IWatkins

Guys,

I was joking (hence the )

There doesn't appear to be any real pattern in No.3 failures that I can see. The only thing that stands out as being almost a pattern is that high revs./high gear seems to be involved. Exhausts, filters, ECUs etc. do not seem to be involved in these events.

Adam, yes, MoTeC can have a blanket fuel enrichment placed on No.3 cylinder, and has been done for quite a while by many people.

I've always felt that oil temperature/oil pressures have had nothing to do with it. My hunch is fuelling. Don't know why, just my own opinion.

Cheers

Ian

bob

Ian/
Basically the same as you first stated.
Motec required then for all Impreza's. Will my insurance company belive this.

When giving it some in my last Impreza as I hit the rev limiter in all gears the knocklink would light up like a christmas tree. Is this due to the fuel pump S/D and it going very weak?

carlos_hiraoka

Bob,

the correct link should be:

carlos_hiraoka

Bob,

If you managed to get to the web page, click on the left side:
"Partes y Servicios"
then on the centre of the screen click on(and blue):
"Servicios"
and then on the bottom right hand corner click on:
"Banco de pruebas" ..... to see their rolling dyno

Carlos H.

rsquire

Errrmm.. then why do Non Turbo Subaru engines also go "bang". As has as happened to a couple over here that I know of.

Has anyone considered the fact that #3 and #4 cylinders share a single oil feed while the other's don't.

I simply can't believe that one 22B engine will die because it held redline for 3or 4 minutes maximum ( I have it on video coz I was in the car behind about to overtake )

While another will go at 160mph on the way down to the Spanish Rally until the fuel is about to run out.. Very confused little bunny...

Richard

However.. I do know that the rebuild of my engine has included improved oil flow to #3 and a sump pan that has improved baffling..

However, As they say in the movies time will tell...

[This message has been edited by rsquire (edited 30 June 2001).]

R19KET

Carlos,

Now that's immpressive......

Wish some of our tuners would make as much effort

Mark.

bob

Carlos/
Had a look at the Rolling road and workshops etc. Very nice, does this place double up as a hospital?
This could teach our hospitals a thing or two let alone one of our workshops.

rsquire

Nice website..

But when are the En Constructione pages being activated ???

Richard

*Just enough Spanish to be dangerous*

rsquire

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Sustained high revs in top gear(yes v naughty I know!)killed my STI 2 engine, <HR></BLOCKQUOTE>


Yes...... but for how long, how much mileage on the engine, what mods to the engine and so on and so on.....

When are we going to get down to <B>HARD</B> facts on the inherent design flaws that exist in the Subaru flat four ?????

Richard

carlos_hiraoka

Guys,
I had the chance to talk with Fabian Barattero (for a few minutes), when he came to Lima, to look after a Grp. A WRX (really a Grp. N car just with a 34mm air restrictor). Although in rally cars, you never really change gears above 6000 rpm because of the air restrictor, we have this one big race in PERU, called "Caminos del Inca" which covers aprox. 4000 kms of rally stages and the last day (of this 7 day rally) there are two 400kms. Tarmac stages were high speeds are sustained. Barattero recommended the WRX rally driver to not get pass 7200 rpm (he limits the MOTEC ECU's to 7300), he told him that the WRX engine was all about midrange power and not high rpms .....
The WRX won this two last stages, in front of a TTE Celica and a Escort WRC, since the WRX with the 4.11 and stock gearbox (5th gear is 0.825) had far more final speed than the other cars. The engine was checked after the race and was perfect.
So what Barattero told me when I asked him about high rpms, was that if I wanted more speed then I should change my 4.44 diffs for the 4.11 ones.

Carlos H.

carlos_hiraoka

Check out the specs of both the Group N and Group A engines (not WRC engines, both with TMIC)

- Click on "Fichas Tecnicas"
- Click on "Datos IMPREZA Grupo A" for Grp. A specs
- Click on "Datos IMPREZA Grupo N" for Grp. N specs

For Group N:
47 Kgm/3700 rpm
270 bhp/5000 rpm

For Group A:
48 kgm/4000 rpm
320 bhp/6500 rpm


Carlos H.

carlos_hiraoka

sorry double posted .....

[This message has been edited by carlos_hiraoka (edited 30 June 2001).]

scoobycar60

Sustained high revs in top gear(yes v naughty I know!)killed my STI 2 engine, of that I quite sure! Most of those who will admit it probably killed their engine the same way,

The engine either does not like high revs full stop or like a lot of engines it leans out at the top of the rev range....No 3 the most, result det + heat = increased bearing load = bang, either way I will avoid excessive sustained high revs in future.

