Head gasket sealing issues
 

I already had a thread in media section, but it was kind of off topic and not getting much traffic, so I will make a more on topic thread here.

Short story. I have a major issue with getting the head gasket to seal. Upon first start up I end up with coolant smoke in the exhaust and a car that stumbles on accel. Pulling the spark plugs, you can see pools of coolant and/or oil on the pistons.

I pulled my engine and trans about a month ago with plans to "freshen up" the head and replace clutch, along with other little things. The biggest reason I pulled it was to replace the head gasket, since my compression check showed me that 1 and 4 were high, in the 180-190 range, but 2 and 3 were only 150-160. So I assumed it was most likely the head gasket, since I was slowely losing oil and coolant to somewhere internally.

First inspection with head and block apart did not show any visible signs of a gasket failure. I went ahead with the rebuild as planned. Had the head machined, I asked for them to remove .015 of material from the head. The work looked good but Im not thinking the head machine work has to be at fault for my head gasket sealing issue. I have gone through 2 head gaskets and will be doing this for the 3rd time next weekend. I dont have an accurate enough straight edge to check, but Im confident the block is fine, since I havent heard anyone say they have ever actually seen a warped BP block.

Anyway, my qeustion now, having exhausted every other option that I know, how badly would a crooked head hurt compression. Im guessing having the head bolts not pulling evenly would cause a difference in force from side to side. Im thinking the intake side had more material removed than the exahust side, which would explain the head gasket looking less compressed on the intake side. Here are 2 pictures that kind of tipped me off to this latest idea.

Here is pre machining, you can see the casting marks on the intake stud castings. My focus is mostly on the lower circle.
Attachment 210395

Attachment 210396

Here is post machining, you can see the lower circle hasnt seemed to change, but the intake stud castings have had what looks to be a pretty major amount of material removed. Would you guys agree with this assessment or are my eyes fooling me? The machine shop I took the head too was called "Engine Rebuilders" and it looked to be a small 2 or 3 man operation, and Im not sure how good of work they do. It only cost me $30 and what was pretty much a handshake transaction, no paper work. Maybe that should have been a warning.

Here is one of head gasket #1, failure after only about half an hour and 10 miles of driving. The 2nd gasket failed right away and never held at all. You can see the rings around the cylinders have dead spots on the intake side, which to me means less force on that side/more contact on the exhaut side of the head. Make sense?

Attachment 210397

I can see no other reason this damn thing would be leaking like it does. If anyone else has any ideas, please feel free to chime in. I cant get much help at miata.net, people seen to have gotten tired of my problems :)

(Sorry the pictures are a bit small, but maybe you can see what I am talking about. I also have more build info/progress in my cardomain in my signature link if anyone cares)

Are you re-using the old head bolts, or do you have new bolts or ARP studs? Head bolts are generally TTY (torque to yield) bolts that stretch out when they reach the proper torque, and therefore are supposed to be replaced after each use. Also, if the threads in the block are dirty or have crap in them, it can also make the head bolts bottom out too soon or give inaccurate torque readings. I am assuming that you are using the torque specs and tightening sequence in the FSM, correct?

As far as I know the Miata does not use TTY bolts. TTY have to be tightened using a different method, not just a wrench. They go by degrees and a complicated magic dance to get them right. Yes I am torqueing them to correct factory specs using the recommended pattern. Wether or not my wrench was 100% correct is still a debate. Im going to try and barrow a good dial type wrench this next time to be sure.

you need to get ARP headstuds.

/thread

This is not accurate.

TTY bolts can be torqued as you allude to, or they can be torqued to a specific yield and -then- turned a number of degrees. I would not be surprised if there was another manner for them to be installed, it depends on the material of the bolts, the H/G, and what the manufacturer believes is ideal.

AFAIK, damn near everything has TTY headbolts.

the miata ones are supposed to be reusable. That being said, I went with ARP's, freshly resurfaced block and head, OEM headgaset, and never had an issue.

