FredR

Here's one for the chaps who know what they are doing when it comes to leakdown testing.

My motor should be in good condition and most indicators suggest it is but I still have some nagging doubts therefore want to do a leakdown test while I am on the scrap heap and have time on my hands albeit it is now getting somewhat hot again.
A short while ago I purchased a leakdown test rig and today I went down to the Souk [market] to see what I could get for my money in terms of an air compressor. Low and behold they had a 3HP model that can deliver 50 litres a minute at 8 barg [120 [psig]. Needless to say made in China with an integral motor/single stage compressor unit. The kit looked reasonably well made and is probably similar to the kind of cheap stuff you get at K mart or on Amazon for $100 or so. I have always fancied an air compressor in the tool chest so why not. Needless to say it is not going to be something used all day every day so for my needs hopefully a proportionate investment.

My queries for the enlightened:
1. Does the capacity seem reasonable for the task at hand.
2. Would such a unit be suitable for modest attempts at spray painting [not the 928].
3. for the leakdown test I have done some research on how to carry it out. Most say it is better to be done with the motor warm-I intend to try this- any thoughts?
4. I do not have a tool to tighten the rig into the spark plug boss- is hand tightening using the rig enough? I do this when using the compression tester and that hits a higher pressure [185 psig].
5. Any tips/ suggestions to optimise my chances of success most appreciated. Last thing I want to do is have a heart attack because of poor results due to inept technique.

I have done similar tests many years ago but on furnace tubes to ensure the catalyst was packed correctly so I am not a complete novice- the air compressor then was a diesel driven beast of a thing used by road gangs to drive windy picks.

Thanks

Fred

Ad0911

I made a leakdown adapter set for a 928 for my compression tester. The basic ingredients are: broken spark plug (gutted), some copper waterpipe and T-spiltter, air valve, some pieces of gardenhose. I can fix the compression tester on the side piece, fic it with a clamp. Put my airpistol for compressed air on the top part and close the valve when everything is under pressure. Whait a while then push the blow off valve on the compression tester to see what's left of the pressure.

WallyP

My queries for the enlightened:
1. Does the capacity seem reasonable for the task at hand.
I would think so. The biggest problem that I see is that the compressor is noisy, and one of the things that you can do during a leakdown test is to listen for leaks at the MAF inlet (leaking intake valve), exhaust pipe (leaking exhaust valve) and oil filler (leaking piston rings).
2. Would such a unit be suitable for modest attempts at spray painting [not the 928].
Sure, if you use the proper type of spray gun. I have what is called a "detail gun" or "touch-up gun. This is a small unit, and give a spray pattern smaller than a standard spray gun, but better, larger and more controllable pattern than than a rattle can.
3. for the leakdown test I have done some research on how to carry it out. Most say it is better to be done with the motor warm-I intend to try this- any thoughts?
That is the normal consensus. One of the problems is to find exact TDC-Firing on each of the cylinders. Start with the engine at TDC-Firing on #1. If you are exactly on TDC, the pressure will not make the engine spin. When you finish #1, go to #3. Use a long wooden dowel in the plug hole to show when the piston stops moving up. It helps a lot to do this with two people - one to carefully turn the engine via the crank bolt, the other to keep the dowel from binding and breaking by getting crooked and to tell when the piston stops moving. Proceede thru the firing order. (1-3-7-2-6-5-4-8)
4. I do not have a tool to tighten the rig into the spark plug boss- is hand tightening using the rig enough? I do this when using the compression tester and that hits a higher pressure [185 psig].
The compression gauge usually is sealed with an O-ring. Your test rig probably uses an O-ring as well.
5. Any tips/ suggestions to optimise my chances of success most appreciated. Last thing I want to do is have a heart attack because of poor results due to inept technique.
Remember that this is a diagnostic tool usually used when there might be a serious engine problem, and the results may not be exact. As with compression, the differences are probably more important than the actual values. The values may depend upon the test equipment and your skill more than the actual engine condition.

dr bob

Fred--

Everything Wally shares is spot on (of course...). I'll add that you might consider an oil-type compressor rather than the oilless, for both efficiency and noise. I've used a large conventional oilless from a big-box store for many years, and it "lives" in a sound-insulated cabinet. For a reason. Meanwhile, the construction contractor who was doing our home remodel uses a 2.5hp Makita portable that's quiet enough to stand next to and hold a conversation. Very impressive, very quiet. They cost more than the little throw-away compressors, about $300 in the U.S., but the durability and the noise factor make it a good investment at least for me.

