Rep Power: 0

I just tested my daughter's '04 Civic with this kit: https://www.napaonline.com/en/p/BK_7001006

The coolant level was already low, so I didn't need to drain any out as the instructions say to do. I let the car warm up while I ate lunch - radiator cap off. When I came back to the car after lunch there was steam coming out of the radiator where the cap would be if it was on.

Within 10 seconds the fluid in the test kit changed from blue to yellow/green. I think this is a positive indication of a head gasket leak. Am I right? Would steam blowing out have possibly given me a false positive?

Is there a write-up of how to change the head gasket on this car?

Thanks,

Matt

 
Rep Power: 210

Normal: As long as it was just steam and a bit of spillover and not gushing out. The coolant system gets over 190 degrees at operating temp so it builds up heat/pressure.

More than likely a false reading because from your description it sounds like coolant mixed with the test fluid. The idea is to draw in coolant system vapors only into the block tester. To test properly: at operating temp and idle coolant level should be at level of radiator cords (about 3/4"-1" below top of radiator/radiator neck.

If properly tested blue test fluid turns green or yellow then it's definitely a head gasket breach that is allowing exhaust emissions to enter the coolant. However, if test fluid remains blue there can still be a head gasket breach (there's several types and symptoms), cylinder head crack or block crack (rare) that is not detected by a block test.

What lead you to believe there may be a headgasket breach? Typical symptoms for honda engines include:
Coolant loss in radiator combined with coolant gain/overflow in radiator
Intermittent or complete loss of interior (climate control) heat

First reference for repair is the service manual. There's also DIY write-ups and Youtubes on the subject. Head gasket issues are seemingly a daily topic here so do a search here as well.

 
Rep Power: 0

Thanks for the reply.
I could see the top of the coils, or cores, or cords, or whatever they are when looking down through where the cap was. Coolant was low. I had also just changed the thermostat and not refilled. After this test I added nearly a quart of water.

We had loss of coolant in the radiator as well as loss of cabin heat unless the car was driving (heat didn't work at idle).

Also, I had a cylinder 3 misfire while I was originally dealing with this. The cylinder 3 plug looked different than the rest. I [thought] I fixed that problem with a new coil and plugs. I now think that the head gasket was leaking into #3.

I *thought* I had fixed the coolant leak at that time because I found a crack in the radiator, so I replaced it. But, I'm still losing coolant. And now, this test.

Here's my test tube, and the original blue liquid:

 
Rep Power: 210

To note: even being a quart low and /or air in system can cause an overheat so avoid that situation. You should have filled the coolant system and bled it of air after thermostat replacement regardless of block testing.We all make mistakes so please don't take that harshly.
I suggest you fill up the coolant system, bleed it of air, drop coolant a bit and do another block test on it. Or disconnect the overflow that attaches to the reservoir from the radiator and test as shown in video below to ensure you did the test correctly. Even so, by your description below of symptoms it's very likely a head gasket breach.

Very possible. Since it's losing coolant: If you did a coolant system pressure test or went under the car while it was idling at operating temp (under pressure) and didn't find any external coolant leaks then it points to an internal leak into the cylinder as you mentioned. Yes, coolant getting into a cylinder can cause a misfire which is usually more noticeable at cold starts.

If it turns out to be a head gasket breach (and most likely is) and you want to do a headgasket repair you may want to consider getting a leak-down test on it. A leak-down test will not only verify a headgasket breach it will verify compression loss through piston compression rings, intake and/or exhaust valves. A compression test is also advisable to find if there's compression loss at, in one, two or more cylinders. If there's compression loss in two adjacent cylinders then it points to a headgasket breach located between those two cylinders.

His is a 6th gen:

 
Rep Power: 367

pictures might be missing due to Photobucket policy changes...

https://www.civicforums.com/forums/3...arts-list.html

https://www.civicforums.com/forums/1...t-diy-2-a.html

https://www.civicforums.com/forums/1...-belt-diy.html

 
Rep Power: 367

testing - chemical tests are notoriously misleading in these engines.

use test proposed in the videos in the first post.

https://www.civicforums.com/forums/3...reference.html

 
Rep Power: 512

First off, pictures in those threads are still up, so you're good to go on that.
Second, while those threads are helpful, I'd still pick up a Haynes manual for the car, especially if you're a DIY kinda guy. Lots of invaluable information in those things. The one I had for my 2002 got beat to crap, but it served it's purpose extremely well. They're only like $30, so why not?

 
Rep Power: 0

I can see where that would work. I used what appears to be the same tester. But that is not how the instructions that came with the tester said to do it.

My understanding is that the tester changes color in the presence of carbon monoxide. How could you get a false positive? Where else could CO be coming from in the radiator?

 
Rep Power: 210

A false positive can happen if coolant fluid (not vapors) mixed with the test fluid. It should say that on the directions. Green coolant dye + Blue test dye equals yellow. Retest drawing in coolant system vapors only.

 
Rep Power: 367

a word of warning.

there were several cases that the chemical testing did not show any issues, but the gasket was still breached indeed.

D-series engines would develop a small gasket failure, like a hair line, and from there combustion gases push through to the coolant and pushes the coolant out of the overflow reservoir.

our member ezone created the test with shop pressure into cylinders, he vouches that detects 100% of the cases.

if that test does not show any issues, then keep the gasket unless you remove the head or timing belt anyway.

 
Rep Power: 119

All these head gasket fails makes me wonder if a installed air bleeder might work.

something like https://www.supplyhouse.com/Bell-Gos...SABEgIbQvD_BwE

Allow the combustion air to escape before it builds up pushes out the coolant..

 
Rep Power: 367

the reservoir getting full and the radiator getting empty is a good warning - i would rather have that happen for me to detect before it gets worse.

 
Rep Power: 210

The Dong wised-up and attached Ezone's headgasket testing video in my signature. Also, in my signature is a link to types of headgasket failures.I got tired of searching for them to link in all the damn headgasket posts that come up on an almost daily basis..lol

What would have been cool for all auto manufacturer's to do was to include an audible alarm when engine/coolant system temp rises above normal operating temp. Along with the alarm, once the alarm sounds the ECU would automatically set climate control to max heat and fan on high to help engine to cool down.

 
Rep Power: 0

Thanks for the help, all. I looked at a video for doing the head gasket and decided that I'd have to take at least 1 day off work to get it done. My daughter needed the car to go back to school so I paid for it. It was leaking around the #3 cylinder. Which also explains the cylinder 3 misfire I had.

Got the car back today and immediately pull the driver's side lock out of the door to get it re-tumbled (I think I started another thread about that).

 
Rep Power: 518

Erm, not a good idea IMO. Eliminate one symptom but not fix the cause, which continues to get worse and worse.

Continued combustion leakage will eventually flame-cut through the head gasket at the breach, and if the driver really sticks their head deep in the sand, the head and/or block surfaces can be flame-cut as well.


A short while back someone posted pics of a MLS (steel) head gasket that had been cut all the way through as described.

 
Rep Power: 210

Thanks for the follow-up. Glad to hear you got it sorted out.