Compression and Detonation Testing, #1

Builders,

I am going to cover this test project in a short series of articles. Below are some photos of the test hardware we will be using.

.

As the title implies, we are going to run two separate tests, and share the data. I am going to put the results on You Tube also, because I want builders to see the results in action.

.

The Compression test will be done first, it is fairly straight forward. I am building the engine, giving it a break in run, and then checking the differential compression, and running a base line full power run and noting the full RPM the engine makes.

I am equipping one head on the engine with small threaded ports next to spark plugs. For the base line run, these will be sealed. For the second test, I am going to open them up so the engine differential compression tests near zero 0/80) on those cylinders, and run the engine again and check the full RPM in comparison to the base line.

There is a widely held, but mistaken belief, even among aircraft owners and mechanics, that a cylinder with a differential compression of 40/80 makes half power, and one with 0/80 makes none. Neither of these are remotely true. The differential compression just measures flow against a known orifice, and does not correlate to any percentage of power out put. If the myth were true, the test engine with the three ports open would make less than 50% power output, but as we will see, it will do far better than that.  From professional dyno testing done by Dan Weseman and myself last year, we will be able to put HP numbers on the test stand RPM outputs, and the comparison will be enlightening to many people.

.

The second part of the test is to show how rapidly, and how much power is lost when an engine is detonating.  Again, I will establish a normal running base line for the engine, and run it at full power for one minute at that setting.  This will be immediately followed by intentionally over advancing the timing to make the engine detonate. Several things I want people to see:

The engine will actually appear to run better at idle.

It will not detonate at modest power output.

When the throttle is advanced to full power, it will run normally for a few seconds, and when it reaches ‘kindling temperature’ it will suddenly begin to loose power.  This power loss will be far in excess of the engine running with 3 cylinders with zero differential compression.

If the throttle is not retarded, the power will continue to drop, to well below the level required for flight.

.

We are working to get the engine instrumented with recording CHT’s and EGT’s for the detonation tests.  These will show a very sudden spike in CHT’s and a decrease in EGT’s, as all the BTU’s in the engine are destructively going into the heads. I will be working to capture this on video, so everyone can study the effect, know it when they feel it in their plane, and abort their test or take off. Again, a person with a correctly built engine and set timing will not experience detonation, but over the years, I have had a shockingly high percentage of people refuse to set their timing with a light, and I have people who never took advantage of my 39 free colleges to learn it directly from me.

.

The last part of the tests will be a disassembly of the engine, after it has been detonated to the point where it would no longer make enough power to sustain flight, to show the internal damage done.  Because we teach people to use forged pistons in Corvair motors, and because the head studs in the engine are 9″ long instead of 1″ like a Lycoming or Continental, the visual results will be different.  I will carefully document the changes on the engine as a reference for people inspecting engines.

.

The engine in the tests will not have a fifth bearing, nor will it have new seats and guides, and afterword, I will salvage some items like the prop hub and starter, but these test will still be expensive, costing several thousand dollars and perhaps two weeks of labor.  I am conducting them so builders will have a chance to learn something. It this happens the time and the money will have been well worth it.

.


.

Test engine, 2,700cc’s. This bottom end came out of one of the two core motors we found. The plastic orange plug set is a kit sold by SPA/Panther specifically for Corvair storage and painting. Note, old style bell housing cover in place of 5th bearing.  This engine isn’t going flying, and the tests are all taking place in the combustion chambers.

.


.

Heads getting reworked at SPA/Panther in advance of the tests. Travis Young (aka “Retro Black”) took care of the machining operations.  Their machine shop is set up for both engine and airframe parts production. The heads are getting perfect head gasket machined surfaces, and welded on intakes, but the valve job and the guides are just street car level work. Again, the motor will run for less than one hour, it doesn’t need new guides.

.


.

