vagalati

I have a cyl 1 misfire CEL. Sometimes when the engine misfires it smells pretty bad (like sth is burnt) and the exhaust has white fumes. I'm not sure if the smell is like "burnt" coolant because the honda coolant smells funky either way (not the very sweet smell i am familiar with).
Furthermore, seems like a bit of coolant has escaped from multiple points (where hoses connect) possibly indicating too much compression inside the cooling system...

I checked the engine compression with a gauge and seems to be pretty good:
Cyl 1: 190
Cyl 2: 190
Cyl 3: 190
Cyl 4: 185

Then I used a combustion leak tester:

The blue liquid turned into yellow after few revs. Bummer... so I expect to have a blown head gasket. Fortunately the engine never overheat (it has always been operating under normal temp) so the chances of head warping are very small.

How come compression is so good if there are leaks??

revr63ed7

Most honda head gaskets develop a tiny breach from the combustion chamber into the cooling system. This is enough to slowly put compression into the cooling system but rarely lets coolant into the cylinder. It's also not enough to show on a compression check, use a leakdown test and whatch the coolant in the radiator. It may take a few minutes but when the cylinder in question is pressurized the coolant will rise or bubble.

vagalati

Interesting... This is a felpro gasket though; I had changed a gasket in June. Back then, the engine overheated badly so I had the head machined (+valve job) too.

revr63ed7

what model and engine are you working on? Als, felpro doesnt impress me as im in the factory only opinion club. Like ezone said in another post, factory parts last 100+k and aftermarket last much less.

vagalati

It's a d16y7 without vtec.
I stressed the engine at the very first day after i changed the gasket in June. I went on a road trip which included going up on a very steep hill with the car fully loaded. I was told it was a bad idea... and well... It was running fine until last week though, when I had the repeated misfires I mentioned before...

revr63ed7

i dont think you hurt it on that trip. if its fixed right its fixd. Some things need to break in but head gaskets arent one of them. do you have coolant being pushed into your overflow bottle? Do a leakdown test to check it, the chemical test may be reading carbon traces from last blown gasket. Its nearly impossible to get it all out.A leakdown will tell you which cylinder is breached too.

vagalati

Yes, when I checked last week the coolant in the bottle was higher than MAX and I needed to add some coolant. A good indication that the pressure from the combustion gases had obviously pushed coolant out.

When I changed the gasket back then I also changed the radiator (was cracked), hoses, water pump and a bunch of other stuff and I bled the cooling system (new coolant). I don't think it's gases from last time. I'll do a leak down test once I have the tester in my hand, but I'm 99.99% sure it's a head gasket.

I'm just wondering if I did sth wrong, and if I did, what was it...

ezone

The good ones will take it, the bad ones won't.
You didn't force it to do anything it wasn't designed to be capable of.
If driving a loaded car up a hill was off limits, there would be an awful lot of people stuck at the bottom.

The only way YOU would hurt it driving would have been to overheat it again.
Or something wasn't perfect on the previous work, there's no way of knowing from here. Or there is an underlying cause that wasn't found, like having 150,000 miles worth of bugs and dirt stuck in the radiator and it can't cool efficiently, something like that.

I know I usually trust Felpro for most other applications so I probably wouldn't hesitate to use theirs in this application either IF I was in an indy shop. I'm at a dealer, so it is OE parts only.

gearbox

iTrader: (95)

i had the same exact thing happen. perfect compression but blown gasket and burning coolant with misfire.

vagalati

I meant to say if I did sth wrong while assembling it.

Since when I replaced the head gasket I've been watching the engine temperature and I've never seen the gauge going more than the optimal temperature (about 1/3rd to the top), so pretty sure no overheating happened this time.

gearbox

iTrader: (95)

mine never overheated either. sometimes they just go for no real reason and can be small enough that nothing terribly bad happens.

vagalati

So, I bought a leak down tester (Otc 5609) and I did a leak down test:

cyl1: 90 PSI regulated, 84 PSI cylinder, 7% leakage
cyl2: 90 PSI regulated, 84 PSI cylinder, 7% leakage
cyl3: 90 PSI regulated, 90 PSI cylinder, 0% leakage
cyl4: 90 PSI regulated, 86 PSI cylinder, 5% leakage

Aren't these leaks very much acceptable? I read on different sites that up to 20% is ok.. however:

Now, cyl1 & cyl2 were bubbling the coolant from the radiator + I think I could hear a slight hissing noise from the spark plug of each other when I was testing them, clearly there is a leak

cyl3 & cyl4: No visible pressure increase in the radiator

ALL CYLINDERS: I could hear some air flow from the oil cap, when open (and maybe from the dipstick tube). Is that normal? I know that no engine has perfect sealing and air will leak into the crankshaft through the cylinder walls but is it because of that or is this a sign that rebuilt is needed?

