m43b16 mysterious overheating

[kukkakeppi]

Hi

So last fall I bought an e36 equipped with m43b16 engine as a sliding missle. However, few months back I noticed engine running pretty cold and heater not blowing hot air (it was about -15 °C outside). I figured maybe thermostat had failed in open position. And that was the case, one of those metal clips holding the thermostat in its casing had snapped causing the thermostat to rest in incorrect position allowing the coolant flow all the time. So, I replaced the thermostat and thought that problem was solved. But now the problem had turned the opposite, now the engine was overheating. The temperature cauge rised to normal (in the middle) and rest there for a while. But then it kept rising up to almost to the red. (I've heard there's some kind of built in retarding system in e36 instrument cluster which causes the temp cauge to rest in the middle position for a while even though the engine temperature keeps rising linearly).

The upper water hose gets very hot, but the lower hose stays very cold (matches pretty well the ambient temperature according to ir-thermometer).
When idling for a while, the lower water hose starts to heat up slowly (not sure is that due to heat conduction in water inside the radiator). But however the coolant seems to be flowing very poorly.
But the heater works fine. Keeps pushing hot air like it should. So I figured maybe the waterpump had failed this time or were already in weak condition. But it seemed to be fine (metal impeller), so I replaced only the seals.
I've also replaced following parts:

-radiator (the old one was leaking a bit), expansion tank cap (probably 3 different ones, maybe 5 different bleeding screws D)
-plastic water outlet on the back of the engine (old one leaked too)
-waterpump checked (as told above)
-thermostat (tried 2 different ones, meyle and topran so it can't be stuck in closed position)
-both temperature sensors (so the temp cauge shouldn't be lying)
(all of these parts brand new)

I'm also aware that these engines are sometimes pretty difficult to bleed. So I've introduced myself to several methods, and any of them makes no difference so I believe problem is not bleeding related (and improper bleeding would cause the heater not to work as I've understood).

Now I'm running out of ideas. It seems like that water circulates only between the engine and the heater, and it pushed hot water to the radiator, but doesn't accept cold water in (instead keeps circulating that same hot water) I've thought about headgasket, but even a bad headgasket shouldn't prevent coolant from flowing? And there's no white smoke nor coolant loss so I think the headgasket is fine.

But any ideas what to do next?

[Eric93se]

I think you should have replaced the WP while you were in there, its only ~$30. I know its metal but its failure modes are different than plastic, it can also slip on its shaft. You should also test your coolant with a coolant hygrometer, also if it's more than 3 yrs old it should be replaced.

You went through 3 tank caps in what amount of time??

I have the composite WP for a long time now and its been fine.

[samy01]

Are you sure you put the new thermostats in the correct way around ?

[kukkakeppi]

I think you should have replaced the WP while you were in there, its only ~$30. I know its metal but its failure modes are different than plastic, it can also slip on its shaft. You should also test your coolant with a coolant hygrometer, also if it's more than 3 yrs old it should be replaced.

You went through 3 tank caps in what amount of time??

I have the composite WP for a long time now and its been fine.

I inspected the waterpump pretty thoroughly and it wasn't slipping on its shaft. And yeah, I know that new waterpump wouldn't have been that expensive, but I need to keep in mind that I can get a like working m43 engine for about 50E so investing "that" much in part that seems to be fine is not worth it.

And the caps. I've just tried 3 different ones since my friend had those lying around (brand new though, but not OEM)
And the coolant. I've been through probably 40 liters of coolant while trying to get that system work, so coolant has been changed many times in short period of time

- - - Updated - - -

Are you sure you put the new thermostats in the correct way around ?

It's not possible to mount the thermostat incorrectly since it's integrated in its housing and the housing is bolted to engine with 3 bolts (one corner empty) so also it mounts only one way.

- - - Updated - - -

[samy01]

Okay well if you get heat to the heater but not to the radiator it means the big coolant circuit isn't working. Usually that's because of the thermostat, but you say it must be working since you replaced it. I assume the waterpump works, otherwise you wouldn't get decent heat out of the heater inside the car.

If i was you i'd let it warm up, disconnect the upper radiator hose and see if the pump is pumping out fluid while at the same time constantly filling up warm water from a bucket into the expansion tank to compensate for the lost water (i'd simply use the garden hose but someone surely will complain that the cold water will "shock" the engine, but i wouldn't care on a 50 euro engine). Then you at least know that your pump is working. If yes, i'd remove the lower hose and see if water can pass through the radiator freely.

[Eric93se]

The pump may only be failing when the car is hot.

Unhooking the upper heater hose when the engine is running and warmed up is a retarded idea (sorry samy). The engine has to be really hot for the thermostat to open and fluid to flow.

- - - Updated - - -

Can you remember if the current wp had splines on the shaft or was it a smooth shaft?

