Gez

This should be an intersting thread!
After reading many threads of scoobs blowing to smithereens, i want to get a genral idea of why they do it. Many people have told me that the top part of the piston doesnt get much oil, therfore friction heats the piston too much, others have told me its the fuelling, etc etc etc. You always get different stories from different people. Is it a manufacturing fault? Why hasnt subaru looked at this problem? Surely there is enough evidence now for some action!
Is there any way of minimising the risk thus saving ££££ on rebuilds?
Any input appreciated.

Regards

Gez

[Edited by Gez - 8/5/2002 12:47:04 PM]

cwal1

Gez,

This has been the subject of much debate over the past year or two. Have a look here if you have time.

Chris.

[Edited by cwal1 - 8/5/2002 1:14:37 PM]

WREXY

Gez,

Well thanks m8. Full sunny and 39 deg Celsius today, tomorrow and everyday. Gets a bit too much when you cop it everyday. I have forgotten what rain is like.

Are you getting ready?

Cheers,

George.

[Edited by WREXY - 8/7/2002 12:48:19 PM]

gravelexpress

Dare I suggest that because Japanese legislation requires that a Speed Limiter is fitted to Cars

I still havent removed mine from my WRX so cuts out at 112mph.

Big Long shot - that Subaru designed the domestic model to only run up to this speed. However, at speeds above this the risk of det increases?

also

That the Uk model lower power output - was designed again not to risk det.

I gather that det is more likely at high rev, high gear?

Very poor theory I know, obvious flaw is that cars without modifications have failures as well. However, may explain why failure within Japan is low?

edit
There is no technical merit in the above! I was just sort of hoping someone was going to say, that JDM cars run in this country on Optimax, with the limiter still fitted, and not abused are not prone to big end failure!! (fingers crossed)

[Edited by gravelexpress - 8/7/2002 7:43:02 PM]

Pavlo

high tech racing engines canhave individual cylinder cam timing, ignition timing and fueling.

Result on an indycar motor lifted 30hp or so, just by optimising running for each cylinder.

One could easily measure EGTs for all 4 cylinders, afr measurement is more difficult as you need 4 sensors and they are quite bulky and each need a box of their own. But you could do it with off the shelf bits with enough money.

Paul[/i]

[Edited by Pavlo - 8/8/2002 9:54:47 AM]

Andy.F

I opt for the lower fuel pressure at No3. Yes, that's where the FP reg is but we all know how good the std fuel pump is the regulator will just be shut at anything over 280 bhp on most std pump equip't cars in an attempt to maintain the flow.
Hence lowest pressure there...heat....melt...bang

[Edited by Andy.F - 8/7/2002 9:36:47 PM]

nom

Lots of evidence that leads to no conclusion! So debate still rages...

Adam M

an old one with no conclusion.

I can tell you that no. 3 goes just as often as no. 4, and slightly more often than no. 2.

From this it is unlikely to be an oil feed problem.

possibly a sticky oil pressure relief valve? who knows.

submannz

It is mainly due to under fueling, due to the design of the Subaru fuel delivery, best way to fix is place a t junction and feed the fuel to both ends of your fuel rails.

Or even converting the stock fuel system from series to parallel.

Adam M

surely feeding fuel to both ends of the fuel rails would be converting from series to parallel?

plus dont you need to explain what to do with the fuel return before telling people this?

Gez

I thought it was to do with oil starvation more than fuel. Doesnt it make sense that if there is no fuel then the engine just stops. If there is no oil the engine keeps on running but pistons, bearings etc get so hot they melt.
Please correct me if im wrong.

Regards

Gez

WREXY

Maybe it's due to oil surge, leaving the bearings dry when going quick around bends, when the oil is a little under the full mark on the dipstick. Just a thought.

Cheers,

Wrexy.

nom

Gez - if one cylinder is getting less fuel than the others, then it may be causing lean running, leading to det - which as it's only one cylinder may almost be 'ignored' - at least enough to weaken the piston to breaking point.
One additional point with the det thing (and the fact that the one constant in all these cases is a held high rpm run) is that - I have heard - the knock sensor is unable to distinguish knock above engine noise over 6,000 rpm. So the engine may be knocking to bu99ery (especially if the intake temps are high - which they would be at sustained high rpm).

scobymars

I had the number 4 making a hole in piston and you could see it had been lean don't know why MAF or fuelling

Gez

So wouldnt it be possible to map the ecu to provide more fuel to poor old no3????

Ps: How you doin Wrexy? Weather nice in Greece?

nom

Nope - fuel is expected to be the same everywhere (I think ) but anyway it would onl drops when the fuel pressure drops, which makes things rather complicated. That's why the 'parallel rather than serial' idea is there - but the thing is, the cylinder at the end of the fuel line isn't necessarily the one that goes.

Gez

Yeah mate, cant wait!
Been having terrential rain back in London. Im just dreaming of going for a swim at Tolo.
You may like to know that i dont need to sell the Scoob anymore so ile be enjoying a few more years with it. Thats unless No3 goes :-)
Will be replacing ecu with either Autronic or Gems sytem. Only problem is that i dont know anyone that mapps Autronics stuff in UK.
Anyway, look forward in seeing you next week mate!

Cheers

Gerry
PS: out of curiosity, how much is your insurance in Greece?

Joules

I am running a GEMS ECU and have trimmed the fuel so it is richer on 3

Joules

Adam M

couple of points.

