So my friend's 2008 STI blew up the motor last week.

In reply to HiTempguy:

I've seen ringland failures on cars that were Cobb tuned to hopefully prevent that problem. Its not a tuning issue causing the pistons to fall apart.

Tunes don't fix oiling issues, and the #3 bearing is a repeat culprit for failures.

idk why people use download tunes anyways. Getting an opensource tune coast the same as getting all that cobb E36 M3 and it's far better.

What local shop is he looking to have rebuild it?

HiTempguy wrote:

dculberson wrote: (and the ECU is part of the engine)

No, its not in any way, shape, or form. The motor itself is not an issue. Its the control of the motor. What you are basically saying is the operator of a motor is part of the engine, because an operator can also asplode an engine too

Nah, I'd rather not play your games.. I choose to exist in reality. It's nicer here.

A manufacturer makes a product that fails, when it isn't abused but is used as intended, that's the manufacturer's fault. Your position is one of intentional blindness / denseness. No thanks.

dj06482 wrote: What local shop is he looking to have rebuild it?

These guys.

http://www.efilogics.com/

There are others out there I'm sure.

I had a nice conversation with the guys at COBB Tuning earlier regarding these engines, failures, etc.

Needless to say, he made me feel much better about what the issues are and how to prevent them.

z31maniac wrote: I had a nice conversation with the guys at COBB Tuning earlier regarding these engines, failures, etc. Needless to say, he made me feel much better about what the issues are and how to prevent them.

As I drive an '07 Legacy GT spec.B as my daily and have 127k on the clock (original owner, no mods) I'd love to know what they shared with you...please

…a four page thread about engines that E36 M3 all over themselves and they aren’t rotaries; IMPOSIBRU!!!

wawazat wrote:

z31maniac wrote: I had a nice conversation with the guys at COBB Tuning earlier regarding these engines, failures, etc. Needless to say, he made me feel much better about what the issues are and how to prevent them.

As I drive an '07 Legacy GT spec.B as my daily and have 127k on the clock (original owner, no mods) I'd love to know what they shared with you...please

Yes please let us know.

irish44j wrote:

Knurled wrote: That's not what I heard. What I heard was 17:1 at 3-4psi boost at some certain RPM. But regardless, lower RPM is exactly where Joe Applianceguy is going to be driving it...

for anyone who cares, here's the stock A/F on my 09 WRX... even when brand new there was a noticeable hesitation under boost in midrange as the engine pulled timing. But I had no idea how it dropped off such a shelf. stock tune: OTS Cobb stage 1 .. big difference, if still not entirely ideal (totally stock engine)

My totally un-proven theory behind all those exploding Suby engines is that all those aftermarket tunes run too rich. And that stock graph sort of confirms my suspisions.

Here is what I think happens. At low rpm, the air velocity is probably too slow to acheive good atomisation of the fuel. Fuel that isn't atomized properly would "pool" in a very localised spot towards the bottom of the cylinder because of the flat engine. Un-burned fuel dilutes oil accentuating the uneven lubrificatiuon of the cylinder causing various lubrification problems and inducing side loads on the pistons.

To avoid this, you would need to keep the AFR lean and rely on the knock sensor to avoid detonation at low RPM. Which is exactly what the stock tune does. And turbocharging would simply compound this problem.

This problem combined with a mechanically fragile engine would make for a very blow-uppy engine.

Feedyurhed wrote:

wawazat wrote:

z31maniac wrote: I had a nice conversation with the guys at COBB Tuning earlier regarding these engines, failures, etc. Needless to say, he made me feel much better about what the issues are and how to prevent them.

As I drive an '07 Legacy GT spec.B as my daily and have 127k on the clock (original owner, no mods) I'd love to know what they shared with you...please

Yes please let us know.

Good oil with lots of Zinc he recommended Motul 8150 IIRC, frequent changes, watch the oil level carefully (low capacity given the heat/power these engines make).

They said there experience is guys run the oil too low, get a bit of bearing damage, particulate circulates through the engine and scores cylinder walls, creates hot spots, bang, broken ring land.

I will say, I'm not going to get into a pissing contest with anyone about if they are right/wrong/etc. But I'll take a trusted tuners word over people with anecdotal evidence at best.

I'm not sure why a tuner should be trusted on some pretty complex engineering issues like that, but to each their own. Given my friends experience low oil seems to not be a factor at all in some cases and in fact the performance shop he contacted to do the rebuild just seemed to consider that it was a matter of course that the stock engine would blow up and were surprised it lasted as long as it did.