Incidently Prodrive seem to go more for mid range grunt at sensible revs perhaps they know more than most about the character of the engine. Any PPP cars gone bang recently....Sorry Prodrive but in light of recent threads its a fair question.

[This message has been edited by scoobycar60 (edited 30 June 2001).]

Greg115

Clearly we have struck upon something here. Subaru would have not changed the timing of number 3 cylinder if the problem was not with this unit.
I agree that the location of the turbo should not be a factor in the running temp of number 3 cylinder unless it is causing a different scavenging effect in the engine due to the exhaust manifold shape. This combined with an unefficient fuel flow due to fuel manifold may be causing the mixture to be totally screwed up at high revs.
It is interesting that there is only one oil feed to number 3 + 4 bearings. This is an extremely important finding as it concerns the bearing that we know to fail.
Basically as I see it, we have two known problems - lubrication failure to number 3 bearing and detonation of number 3 cylinder. Sometimes these occur together, sometimes only one of these factors occur. Therefore we may be trying to tackle two problems at once when we are only looking for one.
I think this thread has been the most interesting I have read for a long time and is really pulling the power of the internet into the forum.
Greg

Adam M

I can accept the fuel thing, there is no harm doing this mod and perhaps upping number three fuelling on the motec but I have concerns with the oil.

Richard can you find out exactly what mods have been done re number three/four oil supply.
If there was a temporary oil supply problem, would it not show up on an oil pressure gauge?

a restriction would show up, or are we talking more of a sustained restriction, that simply doesnt supply enough?

Increasing flow will reduce oil pressure???

meaning the pump has to work harder if you want to sustain more. So combining richards mods with an uprated oil pressure regulator and oil cooler also for safety and perhaps there is a chance we would have covered all the bases.

robski

Ian, doh missed the winky

Carlos, excellent, this makes a lot of sense, as I hinted at, and people like rich have indicated, maybe we do have 2 issues

lubrication and fueling

I can see a link between high rev continuous and both these faults, poor fueling and bad lubrication isnt a good thing really is it!!

Im sure Stef mentioned at some point a revised oil flow that he was having done, or was contemplating

robski

CraigH

So if the oil feed to number 3 is modified by taking/rerouting/sharing or whatever another oil feed, surely this will have the effect of dropping pressure elsewhere in the system, so it's not removing the problem, just moving it to a different area?

Greg115

The oil pressure is controlled by a relief valve on the discharge from the pump. Therefore if the oil supply to number 3 bearing is improved, the same supply pressure will be obtained at no detriment to anything else providing the new routing is efiicient for the rest of the engine.
Greg

Bob Rawle

Anyone care to consider and comment on where the oil feed to No 2 big end comes from ?

No1 fed from front crank bearing.
No4 fed from rear crank bearing
No 2 fed from ??
No 3 fed from ??

No 2 & 3 are BOTH fed from the centre crank bearing.

So can any light be shed on why No 2 big end is not as well reported on as No 3 big end.

Logic says that it should be if its an oil feed issue. So maybe its not.

Temperature, I monitor the oil temp in the gallery just above No 3 cylinder ... its no hotter than any where else.

Mixture, up to and including MY 98
No4 is richest.
No2 is leanest.
No1 & 3 run the same.

This is mostly down to the inlet manifold design.

Its also what contributes to No4 cylinder/pistion going "slappy" as it will wear at the highest rate given its getting the richest mixture.

MY99/00
All seem to run about the same, the inlet manifold design has been improved on the Phase 2 engine and there is no ISC to get in the way of No4.

Fuel rails, pressure drop across the injector induces flow. The pressure reg is right next to No3 so No3 will see the regulated pressure for sure. All injectors will see similar flows given that the pressure drop is similar across each, its easy to calc it out, simple fluid dynamics.


Just some more thoughts for debate and consideration.

BTW the oil pressure regulating valve in the oil pump gets a lot of attention, how many people consider that the oil filter has a bypass valve built into at which is "only" set to 1.6 bar or so, given the oil pressure is well over that a partially clogged or poor quality filter could be letting debris through, could this contribute to the failures that we see "just after a service"? If oil type is changed, oil not drained properly or filter not changed for some reason then the filter could maybe get obstructed enough to cause unfiltered oil to get to the crank and big end bearings.

When a bearing fails it does so due to "pick up" in a local area, this then propogates (or not) into failure. Sometimes a "pick up" can bed back down and no problem, prolonged high revs accompanied by "pick up" equals bearing failure though.