Just to remind that cylinder pressures with the pressure gauge where:
Cyl 1: 190
Cyl 2: 190
Cyl 3: 190
Cyl 4: 185

ezone

That sure seems like proof enough to pull the head off.

(Someone else had posted about an older engine that had serious erosion of the aluminum casting allowing coolant leakage.)

Make sure to have the head checked for flatness, and if the rest of your post is accurate you may need some valve work too.

You have to make sure the valves are all closed when you do this test, that usually means pulling the valve cover so you can feel that the rocker arms all have clearance (play) on the cylinder being tested.
Any leaks through valves are usually heard and felt as air coming out of the other spark plug holes, because those other cylinders have their valves in various stages of the cycle (valves open).


If the valves are verified to be closed, then you have yet another problem...Valve leakage. Figure out which valves: Screw the plugs back in to cork those other 3 holes, and see if the air is now going out of the exhaust pipe or the intake.

A HG breach between adjacent cylinders will allow the air to come out of the plug hole next door too, but it won't be leaking through the valves of the cylinder you have the pressure on. At this point if I thought this would be the issue, I would be lifting the entire set of rockers so that ALL the valves in the head can be closed at the same time.

Another issue: The leakdown test readings don't agree with your compression test numbers. Why not?



Found this in the first post....The radiator cap regulates the max cooling system pressure to 13-16 PSI. If the rest of the system allows leaks before that pressure is reached, that is yet another problem. They need to be able to withstand well over the rated cap pressure without leaking.

vagalati

Alright, I repeated the measurements after heating up the engine to operating temperature.

1) I measured the cylinder compression 3 times (rotated among the cylinders):
cyl1: 230, 240, 250 PSI
cyl2: 215, 230, 240 PSI
cyl3: 210, 210, 210 PSI
cyl4: 210, 215, 210 PSI

while cranking up the engine to do the above measurements cyl1 and cyl2 were "spitting" coolant and steam from the spark plug hole. I was surprised by their compression performance, (especially since it increases with more tries!), even though there is at least a gasket leak between those two. Is it possible that the leaking coolant acts as sealant in the cylinder walls and/or valves??

Also, these compressions differ from what I had reported in my first post because then I was using an adaptor/extension with the gauge, which this time I removed. I repeated the measurements with the adaptor and I got the same results as in the first post for cyl3 and cyl4.

2) I popped up the valve cover and checked the valve clearances when the piston was at TDC before redoing the leak down test. Here are the results, when I applied 90PSI compressed air:

cyl1: 82, 9% leakage
cyl2: 86, 4% leakage
cyl3: 89, 1% leakage
cyl4: 84, 7% leakage

Again, I was expecting big leakage numbers for cyl1 and cyl2 since whenever I applied compressed air the coolant would bubble in the plastic container. Any thoughts??

ezone

That was enough to pull the head.

If there was any question about valves, send the head to the machine shop and let them check it.

Rings questionable? Now is the time, while the head is off.



I'm still wondering how you got 0 and 1% leakage on #3 yet compression doesn't agree.

I wouldn't expect to see 250PSI unless the cylinder was oiled (or some other liquid?) or maybe full of carbon. Maybe your gauge wasn't 100% accurate/consistent? IDK. 210-230 is my average IIRC.

vagalati

How can I know if the rings are questionable?

The compression for cyl3 and cyl4 seems to be reasonable and consistent among the tests (~210 PSI). I think the measured compression for cyl1&2 is artificial, especially since the more coolant was pouring into the cylinder the more the numbers increased (e.g. from 215PSI to 240PSI by the third measurement).


Good question, however I double checked the leak down tester gauge: I sealed the output and fed it different pressures: it was consistently showing 3% more than the input. I accounted for this small error in the numbers in the second leak down test: if you see they are almost the same as before but decreased by 3%).

ezone

If they were ever questionable, now is the time to fix it.

