birkhoff

Good news Now you can focus on the heads, which are increasingly looking like the culprit(s).

Depending on what you mean by this . . . Measure flatness by putting your straightedge on TOP of the feeler gauge. Then pull on the gauge. If the straightedge follows the feeler, the gap, if any, is less. Do this in a few different directions, measuring at various points along the straightedge. I don't think you need to go any finer than 1.5 thou, if I am remembering the specs correctly. Myself, I've never had much luck stuffing a 0.0015 feeler gauge under a straightedge. Not with any confidence, anyway.

The valves were in water for a while, when the OP `wet tested' the heads. Could that be the source of surface rust? There doesn't seem to be any other evidence consistent with a HG problem or cracked head, unless pretty much every cylinder has a crack. Any pitting on the valve sealing surfaces, however, is just plain bad. It is hard to imagine how the engine was running without lots of problems with chewed up valves. Is this also surface only, or can you catch a fingernail on the pitting? As for distortion, I think it would be tough to measure valve length accurately, and that might not be a very useful number anyway since the important measurement is the length from the seating surface (midpoint, say) to the stem end. Brand new, the overall length might be a satisfactory surrogate for effective valve length. Eyeballing the remaining valve margin might reveal something. Does it look uniform? Are they all more or less the same (exhaust to exhaust, intake to intake)?

Let's say you find a hairline crack in a piston. Would you consider that significant? Unless a chunk falls off a piston during inspection, I would focus on the valve issues you seem to have uncovered. Those poor subaru pistons. They never get a break!

In a word, no. If you are seeing that on the exhaust valve seats on cylinder 2, along with the liquid test, you probably have found your problem. You should be happy! Your car just needed a `valve job' Early in the life cycle of these cars, broken ring lands were all the rage. Now we are moving into the late stage of designed lifetime, I suspect we are going to see more top end problems. That was the case with my engine, and perhaps yours. Careless mods and hard driving can break pistons no matter what the condition of the bottom end. But with 150K and no attention to valve clearance, top end problems will start to appear. I expect we are going to see more and more valve issues on top of the usual broken pistons. I was going to suggest you get some lapping compound and see how uniform your contact area was on the #2 exhausts, but you have kind of answered that already if I understand what you are seeing. Another possibility, that could be a parallel problem, is a hairline crack in an exhaust valve that may have distorted the valve head, or opened up at operating temps making #2 compression marginal at hot idle. All speculation of course. As far as the SB, rebuilding what you have now is even more viable than when you started since nothing failed. A new SB is still an option, just less necessary than when you started. I would see that as a good thing, given your plan!

Diesel/Kerosene is a better bet. Its incredible creep -ability will find any crack eventually, no matter how small. Surface tension in water is going to limit this test. But anyway, you have some information there that may indicate something up with Cyl #2. Re: dropping parts. Better to get this out of your system now. We've all have been there. Wait 'till you drop your $150 micrometer on the floor

Just saw your pictures of the internals uploaded in the first page. Everything looks good (and normal) with the bearings and presumably the crank as far as we can see from the pictures. Congratulations, you got everything broken down in record time! I asked about the oil and streaks in my previous post because I didn't see that in my first teardown of a relatively well cared for SB, but I saw lots of it in my 180K engine that had a couple of years of poor oil change history. It was also burning a LOT of oil and all the piston tops had the telltale signs of oil wash around the edges. You didn't mention anything about oil consumption -- was that also a problem?

True enough, but save that for after the forensics. At this point, there is still no obvious reason for the OP's misfire. The plan is to use new OEM pistons anyway, as I understand it.

Deposits on piston tops looks on the heavy side of normal. The flushing action on the piston edge could indicate poor oil control. Check for free movement of the oil rings. Skirts look pretty worn. Not sure what that streaking on the skirts under the oil return ports is about. Is it possible this engine was run with a lot of dirty oil at some point? There is a dimple on each piston top. The dimple faces front on assembly. They are not showing in the photos, but you should be able to find the dots. That will tell you the orientation. I wouldn't completely rule out ring land issues until you pull the rings and clean things up a bit. Sometimes the cracks are sneaky.

I KNOW you shouldn't remove it. If you damage it, not available without buying a new head! Leave it in place. I know this because I bought a used head where the PO had pulled it out for some reason and damaged the pipe. Tried to source a new one but in the end, had to swap mine over. Major pain.

