Need help interpreting leak down test vs. compress

[Joe R.]

Hello All:

I need some help interpreting the results of a cylinder leakdown test.

I have been performing various diagnostic tests on my engine to determine the cause of a poor idle condition. As you may have seen from an earlier posting, one thing I found is that my 300 HP engine has the 350 HP cam in it, which is not very compatible with my Powerglide transmission. It is possible that just a cam swap (which is now underway) will completely fix the problem.

I have also been looking at compression and leakdown. I read about the leakdown test on this board and decided to try it along with a conventional compression test. I purchased what appear to be decent compression and leakdown testers from Jegs, and ran the tests.

The compression test results look pretty good, ranging from 158 to 165 psi. The leakdown tests range from 20% to 30%. On the leakdown tester, the range of 0% to 40% is marked "low" and the range of 40% to 70% is marked "moderate." I'm not sure how the percentage is calibrated on a leakdown tester, so I don't know exactly what a result of 20% means other than it is reasonably good (at least according to this brand of tester). I wonder whether another brand of leakdown tester would yield different numbers for the same leakdown rate.

If I just looked at the compression and leakdown numbers I would conclude that things are in reasonably good shape. However, what bothers me is that on one cylinder that reads 158 psi compression and 30% leakdown, there is a very clear leak in the exhaust valve that I can hear and feel (the exhaust maifolds are off). On all the other cylinders there is no perceptible leakage in the intake or exhaust valves, with all the leakage appearing to be past the rings.

In the old days I would have simply run a compression test and would have concluded that things are okay. My problem is that I'm not sure how to interpret the leakdown test numbers. In particular, I am bothered by the perceptible leakage past one of the exhaust valves. Is this "normal" for a stock engine with maybe 40K miles on it, or does it clearly point to a need for a valve job?

Can some of you who have more experience with leakdown testing help me get calibrated on what is normal for a leakdown test and what is abnormal? Any help would be appreciated.

[Clem Z.]

to do a leak down test you must have the

piston at TDC. if you have a valve leak you can "smack" the leaking valve stem with a plastic hammer while still under air pressure to remove any carbon that could be stuck in the valve seat.

[Joe R.]

Re: to do a leak down test you must have the

Hi Clem:

The piston was at TDC for all the leakdown tests. The only "rule" that I knowingly broke was that the engine was not hot.

I wondered about the possibility of a piece of carbon on the seat, so I put all the spark plugs back in and ran the engine for a few minutes to try and clear any possible debris from the valve seat. This also warmed up the engine somewhat.

The leakdown and compression results on the cylinder in question were pretty much unchanged by this. So, either I have a stubborn piece of carbon under the seat or I have a marginal valve situation in that cylinder.

Later today I can repeat the test using your "smack" technique to see if the results are affected.

I have no experience interpreting leakdown results, but it seems worrisome to me that there is detectable air getting past the exhaust valve. On the other hand, the overall leakdown rate of 30% is within what my leakdown tester considers "low."

I have been told that race engine builders look for less than 10% leakdown in a fresh engine, but I don't know what is considered normal for a general purpose street engine, especially one with perhaps 40K miles on it.

[Joe R.]

Re: to do a leak down test you must have the

Hello again Clem:

I repeated the leakdown test on the questionable cylinder and tried "smacking" the exhaust valve as you suggested. That seems like a good technique for releasing debris on the seat, because it lifts the valve just slightly while there is 100 psi in the cylinder. Makes a loud "pop."

Anyway, despite repeated smacks, the leaking through the exhaust valve remained pretty much unchanged. So, I guess what I am seeing is not due to debris on the valve seat.

[Duke W.]

What input pres. are you using for leakdown test *NM*

[Clem Z.]

Re: to do a leak down test you must have the

sounds like there is a problem with the valve because that should clean out any debris holding it off the seat

[Clem Z.]

Re: What input pres. are you using for leakdown te

looks like he is using 100 PSI which should be fine

[Joe R.]

Re: What input pres. are you using for leakdown te

Hi Duke:

I think the test pressure is about 100 psi. The leakdown tester consists of an adjustable regulator followed by a "leakdown" gauge.

The instructions say to connect a compressed air source to the regulator with the "leakdown" outlet closed, and then adjust the regulator so that the leakdown gauge reads at the "set" position. I have determined empirically that this corresponds to about 100 psi.

Once the pressure is set, you connect the leakdown gauge's outlet to the cylinder under test. The leakdown gauge responds to the amount of air flow out the outlet side. If no air flows the gauge remains at the "set" position, but if air flows the needle moves counter clockwise on a "percent leakage" scale.

I understand qualitatively what the leakage tester is responding to (the amount of air that flows from a controlled 100 psi source), although the relationship to a "percent leakage" reading seems like it could be arbitrarily assigned. Hopefully, there is a standard for leakdown testers so that a reading obtained with one manufacturer's leakdown tester corresponds to what another brand of tester might read.

[Duke W.]

