Big end failure, det or oil surge? (long post)
Posted: Mon Sep 07, 2009 8:16 pm
Well, after 5 years in 3 Imprezas and loads of trackdays I've suffered the big-end death rattle. Had to happen at some point I guess, but I'd like to try understand more precisely what caused it...
Background: STi6 RA Lim with Cosworth baffle plate, breathers going to RCM can returning to sump, no oil cooler yet. Road tyres (Marangoni Zeta Linea). Standard 2.0 block, Zerosports hybrid VF28, Samco inlet tract, hybrid FMIC, ported headers, 550s, Apexi PFC, around 340bhp at 1.5bar.
Suffered a big end failure after 7-8 sessions at Mallory on Saturday, engine still runs (idles fine) but has the characteristic rattle as revs die away. Exact cause not clear to me at moment, maybe someone here can advise. This may read like a catalog of errors, but bear with me.
In reverse cronological order:
1. made aware of bad mehanical noise by the detalarm, all LEDs lit at slightest boost, but no significant knock count on the PFC hand commander; retired to investigate problem.
2. couple of laps earlier, filled up with half-tank of 97RON on top of existing half-tank of 99RON plus booster. Although car is mapped for 99, we were monitoring carefully for det and didn't want to risk fuel surge. Don't think there's any evidence of det on that fuel.
3. one session earlier, stayed out a bit longer than earlier whilst reeling in a 911 GT3, and reached 130deg oil temp. Whoops. Earlier sessions were less than 10mins, reaching 110deg. Millers CFS 10w60 which has done one previous trackday and 4k road miles. Datalogit trace shows 2 bouts of 'knock' count of 15-22 at 7k rpm, whereas engine normally runs a count of around 10.
4. couple of sessions earlier, reduced boost to 1.3-1.4bar on AVCR after seeing knock count around 20 on hand commander after a session, although detalarm remained quiet (green). Started logging to see what was happening during session.
5. error of judgement led to a spin exiting Gerrards. One wheel onto grass at top of 4th gear (RA gearing, 110mph?) and 720deg spin narrowly missing barrier. Foot on clutch and engine still running. Back to pits to check and calm down.
6. first time at Mallory and realised it's v.quick except for the hairpin. Fuel consumption through the roof. Gerrards long right-hander (like Coram or Chris curve but faster) could see oil pressure drop from 6.0 to 5.5 bar on some laps. Had fuel surge once (I think) at Devil's elbow, a 4th gear lefthander that drops away.
7. before the trackday had various short bouts of det due firstly to dodgy crank position sensor (now replaced) and to compressor surge (Paul at Zen's assessment, from the logs). Boost reduced in midrange to avoid the surge.
8. Oulton trackday a few months ago, before compressor surge problem was realised, with brief bouts of knock count 30-40. All apparently ok subsequently, but perhaps the seeds of damage were sown back then?
Apologies for going on a bit. Anyone care to comment on relative importance of the possible causes?
Andy
Background: STi6 RA Lim with Cosworth baffle plate, breathers going to RCM can returning to sump, no oil cooler yet. Road tyres (Marangoni Zeta Linea). Standard 2.0 block, Zerosports hybrid VF28, Samco inlet tract, hybrid FMIC, ported headers, 550s, Apexi PFC, around 340bhp at 1.5bar.
Suffered a big end failure after 7-8 sessions at Mallory on Saturday, engine still runs (idles fine) but has the characteristic rattle as revs die away. Exact cause not clear to me at moment, maybe someone here can advise. This may read like a catalog of errors, but bear with me.
In reverse cronological order:
1. made aware of bad mehanical noise by the detalarm, all LEDs lit at slightest boost, but no significant knock count on the PFC hand commander; retired to investigate problem.
2. couple of laps earlier, filled up with half-tank of 97RON on top of existing half-tank of 99RON plus booster. Although car is mapped for 99, we were monitoring carefully for det and didn't want to risk fuel surge. Don't think there's any evidence of det on that fuel.
3. one session earlier, stayed out a bit longer than earlier whilst reeling in a 911 GT3, and reached 130deg oil temp. Whoops. Earlier sessions were less than 10mins, reaching 110deg. Millers CFS 10w60 which has done one previous trackday and 4k road miles. Datalogit trace shows 2 bouts of 'knock' count of 15-22 at 7k rpm, whereas engine normally runs a count of around 10.
4. couple of sessions earlier, reduced boost to 1.3-1.4bar on AVCR after seeing knock count around 20 on hand commander after a session, although detalarm remained quiet (green). Started logging to see what was happening during session.
5. error of judgement led to a spin exiting Gerrards. One wheel onto grass at top of 4th gear (RA gearing, 110mph?) and 720deg spin narrowly missing barrier. Foot on clutch and engine still running. Back to pits to check and calm down.
6. first time at Mallory and realised it's v.quick except for the hairpin. Fuel consumption through the roof. Gerrards long right-hander (like Coram or Chris curve but faster) could see oil pressure drop from 6.0 to 5.5 bar on some laps. Had fuel surge once (I think) at Devil's elbow, a 4th gear lefthander that drops away.
7. before the trackday had various short bouts of det due firstly to dodgy crank position sensor (now replaced) and to compressor surge (Paul at Zen's assessment, from the logs). Boost reduced in midrange to avoid the surge.
8. Oulton trackday a few months ago, before compressor surge problem was realised, with brief bouts of knock count 30-40. All apparently ok subsequently, but perhaps the seeds of damage were sown back then?
Apologies for going on a bit. Anyone care to comment on relative importance of the possible causes?
Andy