If the knock sensor cannot distinguish engine noise from det above 6000 then how comes the knocklink doesnt light up like a christmas tree when running at these engine speeds?

several tens of engines failures have been catalogued by bob rawle, many of the big end bearing failures have happened wihtout any evidence of det first, certainly mine had no signs of det on any of the pistons.

It did however show evidence of running lean on numbers 3 and 4, the last to receive fuel in the chain along the fuel rails.

this can be rectified by running modded fuel rails which are less restrictive and running parallel feed so that no one injector is fed after the other three.


some ecus allow individual cyylinder trimming such as motec gems and pectel.

I am fairly certain the jecs will not and am not sure about the link.

WREXY

Gez,

950Euros or 600GBP and that's everything, theft, fire, natural disasters, free yearly windscreen, etc, except if an accident is my fault, they only fix the other car. I have to fix mine. If I would of had full comprehensive insurance, it would have been 2000Euros or 1264GBP.

Cheers,

George.

Gez

And i thought i was paying more than the Greeks! I must be getting a good deal then. I pay £1200 and that covers all accidents whether or not its my fault. As well as that i get track cover but then i decide to smash it up then i have to pay the first £1000 where its usualy £250 when driving on the road. So life in Greece isnt as cheap as it used to be.

Are all those clubs still near Nafplion along that beach road?

Gerry

WREXY

Yep. The road that goes to Nea Kio.

nom

Don't know how the knocklink works or if it's just the Scoob knock-sensor that has the problem. I thought it was all, so they just blocked it off... But anyway, from what you say about Bob's collected info, the detting isn't a likely (specific) cause.

Fuel does look a possible one - but why would that be affected by a recent oil change (or is this just coincidence) and has any engines with the fuel-mod gone? I imagine not, but that wouldn't prove much as it doesn't appear to be that popular a mod, so statisitcally wouldn't mean much...

Gez

Im wondering if engines blow up in Japan as regularly as they do in the UK. I think not because Subaru would have been more aware of the problem and would have dealt with it.
If the cause was related to oil changes then Japan have Subarus blowing up VERY often as they give them an oil change every 4000KM (2500miles). I think it has something to do with the way the oil is changed. If memory serves correctly i remember there was a thread a while back on how to change oil on a subaru. A special technique was used to prevent damage.

This thread drives me mad. Its enough to drive anyone away from buying a subaru.

All we can do is collect data and hope that we can minimise the risk of failure. I think this topic has had an effect on Subaru residual values.

Cheers

GEZ

Pavlo

see here for some autronics info:

http://www.lancerregister.com/bbs/sh...threadid=10399

Paul

nom

If you're thinking of the same 'special technique' that I am then it happens to those cars as well (pre-filling oil filter, turning engine before starting, etc.).
Jap cars are a different spec. Although, as it seems the problem doesn't exist there, what is it that's different in the spec which is so important? I think something to do with the pistons, which would prevent it anyway...
Some enterprising garage should offer a special 'pre-emptive' on changing the things that go on the engine-pop as a one-stop type thing. Might get a few takers...

Gez

Cheers Pavlo
Cheers Nom

Pete Croney

Despite some forceful replies, there is nothing on this thread, so far, which is an explanation of the cause of this problem.

I do think it is det related, but not the sort that makes your pistons look like they've been in robot wars. If the whole lot goes off on the up stroke, it will do no piston damage but will put massive load onto the rod and, most notably, the big end bearing. This would also explain why many people that have this failure also have a bent crank. A bearing that runs out of oil just wears away the bearing shell. It doesn't have enough resistance to bend the crank.

My money is on excess in cylinder heat, causing "true" pre-detonation.

Common (but not exclusive) factors are that the car has been used at very high speed in the days before the failure, although not necessarily on WOT.

Some time ago I questioned the flow of air into the bonnet scoop at high speed, as water spray comes out and goes over the windscreen at very high speed. This may be a factor, but I have not had the time to check the airflow over the bonnet. Of note is that the 01 bonnet is much more steeply raked, overing the scoop a much better airflow. Also of note is that Adam and one other car I know of have suffered a failure and have front mount ICs which would have removed this potential problem.

The "special" procedure for changing an oil filter is to fill it with oil before it goes back on the car. This is first year college mechanics and nothing special.

It isn't oil surge as I've seen it happen to cars that have been on the motorway at speed for hours and then let go. People with oil pressure gauges have reported no problems until THE problem.

Oil pump relief valves have been suspected for some time, but I've inspected lots on failed engines and have yet to find one with bore wear or signs of sticking.

Cylinder number 4 is also a common one to fail, but its the FIRST injector to get fuel.

If you know the subject, I have added nothing here. I am just trying to prevent us going through 10 pages to get up to where we were before.

I should also add that whilst this problem may appear common, it is still thankfully rare and the vast majority of cars enjoy a totally reliable life.

nom

So could it be plain & simple bad manufacturing of a single part?

Gez

Pete, a while back i did post a topic relating to det and other factor, i also mentioned that a friend of mine who is pretty good at conputational fluid dynamics (CFD), has modelled an impreza bonnet on the computer and simulated airflow. I cant remember the exact speed but at just over 105moh the air does not enter the scoop. Instead small amounts of turbulence occured and the rest of the air just passed over the top of the vent. Ile get a copy of the simulation and ile put a picture of it a.s.a.p Tat will be a VERY interesting thread.