[samy01]

Okay i've done it before and it helped diagnose a bad water pump. i'd also replace it but he doesn't want to, so what else would you then do to diagnose a water pump ? Who says you can't do that when the thermostat is open and the engine fully hot? I actually said that's a requirement for the test, and now you turn it around as the reason why it won't work? I can't follow your logic. But it must be because i'm retarded. The thermostat won't close so fast, that's why i said compensate with warm water. That way you would diagnose if it is failing only when hot. Not my problem if you don't understand. But on the other hand you're fast to use strong language.

Think about what words you use and don't excuse yourself right after saying it, it makes you look like it isn't your responsibility that you said it, only cowards do that.

[Eric93se]

Hey I apologized in advance lol. It's a really really bad idea, to unhook a hose with 200+F hot water to come shooting out of the top hose and what? gush all over the engine, give the guy doing it 2nd degree burns. And how does one determine there is sufficient fluid flowing?? Compare to what benchmark lol.

[samy01]

actually you're right, it does sound like a dumb idea. i did that once and it kinda worked and yes it was hot a f, but i replaced the pump anyway. was more like a side of the road kind of repair. There's nothing here to do if he doesn't want to invest in a new pump. he should just replace the pump.

[kukkakeppi]

actually you're right, it does sound like a dumb idea. i did that once and it kinda worked and yes it was hot a f, but i replaced the pump anyway. was more like a side of the road kind of repair. There's nothing here to do if he doesn't want to invest in a new pump. he should just replace the pump.

Yeah that sounds like a bit dangerous, maybe I'll try that some other time
Well apparently I've no choice but replace the WP. I just don't believe myself that's the problem, let me tell something more:

While bleeding the system and mounting a half cut plastic bottle on top of the expansion tank filled with coolant and giving some gas the level of liquid in the bottle drops, and then rises back. I think that is because the WP circulates the water faster on high rpm causing the hoses to expand and hence accepting more coolant temporary. But it can't be going through the radiator since the engine is still cold and the t-stat should be closed.

I did also more some interesting testing yesterday. I started thinking maybe the problem is however within the t-stat. So I swapped the t-stats since I've two of them (Topran and MEYLE). While using the MEYLE one there seems to be no circulation at all (to the radiator, heater works always correctly). The engine temp would rise all way to the red if the experiment not aborted. And even when idling, the lower water hose stays cold. But the Topran one is more interesting. The temperature cauge rises almost to the red but still not into the red. I could drive around, the temperature doesn't rise more. And when idling, the lower hose starts slowly to heat up just like it should.

So apparently there is some circulation with the Topran one. So I compared those t-stats side by side in my kitchen and tried to figure out what's the difference between them. The Topran one seems to let some water through even when it is in closed position. And the MEYLE lets barely any coolant through when in closed position.

So maybe the thermostat stays in closed position all the time even though the engine is hot? How is is supposed to even work since it's mounted in lower waterhose, so when the t-stat opens the cool water enters and closes it right away. And what is the purpose of the smaller plate in the end of the thermostat? I'll write more in the evening with some pictures when I'll have more time.

[samy01]

the level of liquid in the bottle drops, and then rises back.

That sounds like you still have air in there somewhere. the air compresses and then bounces the fluid back.

[Eric93se]

If the thermostat doesn't have a small hole at the 12 o'clock position you can drill a very small hole. This will allow trapped air to be released. It may be difficult to do with your housing, you would have to support the back side when drilling.

When testing the thermostat in a pot of boiling water realize it may not be hot enough to make the thermostat open. You can try adding a large pinch of salt to the boiling water.

[Eric93se]

Take a pic of what you were using that connects to the reservoir when filling, I'm curious. We should get a good design that we can 3D print.

[samy01]

When testing the thermostat in a pot of boiling water realize it may not be hot enough to make the thermostat open.

A thermoastat usually opens between 176°F - 197°F, and water boils at 212.

[ThatsNotOregano]

Just changed the wp a few hours ago. Temp warning light went off last night. Car had the same symptoms as the op's after replacing all pipes/hoses, thermostat, and rad, but also decided to pass on the wp. Changing it was pretty nasty, but found the metal impeller was not spinning along with the shaft as it has discoloration on it. Bearings were fine. Used a carquest part with metal impeller. All seems fine now. Gonna rebuild the heater valves next when the parts come in.

[kukkakeppi]

Just changed the wp a few hours ago. Temp warning light went off last night. Car had the same symptoms as the op's after replacing all pipes/hoses, thermostat, and rad, but also decided to pass on the wp. Changing it was pretty nasty, but found the metal impeller was not spinning along with the shaft as it has discoloration on it. Bearings were fine. Used a carquest part with metal impeller. All seems fine now. Gonna rebuild the heater valves next when the parts come in.

Well apparently I'm left with no options but replacing the WP. I just don't know, how the WP could've failed so suddenly when the problem in the first place was a thermostat stuck in open position.
And I've tried squeezing the hose while revving up, I can feel how the upper water hose expands.

[kukkakeppi]

That sounds like you still have air in there somewhere. the air compresses and then bounces the fluid back.

