Pete,

And on top of all that, a bigger hole will reduce the oil pressure.

Mark.

 

so is it safe to say that the piston slap developes on the rear passenger side and the big-end is failing on the rear drivers side?.

John F don't I get an apology too

 

Rob,

Cheeky sod, the reason you don't take your car out on the track, is because you're too busy showing us how to drive ours

Mark.

 

Hi all!

The oil path to the No. 3 bearing is the longest in the chain, I'll try to explain...

All following references to left / right are from the front of the car, ie stood infront of itl, looking at it.

The oil feed enters the right hand side of the oil pump from the sump. The pump is located at the front of the block in the middle, it is driven directly off the crank. Oil exits the left hand side of the pump and goes to the oil cooler. It passes through this then through the oil filter radialy inwards, then up the middle of the filter through the middle of the cooler and up into the left hand cylinder block. The oilway at this point is quite large, perhaps 15mm in diameter?

Oil goes straight up and then forks, the left hand main gallery going pretty much straight back, and a cross-tube provides oil to the right hand block half, which provides oil to the right hand oil gallery, which is pretty much level / parallel to the left hand one.

The left hand oil gallery feeds (IIRC) Numbers 1, 2 and 5 main bearings. Feeder holes in the crank take oil from No. 1 and No. 5 main bearings up to No. 1 and No. 4 big end bearings. At the ned of the left hand side main oil gallery there is a feeder up to the left hand cylinder head. This is substantially smaller than the main oil gallery diameter, so the odds of this depleting oil is pretty slim. Upon entering the head this oil feed forks off (!) to feed both the turbo and the left hand side camshafts etc.

On the right hand side, the main oil gallery feeds the right hand cylinder head (need to double check, but think it does so from the front of the engine, rather than the back as in the case of the left side) and the No. 3 and 4 main bearings. The oil gallery doesn't appear to extend all the way to the back of the block according to the diagram, but I will verify this tomorrow.

The No. 2 and 3 big end bearings are BOTH fed from the No. 3 main bearing. This is where it gets a bit weird... abviously No. 3 main bearing will require more oil, so it can flow sufficient for both No. 2 and No. 3 big end bearings, perhaps this is why there are problems, BUT why does No. 2 never seem to let go? Weird, maybe it's simply because No. 3 is running hotter (but, the oil feed length to No 2 and 3 big end bearings are the same). No. 2 and No. 3 crank throws are also in the same direction....

A possible "solution" would be to plug the existing oil feed from No. 3 main bearing to No. 3 big end bearing and drill a new feeder from No. 4 main bearing to the No. 3 big end bearing, perhaps by threading the hole deeply enough to wind an allen grub screw down, threadlocking it in the process?

Interesting also is the fact that the little end bearings on the rods do not have oil force fed to them, rather the rods have two holes which allow passive oil feed, presumably from oil splash off the back of the pistons. The WRC cars do (apparently) run forced oil feeds to the little end bearings, so presumably they MUST feed No. 3 big end from somewhere else, I doubt it could supply enough oil for No. 2 and 3 big and little end bearings!

It would be possible to drill through the main oil gallery on the right hand cylinder block and insert a new feed to No. 3 main bearing. This should not affect the function of the gallery since it effectively "stops" there anyway. Unfortunatley, getting clean oil to it will be difficult. Although there are plugs on the high and low pressure side of the oil pump, to which one could attach a news feed, this feed would be pre-filter, so it would make sense to not only install the feed but also a separate filter.

None of this, however, has answered the underlying question of why it is almost always No. 3 that lets go, when it has an "identical" feed to No.2; perhaps No.2 is a "greedy" bearing and robs flow from No.3? Perhaps it is a combination of higher-than-average temps in No. 3 cylinder, coupled with slight det causing the oil film to break down, coupled with the fact that oil is shared between No. 2 and No. 3?

It is also quite interesting that "No. 3 runs lean" on full chat. It is true that it is at the end of the fuel rail, but it is also closest to the regulator! Also, its manifold branch is the closest to the turbo (and therefore presumably the hottest?) which would tend to make it run richer, due to the same amount of fuel being delivered to a less dense charge. It is also possible, that the fuel is being heated on its way round the rails and that this is in some way affecting the mixture.

With regard to punishment, it is certainly possible to exceed track conditions on the road. It takes a long time for my tyres to overheat on the track, yet I have had a set go off on the road within 15 miles! It's also not possible on (most) tracks to be bouncing off the rev limiter in 5th, but perfectly possible and legal on the road (Autobahn). I recall doing a 30km stretch like this but gave up because the fuel economy was naff (hey, big surprise, Scoob has drag coefficient of a block of flats!).

