Help! car overheating!

I have a 1990 Dodge Caravan that I recently got from BIL. It has the Chrysler OHC 3 litre V6, from A 1988 van. These vans don’t have a temp gauge, just an idiot light.

Last week, when I was out of town, it started overheating. My wife drove it straight to her dad’s house, and he changed the thermostat. Ok, good. Three days later, when she was pulling into our driveway, she noticed it steaming. She turned it off, called me out to look at it. There was water dripping down onto the ground in several spots, and that little pan that the radiator sits in was full of water. I could hear the water in the radiator boiling (or at least bubbling).

My FIL says he didn’t check the thermostat before he put it in. Even if its a bad thermostat, would that cause the radiator to boil? Wouldn’t it trap the heat in the block? In tempted to pull the thermostat and test it; I don’t know what else it could be. Once it cooled down I checked the radiator and it was empty.

If I pull the thermostat I’ll have to get a new gasket, which isn’t a big deal at all, but if there are other, more likely things causing this I want to test them first. Also, my wife drove it for several days before it overheated again. I don’t know if a thermostat has a three-day lifespan.

Any mechanics out there have any ideas what could be causing this overheating? And how do I test them?

What I’d consider the first steps:

  1. Test for cooling system integrity (leaks). You’ll need a pressure tester.

  2. Is the electric fan activating? I assume it’s terpereature activated so you can just observe as engine warms up.

  3. Water pump any good?

  4. Radiator clogged or blocked?

Gotta be one or more of the above.

The next thing to do is test for leakage. There’s a “chicken or the egg” aspect to overheating and coolant loss: overheating can cause loss of coolant (but generally it’s lost only through the radiator pressure cap, possibly working its way back to the overflow jar), and loss of coolant can cause overheating. It’s also possible that the system was not filled properly after the thermostat was replaced, but the fact that it took three days for symtoms to show suggest that leaks (but not huge leaks) are the most likely culprit. The wise thing to do is find and fix any leaks first, then fill the system properly so there are no air pockets, then evaluate system operation.

This is often best done by a well equipped repair shop. The simplest and least messy way to find leaks is with a cooling system pressure tester, which most home mechanics don’t have. Likewise the simplest and least messy way to get the air out while refilling the system is with an air evacuation tool that hardly any home mechanics have. These things can be done without the tools mentioned, but the risk of a misstep is high.

If you want to try taking care of this at home, I can give you some tips, but it’s difficult if not impossible to get to the point where you KNOW there are no leaks, and sometimes it’s tricky and tedious to get all the air out of the system.

thermostats are typically designed to “fail safe.” That is, they are designed so that when they fail, the provide the maximum cooling effect, rather than the minimum: your engine runs cold instead of overheating. It’s not very good for an engine to run cold for long periods of time, but overheating is much worse.

The thermostat is probably fine, and even if it’s not, it’s probably not the cause of the overheating - especially since the overheating existed before installing the new thermostat. In fact, your original thermostat probably wasn’t the problem, either.

Where is the coolant leaking from the radiator? Is it coming out of the cap, or somewhere else?

Wait, so with the most recent driveway incident, are you saying it was steaming and boiling and puking coolant, but not actually overheating? (i.e. the temp light wasn’t on)

Another “worth a try 'cause it’s cheap” part is the radiator cap, especially if the above is true.

I don’t know where you heard this, but it’s not true. A thermostat’s rest position is closed. In the field, they can stick closed (will overheat), stick open (will overcool), or stick somewhere in between (could overheat or overcool, depending just where it sticks). In my experience, sticking closed is more common.

I’ve heard of pressure testing the system, but don’t know much about it. Can a home mechanic go to Napa or O’reilly or somesuch and rent the tester for an afternoon? If not, any idea what a shop would charge to pressure-test the system?

The idiot light never came on, so I can’t really say if it actually overheated or not. The light does come on when it is first started, and it does the self-diagnosis of the dash lights. I know this doesn’t mean the sensor is good, but without an actual temp gauge I don’t know how hot it really got. My wife said that when the thermostat failed the light came on intermittently, but was not on when she had to stop and there was steam pouring from under the hood.

As far as where the coolant was coming from after the driveway incedent, I’m afraid I don’t know. It was not spraying from some obvious place, but the pan that the radiator sits on (looks kinda like a long, shallow bread pan that encloses the bottom inch or so of the radiator) was full of coolant and was overflowing. I could hear the coolant in the radiator bubbling, as well as the coolant in the overflow tank (escaping pressure from the radiator?)

The idiot light is appropriately named. While it may be better than nothing, it’s tough to diagnose a problem without a guage.

On the subject of thermostats, I always test them before installing. Have done so for several years, especially as so many are made off-shore (China, Mexico, etc.). I hang them from a string in a pot of water on a stove, turn on the burner and make sure they open as the temperature rises. I’ve only caught two over the years, but the ten minutes I invested saved me many hours of labor, to say nothing of grief.

(not a mechanic, unfortunately, but I’ve seen a car overheat before)

Does it primarily overheat when the car is running, but stationary? I’ve seen a 1995 Ford Escort that would overheat (and the coolant would actually boil!) when idling or in slow, heavy traffic but was fine driving at 40 mph, the problem was the engine fan. Just driving it at speed was enough to keep it sufficiently cool via draft, but as soon as you stopped, the temp started to climb.

