Check my assumptions on Leak test please

[Elombard]

I bought a cheap leak down gauge on Ebay.

When I hook air up to it ( not hooked to car) and adjust the regulator (in put gauge) to 50 psi the second gauge (leak gauge)reads 40 psi. Very linear relationship , lags by 10 lbs all the way up to 80 or so (as hi as I tested it). No leaks in the assembly (that I can hear). At lower pressures its not so easy to read.

So I plug it into the motor. and on the first cylinder it shows 38 PSI on the second gauge (output), 50 PSI on the regulator gauge (input).

So 4 % leak (on this gauge). I figure the second gauge reads 10 pounds low and its leaking 2 psi at 50 psi so at 100 it would be 4 psi or 4%

Any problem with my logic?

I realize its not a definitive number with the low quality of the gauge. But I think I am safe to say the cylinder is less than 10 % leak which is OK for a high mileage motor??

[asphaltgambler]

You should use 100 psi inlet gauge. Easy to read and easy to do %. The other (measured) gauge should read exactly the same. Keep in mind you will have to block off the outlet completely in order to know if it is correct.

[cgarr]

What size bore and what orifice does the gauge have in it? You need different orifice for different bores.

Craig

[Elombard]

Thanks

The inlet gauge is 100 psi but I set the pressure at 50 using the regulator on the tool, I dont know if my little compressor will work at 100 for very long.

When I block off the outlet completely the second gauge reads 10 lbs low at all pressures.... like its calibrated incorrectly.

I guess I am wondering if there is any black art behind gauge calibrating that I dont know that would make my readings way off the mark.

I really just want to verify that the cylinders are less than 10% on a leak down.

[cgarr]

with it shut off the gauges should read the same pressure, when you open the valve both will drop but the second should drop more and more as you open the valve, you read the differance between the two, I use 80 psi input.

Craig

[hcoles]

A while back I attempted to find out about leak down testers and see if there were any specifications that they are made from or conform to. In general they are all different, there is no automotive industry spec. for these. The airplane engine industry has the most regarding the specifications for these. For example you can find for a given cylinder circumference the leakage should be X.XX or less at say 100 psig. Smaller pistons have lower leakage all things being equal and the valves not leaking much.

I would like to start a thread or we can continue on this one with a discusion on leak down tester specs./etc. The result will be that people can easily build their own tester and also understand what the numbers really mean.

Here is my take to start with....

Leak down testers come in two styles: - one gauge and two gauges.
The "tester" is really only a device that measures pressure drop across a restriction. The restriction is usually just a fixed orfice in the "leakdown" tester. When pressure is applied against the cylinder there is flow across the restriction and past the rings and valves and a corresponding pressure drop compared to no flow.

What you really want to know is how much air volume flows at a given pressure that the cylinder sees. The fact that everyone talks about this or that % leakdown doesn't really provide too much information because all of the testers are made different.

History - my interest started when a friend purchased a BMW 530 (V8). As you may know there was a big reall because some of the engines were not made with Nickicil cylinders. The check that BMW did to determine if you get a new engine was a leak down test. People all over the world started getting thier engines tested and then wanted to verify the test that the dealer did. It turns out that BMW wouldn't let out exactly how thier tester was constructed or calibrated and what the numbers really meant. So my friend did quite a bit of investigation and built a few testers to see how they match the factory numbers/etc.

Anyway...if there is another thread or past set of threads I'll go look at those.

Anyone interested on taking off on this subject...let me know and I'll start a thread..if you think I'm mistaken on what I've said...by all means please respond also.

Cheers,

-Henry

[Elombard]

Interesting, There is a pretty good thread on this where Grady Clay talks about the gauges etc in great detail. I wil look for it too.

[hcoles]

I found the thread where Juan (is that the Grady thread?) responds and goes back and forth with Wayne and others. I don't disagree with anything Juan says...the bottom line is there doesn't seem to be a spec. that anyone in the auto industry is using regarding the orfice and what the guages really read. I do agree with Juan that using 100 psi makes more sense than say 36 psi. I emailed Wayne....to ask his tester supplier to respond back with what spec. he is using to build and test to. I bet him a cup of coffee that he hits a dead end. We'll see. I think there is value in understanding this or at least verifying the statement that all testers are made different and are going to read different.
-Henry

[Elombard]

DID you notice that Grady recommends using 5 PSI!!! I was amazed you could do it that low. I guess a gap in the ring to wall or valve to head is the same no matter what.

I dont understand why the orrifice matters?

Pressure in and resultant pressure, isnt the calibration in the gauge?

[dad911]

Originally posted by Elombard
Interesting, There is a pretty good thread on this where Grady Clay talks about the gauges etc in great detail. I wil look for it too.

Testers were discussed in this thread: leak down test

I took mine apart and measured, hole in orifice is about .040

[snowman]

My understanding of this test is that the percent leakdown should be constant no matter what the orfice or guage or pressure used.

[Elombard]

Thats what I am thinking. We are talking about a huge amount of air that can move through the orifice relative to the amount escaping past the piston rings. If you have a realitvely small amoun of leak I dont think the orifice comes into play?

[cgarr]

Just remember the larger the orfice the less the percent drop, the orfice must be matched to the size bore being check otherwise why use any orifce at all?