Engine getting decked in the mill. These surfaces are now perfectly true and flat. After severe detonation, warping in the heads can often be measured directly with a straight edge. The head gasket areas are often impressed into the heads, and the lower row of studs can become misaligned. Detonation puts a tremendous amount of heat into the heads, enough that they will plastically deform.  Most people think of detonation putting holes in pistons, but it only does that on cast pistons. On forged Corvair pistons, the damage is more subtle, the oil control rings become stuck in the grooves. I intentionally specified forged pistons because the engine will actually tolerate a lot of abuse with them. They are a margin  of safety on small mistakes, but they will not provide immunity for fools.

.


.

Head bolted to a fixture plate in the Mill. The unit is a very nice Chevalier with DRO. It is in good shape, having lived in a tool room most of it’s service life. Next time you See Dan at a College or an Airshow, you can get him to share the story of how he bought this piece of equipment for $6.

.

wewjr.

.

About William Wynne
I have been continuously building, testing and flying Corvair engines since 1989. Information, parts and components that we developed and tested are now flying on several hundred Corvair powered aircraft. I earned a Bachelor of Science in Professional Aeronautics and an A&P license from Embry-Riddle Aeronautical University, and have a proven 20 year track record of effectively teaching homebuilders how to create and fly their own Corvair powered planes. Much of this is chronicled at www.FlyCorvair.com and in more than 50 magazine articles.

12 Responses to Compression and Detonation Testing, #1

  1. michael mckosky says:

    Ah, good technical stuff! See, I always thought you were a good guy. Thanks for the article, very valuable for people like myself that don’t have a background in these matters.

  2. TONY CRAWFORD says:

    Awesome!! THANK YOU!! I still need to attend one or more of your colleges and build my motor. At this moment i am laid up after my recent surgery. I just went through a Total Ankle Replacement surgery. No weight bearing for 10 weeks. Again. Thanks for all you do for us aviators!!

    Tony

  3. Wesley Ragle says:

    William,

    Quick question for you: Is there a suitable sensor for directly detecting corvair detonation?

    I understand the rise in CHT and drop in EGT, but was just wondering why there is not some sort of “knock” sensor on all aircraft engines.

    Thanks,
    Wes

  4. Harold Bickford says:

    William,
    This is a very proactive effort. People should pay attention but of course we know some resist learning. It is mystifying yet not hard aside from admitting that prior notions or conventional wisdom might be less than accurate.

    Harold and Edi

  5. Lane Seidel says:

    I’m with Mr. Bickford! It will be a great learning experience for the people that love to learn things about their engine. I have been to I think 7 or 8 colleges, have a great running engine and I still learn something new every time I attend! This is a fantastic experiment. Well actually it isn’t an experiment but a demonstration. The outcome is already known. Hey tell Dan I’ll give him $12.00 for the machine. That’s 100% profit! Still I wouldn’t know how to use it!

  6. jaksno says:

    Another demonstration of the ‘above and beyond’ nature of your life’s work…well done! Thank you for your stubborn pursuit of safety and best practices evidence based education for homebuilt aviation and more!

  7. mark says:

    can you attach a motion sensor as well so people can see a graph displaying the difference in vibration? all the best with the testing.

  8. Andy Elliott says:

    Note that if you drilled bleed holes of various sizes in sets of the small head plugs, you could probably adjust them via trial & error to get any compression ratio you wanted, and add some study of “this much compression loss does this…”

  9. George Willenbrock, small time customer, thankful reader says:

    Cannons and Flintlocks, fuses that BURN; all holes into the combustion chamber. All leak. All worked. Just a question of how big a hole gives how much pressure (till it bleeds off) for how long, will results be useful. Note: big leak happens at the muzzle, don’t look down the barrel trying to see the leak.

  10. Mike Crowley says:

    What a fantastic test. Thank you so much for going to such lengths to help others learn, build, fly!

  11. David Swann says:

    Hey William you should try to get hold of a multipoint data recorder, that way you could look at all your data points simultaneously. Id bet Paul could fine one?

  12. Pj Jacobsen says:

    When will your test results be out ?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: