1993 Civic DX 1.5L Auto, 120K miles

Car had been blowing white smoke & overflowing radiator coolant. Could see bubbles in coolant w/ radiator cap off. Oil appears clean, no mixing with water. Assumed a blown head gasket, so I removed the head, had it resurfaced, & reinstalled with new gasket after cleaning all mating surfaces.
Started car & it still blows white smoke & bubbles in coolant. Compression checks OK, 185-195 psi dry. So assuming new head gasket was installed correctly, then I expect it means crack in head or block. Anyone know if this is a common failure for these engines, & maybe could identify likely locations where the crack might occur. I'd like to fix this, even if it means swapping in a rebuilt long or short block, but would like to get a bit more confident as to where the failure is, block or head?
Any common failure locations?
CRuby

 

hate to break it to you but, sounds like a cracked sleeve or head. most likely going to have to pull the head again.

 

Before removing the head again rent a pressure tester and see if any hoses, radiator could be the problem.

When you had the head resurfaced did they notice anything or mention the heads shape?

 

Did you request any other work or tests done on the head when you sent it in to be resurfaced?

Did you check the block with a precision straight edge and feeler gauges to rule out a warped block?

These are just a couple of other possibilities in addition to the cracked sleeve or head. (Pressure testing and dye testing the head would have ruled out the cracked head scenario).

 

Radiator & hoses are ~6 months old, & I don't see how they could cause coolant leak into exhaust (ie. white smoke). I expect the failure is between the combustion cavity & the coolant circuit. I think that means:
1) Bad headgasket seal (I wish)
2) cracked head
3) cracked block.

Offhand I'd think that 1 & 2 are more common than 3, because I think it must be uncommon to crack the steel cylinder sleeve, as well as the Aluminum block.

When I had head resurfaced, I hadn't asked them to do anything else with it. I pointed out some slight pitting on the surface somewhat between coolant cavities & the owner said that it was OK, as long as the pitting didn't get close to oil cavities or mating cylinder surface. It didn't. Surface on head adjacent to mating cylinder surface was machined smooth. Also, I did not have block resurfaced, as I never pulled engine. I just cleaned up the gasket surface carefully using a razor & lacquer thinner. No pitting noted at all on the block surface.

I'm OK with pulling the head again, but trying to do my homework to know what to look for & where to look, if cracked block or head are the culprits.
I'm hoping there are some on this site that been there-done that, & can give me some pointers to likely locations, or other diagnostic checks to do.
Next time around I'll get a straight edge to check flatness of block, which I hadn't done this time.

I do have a leak down tester which I haven't used much in the past, so if there is some knowledge to be learned using that, I'm all ears.

Thanks for the feedback so far.

 

Leakdown testers are easy to use - hook it up, and listen for where the air is escaping. Do that.

 

As the compression numbers look very good, I'm not sure that the leakdown test will yield any answers.

What head gasket was installed? How?

New head bolts? How did you torque them?

Otherwise, I think checking the block mating surface for flatness is something important to do.

 

I'm leaning towards a block or head issue. Nothing else will allow compression to enter the cooling system besides the 3 things you listed.

 

Did you check the blocks surface for warpage before reinstalling the head? Were the mating surface cleaned?

 

I used a graphite gasket from Domestic Gasket kit. Head was resurfaced, but I did not check flatness. I cleaned block of old gasket carefully using razor (no scratches, or obvious cracks) & wiped with lacquer thinner. Chased threads in block using tap, but they were clean & no real hangups. With head off, bolts threaded freely into holes by hand, & I kept bolts matched with their original holes. Boltholes clear of fluid or debris.
I did not check block for flatness. Won't make that mistake again. Just before gasket install I wiped mating block & head surfaces with brake cleaner damp paper towel. Applied clean engine oil to bolt threads & washer face. Installed in proper holes & 3 step torque process following proper bolt torque pattern. 1st) finger tight, 2) 22 ft-lb, 3) 49 ft-lb . FYI, block is D15B7.
I did not use new headbolts, but may reconsider next time around. Didn't think these were torque to yield bolts, so they should be reusable...right?

So, my current plan is to read up on leakdown tester & do it just for information. I've read about running car until engine warm & then pulling plugs & use mirror over plug hole to check for condensation. That might help point to which cylinder has the leak. Next I think the heads coming off again. Much easier the 2nd time around. Then I'll visually inspect head & block surfaces, specifically looking for crack. Maybe disassemble valves again & bring head back to shop for other possible checks/tests they can do. I will definitely check block surface for flattness this time around. Haynes says block deck warpage limit is .004" . I guess I'll need to get a good flat edge to make this check.

 

So you didn't check the blocks surface for warpage, correct?

 

#4 Cylinder leaking coolant based on the mirror test I mentioned above. It actually works. I think I run engine a bit longer until it's really putting out the fog to see if I get any indication on other cylinders. Then it's on to disassembly.

Correct, I did not check block for warpage. Don't have a straight edge necessary for the .004 tolerance. But I won't make that mistake again. Already googling 24" straight edges. Steel ones are a bit pricey. Are Aluminum ones OK?

 

As long as its a straight edge, and you keep it stored properly, than it won't be damaged.

Did you buy this car knowing that it was blowing white smoke and assumed it would be an easy fix or something?

