I have a 1993 316i Auto Lux (M40 engine) with 87,000 genuine miles. I have owned th evehicle for since Sept 03.

My car overheats on warm days in stationary traffic. I have  :-

1. Replaced the fan clutch without success.
2. Bled the system from the stat housing several times.
3. Did a basic rad flush with rad in-situ

The car is defintely overheating cos both top and bottom hoses get very very hot and on one occasion it did slightly overboil. Otherwise the car runs beautifully without any oil use and the power is ok.

However (this may be be a clue and I was wondering if anyone with a similar car could tell me if its normal with their car) on the very top quarter of the rad is getting hot, towards the bottom is cold and slightly warm in the middle.

Is this normal ?? If it was an air-lock in the rad surely the air should rise to the top of the rad and hence only the bottom of the rad should be hot and top cold.

Any clues or sujestions would be very very welcomed.

Many thanks

Dal

My appolgies, I forgot to mention that I did replace the pump as well, again without success.

Sorry

Cheema,

The radiator's working fine, hot water enters the radiator at the top, through gravity and pressure it is passed down through the radiator, as it runs down it is cooled in the core, so you should have hot at the top and cool at the bottom.

Are you getting flow through to the heater? Does it stay hot when the engine is warm? Other option could be the radiator cap, maybe not allowing the system to vent? You say the hose is hot, does it reach red on the guage?

Themostat changed?

Rob

Hi Rob,

Many thanks for your reply,

Doh, I should have thought of that, ( I just bought some rad flush this lunch time) the rad should be cooler at the bottom. The stat was replaced about two months ago, since then I have removed it and checked its operation in a sausepan with boiling water and a termostat. Everything with it is ok when I tested it.

The rad does stay hot when the engine is warm (again only at the top). The hose is very hot and the temperature guage does hit red on a hot day.

You mentioned the radiator cap 'not allowing the system to vent'. I don't understand what you mean. I didn't realise that the radiator cap is supposed to do anything when working normally. Is it worth me trying a replacemement.

Many thanks

Dal

Far in excess of 80% of a cars cooling process comes from the draft caused by forward motion.  The fan assists this process when idling in traffic.

Hoses do get too hot to touch in a perfect system.  So don't use that as a guide.

It could be the guage or the sender unit.  If it is *actually* overheating, and you have changed all the possibles...it could be a problem.  Are you running Super U/l fuel?  That can add a few degrees.

Other than that, I can only hope it's a blockage or an air-bubble.  Blockages will be in the most inaccessible areas, like the radiator or the heater matrix...where the pipes get down to small diameters.  Flushing with  a hose-pipe will not show the problem, as the water will simply pass through the clear bits.

Old shape wrote:

'Far in excess of 80% of a cars cooling process comes from the draft caused by forward motion.  The fan assists this process when idling in traffic'.

This is why I said it sounds like a fan problem. All the air flow over the radiator is caused by the fan when the car is running but stationary, so if it overheats then it is because the fan is not causing sufficient airflow. Surely if anything else was at fault it would show up when the car was moving as well.

Hi all thanks for the responses.

I have done the newspaper test, but I'n not sure on how many sheets of paper to use and how tightley to roll it up.
last time I rolled up about 3/4 of The Sun newspaper and I was able to stop the fan when the engine was quite. The fan did not tear the newspaper in the process. I bought a new fan clutch about 4 months ago from GSF. I did the paper test then and was still able to stop the fan, but then winter set in and abviously the car was not warming up as much in stationary traffic.

It does defintely overheat, I can safely dismiss the temperature guage being faulty.

If the Viscous clutch is knacked, the fan should still spin of course, (Unless there is absolutely no friction left in the clutch - which I have never heard of) and even that basic spinning should prevent overheating to the red-line.

I bet it's a gunged up radiator.

Hi,

Just a few quick pointers for you.