Well if that's really air then I need to admit I can't remove it. The coolant's been out so many damn times (like at least 10 times, believe me) including removing the waterpump and intake manifold so I don't know how to get rid of that air. I've always bleeded the system by jacking the front up and pouring the coolant in until it comes out from bleeding screw. Then I've brought in that half-cut cola bottle and mounted it into radiator. Then squeezed both houses, a vast amount of air bubbles comes out.

[kukkakeppi]

Take a pic of what you were using that connects to the reservoir when filling, I'm curious. We should get a good design that we can 3D print.

Well it's pretty simple it is just a regular cola bottle with the bottom removed. Then applied some electrical tape to the other end to make it seal. It fits pretty well in e36's reservoir

It looks something like this:
Attachment 671972

[samy01]

i don't know why you bleed it like that, but i'd go with the standard procedure (google it) and don't use that bottle thing strategy.

[Eric93se]

If the pump is messed up then you will never get an air pocket out.

[ThatsNotOregano]

Well apparently I'm left with no options but replacing the WP.

FWIW I replaced mine because there was no previous history. Found out the hard way when the housing was seized to the block.

If the pump is messed up then you will never get an air pocket out.

+1

[kukkakeppi]

i don't know why you bleed it like that, but i'd go with the standard procedure (google it) and don't use that bottle thing strategy.

Because I've found that standard procedure leaving still a vast amount of air in the system. Using that bottle I can get that almost 0.5l of coolant more into system.

- - - Updated - - -

FWIW I replaced mine because there was no previous history. Found out the hard way when the housing was seized to the block.

I'll replace the WP then. But I've taken mine out once and in my opinion it wasn't that hard. Removing that visco flect pulley was a bit tricky but got it out easily by using and old serpentine belt. The WP itself got out by mounting two bolts in those two empty threaded holes, which pushed the WP out without any problems.

[BimwadM42]

I've flushed the cooling system on an M42 more times than I care to remember, including draining the block, and have not had an issue bleeding the system by following these steps:

1) HVAC controls full hot; ignition ON (electronic valve)

2) open radiator bleed screw

3) fill expansion tank until it is full, without spilling over

4) start engine, and add coolant as allowed as it is drawn into the engine

5) allow engine to reach operating temperature, again while watching level in expansion tank and adding coolant as needed

6) once up to temperature, with thermostat open, and upper hose clearly hot, rev engine until a mostly solid stream of coolant and few bubbles emerges from the bleed screw; return engine to idle.

7) snugly bleed screw, but do not over tighten (!*), turn engine off.

8) button everything back up, go for test drive, and allow engine to cool

9) top off to the Cold/Kalt line as needed.

Every time I've done this, I've gotten within the ballpark of the stated 6.5L capacity, and no other issues.


Assuming everything else is in order, and given the number of new parts, the logical fault would lie with air in the system, or a faulty water pump.

Unlike the six-cylinder engines, the cooling systems on the fours don't have nearly as many issues. I've replaced a radiator that was original to the car, and recently replaced a water pump that was installed in 1996, with 90k miles. I had some apprehension that the pump might be reluctant to come out, but it went very smoothly. I guess that's the value in regular maintenance, and carefully extracting it by slowly and evenly tightening the bolts, and lateral taps on the horizontal axis to apply even force around the whole of the pump. It was, BTW, an OE BMW pump with a composite impeller, which looked pristine. It needed to be replaced because the bearing was worn and it was weeping from the shaft.


*The bleed screws are fragile, and only need to be tightened enough for the rubber o-ring to seal. Any tighter, and they can strip, and depending on the type of screw/plastic used, you'd better hope it's the screw's threads, and not the radiator's.

[kukkakeppi]

Thank you a lot for this. I'll try this method as soon as a new WP's been arrived and installed. But I've still got few questions regarding your manual:
3) wouldn't the coolant spill out from bleeding screw since it's been removed?
4) how quickly will the engine suck coolant in? Is there enough time to start the car and rush pouring more coolant before the level drops too low and it sucks air in?

And overall running the engine without reservoir cap. Aren't these cooling systems supposed to work pressurized? (When coolant warms up, it expands causing some pressure--> more pressure--> higher boiling temperature). And without that pressure is it possible for the coolant to start boiling which produces more air in the system?

[samy01]

3) wouldn't the coolant spill out from bleeding screw since it's been removed? That's the idea, you want to get the air out of there, so some coolant will come out there as well.
4) how quickly will the engine suck coolant in? Is there enough time to start the car and rush pouring more coolant before the level drops too low and it sucks air in? You'll have enough time.

And overall running the engine without reservoir cap. Aren't these cooling systems supposed to work pressurized? (When coolant warms up, it expands causing some pressure--> more pressure--> higher boiling temperature). And without that pressure is it possible for the coolant to start boiling which produces more air in the system?You don't have to let it get that hot for bleeding.

Don't make it so complicated, just do it.