Erratic overheating, intermittent warm HVAC air.

[andyems]

This pertains to a '05 Civic but figured the process would be the same for any make and I've always had excellent help here. This will get long-winded but bear with me. I can do most anything when I know what part is bad and needs to be swapped. However, I'm horrible at trouble shooting and diagnosing.

Recently my wife tells me her Civic is not blowing warm air, even after the engine gets to the normal temp range. Several days later, before I had a day off to think about it, I noticed an antifreeze smell after driving it. I raised the hood and found the car had been blowing/boiling antifreeze out of the overflow reservoir. My dad said it acted to him as if the thermostat was bad. I replaced the t-stat and the overheating seemed to stop for the past 3 weeks. Wife said it still seems to be slow warming up but no overheating. Fast forward to today when we went for a drive in the mountains. All day of slow, curvy, up hill driving the car again performed fine, although we didn't use the heat since it was warm we had no cooling problems. After stopping to eat, the sun went down and the night cools off, leading to the need for warmth on the drive home. Again we have no inside heat. The temp gauge is in the normal range, but again it's just blowing cold air. After about 20 minutes of driving we notice the air finally begins to blow warm but the engine temperature gauge is climbing sky high. Car reads high for 45-60 seconds and the air from the vents cools back off. Suddenly the gauge drops to normal and the vent air gets warm again. I stop and check, and the car is blowing antifreeze out of the overflow again. The rest of the 45 minute drive home it constantly repeats the same process. The HVAC blows warm and temp gauge is normal for several minutes. The engine temp begins to climb and the air vents cool back off. It only reads high a few seconds and then the process repeats. I was thinking the t-stat is sticking even though it's new. Any ideas or advice would be awesome and very appreciative. I'll try and fix anything but I have no idea how to go about diagnosing this one. Thanks in advance.

[Doug_Scott]

Cars can sense when surgery is about to happen. Don't fall for it, the car is trying to trick you into thinking something magic happened.

 

Keep in mind, this is also the same car that will make any socket you drop into the abyss around the engine find the deepest. hottest, pan of anti-freeze, or oil to fall into, quietly.

[andyems]

Talked to the mechanic again today. He said next week he can do a "chemical test"???? and that will tell if there is a leak also. I think he actually tested the cylinder leak down test. I'm just not familiar with the terminology but he described what you said, he forced air into each plug hold to see if it leaked down and he said it didn't show a leak. As soon as I get the chemical test I'll let you all know what it shows.

Funny thing is that Thursday we took a drive from East TN over the mountains to western NC and it had very drastic elevation changes. When we got home is when I noticed it had overflown again. We took and local drive yesterday that was a 120 mile round trip. No major mountains or climbs. When we got home I checked again and it was dry under the hood. It didn't overflow any at all on the trip last night. Probably the same outside temperature on both trips. One was just a little longer with elevation changes.

[Doug_Scott]

The special tool is a compression tester. Disable ignition, take the plug out, screw a hose in, connect to gauge, hold throttle wide open then crank engine for 4 or 5 hits. This will tell you what the compression is for each cylinder. Head gasket generally will put coolant into the cylinder, sometimes it will just allow compression to enter water jackets, which can pressurize system high enough to get rad cap to vent, but, it usually is pretty obvious by looking at the coolant. Head gasket gone between cylinders will not generally hit a water passage. Looking at the plugs will let you know pretty quick if coolant is going into cylinder. Plug will be very clean if coolant is getting in, compared to the other plugs.

[Doug_Scott]

From your description it sounds like an air lock. Chances are there is air trapped in heater core. To diagnose, I would first make sure rad cap is holding correct pressure, then, I would be checking the heater hoses to make sure they are as warm as upper rad hose. Both hoses should be near same temp by feel. Coolant system flow rates are not as positive as they could be. If you look at the design of the impeller and how it fits in the pump housing, they are not very efficient. But, they also do not require to be. I think it is due to this, that air locks don't get forced out. Stock water pumps will not build pressure like an oil pump does.

 

Has the heater core been changed at any time? Check around the heater box for any anti-freeze. Generally when coolant temp goes high, then just before overheating it drops back down, the thermostat is sticking, or slow to respond. But, an air lock will behave like that as well. For whatever reason, Japanese engines seem to require special procedures to prevent air locks. By comparison, most traditional North American engines will either not get an air lock, or if they do, once it overheats or gets close to it, that will fix the air lock. I know, apples to oranges, but, it is just something I have noticed over the years.

[Doug_Scott]

Chemical test is basically testing the coolant for combustion gases. If the head gasket is causing the rad to overflow into recovery tank, then it is actually the combustion gases pushing coolant out.

 

There is something about how you are wording your comments and questions. Does your reservoir/recovery/overflow tank have just simple plastic cap on it, or is there a rad cap on it? Has the temp gauge ever shown overheating? I don't mean high, I mean in the red zone.

