09 Legacy broken piston

[JohnCT]

Hi, daughter in law's 09 Legacy with the 2.5 and a sudden bad skip at 98K miles.  Timing marks right on, very low compression in #4.

I pulled the engine out and removed the heads expecting to find a blown gasket or sucked valve, but found #4 piston broken at the top edge at the ring bore like the ring was trying to break out.  I've never repaired a Subaru engine,  but did find a video on youtube (where else) that shows the pistons can be removed without touching the rods. Seems there's an access port which allows the removal of the wrist pins and pop the pistons off the rods directly.

Anyone done this?  I guess I should replace all four pistons and rings but is there a reason this happened?  Is there an upgrade?  Any tips or pitfalls to avoid when pulling the pistons out?

Any help appreciated.

John

 

[FerGloyale]

you'll have to pull the engine to get to the #4 wrist pin.

But if you go that route, it's not hard to replace pistons/rings.

 

[1 Lucky Texan]

bad luck but, I have read of the occasional broken ring land. Has the car been well-maintained? (according to the TIME side of the maint. sched., it needs timing belt system service)

how many miles on the engine? If high, you might consider used or JDM engine - could have fewer miles.

Edited March 26, 2019 by 1 Lucky Texan

[JohnCT]

28 minutes ago, 1 Lucky Texan said:

 

how many miles on the engine? If high, you might consider used or JDM engine - could have fewer miles.

 

Less than 100K on the clock.

Talking to some buddies of mine in the industry, they told me that if I buy a used engine, don't install it without doing head gaskets and timing belt/pump/tensioner regardless of how few miles are on it.

If that's the case, then spending over 2K for a used engine that requires another 1K for parts and head machining doesn't make much sense, which is why I'm thinking about putting in new pistons, rings, and of course the head gaskets and timing stuff.

Just wondering if anyone here has done it.

John

[1 Lucky Texan]

you need to do the timing belt anyway - and you'll be doing one HG at least.

I can't help wondering if some other issue is looming, has oiling been poor and is a rod bearing going out 3 weeks after repair?

 

paging @GeneralDisorder

Edited March 26, 2019 by 1 Lucky Texan

[idosubaru]

Yeah that's a tough call.

Any rust - maybe you should just move on now?
JDM has inexpensive solution options sometimes?

were there any check engine lights before this happened?

Low oil, too long between changes, and/or not using synthetic ....?
Cut the filter open and check the oil there and in the oil pan.

I'd get another block under the suspicion that something caused this and may have caused additional compromising of the oil/rod bearings.  But Fergloyale is well experienced and didn't say anything...

[hankosolder2]

You didn't mention if this was a regular Legacy or a turbo Legacy GT.  I'd be less surprised to see a piston breaking up on a turbo engine.

It might be a good idea to send an oil sample off to Blackstone Labs for analysis to see what they say before committing to repairing that engine. They can check for bearing material in the oil and such. Think it's about $40. 

Good luck, sounds like you've got a big project no matter which repair route you take!

 

[ocei77]

get FSM here:http://jdmfsm.info/Auto/Japan/Subaru/

Only remove the outer circlips.

Use a thin stiff rod and you can snake it past the opposite rod and  tap out the king pin.

ie: past 3 and tap 1. 4 and tap 2.

Make a drift out of 3/4" dowel to install.

The block has markings to say whether the piston is an A or B. Spec sheet gives dimensions.

Put pistons back into cylinder they came from.

Make sure piston is installed with dots or markings facing the front of block.

 

O.

[JohnCT]

I've decided not to replace the pistons..

I talked to a guy who used to work at Subaru and now does rebuilds for most independent garages in the area and knows his stuff - he told me to get a new short block, so that's what I did. $2000 for a brand new short block isn't too bad. He said I might get away with just pistons and rings if the bore wasn't damaged, but he always does a clean up hone when he does pistons.

I brought the heads to the machine shop and the guy told me that there was signs of detonation and by inspecting the pits on the valve, correctly identified the cylinder that had the piston failure (although I wonder if the distress wasn't the piece of piston banging around). He told me that even though it calls for 87, use 91 octane in the 2.5L Subaru to prevent this type of ringland failure. He also said valve guides are known to walk out so he machines in a retaining clip.

This engine does feature a knock sensor, and I'm going to change it. I don't know if the car had a discernible spark knock or not (I'll question my DIL more thoroughly), but I'll make sure she uses high octane gasoline. In any case, there were no lean codes in the PCM.

I'll post a follow up when I get the engine together and installed.

John

 

Edited March 28, 2019 by JohnCT

[JohnCT]

On 3/26/2019 at 5:05 PM, hankosolder2 said:

You didn't mention if this was a regular Legacy or a turbo Legacy GT.  I'd be less surprised to see a piston breaking up on a turbo engine.