I just realized that back in 1999 when I put the head back on my vw, I followed the TTY specs in the factory manual for the ARP studs, not the recommended torque. However, everything worked fine.

Like I said before get a straight edge.

Then you can check everything. Then there is no ASSuming anything. ASSumptions will fuck you up every time.

Also, take the head to a shop and make sure it's not cracked. They will chemically check it for cracks.

You don't need studs.

And I don't see why they would cut the bottom of the intake manifold area. That's not level with the deck, though it is on the same plane.

nice headwork :)

Cant tell if thats sarcasm or if you really mean it. I havent done much porting in the past, and it was a pain in the ass using only a dremel and $40 worth of bits. Got the exhaust ports smoothed out, removed a pretty major ridge just under the valve seats and smoothed it out. Opened up the intake side and removed the inside diameter, or the floor of the port where it had a bit of a piece sticking out. Probably did little or nothing for flow, but at least I feel better about it. Smoothed the cc a bit and removed some casting junk. Cleaned valves, new valve seals and other related head junk. If this head does end up being crooked, Im not sure if I can take off enough material to flatten it without having my compression go way up. Its already ~.015 removed, would probably need that much more to level it up. Might not be too much for a low boost application though.

I have a machine shop telling me they have an HLA BP head for sale, already machined, new seats and guides for $200. I might take him up on that. Mine probably wouldnt hurt to have new guides and seats anyway. Trade my core for his. Of course all of the "porting" work would be lost, but I can do it again, but better this time knowing what I need to do.

Straight edge... I can bring myself to find a piece of stock and have it machined just to check my block. Like pretty much everyone I have talked to, there is no way the block could be warped.

Well that's it then the deck can't be warped. Sorry I don't know what what I'm talking about. Just buy another head gasket and that will fix it.

Well, sorry to sound like an ass, but in this case you dont know, nor do I. I will only know once I check the head and have the block checked. I just dont see how an iron block can be warped having never been over heated at only 115k miles.
Im not going to drop the money for a chunk of metal and pay to have it milled to check the deck when I can just pay someone to do it for me. I will rarely need a straight edge, and dont really have a good place to store one where it might not get bent up.

Im still going to check it of course, its common sense that if you have the head fixed to check the block (which I didnt do the first time because of my rush, and expecting it all to go smoothly like it had in the past). Sorry if I sounded unappreciative of your advice, I do appreciate it.

Correct me if I'm wrong:

You bought this car and the head gasket leaked for unknown reasons. So you get the head surfaced and replaced HG. The shop never confirms/denies that the head is even warped. You ASSume that the head was warped and surfacing it will fix the problem.

And it still leaked. But you suspect the staple caused the problem.

So you replaced the head gasket again.

And it still leaks.

Yet you won't accept that the deck could be warped, and won't even buy a straight edge to check it. So now what? Pay to have the head resurfaced, again, third new head gasket, and try again? So if the third head gasket leaks will you consider that a warped deck is a very real possibility? Who said the engine has never been overheated? How do you know it hasn't been? You don't, end of story. Who said mazda decked it properly? Nobody. You don't know. How do you know the deck didn't warp on its own from insufficient artificial aging before being decked originally? You don't. Fact is you have no clue if the deck is right. For that matter, you don't know that the head isn't cracked internally because it has not been checked.

In case you haven't figured it out, you need to stop making ASSumptions and start logically diagnosing the problem. There are many possibilities. You could blindly have the head and deck resurfaced without first verifying they were warped, but then you don't know if your fixing the problem nor would you know if there was even a reason to resurface them in the first place. You could have the head and deck both resurfaced and have the same problem, and come to find out the head is cracked because you ASSumed it wasn't and didn't have it checked. Now you've ported, polished, and decked a head twice that was garbage when you bought the car.

Figure up what your total time porting is worth to you, your total time spent changing the head gasket is worth, twice, and the total cost of all associated machine work and parts used thus far.