I'm not the most mechanically sympathetic driver in the world :-) I killed one gearbox while mapping, and one engine while playing around with a full throttle shifter [very loud pops and bangs, ALS is tame by comparison! ever seen a 3-4 foot blue flame out the back of a car? :-)]. My current engine is poorly, having lost compression after being "tortured" by being run 1000 miles at on no boost and max 3500 RPM, then up to 6000 RPM and 0.4 bar boost, finally when I was feeling brave and raised the boost to 0.8 bar (still only 6000 RPM) it let go.... a stock car runs 1 bar! I think it may have been a duff engine! Fortunately I have a datalog of the time it let go :-) The curious thing is that *BOTH* engine have let go on No. 4!

Suffice to say I will be looking at everything in minute detail when I come to rebuild this engine.... and will be making a few "choice mods" as well :-)

Cheers,

Pat.

 

Just wanted to say this is a truly excellent thread with loads of great info. Just a shame it has to be about something bad rather than good.

Glad i dont take my car on trackdays too.

Sunil

 

Otis,
It is interesting that IM told you the piston slap problem was caused by the short piston skirt.
I don't think anybody else, except for myself on my last posting, had made reference to the fact that piston slap is a consequence of piston skirt design, allied to bore clearance.


Pat,
Thanks for the info on the oil feeds.

I wouldn't think tapping the main bearing journal oil feed to No 3 big end is a wise move.

Surely it would increase stresses within the crank at a position that is taking imense loads.
All corner radii on big end and main journals have to be smooth and of a reasonable size to lessen the chance of fatigue cracks.
Its a long time since I have examined a crank closely, but I'm fairly sure the rads are in the order of 0.060" or larger.

If you were to tap a threaded hole as you suggest, the corner radii at the top of the thread (the top of the 'V') will be in the order of 0.010" or less.

Additionally, the expansion and contraction rates of the grub screw and crank may well differ, which in a static position probably wouldn't cause a problem, but in this application, would be whizzing round at 6/7K.

Unless uou know that this sort of mod is commonly carried out with cast iron or even nitrided steel cranks, I think it may be unwise.

Cheers MTR

 

Pat,
Just hypothesising about the forced oil feed to the little end bearings on the WRC cars.

I presume the STD con rod ha san oil hol;e drilled at the piston side of the big end, to squirt oil up into the bores, once on every revolution, which then also lubricates the oil feed to the little end bearing.
I would imagine a con rod with an oil gallery drilled up its length to directly feed the little end bearing, would drain VERY LITTLE oil, as the gudgeon pin to small end bearing clearance is usually very small.

Either way it would only recieve one squirt on each revolution, as the oil gallery drilled in the big end journal aligns with the drilled gallery up the rod.
So it should be possible to utilise the No3 main bearing feed.

I'll bet Rodger Duckworth knows the answer to all these problems, and more besides.
(For those that don't know who he is, he is Keith Duckworths', the 'worth' out of Cosworth fame, son, and rallies an Impreza Turbo.

Cheers MTR

 

It would be interesting to see what the fuel contamination of the oils recovered from expired engines is. Has anyone had their oil analysed yet from an engine that's gone pop on number 3 cyl?

 

John Felstead,
The undersize piston/oversize bore explanation given by RN Golden, may not be nonsense after all.
I found this article about a Honda Integra VTEC, with oversize bores, and it was from a batch of at least 60 engines.

 

MTR

This isnt helping the debate now is it?

 

Pat,
Further to your drilling the crank idea.
I had vague mmemories o seeing 'cross drilled' cranks for sale, 20 years ago.
I phoned Minisport up, and they confirmed that yes cross drilled cranks are stil sold for the 3 main bearing cranks on Minis.

As STD, No1 main feeds No1 B/E, No2 main feeds No2 & 3 B/E, No 3 main feeds No 4 B/E.
When cross drilling, the big ends between 1 & 2 and between 3 & 4 are joined by drilling through the crank web.
In doing so big end Nos 2 & 3 are efectively fed by 2 main bearings each.

I suspect that if it can be confirmed that this would be a safe mod, it could be carried out at all the big end positions on the Impreza crank.
So utilising your idea of adding another gallery from an adjoining main bearing BUT without blocking the original gallery, create somethinh akin to the Cooper S type cross drilled crank.

I also phoned Burton Performance, but the chap hadn't any experience of Subaru flat 4 engines and cranks.

John,
Given that different Subaru 'experts' are saying different things to different people about the piston slap problem (I was told undersize pistons, Otis was told too short piston skirt, you were told overheated piston, which one shall we pick).

Personally ii would suspect bore to piston relationshipo, and skirt design.

