Is this on the drivers side front of the cylinder going vertically into the head?

 

Here you go. Someone on a Forrester forum complete with pictures and their experience

('06-'08) Oil Consumption or a leak? Help Please! - Page 2 - Subaru Forester Owners Forum

This guy first tried to have someone weld it which was botched pretty badly then ended up getting a new head. If it were me I would JB weld it...

Good luck

 

an oil leak on the passenger side of the engine it looks like it is coming from the sensor on top back of the cylinder head.

It's probably the sensor related to the variable valve lift (VVL) operation. At the back of the passenger side head, there's three related parts -- the VVL solenoid (22), the VVL oil pressure switch (26), and the VVL Oil Temperature sensor (25).

I suspect it's either the pressure switch or temperature sensor. The pressure switch is actually in the head casting, whereas the solenoid and temperature sensor are in valve holder (23), which would make repair far easier.

There have been reports of failures in the solenoid and the two sensors, all of which lead to significant oil leakage. There was also one thread here in which the gasket between the valve holder and the block was, apparently, missing.

[Edit: Woops, doubled up with Glennda5id, but will leave it with the diagram.]

 

It's probably the sensor related to the variable valve lift (VVL) operation. At the back of the passenger side head, there's three related parts -- the VVL solenoid (22), the VVL oil pressure switch (26), and the VVL Oil Temperature sensor (25).

I suspect it's either the pressure switch or temperature sensor. The pressure switch is actually in the head casting, whereas the solenoid and temperature sensor are in valve holder (23), which would make repair far easier.

There have been reports of failures in the solenoid and the two sensors, all of which lead to significant oil leakage. There was also one thread here in which the gasket between the valve holder and the block was, apparently, missing.

[Edit: Woops, doubled up with Glennda5id, but will leave it with the diagram.]

Ok back from vacation its now tims to try and fix this problem. In the picture #26 is the senor that is what I am working with. The cylinder head is cracked on the back side. Is it ok to take off #23 and the parts of that to gain access for welding? I am thinking about welding the top and re taping the hole if needed and some type of epoxy on the side to seal it up and some good pipe dope on the threads. Well that's how I hope it works out and if I can do the work with out pulling the engine it will be that much better!!

 

Good lord, are those tapered threads? You'd never use a torque wrench on a tapered fitting...

Either way, that is the type of setup that screams DO NOT USE A TORQUE WRENCH!!!

 

I'm unclear on this...why would you not use a torque wrench?


-Jeff

 

Timesert might be a good option for something like in that thread, you could maybe install 2 to cover the full depth of the crack.

 

OK, I know what I'd do- drill a hole horizontally through the crack, counterbore to just beyond the crack, thread the rest, use a bolt to pull it together like a clamp, after cleaning the heck out of it and getting JB weld into the crack... but I'm a little gung-ho.

 

Oh wow, there's a whole world of magic out there for crack repair.

Lock-N-Stitch Repair for engines, compressors and other cast iron equipment by Reynolds French

The thread design pulls the cracked pieces together...

 

I have seen the Lock-N-Stitch method done on a 5.9 Cummins "53" block on a 5 inch block crack that was still good until the pickup was sold 100k miles later.

 

if you have any really good automotive machine shops in the area, they are worth it. make some phone calls. a few of them will work on the engine in the car if you bring it to them. they have all sorts of goodies and do this stuff daily. i've done it a few times - keep calling until you score a place that's a good fit. if you were in Westminster, MD i'd know who to send you too.

they use stitch pins to fix cracked cylinder heads - like the old school Subaru EA82 and ER27 heads in the 80's that used to have the mostly benign cracks between valve seats.

i would find a way to fix it. either with stitch pins or something else, i'd probably try a home remedy myself first.

i would avoid automotive shops/garages....most are not going to be economical because they don't do this kind of custom one-off work.

what a bizarre issue....

 

I used to live just down 97 from westminster... hard to believe an area so nice could be so close baltimore and columbia (can't really decide which was worse...)

 

I talked with my cousin he is a very good welder so I am hoping we can get it fixed while in the car. If I have to pull the head to fix it I will.

I figured I would have to pull the engine out of my 2000 outback before I had to pull it out of my wife's 2008. I guess you just never know what you will have to fix next that's why I have a lot of tools.

 

Far far too many variables in a tapered fitting to go by the torque, especially if not using the exact goop to seal the threads.

There are a few small bolts on my motorcycle that I refuse to use a torque wrench on, and I read posts from folks all the time using a torque wrench and snapping them.

 

Far far too many variables in a tapered fitting to go by the torque, especially if not using the exact goop to seal the threads.

Ah, OK, I think I see what you were saying now.

It's not really a problem with using a torque wrench, but rather, with using the wrong sealant, or wrong torque spec, or an inaccurate wrench.

It seems like the OP tried to crank the sensor down with a regular wrench, with no idea of the required torque or sealant.

But your point is still valid.

First, you can't rely on the torque specs found on internet forums. For example, the thread linked by Glennda5id says 18 ft-lbs, but the 2008 Outback FSM actually says 12.5 ft-lbs, for the VVLDOPS. And it looks like a rather delicate, isolated part of the casting, so the spec is more important than usual.

Second, you can't assume that the torque spec applies unless you use the "exact goop" (lube or sealant) that the FSM specifies. In this case, "Three Bond 1324". I dunno how the teflon tape changed the spec in this case. Actually it might have made it less likely to overtorque. But still...

