I took cylinder heads off for some maintenance, grinding valves among others to improve compression. I tested the valve seals by pouring isopropyl alcohol in intake and exhaust cavities and waiting 12-24 hrs to make sure the leakage through the seals was minimal or nonexistent. I have no reason to suspect that piston rings are leaking because the leaks of 20-30% before I took the heads off were consistent with the state in which the valves were and which I serviced, however, I would like to test them anyway because the heads are off as they too can cause cylinder leakdown (as well as bad head gasket). The remainder alcohol in the chambers I can easily suck out with a syringe. Assuming that any alcohol leaked down past the piston rings into the oil pan will drain through the oil drain, do you think this is a reasonable method for testing piston rings?

2 Answers 2


Piston rings don’t seal perfectly due to the fact that they have a split in them, so there will be a tiny gap where alcohol will leak past.

When piston rings are fitted, the splits are placed 180 degrees away from each other on adjacent rings to restrict the flow of gasses, but some still gets passed.

Engines have a PCV system to deal with these gasses.

  • so is that why 0% leakdown is basically impossible to achieve even if the valves, head gasket and the piston ring all fit perfectly?
    – amphibient
    Mar 21, 2022 at 1:20
  • 1
    It is bound to be one reason for some loss. I also don't think you find valves that are perfectly gas tight. As it says on here, - mobil.com/en/lubricants/for-personal-vehicles/auto-care/… , "No engine will have perfect sealing with 0 percent loss. "
    – HandyHowie
    Mar 21, 2022 at 12:22

I don't think this is a reasonable method.

  • To assess a test result, you'd have to know how a healthy set of seals would behave under the same test conditions. Do you have other test results (using this isopropyl alcohol method at room temperature on known-healthy valve seals) to compare?

  • If test results using isopropyl alcohol on known-healthy valve seals are not available, what evidence supports your assumption that valve seals will behave similarly with room-temperature alcohol as they do with higher-temperature oil as found in a running IC motor?

  • I was thinking that if no cylinder significantly differed from others in how much it leaked, ie if they were all about the same (preferably none), than it's good. That's how I did the valve seal testing (after I tuned the heck out of them, they pretty much leaked none)
    – amphibient
    Mar 20, 2022 at 18:40
  • how a healthy set of seals would behave under the same test conditions -- I would go with the assumption of "healthy set --> 0 leak"
    – amphibient
    Mar 20, 2022 at 18:47
  • @amphibient I agree that cylinders returning similar results is a desirable state. But we don't know if that means they're all excellent, they're all adequate, or they're all inadequate. Mar 20, 2022 at 19:59
  • If the level of alcohol remains (nearly) the same over several hours
    – amphibient
    Mar 20, 2022 at 20:00
  • @amphibient "healthy set —> 0 leak" is a desirable result, but only occurs if room temperature isopropyl alcohol behaves the same as engine-warmed oil. Let us know if the method works. Mar 20, 2022 at 20:00

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .