91 civic - tough question about cooling

first i'd like to say thanks to the great people on this forum for all
the help they provide.

i've done a search for my problem and i've found some things but
nothing exactly fits.

i recently performed a head gasket job on my civic. i began this job
because i had bubbling in my coolant resevoir after driving for a
while. i also had some electrical issues, but i don't see those as
related to the bubbles. the bubbling was not boiling. some kind of
gas was pushing through to the resevoir, and i assumed it was
compressed gases of some kind coming from the cyclinders through the
head gasket.

after removing the head, i took it to the shop. they told me the head
had to be machined by 6 thousands to remove the warp, which seems like
quite a bit. he also pointed out where the likely breaches of the old
gasket were.

i cleaned and lapped the surface of the block. there was no apparent
cracking or warping (i checked for warp with a quality straight edge in
multiple directions).

put everything back together to spec.

i filled up the cooling system, on an incline with the radiator high.
bled the system according to haynes.

i've been test driving it for several days, and what happens is a
little weird. after a light drive, i get a few minutes of bubbles
(again, not boiling - bubbles) and i can hear coolant sucking past the
rad cap in pulses. after a heavy drive i get the coolant pulsing, but
much more bubbling - like the bubbling i used to get.

the only thing i have found that i didn't do was crank on the heater
when i was filling the cooling system. is it possible that since i did
not do this, there was a chunk of air in the heater core that is now
escaping when the car heats up? i ran the car tonight at idle with the
heater on full blast after discovering this possibility, but the rad
stayed full (again on an incline) even with the bleeder open.

it seems there are three or four possibilities:
1 - block is cracked
2 - head is cracked and crack missed by shop
3 - heater core had air bubble and now air is working out
4 - head gasket was installed improperly

any thoughts? thanks.

 

how exactly did you do this?

the head could be cracked, but honda aren't renowned for this kind of
problem. much more likely imo, is that the cleaning/machining/lapping
process has left marks in the head surface that make it impossible for a
gas seal to be made.

either way, i think chasing down the problem is a bit academic at this
point. the cost of diagnostics on the head to see if it really is
cracked, labor, etc. just doesn't compare favorably with getting a "new"
jdm engine from japan for $290 plus shipping. if the motor was rare and
much more expensive, sure, chase the problem down, but it's not, so at
this point, i'd just swap it out and be done with it. [replacement also
eliminates the longevity issues associated with all the crud that
inevitably gets into the engine when the head comes off.]

 

i scraped the block surface with a plastic scraper at first. then i
used carb cleaner and a toothbrush to scrub off the remaining gasket
crud. after it was fairly clean i used 600 grit sandpaper tightly
stapled to an index card sized piece of flat particle board. i lightly
sanded the surface.

the head surface seemed near perfect after being machined. there were
no scrapes/scratches of much significance on the block that i could
tell, but who knows. what do you mean by "impossible for a gas seal to
be made?"

if this becomes the route i decide to take, where do you recommend
purchasing a "new jdm engine" (what does jdm mean - i thought the
engine code was DB-13 or something).

thanks!

 

ok. unfortunately, the backing block you need for this kind of
operation needs to relatively large, very flat, and as rigid as
possible. last time i did this, i used a 6" x 8" chunk of 20mm glass i
got for free from a glass shop - it was in their scrap bin. much less
and small local distortion of the backing block means you can get excess
material removed from the "thin" sections, i.e. between pistons, and
almost nothing from the thick sections. unfortunately, the thin bits
are the worst possible place to remove excess as they're needed for sealing.

original surface on these things is near mirror. that way, there's no
small grooves for gas to leak along. if the head was machined on a
traditional milling machine, particularly this alloy head, it's /very/
hard to do without leaving scoring marks in it. last time i had to lap
a head was to cut out the grooves from a bad machining job where the
cutting die was picking up excess material on each pass of an exterior
edge, then dragging that chunk across the rest of the head leaving gouge
marks.

