Rep Power: 191

So I've had on and off problems with overheating since the winter, but there were extended periods of no problems, especially after I did something each time, but this past week it's happening regularly, so there is definitely something up.

Small background: The water pump was done by the dealer when they did the timing belt (OEM parts), which was over the winter. I can't recall if the problems started happening before or after the T-belt/WP change... I think it might have been before, though.


First I changed the tstat and rad cap with aftermarket (Stant) parts and flushed/filled with Type 2 (OEM). That worked for a long while, many months in fact.

The problem came back a couple of months ago, out of nowhere. By this point, the signs started pointing to a Bad Head Gasket. Just in case, I replaced the radiator cap again (this time with a Honda OEM part), I also checked the fuses, and replaced the fan relay (another Honda OEM part) and went ahead and did the "block test" with a loan-a-tool and fluid from AutoZone. Came up negative after a bunch of idling, revving, and more idling... so I assumed HG had to be good. I bled the system and topped with Type 2, and went on my way.

2-3 weeks ago, same crap is happening. I decide, okay, I'm done with aftermarket parts, I'm going to replace everything again and see what happens. I went to Honda and bought, then changed, the following parts:
- Thermostat & gasket
- Radiator cap
- Radiator fan switch
- Coolant temperature sensor
- Upper and lower radiator hoses
- Drained and filled the radiator with Type 2 coolant (which I've been using since day 1, so there's nothing in there besides honda coolant).

Worked fine for like a week or so.

Somehow, air or exhaust gases (or some type of gases) are getting into the system.

A few days ago I decide maybe I didn't bleed it properly, who knows, so I have the car on a slight incline, and I bleed it for 30-45 minutes to be damn sure it's not air in the system from me not bleeding it right.

Yesterday(?) or the day before, it starts overheating again. Great.


More:

- No white smoke from tailpipe.
- No apparent coolant in the oil (not brown, foamy, or milky) nor has it ever appeared to have had.
- I don't believe the heater core is bad, my windows do fog from being in the car with the windows up (especially if it's rainy out) but if I put on the heat, the heat does get hot (when system is bled, it gets really toasty hot, so coolant is definitely flowing through it) and the windows do manage to defog competently with the heat on (the A/C light turns on, but the compressor wont run, I believe that the thermal protector is shot on it, so I don't get the A/C helping to defog). I also don't smell any sweet smell from the vents.
- There IS a light film that looks like it may be oil at the top of the coolant overflow tank (by the way, yes it's filling up, so I know air or something is getting in there).

I'm not sure what my next step should be at this point. Another block test for combustion gases? Pressure test the cooling system (to determine if any leak in it exists)? The first HG test really threw me, I think I'll buy fluid from NAPA instead of using the AutoZone fluid (maybe it was bad... but I think I was the first person to open it).

I dunno, I'm out of ideas for now. Thanks in advance for any advice

P.S. I'm gonna have a 2nd car for this week, so if it happens to be the HG, I need to get it diagnosed ASAP so I can get things resolved. I'd be doing it myself, but really don't want to if it isn't the problem (and if something else happens to be causing the overheating, I want to be sure to fix whatever besides the head gasket is bad, because I don't want to get the HG replaced to find that the radiator is bad, to turn around and damage another HG due to overheating because I didn't fix the underlying problem...).

P.P.S. The times are estimates, so 2-3 weeks could be a month, or 1.5 weeks... I never really kept track of when things happened, I just was fixing things as I went.

 
Rep Power: 811

1. are you losing coolant (burning it)?
2. have you done a coolant chemical test to determine presence of combustion gas?

if those dont turn up anything, im guessing there is some kind of clog in the cooling system. either radiator or heater core, that is causing coolant not to flow properly or fast enough. these radiators are incredibly small inside compared to other cars ive serviced that dont even need to be air bled. just pour in coolant and youre done. ive bled mine for well over an hour (4-5 times) and still couldnt get all the air out after a coolant change. its good enough to where it wont overheat however. im also having to add coolant every so often (not a lot, but maybe 1/4 cup a week) so im getting that sinking feeling there is a second small headgasket leak which i had changed a few yrs ago.

have you tried turning the heater on hot and driving at higher rpms? either on freeway or just taking the motor to 4-5k rpms when driving. that could push some stuck air thru small passages since the water pump is turning much faster than at idle.