High mileage, oil consumption, blue smoke, low compression, overheated, stuck, drinking coolant......

It's a guessing game. And a risk.


Your compression looks fine, but the rest of the questions still need to be asked.
A compression test cannot check the oil control rings.


And how much money do you want to spend, how long do you plan to keep the car?

Customers almost NEVER do rings, but will happily bitch later on about the problems associated with them.



Just how much coolant was getting in there? Must be a lot.

I know an ordinary piston engine never has 0% leakage without artificial help, that's why I questioned it.
My leakdown tester is regulated to about 30PSI.


If the 2 high compression cylinders can be explained by liquid entry, then I'd be ok with that.

sdaidoji

iTrader: (1)

clue 1

clue 2
clue 3 with the coolant evaporating inside cylinders

sorry to hear your almost new HG did fail for some unknown reason.
compression test just confirms, and you could pinpoint which cyclinders, which is more than i knew when i did mine (did not matter to me, though, but yours is a early fail so this info might help you if you ever plan on going nasty to either maker or installer. don't see if you would get any wins tho...)
i back this comment below
^LOL!

ezone

Strongly urge serious inspecting to figure out WHY a new gasket isn't sealing, and avoid yet another repeat.

Maybe use factory parts this time?

vagalati

I pulled out the head. Here follow some pics from the crime scene:

1) The cylinder head, cylinders 1&2 (the bad ones):



2) The failed head gasket, cylinder 1&2. As you can see the sealing coating has been removed between the two cylinders:

And the other side:



3) The cylinder head, cylinders 3&4 (the good ones):



4) The head gasket on cylinders 3&4. The head gasket seems to be almost intact here:

and the other side:



5) The failed fel-pro gasket and a new genuine honda next to each other:




6) Measuring the cylinder block for warpage with a precision edge and a 0.038mm feeler gauge.. I can't fit the feeler gauge between the two cylinders. Same goes for the head. Honda recommends 0.05mm minimum warpage before resurfacing.



7) I saw this gunk on the intake manifold (it's slightly more on cyl1&2 than on cyl3&4). Is this expected?? Can't think of any reason why gunk would accumulate there so thought it might be worth posting these:
cyl1&2

cyl3&4:



I am the installer and I thought I did a good job. I spent hours cleaning the surface carefully with a razor blade, cleaned up the gunk and coolant from every part and especially the bolt holes. I also used the Honda-recommended torque and sequence on the head bolts. However, I am not a pro mechanic and I was just following instructions and recommendations from the forum here and the service manual... so I could have not seen an existing problem or unintentionally created one.

Yes, I would like to know what went wrong. I measured the surface of the block and the head with a precision edge and a 0.038mm precision gauge and both seem to be straight as I can't enter the gauge under the edge at any location.

I really have no clue what caused it!

sdaidoji

iTrader: (1)

wo-ow... the sealant material is completely gone...( me believe the blue material)
the black part on the OEM is the one supposed to seal, but the fel-pro *design* looks... not so good...
at first does not seem install issue...

ezone

IDK why it failed, but I guess this might cross FelPro off my list on these cars.....unless you find the true cause of the failure.


Did you check for flatness at all diagonals on the head and block? Reason I ask is you only pictured one with the straightedge down the center of the block.

Buildup inside the intake is absolutely normal. Clean it out if it bothers you, but it is normal.
'Nuther question, (because I'm pretty sure I can see the machining marks in that head): What is the RA of the machined finish on that head? That sucker should not have a belt-sander finish, it should be nearly mirror smooth when using a MLS gasket....much like the smooooth machined finish of that intake manifold surface.
I like to see well under 30Ra for a MLS gasket.



I actually just found this site that agrees with my theory that these engines using MLS gaskets NEED to be using TTY head bolts:


MLS gaskets are extremely durable compared to other types of head gaskets because of their all-steel construction. They won't burn through, and they won't relax and take a compression set that can lead to leaks and sealing problems. MLS gaskets also reduce blow-by and improve compression by reducing bore distortion in the cylinders (less load is needed to seal this type of gasket.) But MLS gaskets require a special type of head bolt that stretches when it is tightened ("torque-to-yield" or TTY head bolts, which should not be reused). They also require extremely smooth surface finishes on both the cylinder head and engine block to seal properly.