 

No, this is the typical story of car driven by little old lady. My mother-in-law had it since new, but we stopped her driving due to her decreased capability, she's now 94. Gave the car to my son ~ 2 yrs ago & he's been using it. I've done most the work on it myself, even painting it. Couple weeks ago my son noticed it was loosing coolant, but he claims it didn't overheat. So he kept refilling coolant, & driving. By time I checked it, it started blowing white smoke, & I noticed the bubbling of coolant in radiator with cap off. Made me think Blown Head Gasket. So I pulled head, had it resurfaced, & replaced gasket. Didn't fix the problem though. Sons using my '89 Ranger for now.

I just got head off again this evening. Did a visual inspection of head & block surfaces, using magnification, but don't see any cracks in either. Paid particular attention to cylinder/water walls. Checked all cylinder walls by rotating crank to put each @ BDC.
Can a machine shop do other checks on the head to see if it's OK?
Someone had mentioned they can pressure check it?

I need to get or borrow a straight edge to check flatness of block before I even consider putting it back together again. I wonder if this is a tool Autozone would rent/loan out?

I also eyeballed the pitting on the face of the head. It certainly doesn't look pretty, but it does not get close enough to the cylinders to make me think it could be providing the cross talk btwn coolant & combustion. And the machine shop owner checked it too & said it's OK, & it happens due to electrolysis in areas where there isn't a constant flushing of the coolant surface.
Anyway, that's my news for now. I'll be going out-of-town for several days, so the fun will have to wait.

 

How did your son know it was low on coolant? Was it leaking somewhere?

 

You used the original graphite type gasket after you have already had a leak. That is not what Honda recommends on the non vtec motors of the era.

They have a leak gasket which I imagine is a 4 layer multi layer steal (MLS) gasket and they recommend swapping the bolts to handle the more rigid MLS gasket.

They also don't count finger tight as part of the 3 step torque sequence.

Here is the TSB for it:



As for the head tests, it's called magneflux dye testing. It's the most accurate way to find cracks in aluminum. Microscopic cracks might not be visible to the naked eye, but will show up with the dye process.

I think one of the largest contributing factors to the unsuccessful head gasket job is you didn't switch to the leak gasket and stayed with the original graphite gasket after a known leak. There may be other contributing factors too.

I believe autopartsway.com may have the Ishino leak gasket available for a bit less than you'd pay at Honda.

The head bolts on the other hand would have to come from Honda or you can upgrade to ARP head studs.

Here is the "special" head gasket at autopartway.com:

1995 Honda Civic Ishino Stone Engine Cylinder Head Gasket - Buy W0133-1621414 Ishino Stone Engine Cylinder Head Gasket Online

 

Coolant reservior was bone dry & radiator low on coolant. No evidence of coolant drip below car. He claims temp gauge never over heated. FYI, I've got the various replacement temperature sensors, since they've never been changed in life of vehicle. may not have been needed, but cheap investment.

TomCat39, Thanks for the TSB & the info on the leak gasket & bolts. I was unaware of that info. I'll be sure to follow these instructions next time around.

FYI, I brought head back to machine shop for a pressure leak test. So not getting magneflux. But hopefully this will provide me more info that the head is good, before I reassemble...with the TSB gasket & bolts.

 

@tomcat39
so this applies to every honda engine in a civic since 88?

 

Did you order an oem head gasket from Honda or did you go to the parts store? I'm curious to know if the brands that the parts store know about this issue, I'm not questioning the quality of it.

 

No. It only applies to part # 12251-PM5-S02, which is the head gasket for the D15B7, D15B8, D16A6, D15B6, and D15B2.

 

Thanks NotARacist,

As the TSB says, all NON vtec motors from 88-95, which I believe NotARacist just listed.

Also of note, I believe this is a North America TSB only. Probably not the same list for Europe, Japan or Australia etc.

 

My initial replacement gasket was not from Honda, but rather a head gasket kit from an ebay seller. At that time I didn't know about the TSB info. Thought it was a good deal for OEM equivalent gaskets.

 

Problem #1 - trusting important engine internals to an off-brand manufacturer. Don't do that. Was the new head gasket metal, or a weird kind of cardboard plastic paper type ****?

 

New gasket said it's was a graphite material, & it was black & stiff. Not cardboard. Also it had metal seals around each cylinder. I expect it was a good replacement for the original gasket, but not for the MLS called for in the TSB.
Be aware that I don't know for sure it is head gasket leak, but I think it's the most likely.

Hoping to borrow a straight edge from one of my buddies so I check check block flatness. Then, assuming head passes pressure check, I'll reassemble with MLS gasket & bolts. Parts already ordered after seeing TSB.

 

Here is another tip that is not in any of the manuals.....

After you get everything torqued down and reassembled just let it sit for at least one full hour before adding any fluids and running it.

I don't remember where I remember reading it, but that is the best practice for MLS head gaskets. I think it lets the thin special rubber membrane to set up well against the mating surfaces.

Also if you purchased ARP head studs, you will see they specify a higher torque rating for the last torque step. I think it's like 62 ft lbs but not sure. Anyways, I would do the normal process of the TSB to 49 foot pounds and then take it to the final torque rating that ARP recommends with their studs.

I also like to do the last torque rating on all the bolts at least twice, and #1 and #2 three times, following the order sequence every time.