My friend looks to be selling his car to the dealer for $10k.

fanfoy wrote:

irish44j wrote:

Knurled wrote: That's not what I heard. What I heard was 17:1 at 3-4psi boost at some certain RPM. But regardless, lower RPM is exactly where Joe Applianceguy is going to be driving it...

for anyone who cares, here's the stock A/F on my 09 WRX... even when brand new there was a noticeable hesitation under boost in midrange as the engine pulled timing. But I had no idea how it dropped off such a shelf. stock tune: OTS Cobb stage 1 .. big difference, if still not entirely ideal (totally stock engine)

My totally un-proven theory behind all those exploding Suby engines is that all those aftermarket tunes run too rich. And that stock graph sort of confirms my suspisions. Here is what I think happens. At low rpm, the air velocity is probably too slow to acheive good atomisation of the fuel. Fuel that isn't atomized properly would "pool" in a very localised spot towards the bottom of the cylinder because of the flat engine. Un-burned fuel dilutes oil accentuating the uneven lubrificatiuon of the cylinder causing various lubrification problems and inducing side loads on the pistons. To avoid this, you would need to keep the AFR lean and rely on the knock sensor to avoid detonation at low RPM. Which is exactly what the stock tune does. And turbocharging would simply compound this problem. This problem combined with a mechanically fragile engine would make for a very blow-uppy engine.

By the time there's any positive pressure, velocity is plenty high enough to atomize fuel properly when being sprayed at OEM fuel pressures.

You'd have to run REALLY rich with REALLY low fuel pressure for that to become a legitimate problem. Like.... single hole EV1 injectors at 20psi or less and 8.0:1 AFRs.

What you're describing is Cylinder Wall Washing, which DOES happen, but only when someone severely berkeleys up. And even then, i'd probably take it over running lean under boost any day of the week.

Duke wrote:

Harvey wrote: Nothing like getting out of a first year car into another one right? Also he is thinking maybe Audi S4 or BMW 335.

So he's a masochist, then.

you read my mind. Hey, maybe the engine won't blow up, but instead he can drop $1k each time some electrical module fails

fanfoy wrote:

irish44j wrote:

Knurled wrote: That's not what I heard. What I heard was 17:1 at 3-4psi boost at some certain RPM. But regardless, lower RPM is exactly where Joe Applianceguy is going to be driving it...

for anyone who cares, here's the stock A/F on my 09 WRX... even when brand new there was a noticeable hesitation under boost in midrange as the engine pulled timing. But I had no idea how it dropped off such a shelf. stock tune: OTS Cobb stage 1 .. big difference, if still not entirely ideal (totally stock engine)

My totally un-proven theory behind all those exploding Suby engines is that all those aftermarket tunes run too rich. And that stock graph sort of confirms my suspisions. Here is what I think happens. At low rpm, the air velocity is probably too slow to acheive good atomisation of the fuel. Fuel that isn't atomized properly would "pool" in a very localised spot towards the bottom of the cylinder because of the flat engine. Un-burned fuel dilutes oil accentuating the uneven lubrificatiuon of the cylinder causing various lubrification problems and inducing side loads on the pistons. To avoid this, you would need to keep the AFR lean and rely on the knock sensor to avoid detonation at low RPM. Which is exactly what the stock tune does. And turbocharging would simply compound this problem. This problem combined with a mechanically fragile engine would make for a very blow-uppy engine.

er....most of the exploding engines are ones with the stock tune. I don't know anyone locally who has blown up engines with a basic stage 1 tune (Cobb or otherwise), and it's pretty common knowledge on the subaru forums that the stock tune is not good for the engine, running super lean and then pulling timing big time...And I know a LOT of local turbo subie owners between guys I used to hang with, and guys I rallycross with. Then of course guys with high-power tunes blow them up, but that's a completely different story.

I think that the problem is explained if you look at the Subaru pistons pictured earlier and then look at the stock AFR map. The transition from closed loop fueling to open loop fueling occurs right around 3600 rpm but Subaru put in a delay. This also happens to be the point where Subaru's fuel maps lower the AFR to a more reasonable level. In addition the boost is ramped up way too early so that "turbo lag" can be minimized for the average consumer.

So the result is that you have near peak torque at the same time as high AFR, peak boost, and the ignition maps have a spike in timing in a little island right about this point in load/rpm. In addition, because of the delay in switching fueling maps you have all of these things overlapping.