As has been said already, about the only reason there is for the temp to swing high and normal is an air lock. Coolant systems are engineered from the beginning to the end. The coolant flow though the block, head, radiator and heater core all have to be working correctly for coolant to flow throughout the system. The heater hoses are not just placed anywhere, they are part of the coolant flow in most engines(some have the option to shut flow off completely via cable driven valve in heater hose - see mid 70's Dodge engines). Hose temp should be nearly the same on both hoses.

[andyems]

Sorry for my bad terminology and wording. Aside from the regular oil changes and the like I'm not much help on vehicle maintenance.

The overflow/reservoir is just the standard whitish plastic with a simple screw on non-pressurized lid. This is where the coolant leakage always comes from. I don't think it actually boils over, I think it gets pushed from the radiator into it when it gets hot and it just pushes it past the simple lid and overflows. When I stop the car it's never actively coming out, it's usually about a inch below the cap, and the neck and everything around it under the hood is soaked with coolant. I swapped the thermostat myself before I took it to the mechanic but the overflowing continued.

One time I did have temp gauge changes. We were again in the NC mountains with a lot of elevation. My hometown in around 1500 feet. We went to the local ski slope towns in NC with about a 4000 foot elevation difference. On the way home the temp gauge started reading in the red for 30-45 seconds but would drop back to normal for about the same amount of time. It did this for the entire 45 minute drive home. I also had a cycle of warm and cool air coming through the vents when this happened. Since the timing belt change was due anyway I had someone replaced the timing belt, water pump, radiator cap, 2nd new thermostat etc. he is the one who pressure tested the spark plug holes (somehow) and told me the head gasket was good. The problem persists after all this was done.

Yesterday I took the actual radiator pressure cap off yesterday and noticed the coolant looked low in the radiator. I'm not sure how full it should be, but the coolant level was maybe 1/8 inch below the inside metal "fins" of the radiator, yet the overflow was still way too full, like its never drawing any coolant back into the radiator.

Once the motor warms up both large hoses seem equally hot and I'm blowing warm air through the HVAC when I turn it to warm, so I'm guessing it's circulating through the heater core correctly.

I've got an appointment for the chemical test Thursday and I'll ask him exactly what test he did the first time through the spark plug holes.

[Doug_Scott]

Some things to keep in mind when trying to understand what is happening. In order for the coolant to get drawn back into the engine, the engine needs to require coolant as it is cooling down and the coolant inside the engine "shrinks". If the engine was completely warmed up, the coolant inside the rad will have expanded, and that builds pressure since the rad cap is trying to stop the coolant from being pushed out of the rad. Once the engine cools off, any coolant that got pushed out through the cap should be pulled back into the engine.

For that to work, the line from the rad cap to the overflow/reservoir must be in good condition, no crack, cuts, holes, or too soft. It must be tightly connected to the radiator neck as well. At the reservoir end, the hose must be firmly connected to the tank. Inside the tank there must be a way for that hose to be able to pull coolant out of the overflow/reservoir. Any leaks will just pull air into the radiator instead of coolant. This is a hard thing to describe.

If you have ever smoked, you will have seen what happens when there is a leak in the tube. If your cigarette ever got a hole or cut near where the filter is attached to the rest of the cigarette, every time you would draw on it, you would just get air, and no smoke. I would imagine the same things happens on a crack pipe.

You can test if the line has any leaks by applying a small amount of vacuum to the hose from the rad cap, if you can pull coolant up from the tank, then the line is likely ok. If you cannot pull coolant out, then that is your problem.

 

Your description of temp going high then normal then high then normal is classic air lock symptom. Honda's used to have bleeder screws at various points in the cooling system, does your car have any?

 

The two hoses I mentioned were the heater hoses, not the upper and lower rad hoses.

[andyems]

Thanks for all the advice. Other than replacing the thermostat a few weeks ago, nothing else has been changed in the cooling system ever. Heater hoses are equal temp and no leaks or smell around the heater core. Electric fan is coming on, as I saw/heard it running last night when I looked at the overflow.

It's due for a timing belt change anyway. I have a good, honest mechanic I've known for years. He used to work with Toyota and then Nissan and retired/quit to start his own business. He is going to install the new belt, water pump, tensioner, etc that comes in the kit. He also said he had a tool that goes in the spark plug holes (or something of the like) and test the pressure??? to see if it's a leaking head gasket? He also said he'd check for an airlock. Said he'd get it figured out for me. Parts are ordered and should be here mid-week.

I had to give my truck up to the wife today so she could go to work. I've done nothing since parking the car last night. I ran some errands, stop and go in town, some rural driving, totaled about 90 minutes/ 20-25 miles altogether and not a single problem. No overheating, HVAC air warmed up as quick as usual and blew warm air all day long. Performed perfectly all day today with no hint of a problem anywhere. Thanks for all the advice. I'll update when and if I get it figured out.

[Jsdirt]

Sounds like your on the right track.

