K-series coolant issues/hot running

[Smithy77]

Hello POBC,

I have had my new Caterham (DVA tuned 1.8K, 185bhp, BP285H cams, emerald & Jenvey throttle bodies) for a couple of weeks now and have observed some abnormalities with the cooling system...

When I picked the car up it had an R500 triple pass rad fitted. The car was looked after at Tech 7 who put some Radseal in the system 2-3months ago to seal the rad which was weeping at the sides of the core. It also had an 82c thermostat fitted but it was drilled with 2x3mm holes. With this rad and stat fitted temps would sit at 82c while driving but I found at cooler ambient temps (say <15c) it would over cool dropping the coolant to 65-70c. Regardless of ambient temp, coolant temp would immediately and quickly rise the moment you stop, e.g. in traffic, up to about 96c and then the fan kicks in and brings it back down to about 92c.

Over the weekend we swapped the rad out for a standard k-series rad and replaced the thermostat with a new 82c stat with just the recommended single 2mm hole drilled. Only driven it once since, which was last night for a 2hour journey home with ambient temp of 16-17c, and I'm not sure if things have improved or if we have just given the same problem a different guise. It now doesn't appear to "over cool" in cooler temps; sat on the motorway at any speed the coolant temp sits at a very steady 96c - a bit on the hot side? I still also have the problem that the moment you stop (literally), you can see the temp gauge very quickly rise to the mid-90's - takes about a minute.

I understand these kind of unstable coolant temp issues are fairly common with k-series owners with no real known specific cause or solution, and some even just consider this behavior normal. Others observe a rock solid 82-85c in all conditions, perhaps suggesting this behavior is NOT normal.

Could I have partially blocked / inefficient waterways made worse with the radseal? Would a coolant flush (example) be worth trying?

Any other thoughts / suggestions appreciated.

 

[Jonathan Kay]

Similar thoughts to yours, except that I wouldn't be worried anything being blocked... yet.

My K (much lower peak power than yours) also now stays too cool on motorways. I think that blanking some of the radiator is the only solution to that bit.

Any problems with bleeding after that work, and have you checked the level again after the run?

The experiment I'd do is swopping the thermostats back...

Jonathan

[revilla]

@Pete: Are you sure you've got all the numbers right in your post?

You seem to have it at 96C when driving then "the moment you stop (literally), you can see the temp gauge very quickly rise to the mid-90's".

When the car was idling in my garage last night it went up to 96C at which point the fan kicked any very quickly brought it back down to around 92C, the fan then switched off and the cycle repeated. It didn't seem to reach this temperature alarmingly quickly and the temperature fell rapidly when the fan kicked in. This was fairly normal if a little on the hot side, which suggests maybe the thermoswitch could do with being replaced with a cooler variant - they're available with a range of set points.

Before with the triple pass rad you were overcooling on the move .., so what exactly is the temperature doing now when driving gently, at constant high speed like a motorway run and when stopping in traffic?

@Jonathan: As part of the work we did on the car yesterday (it was a mammoth session with Pete, Mankee and myself - we only got Pete on his way back to Southampton at half past midnight!) we redid all the plumbing around the bypass hose (long story - previous owners had installed an additional temperature for a Stack gauge in the bypass hose then added a heater leaving the sender downwind of the heater matrix and recording a 20C drop in temperature when you pulled the heater control) and a lot of the other plumbing too as we had to move the oil cooler and Apollo tank around a bit to accomodate the new radiator and they ended up in the way of the existing pipes.

We ended up with a pretty neat layout without obvious airlock traps, and we filled it carefully from the bleed tee until the radiator drain overflowed, then until the expansion tank level was correct, then to the top of the bleed tee. Filling went normally and on warm up there was hot water flowing as expected in both the main and bypass circuits. Plenty of hot water returning from the radiator once hot so I don't think there were any major airlock issues, although I'm sure some bubbles will come out and it may need a topup.

The bottom hose started to heat up markedly at about the expected opening temperature of the thermostat so I think that was working fine.