In reply to Ojala:

This point where a lot of bad things converge also happens to be close to freeway cruise speed in fifth gear. Picture what the engine is going though on a long uphill freeway climb. Just enough load to build a little boost, but still in the lean part of the map, for an extended period of time. I wouldn't be surprised if local geography was a factor.

I was also thinking today how rare it is to see people checking their oil at gas stations. My Legacy GT didn't use a drop of oil between changes for over 100,000 miles, but I always checked it. Then it started using a quart every 1500 miles or less (broken ringland number 1). Then it dropped again, down to less then 500 miles per QT (broken ringland #2.) The car still ran fine, no smoke, no check engine light, still made good power. I could easily see how a broken ring land could lead to running it low on oil.

Well now my boner for looking at new WRXs and STIs has deflated.

Couple of things:

1 - If we are going to throw some random evidence based on personnal experience, I can play too. I am a two time Suby owner (WRX and Saabaru) and I was an active member of the VERY active local Subaru club. In my experience, the ratio of stock vs tuned engine blowing up is about 1 to 5. Both my Subarus went great with their stock engine. My Saabaru's engine still runs fine with over 275000 km (that's 171k miles) on the un-modified engine. So I have a hard time with the "the stock tune is to blame" routine given by the TUNERS. A conclusion based on general rules of thumbs that aren't backed by any SOLID scientific testing.

2 - Fuel atomization is always a problem, which is why the industry is moving ahead with direct injection and always greater fuel pressures and fancier injector nozzles. And while a little extra fuel landing on top of a piston in a regular upright cylinder isn't a problem, that small amount falling right on the cylinder wall is completely different.

3 - Lean condition does NOT automatically equal detonation. And if you look at most failures, they aren't caused by detonation, but by oiling issues or manufacturing defects. I could believe a holed piston caused by detonation, but any detonation strong enough to caused bearing damage should easily be detected by the knock sensor, no matter how bad it is located.

I stepping down from my crate.

You guys are fun.

z31maniac wrote: So I'm guessing emissions testing happens below 4k RPMs?

Yep. The Federal and California cycles should not run the car into boost, however, so that shouldn't matter.

fanfoy wrote: 2 - Fuel atomization is always a problem, which is why the industry is moving ahead with direct injection and always greater fuel pressures and fancier injector nozzles.

Direct injection has nothing to do with atomization and everything to do with stratified charge. But continuing...

And while a little extra fuel landing on top of a piston in a regular upright cylinder isn't a problem, that small amount falling right on the cylinder wall is completely different.

At 1200rpm, each cylinder hits TDC twenty times per second and fires ten times per second. Gravity simply does not have time to make the fuel "fall", internal airflow keeps things to be quite the tempest in a teapot in there.

There is a slight tendency on opposed engines to get a just-measurable amount of wear on the down side of the cylinders and pistons, but this is far overshadowed by the normally seen thrust wear. Especially on an engine with relatively short rods such as the 2.5.

Ring land failure IS a detonation issue, period. It's not as progressive an issue as pounded-out rod bearings and wristpins. Simply put, combustion pressure forces the rings to expand by design, but detonation puts an intense pressure spike that can cause the rings to grip the cylinder so hard that the crank pulling on the rod will yank the piston apart. Which is why we're concerned with the distance between rings and between the top ring and the top of the piston, more distance means a stronger piston.

Two other points. I do not know what year your WRX was, but the 9-2X was not made in this year range. The poor OEM tune was a 2008 thing. (GE chassis?)

Knock sensing is not a panacea. You have to first knock before the computer can do anything, and the computers are usually tuned to step down gradually instead of clamping down hard like President Madagascar. So you get single knocks (which can still be bad enough to be life-threatening, especially if you get them chronically) or you get mild detonation all the time. This ASSUMES that the computer is even looking at knock. Many computers ignore the knock sensor under certain conditions.

Interesting info from all of you.

I can't believe the independent shop is several months out. The right place with the tools in front of them can have this done in two days.
Good grief its a Subaru motor and I've rebuilt them on the kitchen table.

Quickest route would be to buy a new Subaru short block and have someone install it with rebuilt heads from your local machine shop.

Harvey wrote:

dj06482 wrote: What local shop is he looking to have rebuild it?

These guys. http://www.efilogics.com/ There are others out there I'm sure.

I was hoping that was going to be your answer. Seeing as EFILogics is the shop, I'd have them rebuild it. They're the best Subie shop in the area. The dyno chart I shared earlier was to EFI's Spec.B with the stock tune. I'd trust them over a dealer for good work, they take pride in what they do. There is a wait, but that's because they're good.