Hank Roberts

You'd have to ask someone who knows! I just pointed to what someone said. It sure looks like the description, doesn't it? Is it available somewhere nowadays?

Here's the old post about customizing a socket to make this easier -- it used to have pictures, no longer linked -- it still has the description of how to cut down a standard socket so it will fit in to retorque the heads without having to remove as much else, and a part number supposedly for something Subaru made for the same purpose that might or might not still be available form a third party. http://www.ultimatesubaru.org/forum/showthread.php?t=12239&page=2 and page 3 and page 4 of that.

I once, somewhere, saw a post about actually cutting down a socket, shortening it so it would fit into that space and still hold the bolt. I guess you'd want a six-sided socket and a very precise ratchet. I saw a digital torque wrench used recently (at LIC, in Marin). It spoke aloud when it reached the right torque! Impressive---they said it's a lot more precise than watching the needle on a bendy bar. So, how often can you tell they're loose, when you retorque heads? One on every vehicle, or ....??

Ding. This may be the winning answer. Shudder. How long do you run the engine (time? distance?) before retorquing? That has never been done on this engine, not when it was first rebuilt and not after the two repairs at the rebuilder's shop. I'd asked the rebuilder, they told me retorquing isn't needed; they use Subaru gaskets. Hmmm. The left head has been on 800 miles, since they resealed the engine a year ago -- or else under 2k miles, the rebuilder replaced it the first time it leaked. Is retorquing something that can be done cold? I wouldn't regret paying to get the left head checked now, even before the right one goes back on, just to know. Else I'm going to ask the local mechanic if he retorques routinely, and ask him to check both of them. How long do you run the engine before retorquing? How often do you find some loose, when you do the retorquing? I know, _once_ is enough to convince me it needs doing. I'm getting hopeful again.

That's the only part that applies to my question. I know steel 'loses its temper' (hardness, springiness) on overheating -- you are sure that an engine overheat cannot affect engine bolt springiness? Just puzzling what else could be checked before this engine goes back together--my warranty's got less than a year left on it now. Why do some people retorque the bolts after the engine's been run, and others don't?

This rebuilder did use OEM gaskets (so I understand from the rebuilder); they sent the replacement gasket for the right head being redone here this time, so they know it matters what's used. Well, we'll see how it works. I wonder if this mechanic retorques or not. Guess I'll ask. I gather that's a variable. I don't suppose overheating can make the bolts (I guess I mean studs?) any less springy? I know the torque's supposed to stretch them slightly so they pull down on the head --- does overheating an engine cause the threaded part to lose some of their temper so the torquing doesn't cause the proper amount of squeeze? Or is that not possible in this situation? (Like I said, I don' t know the history of this engine, it was a rebuilt on the shelf there and they swapped it for the one I sent them to speed things along. I hope it didn't come out of a fire or something, since it was meant to be on a hovercraft in Europe, who knows where it'd been. Or can you tell that?) Don't mind me, just daytime waking nightmares ... Most informative topic, thanks all.

Fantastic. Is that last TB 1300 the Subaru tech bulletin number? I think you've just made my mechanic _and_ the engine rebuilder very happy by finding that. I (and they) will still wonder why my engine _did_ leak at several different gaskets, three times (wince) -- but at least we all can cross this other worry off the future problem list. And that puts us back on topic as well. Sorry for the digression, but it's been _very_ educational and will continue to be.