Painting:
Take a look at the spray gun you intend to use, and the expected CFM draw. Then look at what you plan to paint. I use a touch-up gun similar to what Wally mentions, but also use a few different larger HVLP guns for the more serious projects. One of the regular projects was the bi-annual strip-and-refinish of the garage doors you see behind my car in the avatar picture. The larger door is 20' wide, smaller 10' wide, so lots of area to spray. Fortunately, the doors are horizontal panels so they get sprayed one at a time. My compressor is rated for twice what the gun uses per its spec sheet, so one might assume that I could spray continuously with the compressor at 50% duty cycle. Sadly that was not true. Fortunately the large enough tank under the not-large-enough compressor together made it possible to spray a whole panel at a time, then the compressor would recover as I moved the panels around. We've since sold that house so the door refinishing duty will fall on someone else.

For your needs, think carefully about what your painting projects will look like. If they are touch-up-gun projects that are done because spray cans are just a little too small, then a small oilless might be fine. More than that, and a bigger piston-type compressor will do much better for you. And those you can usually rent when you need them for the bigger jobs.

I'm keeping my 5hp (rated) oilless with the 30-gal tank, but will likely grab the small portable Makita to use for the small projects, like filling tires, the finishing nailer, touch-up gun stuff, and small low-duty car projects. The noise difference is amazing, and I might just valve the larger compressor in when I need the bigger tank capacity.

Speedtoys

Thats now how a leakdown test is performed..generally.

It's a constant 100psi fed to an orifice, to the cylinder, to see what the cylinder will hold at a specific volume of air fed to it.

Its not just pressure fed to the cylinder, to see whats left once you turn off the source...but 'what pressure is allowed to build to given a specific volume of air flow to it'.


??

FredR

Gents,

Many thanks for the thoughts- I think I have most of it covered. I will ask to listen to the intended purchase and make a long delivery hose! If all else fails I can stick it in the TV room while I work in the adjacent car port!

Regards

Fred

ptuomov

A dumb question: What is the leakdown test supposed to measure? I was under the impression that it measures mainly the piston ring end gap (unless the motor is broken.) In particular, by measuring the leakdown of an engine when new and with known ring end gap, and then after use, one can estimate the bore wear. Bore wear will increase the ring end gap (about proportional to the bore diameter) which will increase the rate at which air can flow thru. The remaining question is whether I've understood this at all correctly.

69gaugeman

The way a leak down tester works is by comparing the pressure on the feed side of a very small hole (generally 1.0mm in diameter) to the pressure on the other side (cylinder side).

The general premise if the leak in your cylinder is zero them both gauges will read 100% (I use 100psi gauges and pressure. It makes the percentage readings easy to see)

If you were to say drill a hole in the top of your cylinder 1.0mm in diameter all the air through the orifice would then escape into the crankcase. This represents 100% leak (Or leakdown.)

Now if the hole was say, .25 mm in diameter, you would have approx. 25 % leakdown. The feed side would read 100psi and the gauge after the orifice will read 75psi (yes I know it is not linear, but the premise of a smaller hole means the gauges read closer to the same number)

So with all that said, a warm engine is most accurate as the clearances are at operating condition. I generally just am looking for gross leaks. Small ones cold are usually smaller with a warm engine.

A perfect engine will have 0% leakdown (both gauges read 100psi) to around 2% (second gauge reads 98psi.

I know for a fact that engines will run relatively well with up to 15% leakdown. I try and have less than 10% before I will run an engine.

69gaugeman

It will tell you if you have burnt or bent valves as the air will rush out the intake manifold (if an intake valve) or the exhaust manifold (if an exhaust valve). Or the coolant overflow tank if it a head gasket.

It is significantly better than a compression test as it tells you were the leak is coming from.

dr bob

Leakdown is a favorite go/no-go screening test to detect failures on engines you can't crank or warm up to do a normal compression test. Used 911 engine sitting on a pallet, for instance. A proper compression test gives a much clearer idea of an engine's condition, if it's available. Once you decide there's a low cylinder with a compression test, a leakdown test offers you the ability to sort where the leak might be. Listen to the intake, the exhaust and the oil filler, and the noise will tell you if it's an aintake valve, exhaust valve or rings/piston that's not doing its job. A scratch/gouge lower in the cylinder wall, fr instance, will show up as low compression but might easily pass leakdown, done with the piston at TDC on the compression stroke (so the valves will be closed).

FredR

Tuomo,

For leading edge punters like yourself with intimate knowledge of their motors down to the last component a leak down test is probably worthless. My motor is a bit of a half way house in that I have owned it since the end of 1998 [crikey that is 16 years ago when dear Cousin Wally was a relative youngster back then like me today- ha ha].