Then tell me how much it would have cost to have a machine shop rebuild the head and check the deck of the block with a straight edge. Here's it's 200 to rebuild a 16 valve head and that includes, acid vatting it, cut the seats, grind the valves, chemically check the head for cracks, and surface it if it's not perfect. They'll check the deck of the block for 10 bucks or loan a straight edge with a deposit + 10 dollars.

Sheisse! VR6?


The one time I had experience with this, a customer I had sold a set of ARPs (when I worked at Overland Parts ten years ago) did the same thing, but with catastrophic results. Straight-4, crushed the landings where the bolt heads met the head itself. He was not happy, but luckily had a mirror to find the guilty party. :giggle:

Anyway, thanks again. A shop will really loan out their straight edges? I figured with them being so sensitive they wouldnt want them in the hands of people who wouldnt take care of them.

Also, once a chunk of metal is machined crooked, how do they go about making it true again. Seems like it would be hard to do without a reference to what it was before.

I may just buy the head that the guy offered me though. I would lose my work, which was probably 15-20 hours worth, but it might be less hassle. I may save that route for later, if I find out something is wrong with my head, or it will need too much taken off to make it right.

Thanks again.

Not all shops will. Here there's a shop that does, but you have to put up a 100 dollar deposit +10 bucks. It's not their "shop" straight edge, just a straight edge made of like 3/8" plate 5" tall and 24" long. But yeah if you give it back and it's in any way damaged, you loose the 100 dollar deposit.

It's not hard to make something flat. You could give them a huge steel ball and they can make it perfectly flat anywhere you want. You mount the part in a vise and put a flat on it, put that flat on the magnetic vise of surface grinder and make a pass, flip the part and surface again. Two perfectly parallel surfaces.

If the same shop that surfaced your head is the ones offerinig this other one, I'd be careful. First find out if the one they surfaced for you is perfectly flat. If it ain't spot on I'd walk away.

FWIW I'm not talking out of my ass. I'm not a machinist, but I took 3 years of machine shop in H/S. I've surfaced 5-6 cylinder heads before and rebuilt all of them. And they all work flawlessly to this day. And I've made straight edges, "joe blocks", etc. Very familiar with a surface grinder. Even made a con rod w/ butchress threads for a punch press and it works...

I know how they clamp it in and machine it, but Im wondering how they make it... not crooked. It can be perfectly flat from side to side, but it might not be square. If they machined it crooked the first time, how do you go about finding what is square?

Few way to do this. Could remove the cam caps and put the head upside down on the cam caps on the bed of the mill. Or, use joe blocks to space the head from the lip that the valve cover sits on. Either should be parallel to the deck.

Also they can use a dial indicator in the collet of the mill to check the deck for warpage at any pont.

And if they're a good shop they used a fly cutter large enough to cut the entire deck in one pass. And they used coolant to keep the head from heating and expanding, which would slightly cause it to expand and result in it being more and more crocked as it's machined. And they would have secured it very well, and locked the knee of the mill, and slightly snugged all the skegs such to minimize any play in the machine. All little stuff, but it all adds up.

Yeah I was thinking maybe they could use the machined area that the valve cover seals to. Wasnt sure if that would be exact enough though. This might be all the more reason to just purchase the head this shop has. $200, leak/crack tested, new guides and valve seats. I would have to do my port work again, but at least it should be right. They guarantee their work. Not sure of the details yet, but I will find out more soon. I will ask if they will be able to tell if its crooked or not. If it can be fixed I will just stick with my current head, after they check it over to be sure its not cracked or anything else major is wrong with it. Will see if they will loan me a straight edge while Im at it.

Ok, Im about to jump back on the problem this weekend. I THINK I have found the problem, at least I hope. I knew it might be something very simple, and something I would kick my ass over. Im thinking maybe it is the intake manifold bracket from the block holding the manifold up. Since the head has .015 removed, and the bracket has a bit of movement because of the holes being slightly larger than the bolts, it may be hitting early and holding things up. Ive been installing the head with manifold already bolted up, so this in my mind would for sure be a problem. From what I can recall, even before machining the head the bracket and manifold fit very closely and precisely, and a .015-.020 drop of the head would be that much of a gap holding it off.