If you carry out a search on 'piston slap' on the internet they ALWAYS refer to the clearnce between the bore and the piston being too great, as the cause.

As you have stated 'there is too much smoke and mirrors ', so I felt that if I could find case evidence of the same problem, and its fix it may help dispel/confirm the various and sometimes obtuse theories being discussed.

Apart from a Ford truck example, the Honda Intgra was the best one for giving a fairly detailed summary of the problem, and at least its temporary fix.

All info should be treated as potentially useful until such time as the problem is resolved. We non of us have the right to discount any idea, unless we can categorically PROVE what is happening.

The 'experts' dont know ALL the answers to ALL the problems, (one that you spoke to didn't even know where the pistons are), otherwise we wouldn't have the problem in the first place, as they would have already fixed it.!!

By bouncing different ideas of each other, we may stumble on a fix.

Pat suggested drilling the crank, which made me suggest that by doing that, BUT not blocking off the original gallery, it may be possible to create a specil hybrid Cooper S style crank. ONLY if we can confirm that it is safe to do so.

If mutiple ideas had not been put forward, then this suggestion (good, bad or otherwise) would probably have not seen the light of day.

I think my next phone call will be to Cosworth Engineering.

Cheers MTR

 

i could go on for hours about piston slap and it's causes in many diferent engines, competition or road, but its taking this thread off in the wrong direction, so i wont.

The most important thing to crack here is the failing of the bearings, lets stick to the topic because it's wandering all over the place now.

I wouldn't expect burtons to know anything about subaru's anyway, they specialise in ford engines. If you want to talk crank design, give farndon engineering a call, they do virtually all the top competition cranks used in motorsport right up to F1.

 

John,
Your right about the slap issue being secondary to the original point of this topic, and for most people it would be covered under warranty anyway.
But it is still nice to get a clear understanding of whats going on for the benefit of people like Stef, whose car is now out of warranty.


Back to the cross drilling issue, Cosworth are closed today, but I have just found this site,

 

Start a thread on piston slap then martin, lets not go off topic, because piston slap is not the issue here.

 

John,
Found the web site for this Farndon Engineering company. They look like a very high tech concern.

 

Just for interest seem as though you put the link there. the MGF mentioned on

 

John,
Although I have spoken to you once, you may not recollect, but I talk rather quickly.
In fact my nickname at work is motormouth, as not only is it quick, it never stops.

I will let Stef phone them up, best not give Impreza owners a bad name.

Cheers MTR

 

Stef,
I called in to see Billy Jones, the owner of a garage called Autofix, Tel 0161 228 8599, 0161 628 5528, in Oldham.

He currently maintains a Mitsibushi Lancer Evo VI stage rally prepared car for one of his customers.

Previous to running the Evo, the same customer used to stage rally an Impreza Turbo.

I asked Billy, what if any problems they used to encounter with the lubrication aspects of the Impreza.

He told me that they used to run an external oil cooler, use Valvoline fully synthetic racing oil, and the car was obviously fitted with a full dural sump gaurd,and had NO problems.

They had a STD oil delivery system, no additional oil feeds in the block, no additional cross drilling of the crank etc.

He did however, tell me that on Impreza engines, that No 3 LITTLE END was prone wearing , which then transmits vibration down No 3 con rod to No 3 Big End.

To put the reliability aspect into perspective for all you track day boys, the Evo engine gets fully stripped, re-ringed , re-shelled (big ends and mains), and any worn components replaced after FOUR EVENTS, at a cost of about £1500.
A clutch cost £800, and some of the other blokes who compete in the same series, use £80,000 sequential gearboxes, according to one of the other mechanics who works there.

So reliabilty come at a very heavy price in this instance, and the car gets a VERY HARD LIFE.

Give him a bell if you want to discuss any concerns about the future reliability aspects of your engine.

It MAY BE, that my suggestions regarding cross drilling of the crank, although possibly effective?, and perhaps some of the other suggestions, may well be uneccessary, if you fit an external oil cooler, obviously with a stat.

You could be in danger of attempting to crack a walnut with a sledgehammer.

Cheers MTR

 

All these solutions that are being posted are all very useful, but I didn't think anyone actually knows what the problem is yet, so how can you go about fixing it ?

Sorry, I don't mean to be rude, but surely whats needed is we (er Stef) finds out exactly *why* the engine failed, ie a physical analysis on the actual engine as opposed to all this pure speculation.