The fan will not work if the centre of the rad does not get hot enough because the fan control is via a bi-metal unit mounted on the centre of the fan clutch. In such a case you will be able to stop the fan. A blocked rad will usually cause overheating at speed more so than at idle. You mentioned bleeding from the stat, there is a bleeder beside the rad cap which is most important(i assume it is an E36 since it's 93). This is a black cross head that sits flush with the rad body, be careful not to overtighten it. At idle the fan should keep the rad cool which makes it hard to tell if the rad is heating up all over. Remove the fan and run it. Once the stat opens the rad should heat up all over and quickly. The fact that the coolant seems to circulate well in the rad means nothing as it will do with half the core blocked. Disregard the rad cap, this is only a safety feature which is designed to blow off excess pressure when overheating and will only cause problems if it leaks so the system cant pressurise as it should.

HTH

Alan. 

Hi Alan

Thanks for those pointer, I'll run the car with the fan removed and see if the whole of the Rad warms up.
Its actually an E30 Touring 1993 (one of the last ones from the factory). It has a black plastic bleed screw on top of the rad, I refilled and bled the system from both bleed screws witout any joy.

I have heard some bad things about GSF parts (where I bought my clutch fan from) what do people think about the following findings.

According to the packagaing on the box the the viscous is only supposed to be stored in one direction. When the engine is cold (1st thing in the morning), if I try to give the fan a stong push with one finger it spins at most half a revolution, if do the same spin test with the engine fully warm the fans spins 4 or 5 times. I thought it was supposed to be the other way round, ie harder to spin when hot.

Could it be a crappy aftermarket fan from GSF ??

Many thanks Dal

Hi,

Very roughly, this is how it works.

The extra resistance felt when cold is for two reasons. The viscosity of the oil changes as it heats up and the oil is driven from one chamber to another as ithe unit spins. The valve between these two chambers is controled by the bi-metal strip on the front of the unit. When the bi-metal strip is activated by the temp at the rad centre, it opens the valve and allows the oil the enter the first chamber again, therefore raising the resistance so the fan is driven quicker.

It is unlikely that you have a faulty unit. Most times I have seen this, it has been a rad fault.

Another thought, I dont suppose the fan cowling is missing?

HTH

Alan.

Here's the reference I was on about...

"Viscous coupled" fans are a bit of a misnomer. The units usually descibed
as such have two operating modes, controlled electrically. In the "normal"
mode the fan is free to spin on the shaft, but a fluid-coupling between the
fan and the shaft cuases it to spin at a speed lower than the water pump at
higher revs - there is a lot of "slip" in the coupling. The fan is in this
mode for most normal operation and as such consumes less engine power (and
thus fuel) than a rigidly-coupled fan. If the engine coolant temprature gets
above a certain point (usually set at the higher end of the "normal"
temp-range - say 120degC) then a large magnet in the front of the waterpump
is energised, causing a mechanical clutch to engage and rigidly couple the
fan until the coolant temp is within the desired range again. This usually
only engages when stationary with the engine running for long periods
(traffic jams) and in SOME cars it might engage after prolonged periods of
full-power running. As a result it can be difficulty to tell if the
direct-coupled mode is working - the only real way to do this is to pull out
the relevant relay and make up a shorting lead to connect the "normally
open" contacts in the sensor circuit - make it a shortiong lead with a
switch and you can switch it in and out while the engine is running and
watch to see if the fan starts turning. Some suggest that such tests are
best performed while the car is parked with the handbrake on and the gear
lever in the "neutral" position, but personally I like the
"wind-in-the-hair_ feeling of openning the bonnet and doing such tests
whilst driving on motorways - it's a matter of personal preference...

Sounds more like a rad problem then a fan problem to me! quite common for a rad 2 get gunged up especially if the previous owner didn't look after it! Not sure if this is true, but apparently if you live in a hard water area, you're more likely to get gunged up coz the expansion tank is normally filled with chalky tap water which can settle in your rad!!! Whether that's a housewives tale I don't know...But if it anything like my kettle...???

Hi OldShape,

This BMW viscous fan is not electrically operated. The article you have posted would be acurate for other makes, like Mercedes.

HTH

Alan.

Cheema,

I just bought a viscous coupling from GSF, SACHS part, no problems at all with it. Does the temp drop if your parked but then raise the revs? If so then the fan is definately suspect but you have a new coupling...hmmm. Remember, even if the engine is fully warm, if the air through the radiator is cool enough the fan will spin freely, it's only when the air hitting the fan is hot it will activate the viscous coupling and engage the fan, it does this in cycles as the temp rises and falls. 