 
Rep Power: 191

I dont think I'm losing any coolant by burning it, though it does fill the overflow, and I think some may have pushed out of it.

I did the previous block test (blue fluid, and it stayed blue). I decided to pick up a kit and different fluid to re-do the test again tonight. I'm also going to do a pressure test (rented that kit too) to ensure that it holds pressure and there are no leaks.

If I go faster, sometimes that seems to help, but it could be in my imagination. Turning on the heat seems to make it worse.

 
Rep Power: 811

if the overflow is completely full, there is a problem. normally it can go up past the max line but not way up to the cap. unfortunately its a process of elimination and trying things one step at a time. have you looked at the plugs yet?

 
Rep Power: 191

Yeah it definitely goes all the way to the cap. I'm going to do the two tests tonight (coolant pressure, and block tester) and keep you posted with the results.

 
Rep Power: 200

Everyone jumps to head gasket. Not saying that is not your problem at all, but remember KISS "Keep It Simple Stupid". Any amount of overheating or air will do the same thing over in time. You are going the right way about it. If there is no hydrocarbons in the coolant then it IS NOT the head gasket. Pressure test then go from there.....

 
Rep Power: 366

Green01civic,
Yes, you are approaching this right,and we know you doing it right.
I did the hg since I had to do the tb, wp anyway for the piece of mind.
Anyway, whatever the reasons, I do not have overflow or air in the system issues anymore.
My car only overheated on the track, long before the overflow issues.
Running great now.
Cars are tricky. Worse than my ex sometimes, not matter how much love you give to it... Sometimes logic will not explain and it just need the gasket
Your peace of the mind.

 
Rep Power: 191

Went to AutoZone and got the loan-a-tool Part # 9300 "CST/Pressure Tester Adapters" ($75 deposit). Was told that the adapters I need are in the box. Couldn't perform the test last night, because came to find out the proper adapter is NOT in the box (and it has 2 less adapters included than what is pictured).

Went back to AutoZone, got Part# 9124 "CST Import Filler Neck Adapter" ($12 deposit). Despite the name, this in fact does not attach to the filler neck of the Honda (Denso) radiator. It will attach to the radiator cap, and allow pressure testing of the Rad Cap ONLY. Great.

Okay, used that to test the rad cap. Appears to hold pressure quite well (will hold 18PSI for ~20 mins with no noticible drop, I took the cap off the tester after that).

AutoZone does not have a cooling adapter that fits onto the radiator, so do not bother.

Went out to Advance Auto Parts. Picked up the "PowerBuilt No. 70, 22pc Cooling System Pressure Kit. Model #940427" ($169.99 deposit). I think the SKU is 9150049. This has BOTH the adapter for the CAP (Adapter #10) and for the RADIATOR (Adapter #8). So if you want to pressure test both, you need to go to Advance Auto Parts, not AutoZone.

First tester I got ended up being a bum tester (hairline crack by gauge was letting out pressure... and coolant, lol). Returned to AAP, they called around and said they could get one from ~30mi away for me, but would be about an hour and a half wait. I came back 2 hours later, and got the replacement loan-a-tool.

Okay, so attached the Advance Auto Parts tester (car has not been driven in a day, so completely cool), slowly pumped up to 18PSI, and 30 minutes later, it's holding dead on 18PSI still. Checked for leaks on the radiator where I could see, and around hoses and there don't appear to be any (and since pressure held steady, I didn't expect there to be any).

Have yet to do the block gas test again, because I've been running around all day and feel like kicking back for a min and post this.

 
Rep Power: 811

the pressure test really isnt useful if the leak is small enough. my pressure was fine too even tho i was burning coolant every day. the best is the test strips you dip in the coolant and they turn color if exhaust gas is present. that will tell you for sure.