I brought it to a different shop the other day and had it checked, head is perfectly flat and doesnt seem to be crooked. I test fitted the head to the block today with no head gasket or manifold and it sat perfectly. My .001 feeler gauge wouldnt go between anywhere, so that seems good.

I had the shop tell me that they thought the problem lied elsewhere, the oil control rings to be exact. I just cant see this being it. No way that all 4 oil rings just collapsed by themselves after sitting for a few weeks (at least Ive never seen or heard of it before). It ran well before with very little smoke on full throttle accel. Even if it were the rings, I dont see that allowing enough oil through to cause the engine to die at anything above idle, nor would that explain bubbling coolant at the radiator cap. So that idea is out the window, though Im still paranoid about it just because of my horrible luck.

But does my idea seem reasonable? Even if it pulled up at torque, it would still be less on the intake side. Thats exactly what I was guessing by looking at the markings and material left behind by the head gasket on the head. You can see much less material left behind on the intake side, mostly on 2 and 3, which would be right about where the bracket is. So this weekend I will again try it, sans manifold. I will also use Fel-Pro instead of OEM, and go with some copper spray. Ive never used the copper spray, but I have seen some people online say good things about it.

My experience coming from the supra world where the 7m's have notorious hg blowing problems, i have to say that it might be the block. every supra that i have done a HG job on with a mhg HAD to get the block machined as well. I am not too sure if you guys ever had composite style gaskets or not, but i guarantee if you used one, it wouldn't leak if the problem was the block itself.
Now coming to your other problematic post with the bracket, that could very well be the cause. The supra timing covers need to be machined with the head or else, just like what you are considering, stay up just ever so much, and back to a BHG again.
Also, if you want to use copper spray, make sure you remove all viton coating on the stock gasket or it will leak as well because the chemicals will react with each other and cause an improper seal.

I forgot all about this thread. I attempted a clean start again last weekend, and got another smoke bomb. This time it was coolant from what I could tell. When it first started it was oil smoke, but after about 2 or 3 minutes of idle, the oil smoke went away (which may have just been some of the copper coating burning off, which I wont try to use again) and the coolant smoke started. It runs and I can sort of drive it, though I havent been, but it bogs if I cruise and then give it a little throttle.

I decided last week that if I can find a 99-00 JDM engine I will go that route, just to get it back on the road, since I wanted the 99-00 head anyway, plus a bit more pep with higher compression ratio. If I dont find one soon, I have a set of Mahle/Clevite bearings and rings on the way, and I will bring the block to be machined, have the head checked over once more, redo it all again and hopefully this time have the problem fixed. I havent yet checked the block with a true flat edge, so it very well may be the block, though I cant see how with it being iron. It may be some of my doing, I cleaned the block off with a wire wheel, but I didnt think a brass wheel would take off enough to matter. I guess maximum warp on most 4 cylinder iron blocks with aluminum heads that use MLS gaskets is only like .002 over the full length, or .001 over any given 1'' (I think thats correct anyway). So maybe my wire wheel did take off enough to matter.

This time I WILL NOT cheap out or rush. It took me 3 times to learn this, but I just assumed too many things. Block will be machined, head checked to be 100% sure its still flat and not crooked/out of square from the first machine job. WHile Im pulling the block again, no point in not doing bearings and rings. I also meant to drill and tap the oil pan for near future turbo install, but couldnt find the tap in time and wanted the car running again, so I left it alone. Alot of little things I can take care of this time. I also had some throw out bearing noise that had me wondering. New clutch, flywheel, pilot bearing and throw out bearing, but for some reason it had that usual rattle with the clutch up, when I would rev and let off it would rattle as the rpm's began to drop. I have the increased gear noise from the light weight flywheel, but thats not the same noise.