I personally would like to try to understand why some K Series cars seem have thermal issues at idle and some don't. Many people seem to consider this rapid temperature rise when stopped in traffic as being normal and end up installing fan override switches etc. Many others have cars like mine that will just sit at idle all day without any noticeable rise in temperature at all. My car runs at 82-85C. Always. Period. From a fast motorway run in winter to stuck in a traffic jam on a hot summer's day. Just doesn't move.

I'm tempted to start a survey just to see if there is any pattern as to which cars have overheating problems and which don't. The main differences between Pete's car and mine are i) age (so wondering if later castings had better water flow), ii) mine retains the VVC system so at idle has extremely mild camshaft characteristics with very little overlap, whereas Pete's has hot solid Piper 285H cams (so wondering if hot cams might lead to less efficient idle and more heat output), iii) mine has a standard K exhaust, Pete's has an external 4:1 (so wondering if more heat loss through the primaries just leads to higher underbonnet temperatures which compromise radiator performance when there is little airflow), iv) mine has the plenum chamber and a single throttle body, Pete's has Jenvey DTH TBs (can't see much relevance there) and finally v) Pete has an Apollo tank plus oil cooler (so wondering if the oil cooler could be heating the air around the rad, reducing cooling).

Fan was confirmed to be blowing air the right way through the radiator.

Anyone have any thoughts on any of this?

[Burnside]

Hi Smiffy7, 

Funnily enough I've had similar characteristics to that with mine since putting silicon hoses on, replacing the thermostat - my mate, not Dave! (the mechanic) drilled two holes in the thermostat (one by mistake as it was not the top! lol), and 50/50 mix of X Stream G50 coolant and deionised water.

Since this little change up the temp stays low at 60-70 niitially and when I sit still for a short time it rises but sits happily at 80.....anything longer i.e. traffic it goes up, naturally, and the fans starts kicking in thankfully! 

Not exactly the same as yours but similar in it's early warm up mannerisms. 

I am a little suspicious of an air lock but however header tank levels are constant? Need to get a C7 specialist to look at it to put my mind at rest. 

Paul 

[thompster]

I'm in the happy postion of keeping a rock-solid low 80's although I'm not quite sure why.

I have an 82C stat with a single 2mm hole, together with the race fan switch (73469).  According to the archives that will make the fan kick in at 86C and stay on until the rad cools to 76C.  In practise once the fan kicks in stays on pretty much on all the time.

With the fan cutting in so early I'd expect water hitting the thermostat to be cold enough to close it but according to the temp gauge sensor in the water rail it's just about spot on.

(for your survey purposes - 2007 EU3, supersport cams, standard 4:1 and inlet manifold, silicone hoses).

[Smithy77]

Coolant level dropped by about half the expansion tank during the journey home, well actually in the first 30mins and after that it didn't drop anymore, so just the expected bubbles being released. All looks water tight so far.

Andrew, it would probably help if I got my numbers right! I was supposed to write it sits at a constant 86c when cruising between say 50-90mph, then when I stop it quickly shoots up to the mid-90s. I might actually do the emerald fan switch mod. It will eliminate a potential fault source, and well, it just sounds cool.

I think the over-cooling I was experiencing previously was caused by a combination of both the triple pass rad and the fact the old 82c stat had 2 x 3mm holes. Now I've gone to a standard rad and stat with 1 x 2mm hole, it's like I've gone too far the other way, leading to my stable but slightly higher than normal temps. Maybe an extra 2mm hole in the stat and a lower fan switch setting would all that would be needed to make the coolant behave more "normally"?

I think now you mention the different exhaust, I sware this long primary exhaust kicks out a lot more heat than the standard k-series manifold. Perhaps engine spec could be a factor to rising stationry temps...

[revilla]

OK, having worked on your car over the weekend, here is my analysis:

I agree with your comments on the previous overcooling, the radiator was simply too effective when the water pump was up to speed (i.e. not idling) and the additional holes in the thermostat were hardly helping as it couldn't shut off properly. You proved the point by partially blanking off the radiator and getting a noticeable improvement. You didn't need a triple pass radiator for road use and the one we swapped in should be more suitable.