 
Rep Power: 191

I can't find these test strips. The only ones I found specifically for coolant only measure pH, nitrates, and things other than hydrocarbons.

I did manage to find oil-in-water test strips: http://www.mn-net.com/tabid/10450/default.aspx

I think those may work, assuming that there are no non-polar components of antifreeze I don't know about. Kinda pricey though ($45-50), and only in like 100 strip boxes.

I still have yet to use the block tester + fluid, I haven't gotten around to it yet, lol.

 
Rep Power: 191

Granted, as I understand it, CO2 will react with H2O and create H2CO3 which will slowly lower the pH of the coolant (and then show up on a pH test strip), but I'm sure there are buffers of some sort which will prevent any substantial pH change, which is why the "block gas tester" uses air instead (gaseous CO2 is sucked through the test liquid, reacts to create H2CO3 in solution, which lowers pH, and bromothymol blue turns from blue to yellow). On the other hand, the test strips I have linked previously will only wet in the presence of oils... (I think they're a cellulose coated strip, but I'm not sure what indicator is used if any, it may just be an inactive dye that the oils wet by travelling through the cellulose coating).

 
Rep Power: 200

I wouldn't trust the oil test strips for exhaust gas. You said it did have some oil that you could see. The transmission cooler could leak and cause that so it wouldn't be definative. The fluid tester is fine. The strips are idiot proof, but if you know tools the sniffer will sniff it if used right.... The absolute best test for a leak AND to fill a system is a vacuum filler. Doesn't help you find a leak per say, but if you have ANY leak, or a leak to come for that matter (for instance a weak water pump shaft seal), it WILL NOT make a good vacuum. As for filling they are awesome. It doesn't matter what kind of vehicle you are working on, they make air when filling an absolute thing of the past. No "bleeding" is needed when all the air is vacuumed out...

Make sure it doesn't have exhaust gas first off, it's to easy not to, then lets keep going if needed....

 
Rep Power: 191

I love the vacuum fill idea, lol. I can't fathom the cooling system taking a deep vacuum though without screwing something up since it's normally meant to be under no or positive pressure (I guess the A/C can take a vacuum but be under high/low pressure, so it's not unprecedented), but I kinda wish I had the equipment for it!

I'm definitely going to use the block gas tester, but don't want to mess around there at midnight in the pitch black, and tomorrow I'm gonna be at six flags all day . FML. Sorry for not doing it earlier, after a busy day I just couldn't bring myself to it.

edit: And indeed you are correct, test strips WOULD be useless for verifying a head gasket leak, for the reason you mentioned. Trans fluid is petroleum base stock, so coolant would be positive for hydrocarbons if there was leaking of trans fluid into the coolant. I suppose a cross-test of anti-freeze or water in my trans fluid would eliminate the likelyhood of that... but what a PITA. Water-in-oil test strips for the ATF, an oil-in-water test strip for the Coolant... madness!

 
Rep Power: 200

They have been around a while now... I used one for the entire time I worked at the GM dealer to fill everything. The only problems it ever showed was finding that there was a problem to find. Vacuum is harder for a leak to "hold". AC systems are vacuumed down first to remove water and air. If something is under a vacuum the boiling point of water is such that it actually boils at ambient, so it can be vacuumed out (because its gas). The air because it is an extra contaminet in the system. An extra bonus, is that this is also an excellent chance to check for leaks. If any are present it will not hold vacuum.
As for the cooling system. Pressure testers actually work very well. Pump it up and leave it overnight if you need to. If it leaks off something, wether it be the apparatus or your cooling system is leaking.

 
Rep Power: 811

have you looked at the coolant? like take some out and put it in a glass, see if there is any oily film on the top. thats the first thing i noticed. it will not discolor the coolant if there is very little, but you will see the typical "rainbows" on top of the coolant. ofcourse it could be from trans fluid too, so again would be hard to tell the difference. but if there is no oily residue, it would rule out oil in coolant at least.