Same site also agrees with my 30Ra thoughts:

Surface finish is also very important too. For many years, most aftermarket gasket manufacturers said a surface finish of 55 to 110 micro-inches RA (roughness average), or 60 to 125 RMS (root mean square) is acceptable for conventional gaskets. The preferred range has traditionally been 80 to 100 RA. More recently, though, some gasket manufacturers have changed their recommendations because today's engines are lighter, and castings are thinner and less rigid. The latest recommendations for non-asbestos and graphite gaskets is a surface finish of 30 to 110 RA for cast iron head and block combinations, with a preferred range of 60 to 100 RA and 30 to 60 RA for aluminum heads on cast iron blocks with a preferred range of 50 to 60 RA. For MLS gaskets, a surface finish of 30 RA or smoother is usually required.




http://www.rlengines.com/Web_Pages/C...L_Engines.html

vagalati

For the head, yes.
For the block I did the left-right and the diagonals but didn't have enough space (the edge is too long) for front-back like the lines C, D, E in the figure:


I don't know. Need to ask him for that. The head is definitely not mirror-like but so is not the block. The block was not machined and it's pretty much as it was made from factory.

How can I know if the bolts are TTY? I had new ones from the local dealer.

ezone

Just checking, that's all. I know you can't fit the straightedge in there like that.....and the block is almost always ok (I've never seen one ruined anyway).
You can almost always see the machining lines in the block, but you can't hang a fingernail in them. They are smoooooth.

OTOH, the head looks like you could play those grooves like a record. (But I am only seeing a picture, I can't tell that for sure.)

Steel gaskets don't seal well against a bunch of grooves.

The head looks like the machining lines are all completely straight from one end to the other. Like it was put on a belt sander......not a rotary mill.

I would expect a good machinist to have a Roughness Gauge of some sort for this work, and be able to figure out what engines need what sort of finish.

I also know the sort of work that comes from a couple of the machine shops here...that.......don't get our business.
When MLS became mainstream and popular on Asian engines over 2 decades ago, MANY machine shops here were slow to update their equipment. Old school just doesn't work with new tech. Some still seem unwilling to update equipment.....and their people.





Hey, I'm seriously just looking and thinking outside the box here....and I have seen this problem first hand.


The torque specs tell if they are TTY or not. Honda didn't use TTY on Civics until 8th gens. Yours are not TTY type....and yours didn't need replaced IMO. I never replace those, and I rarely find the need to replace Hondas TTY type either. There are specs and measurements listed in the manuals for the engines that use the TTY head bolts.

vagalati

I went back to the car today in broad daylight to have a feeling of the finish of both the block and the head. The head feels and look very smooth to me, as smooth as the block, or maybe even smoother. I think I'll take the head to a dealer and/or other mechanics to hear their opinion.

I also took more pictures.

The finish on the cylinder head (depending on the viewing angle you can see different machining groves, not in parallel to the head):







And this is the finish on the engine block:







In general I would say the finish between the cylinders and the coolant passages is even rougher.

The finish between cylinders 1&2 (the failure point):



The finish between cylinders 3&4 (head gasket intact):

ezone

Ok good. Like I said, I was just seeing a picture.If you go to that trouble, I'd rather a reputable machinist look. Not the one that did the original work. You know, an impartial opinion.


Average mechanics don't know squat about what goes on in a machine shop (other than we send it there for something, then it comes back all fixed) and aren't qualified to evaluate something like Ra. (I probably wouldn't even know squat if I hadn't had to figure out a problem like this once....It gave me a crash course in a few of the finer points of things they do.)
It might seem closely related, but the reality is it's a vastly different field from what we do in a mechanical shop.
Ok, the eroded area looks iffy. Is it in an area that must seal, or is that in an open area? Lining up the head gasket should tell you if it is or not.
Could the erosion be the cause of the leak?


Both of these pics looks like it has a dual finish. One grain running lengthwise, another looks like it's not quite perpendicular to the first. Did you sand it by hand or something?





Another thought: Did you measure for any localized low spots specifically in the areas where the head gasket failed?