"jdm" is "japanese domestic market". due to tax and emissions laws in
japan, cars usually get junked with only 60k miles on them. that makes
them great for import to the u.s. where these engines still have loads
of useful life.

there's loads of places on the net advertising these kinds of motors,
and i don't endorse any particular one, but something like the site
below is worth investigation:

http://home.pacbell.net/timwang1/38021.html

 

-------------------------------------

Keep this in perspective. When the engine gets hot it will force out
some air (if there is air in the system). If there's enough coolant in
the reservoir, (and the rad cap is working correctly), it will suck that
coolant into the engine when it all cools down. The owner's manual is
very clear: Whenever you have work done on the cooling system, you have
to be sure to fill the reservoir to the MAX mark.

Are you keeping the reservoir full enough so the expelled air can get
replaced with coolant? Once the air is all out, the system works
'transparently', but you still have to check it regularly, since some
coolant can escape thru the water pump weep hole. Service manual says
it's acceptable.

'Curly'

 

i can't imagine that, despite my crude backer, my 600 grit sanding
removed enough material to make a significant difference in the mating
of the surfaces. i am a metal fabricator and i am sure i removed less
than .0005 or less total. if a gasket can't make that up... the
gasket itself was, say, about 1/16" thick or greater. are they not
designed to take up some slack in the mating surfaces?

the head was done at a shop that pretty much only does that,
particularly for dealerships. it was perfect.

the heater core hasn't been addressed. could my mistake of not turning
up the heat while filling the system leave a large air pocket in the
system?

thanks. this forum is easily the best!

 

The symptoms don't sound like a classic head leak any more, but it would be
good to do at least the confidence check. With the engine cold, remove the
radiator cap, start the engine and pinch off the hose to the reservoir.
Place the palm of your hand over the radiator cap opening. If you feel
steadily rising pressure or pressure pulses, the head probably needs to come
back off :-( If not, I'd rule it out. I've never experienced a false
result with that test.

My thoughts are turning toward the radiator cap, a radiator hose collapsing,
or even a leaky heater core... but you don't mention the coolant loss that
almost always accompanies a heater core leak. I'd certainly replace the
radiator cap with OEM on general principles unless you did that recently.
And some peeks at the lower radiator hose to make sure it's still plump are
in order.

Good luck!

Mike

 

not really. original honda head/block surfaces are mirror smooth and
dead flat - no "slack" to take up.

ok. that makes a crack more likely.

that'll make it burp, not bubble. bubbling is a dead giveaway,
especially in a honda where the leakage channel is straight into the
water jacket.

 

wrote in

You did this before or AFTER the head was machined to flatten it?
Why would old gasket material need to be removed after machining to flatten
it? It should not BE there after machining. Besides,I believe you do NOT
want the mating surfaces to be polished(there's no need for it) because it
would reduce sealing to the gasket.

The head would not clamp down on the GASKET sufficiently if polished or
polishing affected the levelness recreated by the machining.

 

sealing is much more effective with a /smooth/ surface. look at the
mating surfaces on high pressure valves and artillery. you're right if
"polishing" makes the surface deviate from completely flat, but the
flatter & smoother the better.

 

just to clarify, i didn't polish the head, i polished the block. i
never touched the head after it was machined.

small update:
i drove the car to work today which ends up being about 70 miles round
trip. at the end of my trip home (which is about a 2500 foot climb
into some mountains), i park the car, turn off the motor and i hear
silence for a time. which if i haven't been clear, is sort of normal
but sometimes i would have a little bubbling after parking and turning
off the motor.

it was clear though that this time there was no bubbling, even after a
hard drive. however, after about 10 or 15 seconds i start hearing the
slurping of the coolant past the rad cap. then the bubbling started.
so, *no* bubbling with engine running, and *bubbling* shortly after the
enginie is shut off.

i pulled the overflow tube out of the resevoir and noticed coolant
coming out in small quantities in time with the slurping sound. is it
possible that the recently shut down engine heats up slightly (i think
they do after shut down), thereby heating up coolant in the engine,
pushing coolant past the rad cap, but the flow is momentary (the
slurping) creating a small steam pocket in the overflow tube, which
then is manifested as bubbles in the overflow resevoir?

the rad cap is newish (~6 months old), but i have at least two other
ones that i may test to see if they perform differently.

i am now hopeful that the engine is okay since bubbling isn't ocurring
during engine operation.

thanks so much for all your input.

josh

 

Not when you're compressing a head gasket,not all the time.
With a metallic gasket,you WANT a texture to grip the gasket.
*The gasket is what does the sealing.*

Which do not use gaskets. They are -mating- surfaces.