Dang. I'd sure like to have more than friend-of-a-friend being _told_ that. It'd take pretty good statistics to know it's true. I'd hoped it would be on the record somewhere. Too bad. Ok. I took another look at the cracked head today and asked how the 25mm was measured. He simply measured the vertical distance from the face of the meta. The cracks visible opening goes down the vertical wall between the valves, and widens out into the curve at the base of the hole. They look a lot like the ones your pictures labels bad. Maybe not so wide at the narrow top, but they'd been hammered there. Wider below the hammered area, then narrower as it goes deeper, visibly stained around the cracks lower down, can't tell if it's liquid leaking out from the crack or discolored metal around it, kind of greenish brown color. I couldn't get pictures today, may try again. Both pairs of valves are cracked about the same. I guess there's no way to tell for sure --- X-rays maybe? ---how much deeper either crack goes toward the water channel in the thicker part of the metal. The cracks are certainly too irregular to get a feeler gauge into them. I can just see taking it to my dentist .... does anyone X-ray aluminum castings to detect cracks? When you cut sections through cracked heads, did the bottom of the crack go deeper in the thicker metal, in between the valve holes, below the place the metal curves, do you know? Or was it a straight line across between the valves once it got into the thicker metal there? Or does it crack less deep in the middle than out at the valve hole area? I read up on aluminum cracking, it's not reassuring. Aluminum cracks eventually in use even when it's well made, and the cracks propagate with changes in temperature and pressure. That's different from steel, which can be made well enough that it has little variations that halt crack propagation. Or so Wikipedia says (sigh). So much to know, so little time to learn it.

>TOD Well, I've talked to a lot of local GL owners in the past 2 years and they don't recognize the noise either (when I pull up next to one in traffic, sometimes, when mine was doing it, I'd ask!!). It may be temperate climate here, or people who do a lot of maintenance (serious smog check requirements?). I"m sure there are 4-wheeler Subaru beaters that recognize it. Remember I'm an old fossil driving very gently, except in the national forest logging roads and then in 4wd low range low gear at 8mph. I'm not around anyone who pushes a Subaru hard for fun or beats it up. I'd appreciate a pointer to Subaru's spec on that, out of curiosity. Note, some of this thread is general advice others can use later. To keep clear what's general, separate from the specific for me --- Very helpful and clear about how to get heads welded. I hope we hear from someone who's gotten it done further down the thread with pics! And if anyone's sawed through a welded head later, pics please!! Mine won't get welded -- it's under warranty, what's fixed is decided by the rebuilder; rather than ship it interstate again, they're working with the local guy to do the warranty repair for this third leak. On mine specifically: Rebuilder and this mechanic _both_ agree they don't rebuild engines by his standard; he's, as you said, fanatical. They rebuild them in large volume. He does one at a time. He _does_ need a reliable source for rebuilt Subaru engines that are affordable. I told him about the rebuilder; they told him they quit fixing heads with that kind of grinder a year or two back. They use a rota-something softer tool now. so, they may agree more on what's reasonable for rebuilds. If so he'll buy from them -- he can sell people their cheaper rebuild here and keep more Subarus running, if he accepts it's good enough. He knows his one-at-a-time are too expensive on these old cars. If it works out --- he'll get a supplier; the rebuilder will get a local warranty guy; and I'll have an engine that holds fluids finally. ----> So thanks generally for all the thoughts. Wish us all luck.

>"fanatically maintained" That's right >"not a Subaru specialist" That's wrong. He's a pro, not a home-builder. Different standard of care. He's worked on local older gen Subarus for a very long time. He cuts no slack. As you say, liability insurance requires not cutting corners. The engine rebuilder was honest with me --- they told me they can't afford to do their rebuilds to this guy's standards every time. They've sent at least one other customer to him since I found him. YMMV as always.

daeron wrote: I'm not worried TOO much.. I just cant get the image out of my mind, thats all. I am kinda morbid like that sometimes. Well, I wonder, too, things like: -- When cracks are welded, is the welding actually filling the crack all the way down to the bottom, or just covering over the top of it? Doesn't the crack continue to propagate, or is holding the top edges together with welding keeping it from getting deeper? -- Where does the metal _go_ that was in the cracks? I'm assuming it doesn't just disappear or shrink, but that the hollowed out part of the head holding the valves is changing shape somewhat. So that might argue that welding the top of the crack stops it from getting deeper. But .... I guess using that hacksaw is the only way to know for sure. -- Aren't combustion chambers sensitive to shape such that some fuel and air are in the crack and burn poorly or change the combusion process? My mechanic, who says he'd never reused a cracked head on any Subaru in 20 years, also says the "noisy hydraulic lifter" noise isn't normal in these engines. -- Is the measured oil pressure at the sensor likely to be inaccurate with cracks between valves like those pictured? Or, a better question, have y'all who did find cracks there also had the noisy lifter tapping even with good oil pressure readings? Just wondering.