As an observation, the MY01 ecu has ignition trim via knock control on each cylinder not just No3, Bosch management has been like this for years, Subaru have just caught up so no suprises there and nothing to read into it more than the ecu is now doing what most other manufactures do. When a knock is sensed it also reads the crank position thus determining which cylinder is to be adjusted. Since the MY01's run "on the knock sensor" a la Bosch again then this is all normal and quite logical.

So where does all that leave us ... still with no real answer I'm afraid.

[This message has been edited by Bob Rawle (edited 30 June 2001).]

carlos_hiraoka

"... Fuel rails, pressure drop across the injector induces flow. The pressure reg is right next to No3 so No3 will see the regulated pressure for sure. All injectors will see similar flows given that the pressure drop is similar across each, its easy to calc it out, simple fluid dynamics. ..."
(posted by Bob R.)

Bob, sure someone must have a disassembled inlet manifold, were you can take out the fuel rail (out of a blown engine ....) with the 4 injectors plus the fuel reg. Why not try and check out if the flow is the same on each injector (try to see if the calculations match reality).
If you have seen, Prodrive Grp. N cars used to have a different fuel rail system, they had an independent fuel rail for each side of the engine, each with a std, fuel reg.
On STi engines an injector that flows a little less might be more of a problem since they ran more boost and bigger turbos than the 211 bhp turbo Impreza, and any type of leaning would cause inmeadiate detonation or piston damage, and on the other side the 211 bhp turbo Impreza is well known for being on the rich "safe" side. It is not the same to get 1.0 bar out of the small TD04 than to get 1.0 bar out of a IHI VF22, we all know the VF series turbo flows a lot more than the TD04.

Carlos H.

Greg115

Bob
You make for very objective reading.

It now makes me wonder if debris in the engine is more prone to be fed to number 3 bearing over the others due to the nature of the internal pipework.

I dont know what procedure is recommended for mechanics when they change your engine oil, but I have witnessed the procedure of fitting the filter dry/rather than filling with oil first by one of my local Subaru garages.They then start the engine until oil pressure is acheived, which can take as long as 5 seconds.

If you say that failures have occurred after servicing, then if the above procedure is used, it is probable that a few turns are made of the engine before lubricating oil pressure is available. This will invariably cause a small amount of wear in the bearings as the hydrodynamic oil film will not be enough to keep metal to metal contact from taking place, due to the load in the cylinder.

As the oil filter bypass valve ,whether new filter or not, will always lift with cold oil, then these particles could be finding their way straight to number 3 bearing. This, of course is partial assumption.

I would have thought that the bypass valve on the filter would be set at higher than 1.6bar. If this is the setting then it is likely to be open most of the time isnt it?

Could it be this simple?

However it does not explain that when detonation damage occurs, it seems to target number 3 cylinder again.

Greg

scoobycar60

Bob just for the record..
The bearing failure on my car was No.2!

Bob Rawle

Carlos, I am fully aware of the Group N spec, we are discussing road cars here though.
The rally boys use different cam timing to the road cars, gives them loads of low and mid but reduces the high end so no point even driving the car up there.

Modify the fuel rails but don't use twin regs unless they are matched and have the same pressure curve and base (don't want to encourage different fuel pressure in each rail). Use parallel rails and a single regulator ensuring that each side is the same diameter and length. Maybe the rally boys do this to help reliability as the single rigid rail would be prone to fracture from vibration.

"any type of leaning will cause detonation" ... sweeping statement and not true, det comes from a combination of things not just leaner mixture, the onset of det is triggered by temperature given that the mixture leans and the timing is too advanced, any car will det if the charge temp gets too high irrespective of mixture dependant on timing. No 2 is naturally leanest on the phase 1 cars and that is the cylinder that dets first (in my experience. Phase 2 ... not so consistent but then it wouldn't be.

Greg115, if debris is "drawn" to the centre main bearing then it would still be distributed to No3 and No2 big end, the two big ends are equi-distant from the main bearing, just opposite directions. Differing clearances within the bearing could cause differing oil bleed through rates but given the spec the bearings are its not likely to be that different.

The real way to check this would be to monitor combustion chamber temps but thats not an easy task (or a cheap one). I may try and monitor cylinder head temps over each cylinder to see what that tells me but my current car runs with coolant temp between 78 and 80 most of the time so I don't hold out too much hope of seeing any differences.

