Bottom line up front. It’s really hard to trust tests. Really hard. There’s always a way that a test could fool you. Test the shit out of your test. Don’t trust any test unless it’s repeatable when approached from a couple different directions.
I played around with 2 engine blocks and 1 head. I wanted to see how repeatable leakdown tests were, how much leakage was bottom and top and how much wet differed from dry. Later I’ll lap the valves on the head and see how much that improves the leakdown test, and then disassemble one or both blocks and see if I can understand why some cylinders did better than others.
The inverted drip test. This is a test to see how the valves are sealing. You turn the head over and pour some gas or similar into the combustion chamber and see if any leaks past the valves. When I first read about this, 4-5yrs ago it said that <3drips/min was good. But having done it a fair amount now, I’d say that you pretty much want no drips at all or your valve seal is going to suck.
I did an inverted drip test on the head that came with New #6. I used alcohol. All cylinders had 0 drips. I interpreted this to mean that altho the PO said he replaced the HG, what he really meant was valve job.
Then I started testing the engine block that came with New #6, recall that this engine was weak. I used my headless leakdown tester http://spece30.com/forum/42-engine/68607-testing-cylinder-seal and did each cylinder twice. With one exception all cylinders got consistent results. Percent leakage
8%
5%
28%, 5%, 5%
7%
7%
5%
What this tells us. It was damned strange that I got such an awful result from #3 the first time I tested it. What I found is that a flaw of the leakdown test is that it’s an attempt to use a static test to evaluate a dynamic environment. In order to be more sure of your results you have to move the pistons around a bit, while still under pressure, and see what happens to the leakdown percentage. As I played with this idea a bit I found that darn near all results improved if I move the pistons around. The tests were all done in the same place mind you, a couple mm from TDC, but what made the results repeatable was simply moving the pistons. Note that this is easy enough to do with a 28" long wrench on the jesus nut, but would be harder with any normal length wrench because you’d be fighting the air pressure. If you aren’t cautious your wrench could whip around and give you a good whack.
Then I put the head on, the head that tested so well in the inverted drip test. new results…
32%
6%
58%
40%
78%
9%
Look at that shit. Only 2 cylinders are showing decent sealing now that the head is on.
What does this tell us. It means that the inverted drip test can tell us only about the most egregious valve seal problems. Your valves could leak like a sieve and the inverted drip test wouldn’t tell you. But if your valves do drip, well, then they really suck.
I also tested another block w/o head. This is the engine that I installed in New #6 in the spring but then pulled back out when compression and leakdown tests indicated that #6 wasn’t sealing well. This engine was bored out 020 over.
Yesterday I tested #5 and #6 and got 18-20% leakage. But tonight, having learned that I need to move the pistons around a bit the results were much better
8%
10%
11%
9%
14%
12%
What this tells us. It confirms the idea that you have to move the pistons around a bit in order to trust a leakdown result. It also suggests that one shouldn’t pull an engine out hastily.
Other things to think about.
When I started trying to understand how to do a leakdown test 5-6yrs ago I read that a couple % of leakage was a realistic goal, but now I’m starting to wonder about that.
Supply pressure is apparently very important for an accurate leakdown test, but supply pressure is surprisingly hard to stabilize. Air pump regulators aren’t designed to provide a stable air pressure, they are designed to provide a stable pressure at a certain rate. So if the rate varies, so does the pressure. Therefore what seems to be 100psi with a little leakage out of your fittings, might be only 80psi when connected to a cylinder with a couple % leak. How the heck do I compensate for this? Or how do I create a genuinely stable air pressure source?
Near as I can figure out, the ring seal of an OEM bottom end lasts one helova long time. That’s something for folks to think about when they are contemplating an engine refresh. That high mileage block above did at least as well as the overbored one with a couple seasons on it. If a person has a high mileage block with decent oil pressure, I’d be hesitant to screw with the bottom end unless there was really convincing evidence that the rings were leaky.