Can y'all specify how you measure the depth of a crack? Do you insert a thin wire or feeler gauge, something that can fit into an irregular crack til the crack is the same gauge? Or a straightedge metal ruler inserted into it? Or whether it appears beyond the bottom of the valve seat (seat, the Some of those cracks pictured look (gulp) really wide and ragged, and obviously the bad ones extend past the bottom of the valve seat. This is really interesting. When my right head comes back from being milled, I hope I can photograph and measure it --- I want to know how y'all do your measurements. And I want to send the pictures and measurements to the rebuilder and get confirmation what's left of their warranty continues to be good after putting this back on. I looked back at the old notes and the first time this engine failed, on first delivery, the left head was so damaged by their grinder guy that they replaced it they said with a new head. I sure hope that's right. Two cylinders good ... four cylinders better, eh?

Got it. I'll ask how thick the probe was he used, to measure the crack down to that depth. Hope it was really thin and the rest of the crack doesn't extend much further into that 5/8" remaining. I assume cracks keep getting deeper. Is there any practical difference once it happens, between a leak when that kind of a valve crack gets down to the coolant channel, versus a leaky head gasket that gets to the coolant channel where it crosses the gasket? Any difference in what happens, how it looks, or how to handle the vehicle if it happens on the freeway for instance? I can't say I feel exactly lucky. But I'll have another 10 months or so of warranty on the engine, so want to watch it carefully assuming it gets fixed this time around. Any "look hard" advice to catch problems early, before the big cloud of smoke and steam stage, that I should be checking regularly? I know to look for brown gunk in the oil and oil in the coolant and white fog out of the tailpipe and oil smoke out of the tailpipe and puddles under the car -- seen all that! But is there anything to check to catch an engine seal failing before the public embarassment on the freeway stage I should be doing? Like Mr. Reagan said: "trust, but verify" --- seems appropriate by this time.

>ASV Port Yes, that sounds like what he showed me on the right side head (properly plugged, said the head would've started life on a carb engine). Do you have a picture showing how the usual crack between valves looks? I know the engine rebuild includes new valve seals; don't know how that fits with cracking, can a 1" deep crack allow any pressure to bypass the valve seals? I gather hacksawing through those would be real tiresome to get a cross section?

My mistake--my typo -- not "port"--- I meant "point" -- attachment point. Bolt hole, bracket --- for attaching what goes in front of or on the top of the engine. The car's at the shop I can't describe it any bettter. Don't leap to decide the mechanic's at fault -- I see that a lot on USMB, and for most of y'all it's true you know a lot about these and many mechanics don't This is the opposite. I finally found the guy who's worked on these for decades, in his quiet little corner way across town. I've only had this one Subaru, for, less than three years. So if something sounds wrong, suggest I need to learn better first, it's almost surely my mistake. So -- clearing up my fog: On the right head, last week, the mechanic said this head originally came off an engine with a carbeurator-- no problem, pointing to where some pollution control part would attach to it for a carbeurated engine. That's why I was asking, what kind of engine could this head have come from. Thanks for that reassurance it's not been on a turbo in its past. Now -- long ago: I always look for forgotten hoses, loose wires, mislaid tools, drips, half eaten ham sandwiches and stuff under the hood. I've found all of them in the forty years I've been driving cars. I used to be able to work on my own, too. Some of them were my fault. On the Su, an earlier time, I noticed a part (radiator? fan? something that sits on the engine) had been put back on after the rebuild -- but it had a flange that didn't line up with a bolt hole, sticking out obviously unattached. I asked what they had left out(!). That shop (the old one) said no problem, the rebuilder replaced your left head with one from a different engine -- it doesn't have a bolt hole there. No problem, three out of four bolts lined up [radiator? fan? whatever it was]. Hope that fits what you know. "Attachment point" not "port" -- my typo. This is the only Subaru I've ever had (a hasty replacement for a Tercel 4wd wagon, that got stolen and trashed three years ago). The left head was replaced 2 years ago --- on the first warranty repair--- by the rebuilder; they said they checked the right one as well, that time, 2 years ago. They said that was their shop's screwup, high and low spots left from grinding the head. The car never got off the rack that time, it leaked as soon as the engine was put in and filled, and it went right back immediately. That lasted a year; leaked, went back; repaired, ran another 800 miles, leaked, and that's today. As far as MPFI, i suppose it's possible turbo heads could end up on a NA-MPFI(non-turbo) motor. But 98% of those motors are in XTs or have are transplants into GLs or Loyales. Hank what year Subaru do you have? This is a 1988 GL - SPFI - non-turbo - 4wd - 5-speed dual range wagon. I got it for the low gear range, for slow steep mountain roads doing field botany work and taking kids camping.

