws. It is a mechanical device driven by the laws of physics so the pressure in the cylinder has to start somewhere and I believe that somewhere to be atmospheric pressure.

So I guess at this point, we can agree to disagree........which is OK.

Suffice to say, if my cylinder pressures are between 120# and AMC 150 #, and with little variance between the 6 of them, I'll be happy!

 

[Kane]

Throw this into the mix. Volumetric efiiciency. Most natrually aspirated motors are only around 80 to 85% efficient. Also when dealing with calculating compression you also have valve overlap that is need to help with exhaust scavenging but also reduces the compression in a cylinder and the slower you are turning that cylinder the more air is leaking out during the overlap. Valve overlap is when the intake and exhaust valves are open at the same time. Any machine shop will tell you it is damn near impossible to figure out compression ratio based off of a compression test alone. All factors have to be involved inculding what I have stated above.

 

[BajaEdition]

thanks Kane for reinforcing my statements about Valve overlap
I will say that 85% is a high speed figure
and a low speed, such as a how home mechanics pressure gauge works at is closer to 73%.
Also note at high speed the effects of temperature creating a super high pressure area makes the engineer have to think about the ratio nearing 18 to 1 which will cause dieseling.
So as stated, the formulas to find efficient PSI vary so much from engine to engine, depending on compression chamber shape and size, piston dome shape and size, temperature within, valve overlap, valve size, valve number, angle of valves to chamber, even the angle of the seat of the valve. and the list goes on.
all this happening at combustion chamber temperatures, makes trying to use atmospheric pressure as a base for calculation of what range you will accept as health a mute point. Truth be told the figures given by the manufacturers is more often than not figured out by after finalizing the design of the engine and testing them out, by taking a cold engine and popping a pressure gauge on it.
why? because the mathematical formulas change with each engine and it is just easier that way. Everything they calculate for the engine is at running temperatures, low speed testing is another mater all together.
now if Hutchman wants to accept a range of 20 % as a ball figure ( 120 to AMC 150 ) a difference of 30 pounds, and then a variance of more than 5% (6 pounds in 120) as a healthy engine, that is his matter and I will not argue it, he has done the math. I wonder though why is 20 percent drop from peak good, but only if the variance is less than 5%?

 

[hutchman]

I am just not sure how to respond to your last post other than to say your tone bothers me a little...... This started out, at least from my stand point to be a discussion of the physics involved in the cranking combustion chamber pressure of a AMC 258 i6 / 4.2l . I enjoy a spirited discussion as much as anyone, but your last statement/question borders on inferring the I am somehow not intelligent in what I posted about what constitutes a healthy motor.........

So I am now out of this discussion as it has degenerated and as far as I am concerned it can be locked or deleted..........

 

[CJ]

Never seen so much discussion over a topic that really doesn't require this much discussion.

Follow th FSM and the check will be.OK.

So, we agree to disagree.

Closed.