I have an idea. Before you go tearing down those heads (and you really don't need to do this until you get to the bottom end of the SB -- you will have plenty of time to deal with the heads while you are working on that) tip them on their sides, and upside down. Pour something light (diesel, paint thinner, whatever you have handy) into the exhaust ports and see what happens at the valves. It should leak out slowly, through the seals and past the valves themselves, into the combustion chamber, but what you are looking for is something unusual with Cyl 2. If it escapes real quick from there, something is up. If there is crack, you may even find it as the oil wets its way through. It probably makes sense to wait until you tear down the case and inspect pistons before you do this. If there is an obvious problem down there, then you have your answer. Otherwise, it might be handy to not have torn down the heads just yet.

Heads look great. Maybe even a little too clean, but photos/shadows can make it hard to judge. At 110K miles it is unlikely the heads have been off before. Says a lot about how well these engines can survive if cared for. Any pictures of the piston crowns? Cylinder walls? Before pulling pistons I mean. You are moving quickly. Don't forget you are looking for a mechanical source for your misfire. Check the heads/valves carefully for hairline cracks or any other sign that could account for the symptoms. Especially the exhaust side of things. It is possible that you will finish the teardown and not discover anything obviously wrong. That can be really frustrating, and is something you want to avoid. I just noticed your leakdown numbers entered into an earlier post. They look fine. Just like compression tests, it is generally more important that they are balanced, rather than the exact numerical values. Your gauge may be out of calibration, and the engine was cold. From those numbers I would not expect to find anything seriously wrong with the heads or the tops of the pistons.

That works too Most likely they spun in the right direction.

I think I'm kinda slow getting the hang of this internet forum thing. One sentence is better than 10. One word better than one sentence . . . This is what I meant to write: = tie down one cam, hold the other = let go the clamp = figure out which way to turn the other so as not to go over the cam nose = nothing goes crunch? = release the second cam = I did it.

You need to unload the cams on the #2,4 side before you pull the bolts out. Since you already have an allen socket, put it into the exhaust cam bolt with your breaker bar* and tie that off to hold the exhaust cam from spinning. You should be able to ease the intake cam back to free position with your hands. Once the exhaust is locked down, have someone release your C-clamp while you hold the intake cam gear firmly and then slowly let it go back into neutral position. I believe the intake should be released cclockwise when viewed from the front; the exhaust opposite. What you are trying to avoid is pushing past the nose of the cam and forcing the valves full open. Since you have the rocker covers off, you can just look at the position of the cam to see the correct way to rotate and release. Once the intake is in neutral, you can rotate the exhaust in any direction without worrying about crashing valves. FSM would say clockwise. Chances are, if the gears are lined to the match marks, if you just let them go they will snap back in the right direction. But it does pay to take care if you want to re-use your valves (and you do!). Of course all this is done with the crank mark lined up, right? That puts all four pistons part way down the bores. *Don't use a regular ratchet since you don't know which way the cam wants to spin. You could figure it out of course, but you don't have to with the breaker bar.

Dipstick. As I recall: One bolt at the top, into the head. O'wise, just pull it upward out of the tube in the oil pan. There are 2(?) O-rings in there that I would definitely replace. OOPS: double post. Sorry BBM. Suggestion: get some penetrating oil on those bolts holding the belt guides to the back timing cover. They get corroded and it is very easy to tear out the threaded insert from the plastic cover. Go slowly, or you will be buying a one or more new back covers. Felt strips and sealing tape for the back covers was not in my gasket kit. You may want to order those in advance. Also consider renewing the plastic grommets behind the back cover bolts. They get really brittle with heat.

I have a bunch of pics sitting here ready to go into a `how to split your case' post but just haven't gotten around to making it. It includes pictures of using the crank itself to break things open. Basically, once you have all the bolts out, insert a wooden dowel through the cylinder access ports and the wrist pin bushing on one cylinder. Turn the crank in the direction to press the case apart. Repeat as required in a cross-wise back and forth pattern at the other cylinders so as to walk the case apart. Takes only a minute or so and nothing gets beat on. Here are two relevant pics. Sorry about the poor quality.

When you get to removing the heads, remember to release the bolts in stages in the reverse order of the torque sequence (from FSM). Not a bad idea to follow the same practice with the case bolts. The suggested teardown procedure is in the FSM for both. The case is split at the bench, which is what all the cyl 2 &4 stuff is about, so you end up taking the RHS case off the LHS. In practice, they will stick together pretty well so you can just take all the bolts out and use your preferred method to split.

I think you've done this before Congratulations. Install is just around the corner.

Not necessary to remove the turbo before the pull. It won't hurt, but it won't help that much either. Great to see steady progress.