I don’t know if they’re available for rental. For purchase they seem to be in the 100 +/- 25 neighborhood. If you get one, make sure it has the proper adaptor to fit your vehicle (for many brands of tester, that’s no adaptor, as yours takes a “standard” old-style radiator cap).

I charge 35 for a pressure test. This includes applying the tester, tightening reasonably accessible hose clamps, and visual/auditory check for leaks. Some leaks are not readily visible (some intake manifolds, heater cores, etc.) and there’s an additional charge to track them down. When the system shows no detectable pressure loss for 15 minutes, I consider it leak-free.

Note that in cases where there is more than one leak, you may only find them one at a time. Typically the biggest leak shows up first, and often other smaller leaks won’t show up until the bigger one is stopped.

I know what you mean; I’ve had a jeep that did this. With my van I can’t really tell, because the steam is only obvious when it isn’t moving. I do know the fan, which is electric, does work.

Both my FIL and my BIL seem to think its a head gasket. This has my bullshit meter pinged for two reasons: in my admittedly limited experience, one or both of two things will happen when a coolant channel in a head gasket fails: lots and lots and lots of smoke/steam coming from the engine (around the heads/valve covers/intake manifold), or there will be lots and lots of dense white smoke coming from the tailpipe when its running.

Any thoughts on a head gasket failure?

definitely a possibility. Failure of the gasket between the cylinder and coolant passages can cause hot combustion gases to be forced into the cooling system, which adds excessive heat to the coolant. typically a key sign of this is foamy coolant when you remove the radiator cap. But as always removing the radiator cap from a hot engine is potentially dangerous.

I worked on a friend’s old outboard boat motor that had a badly-blown head gasket. the only obvious symptom was overheating under power, as combustion gases blew right into the water jacket.

ETA:

there are multiple possible symptoms of a failed head gasket, all depending on how large the breach is and where it is.

  • if it fails between two adjacent cylinders, the symptom will be low/no compression on those cylinders.
  • if it fails between a coolant passage and an oil passage, the symptom will be mixing of the two; either muddy looking coolant or milky oil, or both.
  • if it fails between the cylinder and a coolant passage, it can cause overheating and/or coolant consumption. that depends on how bad the breach is. A small leak may not cause coolant entry into the cylinder, but can allow combustion gases into the water jacket (combustion pressures are a LOT higher than cooling system pressure.)
  • finally, there’s a leak from an oil passage to the outside of the engine.

this isn’t exhaustive, but it’s the most common failure modes in my experience.

FWIW my first thought on reading the OP was a possible head gasket.
A possible HG doesn’t ping my bs meter at all.
::: shrug:::
YMMV

I’m not sure how durable the head gaskets typically are on the 3.0 Mitsu engine (yes, the OHC 3 liter was made by Mitsubishi.) The one thing they’re notorious for is oil burning due to crap valve guides.

Well, I don’t have much experience with bad head gaskets, so maybe I should try that if a pressure test doesn’t show anything obvious. Like I said, limited experience.

Getting to the heads is a bitch, and I seriously suck at re-timing an engine.

:says bad words in preparation:

FWD cars usually have an electric fan instead of a fan running off the (what else?) fan belt. What is the Caravan, FWD or RWD? If it overheats on idle, slow traffic, etc. (but is better in fast traffic with airflow) then maybe the electric fan is kaput.

It’s the head gasket. It’s always the head gasket. Go ahead, pull your oil fill cap and just look at the greasy white gunkiness there. That’s what happens when you mix coolant/oil/heat. It’s always the head gasket.

If you filled the engine up with water and it ran OK for 2 days, then it’s probably not the head gasket…but it will be if you keep getting it hot. You can check the temp by quickly touching the top radiator hose. If the top one, after it’s warmed up, is really hot - like too hot to touch - and the bottom one is much cooler, then its the thermostat. A head gasket would take the water into the engine and blow it out the tailpipe, so there wouldn’t be puddles around the radiator. Puddles you can see mean you have a leak in the cooling system and you need to pressure check it and find them. If it’s something stupid like a loose hose clamp or a bad hose, you can fix it on the spot. If your waterpump or cap is leaking, you will need to spend some money. But until you find the water leak, you run the risk of really overheating it, blowing the headgasket and warping the head. Filling it up cold and looking won’t get it, as it needs the 10-15 psi to activate the leak. That’s all the pressure check is. And if you remove the sparkplugs for the pressure check, you’ll be testing the headgasket at the same time. Don’t wait.

Done. No white stuff. Top of the head looks pristine. No carbon, even. BIL says he put that engine in right before we got it, and it had been sitting for a while. However, when he bought several years ago it it had been just rebuilt. Sure looks like it. It’s very clean under valve covers. Will drain the oil and check for water.

For what its worth, the radiator is empty. Still don’t know where the water came out.

I was with you up to the end. Pull the spark plugs? How does that prove yea or nea the head gasket is any good? Not that I doubt you, just don’t understand the physics.

Incedently, I’m willing to guess that the water pump is ok. I’m told these particular pumps have a weep hole in the side, and a failing water pump leaks the coolant behind the engine. Unless the pump can remain tight and sealed and simply… not pump water. Don’t know if that’s possible. Unless the impeller somehow shatters, in which case you’ve just fubar’d the block.