Re: What input pres. are you using for leakdown te

I really dislike those "dumbed down" commercial leakdown testers. I prefer one with two 150 psi gages. There is a standard size metering orifice connecting them, but I don't recall the diameter. So if you input 100 psi and the downstream gage reads 90 psi you have ten psi leakdown or "ten percent" for the dummy gages.

You can also use lower pressure, but the higher the pressure the more radial force pushing the rings against the wall and the better the seal.

Once you achieve the critical pressure ratio of 0.528 flow past the leak paths is sonic and choked. Further pressure in the cylinder will not increase flow, so if you use 80 to 90 psi input the downstream pressure gage should still read about 10 psi less.

You must use absolute pressures to compute the critical pressure ratio, so 14.7/.528 = 27.1, and subtract 14.7, so the pressure ratio is critical when the downstream pressure is about 12.5 psig.

It's best to do compression and leakdown tests when the engine is at operating temperature.

Duke

[Joe R.]

Re: What input pres. are you using for leakdown te

Hi Duke:

Thanks Duke. I figured there must be some sort of standard definition for leakdown. A difference of 10 psi across a standard orifice makes more sense than "10%."

After thinking about the situation some more I decided to just bite the bullet and pull the heads so I can have them rebuilt. It was just too tempting to do it while I have the intake off anyway for the cam swap, and knowing that I had at least one leaky valve was bothering me. It will be interesting to repeat the leakdown test on the rebuilt heads to see how much improvement there is.

[Duke W.]

Re: What input pres. are you using for leakdown te

If you pocket port the heads you will end up with the low end torque of a 300 HP engine with about the same top end power as the L-79 cam.

I recommend retarding the 929 cam four degrees with pocket ported heads, but only with manual transmissions.

Duke

[Joe R.]

Re: What input pres. are you using for leakdown te

Hi Duke:

I have read your previous postings about how the 300 HP engine responds well to pocket porting. Based on your postings, I am considering having this done when the heads are rebuilt.

What I would like to end up with is an engine that has a lot of low end torque but continues to pull strongly to 5500 RPM. It sounds like pocket porting combined with the 300 HP cam (or a slightly bigger aftermarket cam) can achieve this goal.

What I need to sort out is how far I want to go with the original engine. Right now my focus is on NCRS judging so any changes have to be undetectable during judging. Pocket porting the original 462 heads fits this program.

However, at some point I will be *finished* with having this car judged and I could just bolt on a set of aftermarket heads that flow better than pocket ported 462 heads and cost less than pocket porting. Or, I could pull the original engine and drop in a crate motor. I really haven't planned that far ahead, but now that the heads are off and will be rebuilt, I have to decide whether to spend the extra money for pocket porting.

[Duke W.]

Re: What input pres. are you using for leakdown te

With a good pocket port and multiangle valve job, the OE heads will flow about as well as any aftermarket head of similar port volume. Higher port volume heads will flow more, but kill low end torque and throttle response due to lack of port velocity. This is another myth that the OE heads "don't flow". With proper rework they are very good and will yield excellent torque bandwidth with any of the OE cams up to an including a 450+ HP vintage racing FI engine with a 30-30 cam retarded four degrees, but this configuration is very torque shy at the bottom end.

It's a time consuming task, so it is expensive at current labor rates, but it's something the amateur restorer can do himself.

Resist the urge to install a "slightly bigger" aftermarket cam. Any will hurt low end torque and idle quality without necessarily improving top end power because all aftermarket cams have too much overlap. The 929 cam is very good, and if you're willing to lose a little low end torque relative to the OE 300 HP engine, install it with 4 degrees retard.

Your car's performance is compromised by the Powerglide, because it just doesn't have enough reduction spread, so low end torque is at a premium.

A TH350 or 700R4 would be a much better choice, but they are not an easy bolt- ins. A four speed - like a Richmond WR four speed - is another option that is almost a bolt-in. (See Joe Lucia's discussion of this in a recent thread.)

Duke

[Joe R.]

Re: What input pres. are you using for leakdown te

Hi Duke:

Thanks for the info about how pocket ported 462 heads compare to aftermarket heads. I will look into this further. If I was convinced that the ported 462s would work well for future non-NCRS upgrades, it would be easier to justify the extra cost of pocket porting them.

I know the Powerglide is problem but I have to keep it in the car while I am having the car judged. The car has already Top Flighted at the regional level and is scheduled for judging at the National in July. If it Top Flights at the National I may be "done" with having the car judged unless I decide to pursue a Duntov, which would probably add three more years to the judging (and Powerglide) phase.

The Powerglide is definitely coming out after I'm done with judging. My current plan is to swap in a 700R4, which is a fully reversible, bolt-in change. Toward that end I have already put in 3.70 gears, which I would never have done if I didn't plan on having an overdrive transmission eventually.

So, the choices I make now regarding the heads and cam have to work reasonably well with the Powerglide but can be tilted slightly toward the eventual 700R4.