Apologies to all if I'm talking rubbish yet again or offending people, I am stupid and also tactless

 

Some of my observations in case they are relevent. I have probably looked at plugs from maybe 30 ish cars in the last year, in almost all cases it would seem that No 4 cylinder runs rich, No 2 cylinder runs lean and No's 1 and 3 run between the other two at about right. No 3 and 2 big ends are fed from the same main bearing which has bigger oil ways to suit, oil pressure (pump) is circa 6 bar. If No 3 big end has a weakness then No 2 would have the same weakness. As No 3 is closest to the turbo up pipe and turbo then maybe it runs a bit hotter but as Mark as mentioned not that much if any. Bearing failure is caused by bearing pick up (high spots) and breakdown of the oil film. Since the oil pressure warning switch is only set to trip at 2.1 psi then its next to useless and will never protect. Oil quality (as per Johns posts) is king in this sort of application and using inferior grades will/can definately lead to problems even on the road. If a lambda sensor fails or starts to fail the ecu will over fuel causing bore wash and dilution of the oil thus degrading quality thus breaking down thus causing pickup etc. (check engine won't come on if the lambda just gets tired and outputs are low when temps rise). EGT is just as hot under fast cruise at stoich as full boost at 12.6 afr. Difference is that under cruise the closed loop is on.
Modifying oilways and oilfeeds is not the solution (IMHO), it will screw up the pressure drops which will screw up the flow, drilling extra holes in bearings increases the cross section which reduces the pressure drop which reduces the flow (simple fluid mechanics) unless an uprated oil pump is fitted (for flow not pressure). The oil pressure relief valve has sharp corners and is prone to dig in thus allowing pressure to reduce at lower revs, the poor oil pressure warning switch will not show this up, you need a guage for that. Simple mod, just stone and polish the edges off to prevent it.

Finally treat the engine with respect, after every track session let it cool properly and check the oil, it does lose/use it when driven hard. As John states, change oil and filter after every track day and only use high quality products.

Re piston slap, see my comments about plugs, since No 4 seems to be the culprit on MY98 cars ... so how much extra fuel is necessary to cause abnormal piston wear ? Just a thought in addition to the obvious clearance theory, piston slap develops with time and bores don't wear much at all under normal conditions. How many MY98 cars with PPP packs have suffered from piston slap ... the uprated ecu is a bit leaner than the standard one IMHO.

Just some of my own thinking, oil coolers etc all help, so does removing the undertray from those cars that have it. Keeping sump oil temp down to no more than 80/85 is probably a good target.

Couldn't resist ... one of the best threads for a long time.

 

I made the point about rally cars earlier.

Because they use an inlet restrictor, there is no point using over 6,000rpm. You just don't get enough air in.

This means that these engines suffer from much LESS stress than our road cars.

Think about a desiel engine. Massive torque, massive compression ratio, massive stress on the crank and rods. BUT low rpm and they last for ever.

 

I'm a 1998 Subaru Legacy GTB owner in New Zealand and I'm having a similar problem down here. I started a thread under 'Drivetrain' that might be worth a read. Check out the damage at

 

<sob!>

 

Thanks again for all your comments guys.
A little update.....
My engine is currently still in my car awaiting removal. Now that IM have washed their hands of me, the engine will remain property and therefore can be inspected at some length. The oil is also awaiting analysis.
I am planning to give the engine to Ahmed Bayjoo who will basically be rebuilding the engine using stronger internals (Cosworth and Farndon products). I was also intending to utilise his 'direct oil-feed' mod, although some of your comments concern me. The cost of the mod is relatively low though, as the engine will obviously be out of the car already. I'll see if I can get a more detailed description of what his mod actually comprises of.
I intend to keep the PPP ecu in my car, as I am not initially doing these mods to obtain extra power, although that may well be the goal further down the line . With the internals done, only an upgrade of the ecu, turbo and fuel system will be necessary. The timescale for these obviously depends on cost, etc.
CK 1 will now mainly be used for fun , as I have finally sorted out a diesel for work. This will save me a fortune in fuel, as well as give the car some much needed rest .

MTR.
Thanks for the info. You say that Billy used to prepare Scoobs for rallying, but I assume this car was also stripped down and re-built fairly regularly. It is therefore not surprising that he saw no problems.

Whilst this work is being done, is there anything else I should look at doing? Flywheel mods? Things of that nature.

Stef.

 

I am a bit fearful of entering this thread which comes from those with far more scoob experience than me.

But just a word about torque. WRC cars do produce a lot more torque and I would have thought that this would have the effect of showing any weaknesses in the drivetrain.

My previous car was a Range Rover and a popular mod was to fit a GM 6 litre diesel instead of the 3 and 1/2 petrol V8.

Power may have been down down, but torque was up, and a lot of snapped drive shafts resulted.

I guess the same principles apply to scoobs.

JD

(I must mprove my keyboard skills)

[This message has been edited by JayDee (edited 05 December 2000).]