After I lost my S4 some 9 years ago we pulled the motor from the wreck and partially stripped it for inspection as it had lost compression on No7 cylinder. A leak down test at that moment in time would have been very revealing and easier to set up. After we pulled the motor and started repairing the damage to the 1/4 cam cover we found a few teeth missing from the cam belt and that rang alarm bells so we did a regulation compression test that highlighted the damaged area that we presumed to be bent valves [which turned out to be correct]. A leak down test would have told us exactly where the problem was and would have been easier to implement. Nonetheless we pulled the heads and replaced two inlet valves and the rest we re-lapped after cleaning the ports up.

The dilemma I had at the time was whether to replace the rings whilst the motor was out and easy to get at. We visually checked the bores- no signs of wear of scratches and measured them. We made a decision not to disturb the bottom end given the mileage of 120k km at the time was noting really. Had we done a leak down test prior to stripping on the other 7 cylinders it should have given good results and had we done one after rebuilding the motor it would have helped confirm successful work.

My motor has always had a tendency to oil consumption and I was told this was perfectly normal in a 928. That does not make it acceptable and I would like to close that loop by carrying out the leak down test procedure which hopefully will eliminate suspicion in the ring dealing department. To be clear, compression tests reveal a consistent 185 psig across the board with more or less no deviation so I hope to remove a "nagging doubt" rather than find a ticking time bomb as it were.

I now take the view that a leak down test combined with a compression test is the best non invasive non destructive testing one can do- neither being complete on their own.

It is quite easy to fabricate a test rig- I purchased one from Amazon. The rig has two chambers interconnected with a small orifice and a pressure gauge on each side. You set the inlet pressure to 100 psig and if there is a perfect seal there will be no flow and thus no differential pressure. A well sealed motor will hold about 98% of the pressure and general literature suggests readings in the range of 5 to 15% leak down are acceptable with consistency being as important as the leak down number itself. With the cams engaged this test can only be applied at TDC on the compression stroke

Rgds

Fred

ptuomov

FredR, 69gaugeman, dr bob -- As I wrote, I agree that leakdown should flash red light in many ways if something is _broken_. That I agree with.

69gaugeman

You can and I have done the leakdown test through the whole compression stroke. Start at the bottom and using force multiplier just slowly run it up the cylinder. If there is a scratch it will show up.

WallyP

Well, quick and dirty estimates will show how much force might be involved. A bore of 3.94" give about 12 square inches on the piston, times 100 psig gives about 1200 pounds; stroke of 3.11" gives a lever arm of 1.55", which gives about 155 lb/ft. Adding some Kentucky windage for overcoming friction, and you might be close to 200 lb/ft needed to turn the crank at the highest point of resistance. Yep, sounds doable, but I would really be careful to avoid letting it get away from me.

That would help to judge scores and wear on the walls.

FredR

Well I purchased the compressor kit plus 10m of hose and various connectors to ensure I could hook up my leakdown tester that I purchased earlier from Amazon.

The compressor is a 2HP unit with direct drive motor and it is lubricated via a built in oiler. The reservoir is a useful 50 litre job and glad I did not purchase anything smaller.

I carried out the tests this evening and the good news is that providing I did the test correctly I do not appear to have any leakage issues. The compressor cuts out at about 110 psig and cuts back in at about 70 psig and that is not adjustable. The leakdown kit says you should maintain a reservoir pressure of at least 10 psi over the pressure you test at so I decided to carry out the test at 75 psig and let the reservoir rundown to 90 psig before stopping each test.

Initially I let the engine warm up until the coolant temperature was at its normal running level and I tested the leak down rig with the spark plug connector open to atmosphere. The rig held a steady 75 psig as the reservoir ran down from 110 psig to 90 psig and the other gauge showed only a few psig so my conclusion was the rig was working as expected.

I then set the crank to TDC and worked out that it was actually on No 6 firing stroke. I could hear air movement inside the crankcase but nothing from the inlet or the exhaust and pretty much the same on all cylinders. I presume that some leakage is to be expected but I could only measure 1 or 2 psi difference between the 2 gauges. To be fair the gauges on the test rig are not that clever but they appear to be good enough for the purpose giving reasonable repeatability. I carried out each test twice.

Once the air supply was blocked the pressure dropped off quite quickly- I assume that is normal.

To position the crank I used my tape measure to work out the circumference of the harmonic balancer that worked out to be 59cm and divided by 4 to define the 4 quadrants [14.75mm]. I then measured an arc of that distance and marked the position with tippex to give me an idea where to set the crank. I could not be bothered to check for TDC and that may explain why some of the cylinders wanted to walk [specifically 5,4 & 8 if my memory serves me correctly.

So quite a satisfactory day if I did everything correctly. A couple of photos below.

Regards

Fred