Good to know, thanks. I know this rebuild doesn't quite match --- missing one of four radiator attachment points (radiator place said not to worry) and this mechanic said this head has one carbeurator attachment point. So I knew had been somewhere else, but not how to tell. How about for other kinds of engine? --- someone above noted finding the exhaust port crack on his non-turbo --- could MPFI non-turbo engines have gotten turbo heads during a rebuild sometime? Just trying to sort it out. Looking forward to your hacksaw revelations!!

On the machine-gun noise, I found Gloyale has a thread describing something that sounds like mine --- loud on startup til the engine warms up: http://www.ultimatesubaru.org/forum/showthread.php?t=71978&highlight=head+crack Once the head's redone on mine, I'll follow up in your thread if my noise persists. My engine came apart before he had time to think about the machine-gun noise. ----- On the location of cracks == if you don't know where the heads on your engine have been --- a head could crack on a turbo then get used to rebuild a non-turbo, eh? I'll ask for a look at that exhaust area before it goes back on my engine. Just in case. The "mosquito fogger" cloud on warmup is how mine failed a year ago, 800 miles ago. But that time the engine went back to the rebuilder and I don't know what they fixed on it that time around. On the depth of cracks, I found one thread where someone actually cut a head apart to dissect out an exhaust area crack. Amazing pictures.

Hmmm. Another symptom never solved on this engine was a shake on idle, warm or cold. Once the RPM was up, it was smooth, but sitting at a stoplight it had a definite miss or beat or shake, at about heartbeat rate. The kind of thing that makes the rearview mirror look like a double exposure. I'd just had it tuned up and smogged a few weeks before the seals blew. Could that be related to the leak in the head developing or the crack between the valves?

Ok, it's out of my hands now, so I'm just being curious. >if you cannot fit your fingernail into the crack Which way are you measuring your thumbnail? This crack measured an inch deep; length is between the valves. The metal ruler ---- thicker than your thumbnail? --- wouldn't reach the bottom of a tapered crack of course. It went in an inch. How thick is the metal on the head total? Is there no oil flow between the valves through a connecting crack? I guess the hammering fix spreads spread the metal on the face to sort of cover the crack. Is that supposed to support the gasket, so it doesn't sag into the crack? Or to keep the oil from the valve stems from going through the crack and up under the gasket? Because that hammered bit came off when the gasket did, leaving the opening full width. This head has a mounting point for a carbeurator, so it began on some other engine. Any way to tell if it's off a turbo? Why would the rebuilder always put new heads on turbo engines, but hammer cracks shut and reuse heads on non-turbo rebuilds? Is the non-turbo so much less stressful that it won't propagate an existing crack? Should milling take the whole surface down far enough that the crack can be hammered shut again? Is oil diverting from one valve to the other via the crack, or doesn't that matter as long as the gasket doesn't get loosened? Does the crack have anything to do with the loud tapping noise on starting the engine up, for ten or fifteen minutes, really loud "machine gun" noise? That noise has often happened since I first got this rebuild --- was still happening, really noisy, til the seals failed. The rebuilder told me it's a tappet noise and no problem, and others on the forum agreed.