Collectively, how many times have you removed the head and changed the head gasket without checking to see if the deck is warped? And please, tell me why "I cant see how with it being iron". :slap:

Ive lost count actually. 3 I think, this next assembly will be the 4th time. I checked the deck this last time with my straight edge, which isnt a machine shop straight edge, and only saw .001-.002 in some spots, which I passed off as being my straight edge. Why I cant see it being the block is because you rarely or never hear people talk about warped iron blocks. Thats not saying it cant happen, but I just figured with so little talk about warped blocks, how could mine be warped. I know thats not the right thing to do and think when rebuilding an engine, but I learned from it. Wont happen this time for sure. Which is why its being machined. I couldnt find a shop to loan me a straight edge, and if Im going to bring the block to the shop, it may as well be machined wether in spec or not.

It just doesnt make sense to me that a substantial chunk of iron could warp, especially with it having an aluminum head bolted to it. If it were going to warp, surely the head would go long before the block.

Yes I know, shame on me, I should have my wrenches taken away for being stupid so many times. Im still not even sure its the block, but it is the only thing left really. I also havent checked the head for square, but Im not sure that a .010 or less drop on one side would make it crooked enough to not seal. If it is crooked though, it will be fixed, either by remachining it, or by purchasing a 99-00 head.

when the aluminum is overheated, the metal properties of aluminum will twist, and therefore lifting up one side of the head. When this happens, it pulls the hot cast iron block with it. Although it is stronger it shall still warp.

This is the third post I have seen on this issue but nobody has found the solution. I have pretty much the same problem and I can't figure it out either. The car is a 94 1.8L with approx 110K on the clock. The car started to smoke this past summer but would quit after a few min running. The oils stayed clean so I did not think much of it till last month when it started to smoke all the time. The oil was still fine and clean, no signs of water mixing in it. I did a compression test and all cylinders were around 185. The cylinder leak down test revieled that cylinder 3 had the coolant leak as I could hear the pressure loss thru the radiator. The rest of the cylinders appeared to be fine. Off came the head. I pulled the head with both manifolds on. I could not tell where the leak was as the head came off a few weeks it ran last. I cleaned up both surfaces and did not see anything wrong. No cracks that were visible. The gasket surface checked out ok with my straight edge. I put the head on and did a cylinder leak test on all the cylinders. Everything appeared to be good as all were around 20-30%. I put the rest of the engine together and the car still smokes. I let it run for over an hour in the driveway hoping that there was still residual coolant in the exhaust and CAT. The smoke did not appear to get better. Back in the garage, I did another cylinder leak down test. When I pulled the plugs, there was steam coming out of cylinders 2 and 3. The valves were open on 2 and 3 when the plugs were pulled. The cylinder leak down test was the same after the hour run. All cylinders were around 20-30% leakage which is supposedly good. I then pressureized the radiator to 15 PSI and found a leak due to a loose hose clamp at the Throttle body. After tightening the clamp some more I pressured the radiatior again and it lost about 2 PSI in about 10 min.

I am at a loss here....:vash: Since the cylinder leak down test was good, I have to believe the headgasket is sealed. There could be a crack but why is the cooling system holding pressure?

Are there any coolant ports in the intake manifold?

There was a thread on this site where someone replaced a throttle body and the engine started to pour out white smoke. I think I may check mine.

Sorry to hear your having the same problems I was having. I still havent found my problem for sure, but Im hoping this rebuild has fixed it. There is a thing on the bottom of the throttle body that has coolant passing through it, with a gasket. It could cause a coolant leak into the intake manifold. Have you run it and shut off the car and looked through the t/b to see if any coolant is left standing in the plenum from the hole in the floor that goes to the thing that for some reason escapes my mind right now.

FYI - It is very common to "deck" a block to ensure trueness. Ask any machine shop.

Iron blocks are under mechanical stress from the moment the casting is poured. Repeated heating and cooling from 100k+ miles of use will sometimes cause the casting to "relax" a bit from when it was first machined as a fresh casting.

Castings have a granular structure created as iron flows into the mold (much lile a piece of lumber has a ring structure) which can add resiliancy in one direction more than another.