 
Rep Power: 191

Oh man! Sorry for the late reply. The first AutoZone tester was defective (broken seal let air through from outside tube, and fluid leaked out, I had to return it today for a new one).

I'll let the pictures speak for themselves. I want to get a 2nd conclusive test in order to rule out possible testing errors, but here are the results so far (Image, 1.5MB PNG):

http://img801.imageshack.us/img801/4917/testresults.png

Top row is pictures from Test #1.
1. Radiator cap removed, Radiator drained to 1-2" from top with a prestone antifreeze tester, and fluid emptied into the overflow tank (to conserve/reuse antifreeze later).
2. Car was started from cold and idled for 30 seconds, revved to 3k rpm, then held at 1.5k for a little bit, then let idle. Some more fluid filled the radiator, so I emptied it again.
3. Antifreeze tester was placed on radiator fill neck while car remained idling, and tester aspirated for 2-3 minutes.

Second row is a control shot, showing the tester with fluid and no aspiration.
1. Fluid from test #1 was emptied, and fresh fluid was poured into tester.

Third row was from what was to be test #2, but the car started to heat up to temp and steam(?) was being drawn into tester, deforming the plastic tube somewhat, so the test ended prematurely. The fluid did not turn noticibly yellow or green.



So what do you all think? Should I go ahead and order the gasket kit? Should I bother to do a conclusive second test and see if it is a repeatable result? (when the car cools down again)

I'm gonna get slammed with the hurricane tomorrow into sunday, and then I have school beginning next week (monday, I believe classes start), so I can't do the HG for the next few days myself. Should I even try to DIY the head gasket? (Considerations: I'm unemployed, so money is tight. I have the HELM digital manual, my own tools, and my father-in-law runs a small machine shop, so he could probably mill the head for me and lend me any tools I might not have... though he doesn't have the facility for pressure testing the head, he does fabrication mostly... I may have to find another machine shop to test the head.)

Thank you all!

 
Rep Power: 200

If it's green at all you have at the least a head gasket. Could be more, but probably a gasket and a head mill job.

If money is a consideration FOR SURE do it your self. It could easily save a few hundred dollars. It is not hard, just time consuming. Don't even mess with taking the manifolds off unless you straight edge it and it needs milled. The manifolds do not have to come off to pull the head. The only tedious part is to make sure you have the timing belt in time. If you get it way out of whack you can bend valves (even with the starter, the engine doesn't even have to run). The job is not hard at all but count on it being down for a couple to a few days especially if you have to mill it. If not you could easily do it in a weekend. I am a mechanic and have plenty of tools (hand and air), so not a completely accurate gauge on time, but it took me roughly 8 hours. I am meticulous though, I don't try to hurry through. If I did it again and was in a hurry maybe 6. I didn't mill my head, BUT for most diy'rs it is not worth the risk. If you are comfortable checking the head and block for warpage then you may not have to. If you have overheated it then by all means get it milled. If you haven't, you may not even have to mill it. The most important part to keep from warping the head it to first let the engine be bone cold, then take the head bolts out by hand and only a couple turns at a time. Use the reverse sequence of tightening them.

 
Rep Power: 191

I was thinking of getting the gasket kit (which comes with the following, list gathered from the HG FAQ here on CivicForums):
06110-PLD-010 Gasket Kit
12251-PLC-004 Head Gasket
17055-PLD-004 Intake Mani Gasket
18115-PLC-J01 Exhaust Mani Gasket
12210-PZ1-004 Valve Seal A (x8)
12211-PZ1-004 Valve Seal B (x8)
(They lumped all 16 seals in the same bag, and they were all visually the same, so I'm sure the seal a's are the same as the seal b's. From the drawings, it looks like "a" is for the intake side, and "b" is for the exhaust side valves. Regardless, you get 16 total valve stem seals)
12342-P2A-005 Spark Plug Tube Seals (x4)
91213-P2F-A01 Camshaft Oil Seal
90442-PLC-000 Rubber grommet-washers for the Valve Cover bolts (x5)
12341-PLC-000 Valve Cover Gasket
19411-PLC-003 Water Passage Gasket


Am I not going to need all of the above parts (do I need any besides the HG itself)? I know you know a lot more than I do, so my big question is why do people remove the manifolds if it isn't necessary to do so?