For me, the cost of parts is small compared to the time and cost of the repairs -- I remember asking about this. This engine supposedly had been built for a hovercraft -- to aircraft standards -- but never delivered, so was offered to me. I sure wouldn't have wanted to be in the air with it, from the way it's behaved. With cracked heads, is there any difference between what you'd do for yourself, vs. for resale? for aircraft use? Or do you consider it really can't have anything to do with the repeated failures, and shouldn't matter? I'm just trying to grasp how come a rebuilder's work can fail repeatedly like this, is it just 'bad luck' it happens? ------ Edit: well, so be it. A run of bad luck is a lot more common than people think (so the statistics course taught me, thirty-plus years ago). I recall baseball coaches consistently pull pitchers off the mound after a handful of fouls or strikes, believing they must have lost their touch for the day, well within the number of mistakes that happen normally and about the time their 'luck' ought to turn purely at random. So I'm hoping three failures is just that I got the string of bad luck. If so, y'all owe me (grin) for saving you from it. Keep me posted on what you think is right for rebuilding for yourself vs. for resale, though, if this provokes any thought. I've sure learned that, this round, I've found a conservative mechanic. Wish I'd found him earlier. If anyone needs a good, sober, serious-minded and "old school" Subaru guy around the SF East Bay, try Unitech on Folger in Berkeley. Ask for Bundtha. Tell him that gray haired geek who uses the Internet sent ya .... In sum: The mechanic's agreed to go ahead and do the repair exactly the way the rebuilder says they'll allow under their warranty. No other way we would have a chance at using the car by the end of June. The rebuilder's warranty covers all but the fluids. Head's gone off to be milled. I'll try to get a good picture of it when it comes back, if time allows. Wish us all luck. News when it happens.

Hm. This engine is SPFI, non-turbo. What would crack this? Would cracked heads off a turbo be reused on a non-turbo rebuild? I note the rebuilder says they put all new heads on turbo engines. I bought this from the rebuilder (off the shelf, to save 2 weeks vs. getting my own engine back from them rebuilt) (what a mistake that was). That was two years ago: this one has leaked -- three times: at first install; again at 1 year/maybe 1000 miles; now again after another year/800 miles or so. The local mechanic doesn't want to do as little as they'll allow for repair. Maybe just have to return the engine to the rebuilder. It's got about ten months left on the warranty.

Friday AM: Rebuilder called the mechanic and said, just mill the head and stick it back on, that's all the rebuilder will agree to repair under their warranty. Mechanic cleaned up the head, to send out, and found a crack -- depth of about an inch, 25mm deep, between two of the valves. The metal surface of the head had been hammered to close the surface over the crack, sometime in the past. Mechanic called the rebuilder -- rebuilder said yes, they do that kind of hammer repair on cracks when rebuilding heads, and they consider this normal for old Subaru engines, it's not covered under their warranty. Sound reasonable so far?

So, Eye Why -- still here? Did you find your brake problem? I'm back just to correct this: ... Subaru mechanic in Berkeley? --> [not] Autometrics. Their manager told me their mechanics really don't like to work on anything more than 10 years old. Instead -- Unitech, on Folger in Berkeley. Real, serious, good older Subaru mechanic there.

Rebuilder hasn't called the mechanic on Monday, as promised on Friday they would. Waiting to hear. Hard to sleep. Got no place to do this work myself.

I had to do this with my 88 GL wagon, rear hatch window --- the goop had come loose and it was leaking, collecting water down in the wells behind the rear wheels. Same method everyone's mentioned --- good razor knife, some wire, patience. With the rear hatch, I could just prop the hatch level so the silicone wouldn't run til it dried. I dunno, I guess with a front windshield I'd try to put it facing uphill, on a _steep_ hill, and chock it real good. The silicone will run for a while. Porous masking tape will work ok to keep it from dribbling out, the paper kind, it has to outgass some acetic acid vapor as the silicone sets. It took about three days before I dared move it, it's slow stuff. Doesn't leak, though, after 2 rainy winters since I replaced it.