Also think about the differences in thickness of the various parts of the top of the block. Different thicknesses being heated the same amount will expand differentialy. This is why we must cut the flanges on the cast iron turbo manifolds (yes, that is a more extreme heating and cooling, but the principle is transferrable). Notice that the little cut in the manifold flange becomes larger or less uniform (twists slightly) after several heat cyclings. This is why a "seasoned" block is often decked or at least checked for true.

All of this being said, your block may be just fine. Did you account for the fact that decking the head means that the head bolts plunge deeper into the block? They might be tightening to where they had stopped for the previous 100k miles and no further. Always chase the threads with a tap and clean shavings, debris, and gasket scrapings out of the bottoms of the holes.

Additionally, if you believe the resurfacing that was done might not be square, have another machine shop measure it. Taking off .015 is a but much to get it true unless it was really far out. And machine shops have trouble finding good help just like everybody else.

Yeah, this was taken care of a while ago. Im starting to think now after seeing how my rings looked that my head gasket may have been fine the last 3 times. The first time it was probably from the staple stuck between the head gasket layers. The last 3 times it was oil smoke, and I found all 4 of my oil control rings stuck in their grooves. Still cant understand how this would cause the car to run rough and try to die under load, but I guess with all 4 rings stuck, thats a pretty nice 1.8L oil pump.

The block deck was machined true, and the head will be going back to the machine shop either friday or this weekend to have one of the dowels pulled, and another few thousandths taken off to remove some minor scratches that are bugging me. Should be no excuse for a gasket leak now. Bolts still have 5 or 6 full turns before they bottom out, so thats not an issue.

I recommend getting some new bolts/studs. I had phantom HG leak problems that took me a year to diagnose. Turned out that the head bolts were stretching. I put in ARP studs and haven't had a problem for 3 years and 50K miles with a cheap ebay HG.

Anybody know the factory bolt length? I dont quite grasp how a stretched bolt would cause a loss of torque though, unless its stretched enough to be weakened and stretches each time its torqued. Mine, as far as I know, are original, and the head was never pulled before I pulled it. Never really heard of anyone having issues with the bolts. Not saying that wasnt your problem, or its not possible. Id hate to pony up for studs this far into the "build", and on a stock bottom end. But at the cost of new bolts, if I did need new ones, studs would probably even be cheaper. Actually, wonder what bolts go for?

it was much cheaper to get studs than bolts last I looked. it's apparently a fairly common problem amongst people who swap heads regularly (spec miata, etc).

Damn you, now you have me worried again. I was so sure, now I doubt. Hmm, I see some ARP's for $125... just have to wait probably a damn week on them.

arp for everything is a standard rule in engine modification.

Son of a bitch. Well, now that you guys have thoroughly scared me, Im going to be ordering some studs. I just know this is going to turn into one of those deals where the studs end up being a special order item and take 3 weeks to get here, when I will be ready to bolt the head back onto the block Tuesday. Anybody recommend any certain place to order from? I was looking at Alamo in east florida, since thats not too far from me, and they have a good price at $120, vs FLyin Miata at $170. Id guess Flyin Miata would be more likely to have some in stock, but I cant pay $50 for that.

Last I checked the factory bolts are about ten dollars each. The head shop told me the Mazda factory bolts were good for 3 R&R. At this point I'd replace them with ARP. And check the head and block, and install a new HG, and do a visual inspection for staples.
No actually I would have put another motor in 2 head removals ago. Thats why I have 1car and 3 motors. Good luck.

just don't order them from Titan racing... those fuckers gave me the run around for 6 weeks before I finally canceled my order and bought from FM.

Shit, Ill just order from FM, though more $, at least maybe they will have some in stock for immediate ship out. Still probably end up having to wait on them since my other stuff will be ready mid next week. Oh well, what can you do.

How "reusable" are the ARP's? Im guessing you can get quite a few uses out of them as far as removing and re-torquing.

EDIT:

$194, that wasnt too bad, and with 2 day shipping. Maybe they will process it first thing Monday and it will get here before friday.