From pumpgas to E85 - Lambda/EGT problems - engine killed

Hi folks,

some weeks ago a friend tuned his car on the dyno. I assisted him during the session and unfortunately in the end the engine was killed due to pre-ignition.

We are talking about a 2 litre displacement engine with 4 cylinders.

Before the dyno run, the car was hitting about 620 horsepower with pumpgas (102 octane which we have at the fuel stations here in Germany)

My friend decided to go further and switched over to C85, which is similar to E85 but with racefuel quality.

Usually, E85 needs 30-35% more fuel comparing to pumpgas.

But in order to reach Lambda 0,78, only 15% more fuel was needed which is pretty strange.

So on the dyno we started with low boost and trimed fuel to the target Lambda. Allready here we saw pretty high EGTs in the turbo housing of over 1000°C (>1830°F)

I thought that combusting E85 should lower the EGTs.

So we started to add timing and after adding about 8°, EGTs started to lower, but in my opinion still too high.

Then we started to boost up the turbo and we got 730 horsepower... and a killed engine - one spark plug melted and therefore we are sure that pre-ignition was the result.

So now I need your help to sort out the open questions.

We think that air/fuel mixture was too lean, thererfore too high EGTs, therefore too high combustion temperatures, therefore melted the spark plug which finally resulted in pre-ignition.

After the engine has been rebuild, my friend started to trim fuel to 0,70 target Lambda and then we have about 35% more fuel.

In my opinion it is a very low Lambda, but it works! Usually this target is for methanol...

What could cause the Lambdatool to measure such a rich value? Or could it be a realistic value and if so, what is the reason because usually 0,78 is a reallistic target.

Lambda controller and sensors have been changed and calibrated. Everything is ok.

Could an aggressive cam setup with pretty much overlap can affect the Lambda even in high rpm and full boost?

Maybe someone can give us a hint in order to avoid to kill the engine again.

Thanx alot!

Cheers

Cyril

Was knock detection used during this set up? What type of engine management was used? On ethanol usually you can have a pretty good buffer room for the detonation threshold. But when you do reach that point, its usually too late and the damage is done. Was this pre-ignition or Detonation? Its a little suspect to me that the car ran happier at such a rich Lambda target.

Was there a pre-dyno inspection done for boost leaks? What were the IATs and target boost? Where are the egts being measured?

Melted ground or centre, what about the other cylinders? I would have used a spark plug at least one number colder/harder, maybe 2 if it didn't foul - did you, or were you using the stock heat range?

It may also have been a weak injector?

What was the damage to the piston and head - any evidence apparent?