Overheating(ish) and I'm Confused

I have a 2008 Mazda 3 S (2.3l, AT) that is overheating oddly.  Driving home last night, about 15 minutes into my drive, the temp started creeping toward the H, so I turned on the heater in the car, and the temp went back down. Cool. I made a three-minute stop and left the car running. Monitored the temp gauge, and all looked good.  As I left, the needle started moving a little north, but not bad. And then, it spiked up to the H and stayed there. Pulled over, and the overflow/fill tank was very full and gurgling (air?). A couple minutes later, I restarted the car, fan came on, temp gauge was happy, but the air coming out the fan (through the radiator) was cool. Drove the last mile home w/o further incident. (Ambient temp was ~50°F)

This morning, I drove it gingerly to work. As soon as it reached normal operating temperature, I turned on the heater, and watched the temp gauge. It never went to H. However, at one point, the heat in the car stopped being warm. It went back to warm after a minute or so, temp gauge never moved. (Ambient temp was ~30°F)

Return trip home tonight, temp rose to the H, but then moved down a tad, stayed there a bit, and then went back to normal. (Ambient temp was ~60°F)

The engine in the car was replaced in the beginning of August. New thermostat and water pump at the same time. Radiator was reused. I've been driving it w/o issue for two months. Coolant is clear and full. Fan works.
 

This doesn't make sense to me. Help?

TIA,

- Eric

If you are losing heat inside the car, it's not full, somehow.  It's not unusual to have a full overflow tank, with a low coolant level.  Is there a rad cap and overflow tank, or a remote tank with pressure cap?

Remove rad cap cold, and see what's up.

Could be a faulty new thermostat 

I’m thinking plugged radiator

You're low on coolant, that's why the heat is dropping off. 

No hot water in the heater core = no heat.

No hot water touching the temp sender = wonky temp reading.

With how much you've been driving it and the kind of overheat it had (boilover), I'd say you now need a head gasket and your head is probably warped.

Chances are, you had a blocked radiator which cause the overheat and then the boilover (burping in the overflow). Driving it without replacing the missing coolant has probably cause more overheating which didn't show up on the temp gauge because it had no coolant to read the temperature of.

Well, that's certainly the worst-case scenario there, Shawn. 

There's no radiator cap (pressure) associated with this rad setup that I can discern.  Just the fill/overflow tank on the L, and then a couple hoses to the rad. I would kind of expect that with this setup, any air in the system would end up in the bottle. 

Coolant is clear. Engine is not hot all the time, and cools back to normal temps quickly. It's possible that in the last couple of months that the temp spiked and I didn't notice, but I doubt it. When I witnessed it spike and the 'boilover' (no spillage or loss anywhere), I had pulled over and turned the engine off w/in 20 seconds (probably less, but it felt like forever to get to a side street). I seriously doubt that there is any damage.

I lean toward it being the radiator, but it still doesn't make sense to me.

Looks like I'll take it back to the shop rather than randomly throw parts at it. Time is of the essence.

Thanks, guys.

- Eric

Yes, it's the worst case scenario.

I've done it myself but not on that particular car.

It's also the best diagnosis I can give you without being able to put my hands on the car, based on the description of the symptoms you gave.

I certainly hope it's not that bad but modern engines don't handle an overheat well. Hoping the shop finds something simple and you're not stuck with a huge repair bill.

Good luck!

Get all the air out of the system and see what happens.  The not replacing a rad when you replace a motor is something that voids the warranty on most motor replacements.    

I would agree it sounds like a head gasket.  but it could just be a small leak that is letting air in to the system.  I would pull the plugs and pressurize the cooling system and see if it holds pressure (let it sit for a while overnight is better).  If it does not hold pressure then look in the cylinders for coolant with a borescope or just spin the motor over and watch for coolant to spit out the plug hole.  IF you don't have coolant in a cylinder then you need to be looking elsewhere like the rad or the heater core or a small leak someplace else.  

Fortunately, the engine install was done by the shop... They may well have replaced the rad, I just didn't see it listed on the invoice when I looked it up last night.

I sure hope there's no "big repair bill". I bought the car used from a private party at the end of July. 12 days later, it threw a rod out the block. Shop ordered/installed a rebuilt 2.3 — and was confused when parts wouldn't fit from old to new (oil pan, etc.). Turned out the car had a used 2.0l installed earlier this summer.  So, right now I'm into my $3000 car to the tune of $7500. But I've got a 3-year drivetrain warranty now. 

*Correction: After a new transmission control unit, and new shocks/struts, I'm up to $8250. * 

I dont know if this is helpful because it's a small chance that is more likely to make you paranoid than be true. One time i installed a rebuilt engine (myself, i was working in a shop) and it came back overheating within a month or so. The root cause was that the (new) water pump impeller broke loose from the shaft and was able to spin on it.

That possibility had occurred to me, Vigo. It would help explain the works/doesn't work aspect of the cooling.  

Just a quick follow-up: The mechanic was only able to find a pinhole leak in the fill/overflow tank. They replaced it. I still find it difficult to believe that that's the fault. But, no heat spikes on the way home...

I'm now $8400 into my $3000 Mazda 3.  At least it drives way better than the loaner Prius I was saddled with for the last couple days.

@ShawnG, it looks like your dx may have been spot-on. Engine continued to spike in temp (was using ODBII tool to monitor temps) and I brought it back to the mechanic. The shop found a small leak for the coolant feed into the throttle body(?), and had their employees put a bunch of miles on the car w/o apparent issue. Temp spiked and heat stopped w/in two miles of my getting it back. I opened up the rad cap 3.5 hrs later, and coolant spewed out — coolant was under pressure.

Having a Subaru that's eaten three sets of headgaskets, I expected that i'd see the usual signs: oil milkshake, steam from the exhaust, etc. Other than the hot engine/no heat, there weren't any visual signs.

Anyway, repairs/replacement is occurring under warranty...

Had a very similar situation with a bad water pump on an old suby.  Did EVERYTHING. HG, rad, thermostat, rad cap.  Gave up and got another engine.  While changing over the timing belt the impeller fell off the shaft.  It was a year old and was progressively loosening on the shaft so the overheating got worse and worse.  Would idle cool as a cucumber but never put heat into the cabin.

In reply to Crash Enburn :

Crappy news but I'm glad it's getting taken care of in the end.

I had an old Jeep pickup with a questionable cooling system that would give me those exact symptoms when it wanted some attention.

I would know that the coolant was getting low because the heater would stop blowing warm air. Heater core was the highest point in the system. Temp gauge would read low because the bulb wasn't actually touching any coolant. If I ignored it, it would boil over.

Wasn't much of an issue for me because the engine was all cast-iron.