Third, you can't use a cheap consumer 3/8", 80 ft-lb torque wrench with any hope of accuracy on a 12.5 ft-lb fastener. For that you need a 1/4", 200 in-lb torque wrench. Or the skill & experience to "feel" the correct torque. But of course most people do not have that skill&exp, which is why wrenches come in shorter lengths for smaller drive sizes.

Oh and fourth, with any threaded fitting, you need to consider the possibility that it was overtorqued by someone else in the past, exceeding the yield strength, aka stretching the threads, which means you need a new one. The FSM is pretty good about pointing out when fasteners should be replaced in normal use, but it doesn't know if someone else has screwed it up before.

Overall I agree with you. The average home mechanic should not rely on a torque wrench with the wrong range, wrong spec, and wrong goop to tighten a low-torque fastener. But if you can get all of these things right, using a torque wrench is always better than just guessing.

:7:
-Jeff

 

I contend there are situations where using a torque wrench is the wrong choice, even given the exact correct wrench and application.

There are situations where the cited torque is great if something gets disassembled/assembled, say, 3 times, but on the 5th time, breakaroony.

Even Honda uses butter-soft 6mm hardware sometimes...

 

Totally agree.

Factory manuals give torque specs for first assembly or repair. They usually tell you which fasteners must be replaced (in the Subaru manuals, with a star on the "general description" blowups), but they don't know how many times each fastener has been used or abused.

This is one of the reasons that I only rotate my tires when necessary. On my Outback with the 45/55 VTD tranny, just once for each set of tires, around 25k miles/2.5 years. Would be nice if that was synched up with brake jobs or boot replacements, but of course it isn't.

In this particular case, I'm pretty sure the problem was not using a torque wrench, and instead trying to "snug it down" on a delicate fitting, with no idea of how it should "feel".

YMMV.

:7:
-Jeff

 

teflon and oil do not mix. Oil deteriates the teflon, and over torque is easy. Did not use proper thread sealant, likely old permatex would have been ok. Best of luck.

 

I'm fairly sure that teflon is inert and does not deteriorate in the presence of oil. That's why we use it to tape the threads on natural gas and corrosive gas fittings. Or to coat frying pans that will be used with hot cooking oil.

But yeah, the effect on required torque is still an issue.

Teflon tape reduces friction, and also increases the thickness of the threads, so presumably the required torque is less. Hard to say for sure.

Usually you can get away with over- or under-torquing on most fasteners. It looks like the real problem here is that this particular sensor mount is a delicate piece of the head casting, not very forgiving.

:7:
-Jeff

 

Is it ok to take off #23 and the parts of that to gain access for welding?

Can't see any reason why not, but will need a new gasket, and care with the bolts (but you know all about that). Also, be extra careful when drilling/tapping -- you don't want to get any bits into the oil galleries in this system -- this could damage the valve lift mechanism.

See attached.

 

cracked 2007 2.5 sub head at oil sending switch

I too, messed up big and cracked the head at the oil sending switch location on the passenger side head. I spoke with a machinist who indicated that he could weld it back on if I bring him the head. Does anyone have any experience with this issue. I saw a post from May 2012 indicating that this had happened to someone else. Any information on how this was remedied would be greatly appreciated. Certain people here in the household are not exactly very happy - me being one of them. Thanks.

 

I don't recall seeing any reports here that have a fix other than to replace the head. (Which is also the "fix" in the thread linked in post #3 above.) However, there are a number of subaru-related forums on the Internet. Perhaps someone was able to work out a fix and posted it elsewhere. Here's a few:

Subaru Forester Owners Forum
Ultimate Subaru Message Board
Subaru Legacy Forums
North American Subaru Impreza Owners Club

 

Tapered thread are usually supposed to be tightened using hand-tight plus 2P (2x pitch, AKA 2 revolutions), sometimes +3P (from Machinery's Handbook), but especially on the smaller sizes, should be done by feel; following those guidelines exactly can break fittings occaisionally. A tapered thread in an aluminum head seems especially vulnerable, if that's what this is (can't see exactly what's going on from the diagram).

I deal with specifying torquing instructions for tapered threads on an assembly line for production, and it really comes down to proper training. Tapered threads by nature vary in depth from variation in manufacturing, so what works one time can break another. I've never seen tapered threads with a torque setting, but automakers usually set a torque value for everything, so I'm not surprised.

Personally, on any tapered threads I use the hand tight plus, ahhh, about that much, lets see if it leaks method. Leaking? OK just a bit more... especially the smaller sizes.

 

Hey thanks for the information on the oil switch cracked head issue. I have it at a shop, just don't have time right now to pull head -trying to see if it can be welded on car first

 

^^^ Not gonna happen. They will damage something else trying it.

 

Not going to recommend trying to weld it on the car, too many risks. The firsts is getting something into the oil line and restricting or plugging it up. The more expensive risks are anyone using a high voltage/high frequency welder which can literally damage sensitive electronic components, ECM, TCM, ABS HCU and or BIU. The components your describing are part of the active valve control system (AVCS) which applies oil to a small piston which locks cam follower to a High Lift follower for more performance. It isn't any fun and not cheap, but the right fix is a new/used replacement head.

 

If it's only a pressure sensor, there isn't any flow through the orifice and it's size isn't critical. Can it be possible to install a reduced-size fitting into the hole with some sort of heavy-duty epoxy, then connect the sensor through some additional fittings? The fitting installed into the hole could have a hexagonal portion to hold it with a wrench when installing the sensor. Just thinking.

 

That sucker could be holding in 80 psi or so on a cold start, so who knows.