These are all just suggestions and guesses.
I'm still kinda grasping at straws here. Looking for ANY reason for the failure before it goes back together.

vagalati

It is the area between the coolant passages. Look at the first picture of the head that I posted, upper left corner. The holes and the area between them correspond to the coolant area on the block. I don't know why Honda has the whole coolant area around the cylinders uncovered instead of just holes-passages (as in the head) but I don't think the sealing here is an issue.

I didn't sand it or anything. I only used paper towel or cloth towel to clean it up. The job was done from a very reputable machine shop in the area. They are the preferred machine shop by racers, so I decided to have it done by them.

So I took both the head and the gasket to three places: The honda dealer, a good-score-on-yelp repair shop and my machinist. All three asked me about the bolts, threads, torquing, clean surfaces, overheating, and generally things we've already discussed here. All said that the head surface is good enough for the gasket.

The honda dealer mechanic said that I should have the head pressure tested. Since the head has been recently machined I doubt this will help.

The repair shop mechanic said that I should give a try to the Honda OEM gasket which is MLS with graphite. He said that the material on the Felpro (he knew it was felpro by the first look) feels kinda like plastic. He told me to remove the hard-to-remove gasket parts with a sand-block (a straight metal piece with some sand paper or sth). I thought that was a no-no!!

My machinist said that the gasket didn't seal properly. He said it should have "melted" on the metal surfaces and clearly that has happened only to a small degree. He didn't want to blame the gasket though. He said it should be the block and that I should check the block surface. He said if the temperature was too high and I kept driving for long (which I did) I could have "damaged" the areas where the gasket seats between the cylinders and the main block.

I am not next to the engine block right now but I'm showing you a much higher resolution picture between the cylinders 1&2 where the gasket failed:

Can you see the lines around the cylinders? They feel like different metals put together. I can feel the lines with my fingernails. The machinist said I shouldn't feel anything. This is my main concern now. However, the area between cylinders 3&4 looks and feels exactly the same, and the gasket sealed well there:



In any case I'll start cleaning up the remnants of the old gasket. I will try to be careful and spend plenty of time with razor blades. Any recommendations for other gasket chemical or mechanical removals?

ezone

I use a razor scraper, maybe chemical gasket remover. I don't like using power tools or anything that could possibly remove metal.

This.



That's obvious.

"melted" is a new term to me when it comes to head gaskets.....unless that blue coating is a layer of plastic.....sounds damn flimsy if that is the case.
Man, I have seen some pretty bad overheats that lived just fine with a milled head and factory head gasket......but if it overheated bad anough to warp the head so far it lifted and lost compression, they needed an engine (used is cheaper than real overhaul).

Yes, there could be flame etching (cutting) under the fire rings if it was leaking bad enough. Same cutting action that burns valves.

The machinist is right about this.....and I think you are on to something now.


Second pic shows (what looks like) an impression left by the fire rings (I can't tell which side is higher or lower though). Any idea how deep that impression is, if it really is an impression?

The fire ring must be compressed in order to seal properly....if it sits in a low spot, then it can't be compressed as much as it needs to be.

Might be worth offering the machinist a few beers to drop by and look at it if you can't scrape it smooth.

Hard scraper honed to razor sharp on a flat file or sandpaper on glass....to take off the big chunks from composite gaskets...
Then single edge razor blades.

Check parts stores for aerosol can gasket remover.

Factory MLS usually only leaves black "paint" that comes off pretty easily and isn't a big deal if some is still stuck.

vagalati

I guess the proper word would be "stuck". The outer material of the gasket should stick on the metallic surfaces of the head and the block. Clearly not here: see how little blue "paint" is left on the head and the cylinder walls

I'm not sure I understand what a fire ring is.
For the impression, I feel that inner part of the cylinder walls is slightly higher than outer part, but I may be wrong. it's very hard to tell.

ezone

Ok, bonding. Makes sense. I could NOT think of what would be expected to melt in there.

AFAIK it takes time and heating cycles for the gasket materials to bond, maybe 6 months wasn't long enough. Most that get taken apart have been together for several years before they fail.
Regardless, the bonding process should have no bearing on its ability to seal immediately upon installation."Fire ring": The section of the head gasket that seals the combustion chamber; the thicker steel ring that immediately surrounds the combustion area.
It is usually much more obvious on a composite gasket so here's a pic of that


A MLS gasket may just have a creased area in the metal as opposed to a thicker ring embedded. Designs vary.

HTH