 

Must be the moon phase - my wife had to take my truck the other day as well. HEI coil failure in her '89 Blazer that I threw a '00 Jimmy engine in. Once I finally got a minute, it took a whopping 5 minutes to R&R. Try that on a '14 ...

[txab]

Maybe you have trapped air in the system

[andyems]

Thanks. Ive been reading a lot tonight and saw that mentioned elsewhere, too. Tomorrow, in the daylight, I'm going to try and bleed the system. I thought it would have shown up before now, though, but I'm gonna try it just the same. A shade tree mechanic friend also just suggested I replace the radiator cap. Gave me a long theory of the drop in pressure from a bad cap can change the boil point of the antifreeze. Something to that effect anyway.

[andyems]

It's been almost 2 months now, and still having issues. When I picked up the car from the timing belt change, the mechanic told me there was no head gasket issues. He used some tool/tester to check the compression in each spark plug hole. I have no idea what he used or what it's called. He told me the head gasket was good. He thought I got air into the system when I replaced the thermostat the first time. As I drove it the 30 minutes back home, I smelled antifreeze and hoped it was left over after he refilled it when he was done. No suck luck. I found it was still running over in the overflow reservoir. I then hoped it was just overfilled. I pulled out enough to put it between cold and hot and hoped for the best. Over the next few weeks the coolant slowly edged up until it was running out again. I was told by 4 different people at 3 different parts stores that my next option would be the simple radiator cap. Told me it would malfunction and affect the boil point of the coolant and also would not allow it to flow back into the radiator correctly. That fix was a bust, too. No luck. I then had a friend who had a kit to suck all the air out of the system, figuring it was again air locked. (All I know is that it hooked up to an air compressor and sucked everything out, collapsing the hoses, and he told me it would get any extra air out of the system.) We again filled the coolant line to between cold and hot. I drove it for about 30 minutes in stop and go traffic and it did not move any. That was last week. Since then, it as been slowly creeping up in the overflow. Every time I drive and check it, it's higher than it was before. It never seems to go back down. Yesterday afternoon I again smelled antifreeze when I got home. Again, it's throwing it out of the overflow. I don't know if it's boiling over, or just pushing it out. The car is running fine, the heat is working normal and the temp gauge is not moving from normal. I'm at a loss. I don't know what to try next. Could there be a blockage in the return line somewhere that is keeping it from returning into the radiator? It's got a new t-stat, rap cap, water pump, the cooling fan is running, and it's ben bleed twice, once manually once with a kit. The compression test revealed the head gasket is not leaking. Any idea of my next step would be greatly appreciated.

[Doug_Scott]

When the engine is cold, check the coolant level in the radiator. Is coolant visable in the radiator? The coolant cannot magically increase in volume. If you put 8 quarts of coolant into the system, then the maximum volume you can have at the same temp it was put in is 8 quarts. If it pushes 3 quarts into the resevoir, then there is still 5 quarts in the engine. When engine is hot, and not running, it will start cooling down. As it cools, it will contract. As it contracts, the pressure you had in the system will drop, and once it drops to zero, if the coolant still has some heat in it, the pressure will actually drop to less than atmospheric pressure. On a normally working system that will pull coolant from the resevoir back into the radiator to take up the space that occured from the cooling off of the coolant.

 

The resovoir must have a tube that goes from the rad to the resevior, that tube CANNOT have any leaks in it. Not even a pin hole. That tube must also extend into the coolant in the resevoir, and again, CANNOT have any leaks in it. For it to work correctly, the rad must be able to pull coolant from resevoir into the radiator. Any pin holes in that tube will let air in instead of pulling coolant from the resevoir.

 

It is completely normal for the coolant to transfer from radiator to the resevoir and back again. The most common reason for the coolant to not return to the engine, is a leak in the hose/tube between the two spots.

 

Coolant boils at 212f at sea level, and boiling point increases with pressure. If the cap is weak it will not hold enough pressure to keep coolant from boiling. There is also another valve in the cap that allows coolant to flow back into the radiator when coolant is under less pressure than the resevoir.

[Gone Again]

Take it to a shop that can do a cylinder leak down test. Air is forced into each cylinder through spark plug hole and then you watch for air in the cooling system. The engine is manually turned over during the test to possibly check for cracked cylinder/block. This is not a compression test. I'm still betting on a bad head gasket.

[Jsdirt]

That's true ... but doesn't account for your lack of heat. Usually when you have no heat, it means there's no coolant moving .. especially when accompanied by a temp gauge in the red zone.

 

How's the water pump? I've seen some of those small cars lose the impeller - the engine turns the shaft, but the impeller just sits there ... not likely on a Honda, but ....

 

Also a common failure point is the electric fan ... but you'd still have heat in that case.

 

Best way to figure out what's happening is to feel all the hoses at operating temp - if everything's normal, then you might have a blend door issue. If it's indeed the T stat, the upper/lower hoses and radiator will still be cool ... or warm but not hot. If everything's hot except the heater hoses, there could be a blockage in the heater core. Also make sure that electric fan is coming on when it's supposed to.