Thank you again, you're a big help!

 
Rep Power: 191

Also, I'd have my father-in-law check the head and block for any warping, he does precision work and has very precise measuring equipment, so I'd trust him to make that call.

 
Rep Power: 811

like jdb said, if you never overheated, it will be a much bigger pain to remove the head and take it for inspection if you dont have to. it would also mean you would need to replace intake and exhaust gaskets, which you would not have to do if the head came out with them still attached, if you are only changing the head gasket. i did not need any of those extra things, only the head, intake, exhaust gaskets (i had to get the head resurfaced due to some overheating that slightly warped it). the rest we reused or didnt have to remove. but its prolly worth having them around in case something happens and you need the extra.

 
Rep Power: 200

People pull the manifolds either to replace the gaskets, or send the head to the machine shop. It does give you more room to work and makes the head considerably lighter and easier to handle. If you want to replace all those gaskets that will not hurt at all. You have to pull the cam and break down the valves to put the seals on. Adjust them too while you are there. I would say if you don't have a problem why waste the time? Adjust the valves any way while you are there though. May help the engine run smoother and get better mpg. BUT that is assuming they are too tight anyway. While you are there put a new timing belt and water pump in it. The pump runs off the timing belt so you have to get to that point to replace it, and they are cheap enough. You also have to take the belt off of the cam gear so why not replace it too? I have had my valve cover off 5 or 6 times and the gasket and plug wells aren't leaking. You can replace it too if you want, easy to get to there, but it is rubber and they aren't replaced every time like cork was.... The main thing is get all the old RTV off the corners by the cam holders and put new on. That is the most likely place to leak on the gasket itself. You can't really do anything to ensure the rest won't leak, except wipe it off, wipe it off, wipe it off. OH YEAH, wipe it off. It can't be assured to seal up again if its dirty. Otherwise your list looks ok for this job. What all is in the "kit"?

 
Rep Power: 191

Thank you Gearbox and JohnDeereBones.

@Gearbox: It has overheated, and hit close to the H at least a few times since the odyssey has started (engine was turned off immediately and let cool down any time it has, but it's still gotten quite hot). Has happened a couple times while the wife was driving it, and like twice that I've driven it... so I'm counting on likely having to mill the head.

@JohnDeereBones: Everything in that list comes in the kit, it's $133+SH ($30 for overnight shipping) from bernardi honda.

Will I need to, or would it be a good idea to replace the head bolts? And are the torque values wet or dry (I've read to soak them in 30wt oil, but also worry about things coming loose. Should I apply anything to the threads at all?)

I've got leaf gauges, so I'll definitely do the valve adjustment. As for the T-belt and Water Pump, it was already replaced (along with tensioner) at Honda about.. oh, I don't know, maybe 10k miles ago, so I was just going to reuse everything.

 
Rep Power: 811

just soak the bolts in oil. you have to be very careful reusing the old bolts. the honda manual says during the three stage tightening sequence and order, if any of the bolts make sound while tightening, you have to remove them all and start over. if it gets to a point where the old bolts are making noise and you give up, its good to have new on hand. i used new for mine. if you dont follow the order and sequence, you will be doing it again real soon lol. is this for civic or another honda?

 
Rep Power: 191

This is for my 2001 Honda Civic EX. . It's in good shape other than this overheating/head gasket problem.

Perhaps I'll buy the new bolts and be done with it. I'd rather have new parts on hand than run into problems when I'm doing it.

 
Rep Power: 811

here you go then. make sure you read and understand what to do. there are three torque steps, and a tightening order. for example, put in bolt #1 and tighten to the first torque step, then repeat with bolt #2 and the rest. once you have all the bolts torqued to the first step value, go back to #1 and tighten it some more to torque step two. do that in order for each bolt, and then move on to the final torque step.