 

there's only one situation where bubbling in the expansion reservoir is
"ok", and that's the first time after the coolant's been refilled.
that's the "burp". after that, all the air in the system should be
expelled as it rises to the top of the rad and expands out as the system
warms up. [it won't do it if there's too much air in the system, but
i'm assuming everything was refilled ok.]

if the gasket/head/motor is still leaking, what you'll then find is that
excess fluid gets pushed out to the expansion reservoir, and it stays
there as the motor cools overnight. if there's an air gap under the
radiator cap in the morning, it's either a radiator/hose leak, /or/ its
gas from an engine leak having over-filled the system.

at this stage, you have to leave it to do its thing undisturbed
overnight after it's been running to full temp. if you want to spend a
little more money, there's a hydrocarbon test you can do on the coolant
to see if exhaust is leaking into it, but i'm not sure how much that is
offhand.

i have my fingers crossed for you. i've just been through all this
myself, and mine wouldn't bubble while i watched it. gunning it up a
big multi-mile hill at 105 degrees sure did the job though. it would
blow coolant, then not suck back to refill completely. having changed
/all/ the hoses and the rad and the water pump, there was only one other
thing, the gasket. and as if to confirm, the last time i looked under
the hood just before performing surgery, sure enough, there was a steady
stream on bubbles blowing into the expansion reservoir.

good luck.

 

with respect, you're stuck in detroit there jim. "grip" on a gasket is
simply grooving along which gas can escape - you really don't want it.
that's why your old ford blew gaskets every 60k and a honda [typically]
will go hundreds without being touched. i promise you, oem spec honda
heads are near mirror finish, and that's for good reason.

they'd use them if they could! gaskets are single/low-cycle use.

there's /two/ mating surfaces when a gasket is used, e.g., block to
gasket, gasket to head. /both/ need to seal, not just one.

 

--------------------------------

Here's another tip: You can't really overfill the reservoir with
coolant. Any air the system displaces will just bubble out since it's
not a sealed (pressurized) reservoir. Check it every morning when the
system is stone cold. Make sure it's at MAX. Check it next morning,
refill to MAX. Once ALL the air is out it will stop 'sipping' from the
reservoir. That will tell you how 'healthy' the cooling system is.

'Curly'

 

I'm solidly with the "smooth surface" view on this, but the question as it
relates to this thread is whether the head gasket is leaking or not. I
suspect not, but the OP really needs to check it out.

Mike

 

yeah, if the head was done right, it's not the gasket, or at least, not
this quickly. more likely it's a crack. that's pretty unusual for a
honda, especially this vintage when their q.c. was really at its peak.

 

Try using a single-blade cutter, instead of the two-bladed cutter. This
is slower but yields a cleaner cut.

If they're caused by deposits, try removing the hard calcium deposits around
the water jacket openings. The deposits can be picked up by the tooling
and drug across the surface leaving a groove.

 

The extreme smooth surface may not provide enough bite to hold the
gasket securely. Too rough can also be hard on the gasket itself or even
won't seal properly and the head gasket will leak. One way to judge
surface finishes is with a comparator gauge and follow the manufacturer's
surface requirements.

Laminated steel or rubber coated type of head gasket would require
an almost polished surface finish. But most older milling and grinding
equipment have spindle and bearings in weren't designed to meet these
requirements.

 

ok, question: where do you get this information? both you and jim
clearly heard it from somewhere. what's the source?

there is no lateral load. no "bite" is required.

precisely. think about it logically: big grooves leak a lot. small
ones leak less. a smooth surface leaks not at all. make sense???

and with a honda, that's a near-mirror finish!

no joking. that's one of the reasons why machining honda heads is such
a hit and miss operation.