Craig

[hcoles]

Snowman - interesting idea. When "they" say percent leakdown I assume they are describing:
for a one gauge system - no flow pressure vs. flowing into cylinder pressure, e.g. 100psi and 95psi is 5% leak down
for a two gauge system - e.g. first gauge 100psi and second 95psi = 5% leakdown

If all other things are equal if the orfice is changed I think the leakdown percentages will change. For probably a number of reasons, maybe for the range of cylinder sizes in a typical auto the orfice size showing up the most is 0.040" dia.

For other pressures supplied instead of 100psi you could be correct the % leak down may be about the same.

-h

[Walt Fricke]

Back to EW's original issue: his second gauge on the tester instrument doesn't read the same as his first gauge when he does his initial calibration.

Solution: buy another gauge (or better, two new ones). They will all have the same threads for the size of gauge involved. You can purchase class A gauges from a variety of sources (Northern Tool is where I get gauges to replace the ones in the tire gauges I drop and break, but there are plenty of sources) for little more than the lower classes, though I think none of these are true instrument quality referenced to national standards and so on.

My leakdown tester has this problem - the second gauge reads about two pounds higher than the first at 100 psi. Makes 0% leakdowns quite common though I can hear a little air escaping (some always leaks by the rings). If I cared I'd purchase two new gauges and screw them in (since I don't know which gauge is off, though I suppose I could compare them with other gauges - also of uncertain calibration, of course). And I bet that for a lot more money really accurate gauges could be purchased. That might make sense for shop owners, but probably not for the rest of us.

But I don't care. I'm looking for large problems. I could probably do with no gauges at all most of the time. There should be zero air leaking past valves, and when one is even a little bit bent that cylinder stands out like a sore thumb. Ditto with broken rings (though the gauges are needed here as there is always audible ring leakage).

Now were I checking out used engines for suitability and worrying about ring seal, I might want more accuracy. It would be nice to see results of dynomometer tests which showed how much horsepower one lost for a given leakage past the rings (assuming the instruments all used the same convention for the restrictor size and configuration - flow is affected by more than just hole size, seems to me, or we wouldn't worry about smooth fittings and radii and so on in oil and fuel lines - think proper versus drilled Weber jets).

I used to use 75 psi because my compressor could not hold 100. If my head was fuzzy I would use 50 to make the math easier. My current compressor holds 100 so I use that. I am dubious that the pressure makes much difference (though obviously at some point it might, lifting a head up to leak there, though the valves should just seal tighter, and the rings might as well against their lands). It is easy enough to find leaks from small holes in tires at the 15 psi they have bled down to and air is slippery stuff. Of course, it is easier to interpolate the gauge readings as you come closer to full scale.

If a guy cared about the pressure effect in the 0-100 psi range, he could easily test it for himself just by using a bunch of different pressures on the same cylinder to see if the percentages stayed the same.

Walt Fricke

(who bought a used 2.7 with 12% leakdown on a shops fancy looking tester in a console, and it was an acceptable engine for its purpose, just not quite the engine with new Ps and Cs it replaced)

[hcoles]

Walt,
Lots of good points:

Correct - it is quite common, maybe almost always, that two gauges of good quality (oil filled/etc.) don't read the same. In my lab. at work where I have unlimited budget, I use very good gauges and have quick disconnects so I can easily plug the same gauge in both upstream and downstream postiions. This seems to provide good delta pressure data.

Correct - gauges that are certified to national or other standards are pretty rare to easily purchase....I think for what we are talking about good gauges are good enough if used in a thougth out manner.

Correct - orfices made by drilling out with the drill and bits you find in your garage don't have the same pressure drop for a given size made by a "real" orfice manufacturing process.

Note - if you read about gauges you will find that the readings are best when taken say in the middle 60 or 70% of the range.

Good Info./Reminder - air should not leak by valves and piston ring issues show up like a sore thumb...good to remember. I don't have the experience you have on this...

Questions to You - I've emailed back and forth to Logician (Juan) on this, one of my favorite subjects, and he bought a tester from Aircraft Spruce (I think that's the name) and it's supposed to be made to an actual spec. - he was testing a friends 3.2 and discovered that he had to move the pistons just up or down from TDC to get the lowest readings...this makes sense to me from looking at my cylinders and reading about rings/cylinders/pistons most wear is at or very near TDC and BDC where the rings have to move at zero velocity on the cylinders. Does this make sense to you and have you observed this?

I have on my list of projects, I want to build a wall mounted tester where I can easily swap in different orfices and gauges/etc. I already purchased a flow meter. The reason for the flow meter....that is really what we are trying to measure....how much air (measured in scfm [standard cubic feet / minute or other but this is what my meter is marked at] ) is flowing for a given pressure against the cylinders.

I think the best setup is to have fittings that can be connected to a differential gauge where the upstream and downstream, from the orfice, is connected to the same gauge, then the difference in gauges is completely eliminated...this has some issues like over pressurizing the guage if you are not careful if you get a gauge that say reads 0-10 psid and it gets 100psi by accident.

I need to get this contraption built and then start posting info. and a diagram of the setup, probably before I go to much farther.

-Henry
recent 3.2 top end - still running great after 900 miles