High leak down on fresh build before break in

[dhidaka]

Has anyone done a leak down test on a brand new engine build prior to break it in? A customer brought in a 4 cylinder, 4-valve DOHC GSXR Suzuki for me to degree in the cams. This is a higher end drag motor freshly assembled with all new parts from reputable manufacturers including big bore block, pistons, Carrillo rods, etc. Head is ported by nationally known firm, also with new quality parts. Engine should easily run the 8.90 class

I degreed the cams and the customer took it home. He did a leak down test and each cylinder was extremely low registering about 60-65%. Some air could be heard coming from the engine case, but that is understandable with unseated piston rings. But, air was also coming from intake and exhaust ports. Since the customer did not assemble the engine he brought it back to me to take a look.

First, I removed the cams and did my own leak down. Same results as my customer. So, I removed the head to take a look. Valves were lapped and the faces looked good and proper. The seats were done on a Serdi or Rottler and looked good. To test further, I put some Dykem on an intake seat and “tested” it gently by hand with a pilot and 45 stone. Seat looked fine. So, I am assuming that the valve job is correct.

Next, I assembled the valves in one cylinder and filled the port with acetone. I still had Dykem on the valve seat. I wanted to see if the acetone would leak and “wash away” any Dykem. No leak. I then blew compressed air around the periphery of the valve face at the chamber to see if any bubbles would percolate up the port. Very, very little appeared. Now, I can really assume the valve job is good.

I have many hours in this project and still don't have a definitive answer to the high leak down numbers. I told the customer I was going to reassemble the engine and do another leak down. Same low numbers. Same air coming from ports. I have never done a leak down prior to break in, so maybe this is normal. I think everything will be find once the engine is heat cycled a few times and broken in. Any advice?

[dwilliams]

I use a rubber-faced plate and a hand pump to pull a vacuum on the intake and exhaust ports to check valve seal.

If the seats are good, I'd look for bent, non-round, or non-concentric valves. If the chuck on the valve grinder is worn, the head of the valve may be round, but several thousandths off-center from the stem.

[dhidaka]

I use a rubber-faced plate and a hand pump to pull a vacuum on the intake and exhaust ports to check valve seal.

If the seats are good, I'd look for bent, non-round, or non-concentric valves. If the chuck on the valve grinder is worn, the head of the valve may be round, but several thousandths off-center from the stem.

I refaced a used valve and tested for bubbles with acetone in one of the ports. Had a very few bubbles appear. I have a dial indicator on my valve grinder and it will hold .0005" with a collet chuck.

[hoodeng]

When you do a leak down test you can tap the valves and see if that changes your reading ,if it improves to a point of acceptable % it will be ok to run. But your figure is very high , i would not want to see any more than a couple of % for a fresh motor.
You have observed leakage at the case breather, if the bores are true and the ring pack fresh there should be very little blow by, injecting oil into the cylinder and re testing will confirm your bore seal condition.
Are you blocking the cylinder at the top of the stroke when testing? we also used to rock the piston at the top of the stroke with a long bar and could see the percentage drop..
A word of warning when doing this if you have not performed this type of test before ,the rock motion is very small ,if you let it rock too far that sucker will take off with a vengeance, i have seen a few breaker bars shot across the workshop/pit area by the unwary.

But i don't think 60-65% is going to be rectified with incremental adjustments.
Cheers.

[JoePorting]

Is the valve train keeping the valves slightly open?

[jred]

he said he removed the cams,,,, if the head was ported a lot I would start by looking for distortion on the head or block that is happening when your are torquing the head down..if you have air coming out of the valves when the head is torqued and past the rings something is moving around ,,, i made a vacuum tester and we test every port on the head before it leaves i have also done leak testing as you and good luck with that ,,

most engines that i have tested for leak-down had less then 3 percent with total seals rings and 5 to 6 percent with std type rings after running total seals about 1 to 2 percent and about 4 on std rings..

put the head on and lightly tighten the head bolts add some air pressure and then start torquing the head and see what happens look for changes in leak-down percentages,,, we've had problems like this with air cooled VW heads most of the time it was when the heads were ported and fly-cut for a lot of compression it made the head weak and would bend around a lot..

[cjperformance]

When you tested the head/seats it was off of the engine correct?
I know at least with Z900/Z1000 kawasaki heads that a fresh valve job that is a perfect as I can be picky will change when the head is torqued.
It sounds like your valve seal is ok. What are the ring gaps and cylinder finish?

[BigBlockMopar]

Could you try a leak-test with the heads torqued at only 50% for instance and see if anything has changed?

[jdperform]

On more occasions than i like we see leakage behind the seat rings. Hard to find but soap bubbles in ports will confirm if you can see the seat rings. Sometime you can run your finger across the ring and the bowl to feel if there is any subtle misalinement IF its not been blended.

[1972ho]

Then doesn’t this mean the head is cracked somewhere.

[MotionMachine]

I once had a customer leak down a fresh non-run rebuild and it leaked all over, the numbers were in the 70's. He freaked out and was going to take it back apart but I told him to run it, it will leak fine after initial. Which it did. Based on that one experience, I'd say it's typical.

[BILL-C]

It is very common for head to distort a little when it is torqued onto block,especially if aftermarket studs are used. I'll bet bores are distorting when torqued also. The jap bike world is way behind on torqueplate usage. I think you are inventing problems that likely don't exist by doing leakdown before breaking in engine.

[user-30257]

What's the test pressure?

[dynodarrell]

Are the pistons gas ported? I have seen gas ports leak down 60 to 65% from two different companies due to improper ring/ land back space . They said that when the motors went dynamic it would be ok. I did not trust that. I got a different set with proper ring clearance leak . down went to 6% .

[Tuner]

What's the test pressure?

... and what size is the orifice in the leakdown tester? With the same amount of actual leakage, a large orifice will display a small percentage of leak, a small orifice will make the test more sensitive and display more leak.