TOPIC:

Can't find my original newbie post about the rat, so thought I'd better start a new thread instead of hijacking those other discussions I've jumped on...

Did my leak-off test tonight.

To recap on my previous compression tests (cold engine but never mind);
PSI:
20 / 70 / 110 / 140 / 140 / 135

And from the leak-off (30 - 40 psi static air pressure at / near TDC);
1 - Inlet valve leak
2 - Very good seal
3 - Very good seal
4 - Exhaust valve leak
5 - Very good seal
6 - Very good seal

For the leaky valves on 1 & 4 both buckets & shims rotate freely by hand so I take that to mean no binding on the heel of the cam and therefore either valve seat or bent.

So I read this as a classic good news / bad news result, what do you guys think?

4 (valve aside, 5 & 6 were good on dynamic compression and the leak of supports this and confirms head good (ish).

3 was starting to loose compression on the dynamic test, but as leak-off is good, could support very slight conrod bend?

2 & 1 were very weak on dynamic, but as good on Leak-off (#1 inlet aside) suggests strongly bent conrods in #2 for sure and likely #1 although could be 'lucky' and just the valve only.

Either way, it's "off with it's head" as the valves have to be done.

I'm struggling to get an accurate TDC deck height measurement on the piston crowns (to check for bent rods), as the angled plug hole makes it inconsistent from one attempt to the next.

Does anyone know of perhaps a TDC tool that scres into the plughole for stability of measurement?

The alternative of course, as the head has to come off, is to measure the TDC piston crowns against the top of the barrels, much easier and stable to measure.

Ok. First, i'm thrilled you are doing a leak down test. Between that and the good old compression test, you'll have all the info necessary to determine the issues before tearing down.

The leakdown test - to do it correctly  , it needs to be done at 100 psi. The tester is designed to read out a leakdown rate based on a percentage.10 % leakdown is a tight engine. 20% you need to consider rebuilding. More than 20 % definate teardown.

Now where to listen for leaks. At the carbs for intake valve leaks. At the mufflers for exhaust leaks and at the engine oil filler cap for piston rings. At the rad cap for head gasket failures.

The reading needs to be taken at top dead center on the compression stroke and ensure that you don't have it on exhaust/ intake overlap. Be careful. Sometimes if not absolutely at top dead center, the crankshaft can creep and then flip over to bottom dead center. Fingers clear when you apply the air. Some guys like to afix the crank to ensure it doesn't flip over. Also note that the cylinder wear is usuall
y predominant at 20 to 30 degrees before/ after top dead center so measuring at top dead center is the point of most likely the worst possible spot.

If the leakdown test is 90% or better and the compression is low, then you know you have bent rods.

Thanks Kawboy and yes that's the process followed to get those results re the valves and observations.

My leak-down manifold seems to run at a lower pressure than you use?
It requires the output guage to be zerod (max deflection on the guage) prior to connection to the cylinder, and this requires around 30psi on the input guage, then on connecting to the cylinder the output guage 'zero point' at max deflection) drops according to loss encountered.

The 'good' cylinders I observed dropped hardly at all (actually above the 'green zone' on the guage) and the #4 exhaust valve leak hovered around the minimum acceptable (approx 20 - 30％ guage drop) and the #1 inlet leak was similar.

Of course the calibration of the guage is arbitary, along with the colour zones, and as per dynamic compression tests it's more relevant to compare best to worst for variance.

Your point about TDC for both valves open is very valid also, as it appears (by using the timing marks and #1 as a reference) that TDC firing stroke positions the cam lobes facing away from each other and both slightly below horizontal?
It leaves very little room for error before one or the other cam ramp starts to press on the shim, so I double-checked by ensuring both shims & buckets could rotate easily by hand, to verify they weren't touching the ramp.

I got no air leakage to cases or radiator, so I guess one small mercy if the head gasket is OK, is that the head / deck is probably flat too.

 

Check out this setup. Your tester may be different from mine but this setup makes sense to me.
The "standard" compares a supply of air at 100 psi going through a .040" orifice to the amount of air the cylinder is holding after being supplied with the air coming through the orifice. So the setup gets you to set the regulator at 100 psi and the supply end that goes to the plug hole is open and flowing when setting the regulator. Then you hook up the supply to the hose screwed in the cylinder and monitor the gauge after the orifice. If the leak in all escape routes is less than what flows through a .040" orifice then the leak rate is less than 1 %. Any leak greater than the flow through the orifice will cause the downstream gauge to read less than the 100 psi feed. So if the down stream gauge read 90 psi, the leakdown would be 10%.

The .040" orifice was picked as a standard to represent the likely size of most piston ring air gap equivilents. So the air gap is a requirement and anything else that leaks is a measurable leak.

The 100 psi was picked as a standard of a common internal combustion engine under compression on a single stroke.

Anyway that's my engineered understanding of the process. One could watch " Youtube" and see all of the different clowns telling you " how to build your own cheap leakdown testers and I'm telling you the fundemental principals. It's easy to build your own tester provided you maintain the proper standards. A .040" orifice between 2   100 psi gauges and a pressure regulator on the supply end.

Fair enough Kawboy, and my leak-off kit is a commercial one that looks the same as your snapon example, same output guage zoning etc., and the principles of operation and orifice flow control are all the same.

The only difference appearing to be it has been calibrated to only use 30ish psi supply to zero the output guage.

There has been interesting discussion on mechanics / resto forums about the risks of running the test at high pressures, namely the fiddly nature of holding the piston stationary 'up' against those pressure with transmission lash (if locking via the wheel) or not always knowing if the big end might be on the cusp of rocking past TDC and swinging the other way whist the user is applying pressure (crank / alternator or whichever) the wrong way.

This was apparently raised as some mech's have reported being able to bend a con rod using the leak-off test kit at high pressure!
(I assume what they mean is that an artificially consistently high air supply can 'slam' the piston against whatever form of mechanical end-stop the user has in place, exposing the conrod to a shock loading it wouldn't normally see near the top of it's stroke?

Seems kinda ironic and I've no idea if that's possible, but for sure I saw my engine swing (when I hadn't sorted my locking approach quite accurately enough) the instant I connected the air at only 30 - 40psi.

Probably the most balanced conclusion I read from the restorers on those sites was that, as the leak-off is a percentage based analysis as you say, then as long as an input pressure is high enough to allow that type of ％ resolution analysis (which they consider 30 - 50psi to be) then there is no need to go higher / introduce any risk of damage.

Certainly I got definitive results from yesterday's test, so I'm happy the test is every bit as constructive as you suggest Kawboy and yes it's waaaay better to know where you're going in your strip-down work.

The combination of a dynamic compression test and the static leak-off test covers pretty much everthing you can do externally (regarding cylinders!) and for the minimal cost of the hardware is indeed highly recommended.