Burnt piston help

[jcas]

Burnt a hole in my piston (small half-nickel sized, on the exhaust side, right at the edge), just trying to figure why. I did have this piston in the bike for the last 18 months, so I figured this was coming. The bike was running well, made some recent changes, this is what I can think of/what I did:
Changed fuel from 110 to 109 octane (local shop sells cheaper):
Changed dome from pump-gas to high compression:
I bought a Dynatek ignition, got more spark, but kept gap at .018):
I'm running a drag-ported ESR 310, High compression dome, 40.5 mm Keihin carb. Any help would be appreciated. Thanks

[cdrookie]

just a guess but i'll say it was the ignition...

[NJ450rider]

antifreeze leaking into the cylinder will eat a hole through your piston as well.

[jcas]

I did notice that the coolant level was a little low....also I'm going back to stock ignition for a few hours, then pull the cool head off and check things out.

[GPracer2500]

Does it look something like this?

http://img300.imageshack.us/img300/2018/p7090021pe2.jpg

If so, I'd say detonation was the failure mode. And I'll guess one or more of the following caused the detonation:

--too much compression
--too low an octane rating
--too much ignition timing advance
--lean jetting (maybe from an air leak, if you're sure the jetting specs were good)
--incorrect squish clearance
--too hot engine temps will exacerbate any situation that is causing detonation to occur....in cases where the octane rating is barely sufficient for the engine's tune, too hot engine temps alone can be enough to cause detonation


Compression, octane rating, and ignition advance are all closely tied together. The more compression and ign adv utilized the higher the octane rating must be to prevent detonation and pistons that look like the above. You potentially changed all three at the same time so it could be just one of them that is incompatible with the others or all three that are not right.

--How much compression did it pump after assembly?
--Did you check the squish clearance?
--Did you run a leak-down test after assembly?
--I assume you're using a correct spark plug?
--Exactly which fuels where you using (brand, type)?
--Does the Dynatek have adjustable ignition timing?

Those are some of the places I would start investigating.

$.02

[jcas]

WOW, that looks real similar....What I'm going to do after the rebuild is go back to stock ignition AND go back to the Sunoco 110 that I had been using. The fuel was a 109 octane, I believe it was called "magnum" or something like that. It was a little cheaper than the 110, so I thought I'd give it a shot. Thanks GPracer, that was informative. Question: How do I check the squish clearance? Everything else was checked and was good. Again, thanks for the help fellas....

[GPracer2500]

Originally posted by jcas
....Question: How do I check the squish clearance?....

In case you're not familiar with squish clearance, here's a diagram. It's the distance between the head and the piston in the area of the squish band:

http://www.muller.net/mullermachine/docs/squish1.gif

The squish band is the "band" of very tight piston-to-head clearance around the edge of the combustion chamber. That clearance dimension has a strong influence on how detonation prone an engine is (note: engine knock, spark knock, pinging, auto-ignition are all names for the abnormal combustion phenomenon known as detonation). The a/f mixture located around the very edges of the combustion chamber (the so called "end-gases") is the fraction of mixture that can succumb to auto-ignition if it is allowed to reach a certain temperature (note: auto-ignition should not be confused with pre-ignition--that's a separate abnormal combustion phenomenon).

During each combustion cycle there's a race of sorts to consume the end-gases. The two competing elements are 1) the advancing flame front begun by the spark plug and 2) the increasing temperature and pressure subjected to the end-gases. We want the flame front to consume the end-gases before they are allowed reach the point where free radical activity accelerated by heat reaches a point at which the gasoline breaks down into new chemicals that are so unstable they will ignite all by themselves. Got it? Good. ;) In a nutshell, that is detonation and it's bad for engines.

There are a number of reasons why deto is bad. The one that causes top ring lands to get blasted away involves the insulation of the piston by a thin boundary layer of gases that clings to the piston. The gases produced from combustion are hot enough to melt aluminum. But this normally doesn't happen because these hot gases don't ever actually touch the piston. They are blocked by the insulating boundary layer. When deto occurs, this boundary layer can become so compromised that it can no longer shield the piston from the combustion gases. Since deto tends to be confined to the outer edges of the piston, this is where the boundary layer is most often compromised--the exhaust side especially since most pistons naturally run hotter there.