I am sorry about the comment about the rad cap, It should have been, if it is faulty and won't allow the pressure to build correctly in the cooling  circuit then the water will boil at a lower temp therefore it will overheat quicker. You can also check the temp with a remote thermometer, one of those laser thingys, on the hoses, will tell you actual temp the water is at, just 2 be sure?

The other trick, if the fan is suspect, is to convert to an electric fan (Escort Turbo MK1 fan is a goody) on the grill side and then it becomes a blow through rather than a suck through, smutty comments on a stuck down envelope please    Then it's just down to the sensor or the fan and you know when it's working or not and you don't lose the couple of BHP the fan can use.

I would start with the radiator out and flush with water both ways and see if you have residue. Seen this more than once. Black cat is correct as well, in a hard water area the radiator can end up like your kettle without frequest flushes!!!  Check the core as well for dirt and crap jamming up the fins. Wouldn't be the first to see a core that was 3/4 blocked and next to no flow through the core. 

Rob

Hi Rob,

Thanks for your reply, it didn't undertstand the "Does the temp drop if your parked but then raise the revs? " bit in your reply.

I'm going to try and flush the Rad this w/e with some good quality flush. Whats confusing me even more is that my car is fitted with what appears to be two bleed screws, one on the stat and one (black plastic) on top of the Rad. If I follow the instructions as described in the Haynes manual I don't need to touch the bleed screw on the stat.

Has anybody successfully refilled the cooling systems on an M40 engine (without any problems) by following the Haynes instructions word for word. How do yo ensure the system is bled of all possible air-locks. Would be most interested in finding out. I've even been advised to put the fron of the car on ramps before filling the system (weird), I don''t know why/

Thanks

Dal

Hi Dal,

Yhe M40 engine is easy enough to bleed. I have never seen a Haynes manual, so cant comment on that but this is how I would do it.

Move the heat setting inside to hot with the heater fan off. Open the stat bleeder and remove the filler cap. Remove the rad bleeder. Fill with water until it comes out the rad bleeder, then replace the bleeder but only tighten 2 turns. Run the engine at idle and fill with water again.  I would use water as its cheaper than coolant and there is no point in wasteing coolant at this stage. As the engine heats you will see air/water from the bleeders. close them and leave the engine running. keep an eye on the temp, as long as it does not boil it is ok to get as hot as it likes. As it heats, watch the temp of the bottom hose. When the stat opens this will heat up in a few secs. At this stage open both bleeder again to allow the air which was behind the stat to escape. The one on the stat only needs to be opened for a few secs, the one on the rad for a few seconds longer. Be careful not to overtighten the rad bleeder they only need a pinch to seal. You can open either bleeder a couple of times for a few seconds each if you like to be sure. At this stage put the cap back on on drive it for about a mile. Then let it settle to allow the pressure to drop and carefully open the cap. Top up if needed and run for a few mins more, opening the bleeders a couple of times for a few secs each.

You will now be ready to check it for overheating. If it is ok, all you have to do now is drain the system, refill with coolant and bleed it all over again :) 

Bear in mind that a blocked rad will make bleeding more difficult.

HTH

Alan.

Thanks very much for that info Alan, Ill give it go an post some feedback on how it went.

Cheers

Dal

Cheema,

The comment was regading the check on the visous fan. If the car is sitting at rest and starts to overheat, if you raise the revs and the temp starts to drop then it is the fan that may be causing the overheat at idle, usually. If not then a potential blockage/airlock. Autofix's bleeding tips are just about how I do it, minus the radiator bleed, I don't have a bleed screw on my radiator. Just a drain at the bottom. I wouldn't totally trust Haynes as the manuals can be notoriously inaccurate though.

Hope you get to the bottom of it mate.

Rob

Many thanks for everyone who helped me solve this overheating problem. Basically I think the trick is to fill the system really really slowly, leaving both bleed screws fully opened, wait till you get a constant flow of water (without air bubles) through the stat bleeder screw. Tighten the stat bleeder screw. Then continue pouring the water very slowly till it comes out (with out bubbles) from the rad bleeder screw, then tighten the rad bleeder screw. Run the engine and continue bleeding from the rad 2 or 3 times.

and Bingo all air removed. Temp guage no longer goes over the 3/4 way mark no mater how long it's idling.