Today's Research Topic, Leakdown Testers Options

[McMark]

I inherited a Harbor Freight Leakdown Tester (PN: 94190) from someone and was playing around with it lately, checking out a 2.7 CIS 911 engine that will be going up for sale soon ( (IMG:style_emoticons/default/shades.gif) ).

I learned a few things that I feel should be common knowledge when looking at leakdown results and leakdown testers.

Backstory
I hooked up the HF gauge out of the box and immediately ran into confusion. The instructions indicated that you should pressurize the tester and adjust it to 0% leakage, then hook it to the cylinder and read the leakdown percentage. Well that was a bust, right off the bat. Setting up the gauge as instructed resulted in an inlet pressure of around 35psi. Most people on the internet recommend running a 100psi inlet pressure. So naturally, you would increase the inlet pressure to that recommended level, but the HF then drops to 0% leakdown, as pressure increases.

Initially I thought this was a gauge issue, so I ordered a couple quality pressure gauges from McMaster car and replaced the HF gauges with the McMaster units. This allowed me to compare inlet pressure with cylinder pressure, which is what the leakdown tester is supposed to do. But alas, these numbers still didn't make sense.

What I Learned
It turns out that a two gauge leakdown tester is showing you inlet pressure on one gauge, which is essentially your compressor pressure, regulated to 100psi (or whatever you choose). The other gauge shows you the pressure inside the cylinder. But between those two gauges there is an orifice (precisely measured small hole). The size of this orifice is EXTREMELY IMPORTANT.

What a leakdown tester is actually showing you is a comparison between the amount of air that can squeeze through the orifice COMPARED to the amount of air that can squeeze through the engine leak path (ring gaps, burned valves, etc). It's more like a ratio than a percentage. So if you think of it as a ratio (Orifice : Leakage), then it's easy to realize that if you change the orifice size you change the ratio, even though the leaks on the engine haven't changed.

Bottom Line
I completely disassembled the HF tester and the orifice in that unit is >2.5mm. While there is no particular standard for leakdown testers MOST units and most people are using a 0.040"/1mm orifice. This is referenced in aircraft leakdown standards, and has basically been adopted by the automotive industry. Posts I found on this subject confirm that expensive testers (such as MAC) come set up with a 1mm orifice.

The HF tester should not be used without modification. My next step is to try plugging the HF orifice and redrilling it to 1mm. This should make my HF gauge work like it should. I'll then reinstall the HF gauges and see how they compare to my quality gauges.

Be wary of leakdown numbers. I'm posting this because it's imperative that anyone who's shopping for 911 engines be aware that it's not enough to simply know the leakdown numbers. You need to know the inlet pressure that was used and the brand of gauge that was used.

(IMG:http://www.914world.com/bbs2/uploads_offsite/forums.pelicanparts.com-419-1494258059.1.JPG)

[McMark]

I'd like to know the % the HF unit is off.

Unfortunately, it just doesn't work like that. Airflow is not linear. You can't just use a quick formula to translate. The HF orifice @ 2.5mm is so huge and any normal leakage is so small in comparison.

Leakdown percentages are one of those ideas that get passed around, but most people don't stop to think about what the numbers mean. What IS 10% leakdown? 10% of what? You might be tempted to guess that it means 10% of the air in the cylinder leaks out. That's incorrect. If it were correct your engine wouldn't run.

But back to the idea of a scaling calculation-- Using some random airflow calculator I googled, a 2.5mm orifice will flow over 6x as much air as a 1mm orifice at 100psi (in case you doubted airflow isn't linear). So 12% leakdown would scale to 2% leakdown. 911 guys are posting leakdown values of 2%, which even if scaling worked would be 0.333%. Your gauge doesn't read that sort of resolution...

And that's exactly why the 2.5mm orifice is a problem. The orifice is really what sets the scale or resolution. When you say your engine has 10% leakdown, what you're REALLY saying is that your engine has leakage equivalent to 10% of the maximum airflow through a 1mm orifice.

And none of that even accounts for the idea that 35psi is terribly low compared to combustion pressure, which is what we're really trying to get at. How much air/pressure/compression/power is lost to leakage when the fuel ignites. If you're trying to simulate 150-200psi using 35psi you're bound to get inaccurate readings simply because you're not pushing hard enough.