Here's another piston that was on it's way to looking like that first one. Same thing was happening it just hadn't advanced as far. Notice the lightly shaded "ring" that goes around the top edge of the piston? The aluminum was being blasted away by detonation. We're looking straight-on to the exhaust side and you can see the top ring land is starting to erode.

http://img135.imageshack.us/img135/5900/p7090025uh2.jpg


Squish bands combat this situation in (at least) these ways: 1) The volume of the end gas is reduced because a/f mixture is pushed closer to the center of the chamber. 2) The end-gas remaining in the squish band is in very close proximity to the relatively cool surfaces of the head and piston. This lets those metal surfaces absorb heat from the end-gases. 3) The squish band shields the end-gas from radiant heating that would otherwise take place.

Squish band combustion chamber designs allow us to utilize much higher cranking compressions then would otherwise be possible. Squish bands affect engine performance in other ways too. There's a turbulence effect created (which is good). Also, the dimensions of a squish band influence how much charge is effectively utilized (as well as how effective it's anti-deto benefits are). And there may be other elements of a squish band I'm not thinking of at the moment or am not aware of.

A loose, high-clearance squish band clearance defeats it's ability to function in the three ways listed above. For the most part, the tighter the squish clearance, the better. It can't be zero because the parts need some room to grow as they heat up, as the rod stretches at TDC, and as the crank/rod bearings slowly wear out. If one runs bleeding-edge-tight squish, you've got to check the crank a lot and make sure it's not slowly loosening up or the piston will contact the head.



The way I've measured squish clearance recently is with modeling clay. It's very similar to checking the piston-to-valve clearance on a four stroke, if you're familiar with that.

With the head off, take a small ball of clay and place it in the head at mulitiple points around the edge of the dome. Do at least two spots: on the right and left where the wristpin is. If you can avoid areas where the piston sweeps past a port, that's good. The clay can catch on a port window sometimes if the clay sticks to the piston.

I did four spots on this one. You want the clay to go right up to the edge but you don't want it to get caught between the head and cylinder.

http://i34.photobucket.com/albums/d133/gpracer2500/P9120120Medium.jpg

Then install the head (with gasket). Slowly turn the engine over. The piston will smash the clay in the squish band. Remove the head and examine the thickness of the clay. Take a razor blade and cut the clay so you can see it's thickness. Measure with calipers to get the squish clearance.

On this one the clay stuck to the piston instead of staying in the head.

http://i34.photobucket.com/albums/d133/gpracer2500/P9120124Medium.jpg

http://i34.photobucket.com/albums/d133/gpracer2500/P9120122Medium.jpg


Solder can be used instead of clay. That's probably not quite as messy and possibly easier to measure. Use a flux core solder that is close--but larger--than the squish you are expecting. Maybe try .032" solder if you think the squish is tight. If your squish clearance is larger than .030" then obviously you'll need thicker solder to measure it. Check your hardware store for common diameters. I used clay in the above engine because I couldn't quickly find solder thicker than .062". That engine (a 500) has a stock squish clearance of a rather massive .100". I'd like it to be more like .045" but haven't gotten around to changing it. But anyway, trying something around .050" solder is probably a safe bet. Just turn the engine over slowly by hand with the plug out. You don't want to kick it over hard and risk cracking something if your squish is a lot tighter than you're expecting.

Here's an example of how to place solder in there to measure the squish...

http://www.asrcorporation.com/motorcycles/images/03CR250/BigBore/Squish.JPG

Here's another solder arrangement. Multiple pieces of solder held in place with grease.

http://www.nrhsperformance.com/images/squishsolder.jpg



As usual, I've probably gone overboard and gotten into far more detail than anyone actually cares about. No worries though; writing it out helps me keep it straight in my own head....

[wilkin250r]

If you're checking with clay, smear a little motor oil on the head where the clay will contact it. This will ensure that the clay sticks to the piston, but not the head. You don't want the clay to pull apart and give you an incorrect reading.


For a good article on squish bands, read The Art of Squishing Things Till They Give (Power) (http://www.aircooled-rd.com/default.asp?txtPage=squish1.htm)

The article was originally written for the Yamaha RD350 streetbike, but the principle applies to any two-stroke. (The RD350 was air-cooled, the next version was liquid-cooled RZ350, and was the same engine that later became the Banshee )

[jcas]

Alright, thanks for the class, I'll give it a shot...