 

 No turbo.

John

[1 Lucky Texan]

yeah,  possible knock sensor bad but, usually there's noticeable power loss.

[JohnCT]

Update:

Had the heads rebuilt and installed them on the new Subaru short block and put it together Sunday.

Very little power.  Struggles in fact.  It even has trouble building revs in neutral.  I pulled the front of the engine off to reconfirm I have the timing right, and it's on the mark.  Since it had no compression in number 4, my next though was a stuffed cat from unburned fuel, so I unbolted the front pipe from the heads and the power returned.  I then bolted the front pipe back and unbolted the intermediate pipe that holds the back cat, and it seems to have power.  I have a new intermediate pipe coming in this afternoon, so I'll try that first.

I have a bad feeling it might need both pipes/cats.

A locally sourced front pipe/cat is going to cost $400, but I see there are some on ebay for less than $200.  I've used aftermarket cats before with no problems but never owned a Subaru so my daughter in laws is the first in the family.

Anyone use an aftermarket front pipe/converter?

John

[el_freddo]

On 3/29/2019 at 5:43 AM, JohnCT said:

He said I might get away with just pistons and rings if the bore wasn't damaged, but he always does a clean up hone when he does pistons.

Whoa, @GeneralDisorder will have something to say about this! He must be sick of reading this stuff! 

 

@JohnCT - have you also thought about a local exhaust shop to build a replacement piece? I tend not to trust off the shelf aftermarket stuff from the net, especially fleabay! 

Cheers

Bennie

Edit: it could be worth getting the injectors flow tested and serviced - this will insure injector 4 is working properly and not leaning out due to blockage or some other injector related issue

Edited April 8, 2019 by el_freddo

[Mike104]

General consensus on aftermarket cats is they are junk.  You will probably get a P0420 code shortly after installing them but YMMV

[JohnCTT]

On 4/8/2019 at 10:01 PM, Mike104 said:

General consensus on aftermarket cats is they are junk.  You will probably get a P0420 code shortly after installing them but YMMV

Well, I got a P0420, but I'm still using the original cats.

Here's what I don't get; in researching P0420 codes, many people talk about intake leaks and such.  In my experience, and air leak will cause a lean code.  I don't understand how an air leak can cause a normal response for the first O2, but cause an efficiency code.  Am I missing something here?

Because the car had a misfire from the blown piston, I'm assuming the cat on that bank was damaged and that's what's causing the P0420 efficiency code.  I don't mind replacing the exhaust manifold/cats but don't want to if that's not the problem.

BTW, I had to reregister because every time I try to log in, it always says wrong password and locks me out.  So then I have to wait for the time out, and then go through the process of changing passwords.  Unfortunately, it hasn't taken the same password twice..  Hopefully my new account will work.

Thanks.

John

[nvu]

no lean code means the front o2 sensor is happy with the air coming out the engine.  p420 is when that air goes through the cat and hits the rear o2 sensor.  the engine is fine, the ecu is not happy with the air coming out the cat.  exhaust gaskets, bad o2, bad cat.  nothing to do with intakes

[1 Lucky Texan]

if the after-cat sensor's wiring and integrity are good, anything that prevents the converter from 'lighting-off' will return an error. In a good working system, you can even use an infrared therm and see that after cat exhaust piping could be 100*F hotter than the piping at it's intake. So, if the exhaust going into the cat does not support the 'catalytic action' for some reason, the reaction won't happen and the rear sensor won't be 'happy'.

on older cars, that sensor is just a cat conv 'nanny' and could be ignored or 'cheated' with a spacer or a coupla resistors to clear the code, but on most(?) newer cars

( and, it seems, on Gen2 H6es),  it seems to also be used by the ECU to modify a:f .

Edited June 27, 2019 by 1 Lucky Texan

[JohnCTT]

3 hours ago, nvu said:

no lean code means the front o2 sensor is happy with the air coming out the engine.  p420 is when that air goes through the cat and hits the rear o2 sensor.  the engine is fine, the ecu is not happy with the air coming out the cat.  exhaust gaskets, bad o2, bad cat.  nothing to do with intakes

That's what my intuition tells me, but if you Google P0420 codes, every return will mention intake leaks!  Doesn't make sense to me either.

Anyway, the Subie exhaust is a one piece all welded affair, is it not?  I'm guessing a bad cat, but the exhaust gaskets from the heads on out can only cause a lean code, not an efficiency code?

Thanks for help guys.