Pressure in the filter won't be 5 bar, much less than that pressure being related to flow and area the large area of the filter means that there will be alot less pressure in there so it would not normally be open, only if filter became restricted or blocked.
The only safe way to change a filter is to fill the filter, fit it to the engine, fill the engine with oil to correct level, remove the crank sensor connector to stop the engine firing and then turn her over on the starter til the oil light goes out, that can take a couple of minutes. Re-connect the crank sensor connector and start her up, same when a turbo change is done as well, even more important them. I also wonder if some cars aren't sufferening from lower oil levels due to the practice of part filling to stop solenoid contamination. easy to get wrong, for those that are not sure on the dip stick are three marks, two lines and one knotch. Lower line is minimum, hot or cold, upper line is max level when engine is stone cold, knotch is max level when engine is hot, easy to think the oil is ok when maybe its a bit low. Again this still does not provide an answer.

I still major on the pick up theory, a hot spot in a bearing can exist for some time under normal conditions and not become a real problem, indeed, as mentioned, quite often it will self heah, push a car with such a hot spot hard and it will turn into a bearing failure likely as not.

Detonation damage does not necessarily imply big end failure, det can take out a piston and the ends will still be perfect, its clear that if det persists the shock loading on the piston through the rod will cause the oil film to break down, we are then back to bearing pick up and failure. Its more likely to be broken down by the detonation shock forces involved as anything though.

I know of cars that have failed on No2 and No1 separately and together.

I run my car's with some fairly high boost compared to standard and have covered over 70,000 miles with no problems, I do take all the precautions I have mentioned and I change the oil and filter at mid way between service intervals as well, when on the track the car gets new oil and filter before and after. I don't boost the car until oil temp is at 80 deg C minimum and I never ever drive her for less than 5 miles. Even if I have only a 1 mile trip i still do 5. Oil level is always at the top line when cold ( I don't suffer from the solenoid issue at all). I only use good quality synthetic either 10-50 or 5-40, (10-60 best for track imho) I only try and use official Subaru filters as some of the "alternatives" are not so hot quality wise.

No guarantee and i will probably suffer No 3 failure next week having said all that but I am a strong believer in looking after the car means you stand a far better chance of trouble free ownership, we own high performance cars ... they need treating as such even if it costs us.




[This message has been edited by Bob Rawle (edited 30 June 2001).]

carlos_hiraoka

Bob,

Yep I totally agree with you, changing oil and oil filter is definetly a good habit, specially after track days, it is always good to remember that oil is much cheaper than a engine rebuild.

"...Modify the fuel rails but don't use twin regs unless they are matched and have the same pressure curve and base (don't want to encourage different fuel pressure in each rail). Use parallel rails and a single regulator ensuring that each side is the same diameter and length. Maybe the rally boys do this to help reliability as the single rigid rail would be prone to fracture from vibration. ..."

On the Prodrive Grp. N, the pressure regulators were the same, so were the fuel rails, so the fuel pressure on each rail was the same. But on other Grp. N cars, I have seen just the middle section modified (which is the easiest one to copy), or as you said parallel rails, and a single pressure regulator (which has to be the std. one because of FIA Grp. N rules).

Bob, do you sustain high rpm's ? ..... what do you think on working to modify the oil pump, specially on two areas, increasing the oil pressure a little by shimming, and uprating the valve.

Carlos H.

rsquire

Wow.. and to think that some people were worried about the quality of posts on this BBS..

Right, I'm going to have to print this thread out, sit down with a cup of coffee and attempt to digest it all.

Richard

EvilBevel

Squire dude , I was just gonna say the same ... beat me to it I guess.

Excellent stuff guys... need a few days to digest it all.

Theo

Hoppy

Ditto. Excellent thread.

I wish I could say I understand it all, but I've now decided to modify my oil/filter change routine (again!). It doesn't get any cheaper, does it!

Richard.

bob

Great stuff looks like I started another good un…

Bob/
I understand most of what you say. (Even though I am a thickie yes I know) But was just wondering why the oil filter bypass has a valve set for 1.6 bar. How come there is a lower pressure here than in the rest of the pressurised part of the system. Is the filter not under very high pressure when the engine starts from cold?

Greg115

The 1.6 bar opening of the bypass valve on the filter is a differential pressure of 1.6 bar across the filter. i.e. when the filter starts to get dirty the bypass valve will open up.
It will also open when the oil is cold because the flow rate through the filter is increased because of the change in the oil viscosity. The pressure on the inlet side of the filter will be greater than normal as the flow rate through the filter will be less for the same given pressure. The relief valve on the pump is manufactured for a range of temperature in the normal operating range and therefore is not able to control the oil as accurately at lower oil temperatures.
Greg

[This message has been edited by Greg115 (edited 01 July 2001).]

carlos_hiraoka

Squire ..... you've got mail (if your email still is rsquire@microsoft.com).

Carlos H.

[This message has been edited by carlos_hiraoka (edited 01 July 2001).]