Ok, I'll bite. What does humidity have to do with this discussion? We're not (re-)building a piano here. I didn't keep track of humidity anywhere along the way with my build. Does that mean my engine's gonna blow??!

Nope, I am not (an engineer). If your case is re-useable, that is probably because you still have good cross-hatching in place on all bores, amongst other things, which means the best plan for OEM is to replace like with like. A- pistons in A- bores and likewise B. So, once you get that far, you could ask the forum for a set of OEM pistons of selected sizes. I would not do that until you verify you can drop in. That should get you pistons in one or two weeks, which is about the time it takes to pull the rest apart and measure the other bits, saving perhaps $200? You'll be wanting to measure good to a couple tenths (ten thousands of an inch), not thousands, but I know you already know that. With my bore gauges, I saw drift on the scale of the tolerances I was shooting for while going back and forth recalibrating and handling the gauge head. Micrometers have plastic heat shields on them for good reason. Even very expensive ones! Try to keep fingers on those covers. None of this arises in a typical automotive rebuild. Subaru specs are an order of magnitude closer fitting than anything else I've worked on. No criticism intended on the budget thing. We've all been there. It would be senseless for me to seriously claim you can cut $xx from the budget. I'd lose that bet every time since it is you spending the money! And anyway, I'd already be at a disadvantage after those camshaft bolts . My statement was more philosophical, shall we say.

Yes of course. Massive changes in clearances. But that is what the engineers are for. We can't measure things at operating temps, so everything is extrapolated back to standard temp: 20 deg C. Just like valve clearances. We set them cold to the values specified in the FSM so that once everything heats up, they don't close up completely! Subaru allows pistons to be ~0.0005 bigger than the bore. You can be sure that situation doesn't persist at operating temps! As another example. Measuring is an art (didn't someone say this already in this thread?) Probably best to just get started and see how it goes. In the context of the kind of rebuilds we are discussing here, operator error is much more likely to be a problem than temp control, so long as you are working at steady `room temp'.

FWIW, I've found that careless handling of measuring tools makes a bigger difference than a few degrees of ambient, as long as the latter is stable. Eg: bore gauges. Unless you have a setting ring (which I am pretty sure you do not -- nobody does) then you will be going back and forth and back multiple times to check against your micrometer. Setting with a micrometer means you have to handle the measuring head, and there you start to get drift. Since we are not professional machinists, we are slow and klutzy at this, which makes the drift worse. Use a glove to cut down thermal transfer. You will know if things are working if you get repeatable measurements that are sensible relative to factory specs. If you can achieve this pretty quickly, then you have a natural talent and the rest of the build is going to be great! As far as budget, they tend to blow up in steps; plus $100 for this, plus $250 for that, and plus $200 'cause I'm in there already. The in-for-a-penny-in-for-a-pound principle is almost irresistible. However, budgets can be controlled in steps too. Do you really need it? Could it be replaced after the engine is in and you have determined that you don't have a pile of expensive scrap shortly after first startup? One trick I use is to remind myself there is a non-zero probability that the build will fail, and all my money, both essentials and nonessential will be lost in the pile of slag in my engine bay. For my recent build, I assumed 50%, just to keep the reigns pretty tight! If you decide to go this direction you may want to take up some of the generous offers of parts from friends. Someone (I think) offered you a set of OEM pistons. So far I have bought two sets off this forum, brand new pulls for $100 all in, pins and clips. Just have to source rings*. There's a big savings for you right off the bat, if you are already committed to stock pistons. Grab that oil cooler that BBM offered, unless you plan to reuse yours, in which case don't, and save yourself another $150 or so (new is about $225 right now). With some discipline and patience, I can imagine trimming $1K off your budget. That could generate some goodwill that would make purchasing a new set of rad hoses and belts after the car is running a little easier to justify. * I have two A-grade pistons I'd be willing to throw into the project for $50, but the shipping from Canada would be another $50 and then one month wait for shipping/customs. Better if you can find a set of what you need stateside.

If you can't reliably hit torque in the 10-20 ft-lbs range you'll be in a pickle for lots of stuff on this engine. Nothing is over 55 lbs that I recall (your aftermarket head studs and the crank pulley being an exception). Lowest I think is 4 ft-lbs for the pan bolts. Who amongst us has actually gone only 4 ft-lbs with those! I would just try to verify accuracy the HF one you already have, especially in that critical range. Next time you visit your local HF, can you ask to check it against a beam wrench that they (hopefully) have on the shelf?