John

[JohnCTT]

2 hours ago, 1 Lucky Texan said:

 

....on older cars, that sensor is just a cat conv 'nanny' and could be ignored or 'cheated' with a spacer or a coupla resistors to clear the code, but on most(?) newer cars

( and, it seems, on Gen2 H6es),  it seems to also be used by the ECU to modify a:f .

 Forgive my Subaru illiteracy... is the 09 2.5 Legacy one of the newer ones that use the downstream O2 for AF ratio?

John

[1 Lucky Texan]

53 minutes ago, JohnCTT said:

 Forgive my Subaru illiteracy... is the 09 2.5 Legacy one of the newer ones that use the downstream O2 for AF ratio?

John

There has been discussion about this at another Forum by folks much more knowledgable than me and it seems that, maybe starting with canbus (around 05 depending on model I think) the rear-most sensor may be part of the 'algorithm' for adjusting the a:f ratio.

hopefully one of the gurus here can chime in on this.

[nvu]

4 hours ago, JohnCTT said:

Anyway, the Subie exhaust is a one piece all welded affair, is it not?  I'm guessing a bad cat, but the exhaust gaskets from the heads on out can only cause a lean code, not an efficiency code?

https://parts.subaru.com/a/Subaru_2009_Legacy-25L-4AT-4WD-Limited/_54102_6024341/EXHAUST-EXHAUST-COVER-06MY-/B13-440-10.html

There's a gasket 3" away from the rear o2 sensor. 44184.  It doesn't look like the older donut style that tends to leak though.

 

5 hours ago, JohnCTT said:

That's what my intuition tells me, but if you Google P0420 codes, every return will mention intake leaks!  Doesn't make sense to me either.

I tried https://www.google.com/search?q=subaru+2009+legacy+p0420, only thing that sounds plausible is cleaning the maf.  

 

[GeneralDisorder]

You only get a lean code from sustained positive fuel trims of 25% or more. If the fuel trims are a result of an actual lean condition such as an intake leak, then the AFR sensor's corrections are accurate and the converter will still work (if it's good) and the secondary O2 will read a steady 0.7v or so. 

If the front AFR sensor is bad however, it can cause a lean reading leading to erroneous positive fuel trims (with or without a lean code), overloading the converter and driving the secondary O2 voltage high. Secondary O2 sensor voltages that are stuck at 0.9v+ should be a red flag. 

The tell-tale of a bad cat is normal fuel trims (+/- 5%) and a fluctuating secondary O2 voltage. If the secondary O2 switches in the manner of an old school primary narrow band then this is an indication the converter is non functional. 

And for the love of god please would you people stop honing cylinders?!? Damn when will this wives tale die already? 

GD

Edited June 28, 2019 by GeneralDisorder

[JohnCTT]

13 hours ago, nvu said:

https://parts.subaru.com/a/Subaru_2009_Legacy-25L-4AT-4WD-Limited/_54102_6024341/EXHAUST-EXHAUST-COVER-06MY-/B13-440-10.html

There's a gasket 3" away from the rear o2 sensor. 44184.  It doesn't look like the older donut style that tends to leak though.

 

I tried https://www.google.com/search?q=subaru+2009+legacy+p0420, only thing that sounds plausible is cleaning the maf.  

 

Maybe I'm looking at this wrong, but it seems the upstream O2 is in the collector before the cat, but the "downstream" O2 is actually in the **middle** of the cat. {???}

John

[JohnCTT]

12 hours ago, GeneralDisorder said:

 

And for the love of god please would you people stop honing cylinders?!? Damn when will this wives tale die already? 

 

GD

 

Do you mean don't hone at all (leave as is) or do a full cut?

Anyway, I didn't completely understand the AF description, but I had my son check the fuel trim with his scanner, and he said it had a Long Term Fuel Trim 1 of -7.7% and a Long Term Fuel Trim 3 of -100%.  Since this has a single cat after the collector, this is a one bank system even thought it's a boxer, correct?  I assume the Trim 3 is meaningless.

Does this lean towards a bad cat?

Also, you used the term AFR sensor.  Does the Subaru use a full range sensor or a standard binary O2 sensor?

John

[GeneralDisorder]

-100% would result in no fuel being injected so that reading would seem likely to be meaningless. Probably a glitch in the scanners data stream definition matrix. 

The Long term at -8.... but what was the short term? 

The front sensor is a wideband. But it's not a typical 0-5v. The ECU reads milli-amps from it's AFR sensor. It is not a 0-1v sensor. Narrow band sensors are not Binary either. They just have super aggressive slopes to their voltage when responding to other-than-stoich ratios. 

And no I don't hone ever. It's a terrible idea from the 70's and before. If a cylinder is bad you bore and plateau hone it. If it's not you run it. The rings will seat no problem. 

GD