Oh the humanity - 4 blown pistons - I'd like your diagnosis why

[JohnJL]

The motor had less than an hour on it at the track and started making bad noises and blowing smoke out the oil tank.

I wasn't logging at the time so I only have the motor autopsy itself to try and figure out what happened.

This motor was running well prior to its last rebuild, which was purely to fix an oil leak on the case seam over the winter.

The same pistons, cams, etc. its a short stroke 3.2 with 8.0 compression coated JE pistons, twin-plugged, Adaptronic EFI, crankfire, Carrera intake, headers. 12 psi boost, 350 hp and 350 ft-lbs at the rear wheels.

Pistons 1, 3, 4 and 6 were total toast. Burned through or cracked on the exhaust side of the pistons. Three of them are missing chunks from the lip around the intake valve pocket. One piston, #2, has a moon-shaped mark on the outside edge on the intake side where the others are burned away. Only this one has a mark. #5 is cracked through the intake valve lip but not yet broken off.

This was - ahem - clearly a installation/setting error which means its definitely my fault. I am keen to hear theories about what exactly that might have been as I don't just want to slap it all together again and have another engine blow.

Here are my thoughts;

All the parts of the motor worked together before the rebuild.
The knock sensor didn't show any warnings on the earlier logs. I wasn't logging at the time of the failure.
The pattern of failure is consistant across all of the cylinders so I can rule out any random manufacturing problem with the pistons.

THis looks like massive detonation which is due to
1. cam timing
2. a fueling problem
- bad fuel
- underfuelling
- fuel pressure problem
- ecu error
3. ignition problem
- way advanced or
- way retarded

interestingly, the magnetic pickup on the ignition crank pickup was also completely f'd Check out the magnet pickup...its shattered like it was shot...

[JohnJL]

more carnage

[JohnJL]

Amazingly, none of the cylinders were scored. All of the compressin rings were shattered but intact in the lands. Two of the middle rings were also shattered but in situ. None of the cylinders had any pieces in them, seems the debris burned up or slipped out the exhaust valves. Two heads have some dings but I think all are salvageable.

[Lapkritis]

Lean.

[Raceboy]

Lean mixture -> too hot -> pre-ignition -> detonation = carnage.

[K Sykes]

Could the mag pickup been damaged first and thrown the timing out of whack, causing the rest of the damage?

[neilca]

Did you check the piston to valve clearance? #1 and #6 piston looks like it has a strike where all the pistons are failing. I suspect the valve pocket may be too small or too close to the valve.

I have struggled with the same crank sensor using Megasquirt3. That is a VR sensor off of a Ford F150. The VR sensor outputs a small voltage at low RPM and a higher voltage at high RPM. If your computer cannot handle the voltage, like mine, then you loose the signal. My car acted like I was hitting a rev limiter at 6000 rpm. I am in the process of switching to a Hall effect sensor. The output of a Hall effect sensor is a square wave the computer can read directly and it doesn't vary with RPM.

[Mark Henry]

Were you using a wide band?

When accelerating you want to see 12.5 to a max of 13.5 to 1 AFR on an aircooled.

IMHO knock sensors are worthless on an aircooled. I rely on my wide band and head temp gauge for the most part.

[trond]

I'm with K Sykes; the VR sensor caused ignition timing issue caused the carnage

[Raceboy]

VR sensor can be damaged and I have seen them failing with results having different ignition zero point but it has never exceeded few degrees and judging by damage on the pistons I would say pre-ignition was first which caused detonation.
Just my two cents.

[JFairman]

I think the earlier diagnosis are probably correct.

Regarding pre-ignition or detonation the two are different and this editorial explains both and says detonation occurs before pre-ignition.
Engine Basics: Detonation and Pre-Ignition by Allen W. Cline

Most motors will put up with some detonation for a while and Pre-ignition then destroys the motor very quickly.

[johndaroza]

JFairman
trondK Sykes
Raceboy
Lapkritis
Me too think it was lean and mean!

Did it run good for a time? I bet it had some big power before the bad!

[afterburn 549]

Truly it looks like piston to valve interference problems ..

[Mark Henry]

Originally Posted by afterburn 549

Truly it looks like piston to valve interference problems ..

I'd expect to see a shiny area or ding if that was true

[afterburn 549]

well....at this point not being there no touch no feely...just a pic...the valves look guilty..what ever it was , with this board you will get answers if its not the valves
I just looked again
I am voting missed shift (valve float)

[sjf911]

That looks like detonation damage so high AFR's, ignition advance issues, or bad fuel would be most likely.

[afterburn 549]

Bad AFR there should be "burn through".........

[thumbdoctor]

Detonation !, Check the condition of your upper con ron rod bearing shells. They will most likely show some signs of crush. Usually oil temp peaking is a sign of early deto on air cooled. Max combustion pressure will mechanically damage ring lands and valve pockets at the squish band. Note the cleanliness of the edge of the pistons / combustion chambers.

[Bob Kontak]

Originally Posted by afterburn 549

I am voting missed shift (valve float)

Not discounting this at all. I am not a hard core motor guy but I am not convinced I am seeing melting or heat fatigue.

Also, no stress indication on the walls says something.

Regardless, look at the valves closely for any hints.

Also.....

Was cam timing changed from old?

Were the heads kissed to clean up the cylinder mounting surface? Not sure if this is done on the 3.2. My SC has a big crush ring - I remember JW saying it is for oil sealing - where the machine shop cleans up the surface on the heads for a better seal. They took six thousandths of an inch off.

Any change to the cylinder base gasket. They are normally about ten thousandths thick (per the parts manual - .25 MM) Can a thinner one be installed? Was it installed?

[Lapkritis]

The cracked but not yet disintegrated piston indicates this was a temperature/mixture issue not a physical interference.