So you're getting 86°C in normal driving. You have an 82°C thermostat. The thermostat is however in the return from the radiator and is responding to a mixture of cooled water from the radiator and the bypass circuit. Basically it is measuring the temperature of the water entering the head whereas the temperature sender is measuring the temperature of the water exiting the head. So I would expect the reading on the temperature gauge to be a little higher than the rating of the thermostat. I don't know how much of a temperature difference between the sender and the thermostat is normal and it probably depends on a lot of other things such as the flow rate. Also bear in mind that there is a bit of tolerance on the thermostat temperature and the accuracy of the gauge. All in all I wouldn't say that the temperatures you are seeing whilst driving were too far off the mark.

The hole drilled in the thermostat is there for two reasons:

Firstly it allows any air trapped in the thermostat housing to escape as the coolant is filled. Without that hole, you can get an air pocket in the thermostat housing which results in the thermostat not opening (as it is insulated by the air and prevented from feeling the full temperature of the water) and without the thermostat opening the air is unable to escape, so you get trapped in a situation where the engine overheats but no coolant flows and the condition causing the problem is unable to dissipate. Just to recap the filling technique which we used was; with the car level we removed the radiator bleed screw and expansion bottle cap and filled very slowly from the bleed tee in the bypass hose, holding it up as high as possible. When water reached the level of the radiator bleed screw hole we replaced the bleed screw and continued filling. When water reached the correct level in the expansion bottle we refitted the cap and continued filling until it was full right up to the top of the bleed tee hole, then we capped of the bleed tee. The cooling circuit worked immediately as expected without any need to massage hoses or anything. You observed an initial drop in the coolant level as some final air was dispelled but nothing out of the ordinary.

Secondly it allows a little flow through the radiator even when the thermostat is closed, which helps to prevent the radiator from filling up with cold water which is then dumped into the block and head when the thermostat opens, leading to thermal shock and the unwanted cycling of the temperature that is seen on some cars. It evens out the temperature swings as the thermostat opens and closes leading to the much more stable temperature conditions which a K Series prefers.

It is most definitely NOT there to assist temperature control. In fact it degrades temperature control by robbing the thermostat of the authority to control the water flow. With a small hole the chances are that the only effect that you will see is a slightly prolonged warmup, but the more holes you drill the poorer the temperature regulation will become. The hole you drilled already is doing its job perfectly so there is no need to go drilling any more. Don't do it!

So the only real issue is the rising temperature when stationary. The temperature rises steadily (although I wouldn't say alarmingly from what I saw in our garage, mind you that was with the bonnet off so I guess it may rise more rapidly when closely cowled) to an indicated 96°C, at which the fan cuts in and drops it very rapidly back to 92°C. The cycle then repeats. The fact that the temperature drops away so swiftly when the fan kicks in suggests that the water pump is in fact circulating water adequately (so no need to worry about the previously used rad seal and blocked waterways), the two elements to the problem are then i) the whole cycle seems to take place at a slightly higher temperature than you would like and ii) the radiator doesn't seem to be able to get rid of the quantity of heat that the engine produces at idle without some air flow over it, which would suggest that the air around the radiator is getting hot and not being moved on.

As for (i) the standard 823-959-481F thermoswitch is rated 87°C-92°C which I take it means the fan should cut in at 92°C and switch off again at 87°C. So that would mean that yours is running about 4°C-5°C hotter which doesn't seem a big deal at all when looked at like that.

As for (ii) this is not really abnormal. OK mine seems to disperse heat better than yours, but mine is and SV and yours is an S3; there's a LOT more room for air to move around under my bonnet than there is under yours. You also have an oil cooler sitting behind the radiator and dumping heat out into the air around it. You then have four full length exhasut primaries all the way down to the side skin where I have a short collector and single downpipe, so bet your exhaust is releasing a lot more heat into the engine bay. All in all, I would say your underbonnet temperatures are probably noticeably higher than mine when stationary and the radiator will not be able to dump heat into hot air in the way it can into cooler air, so your temperatures will rise a bit faster than mine at idle, but the fact that the temperatures then fall so rapidly when the fan cuts in suggests that everything has the capacity to cool the engine as required once the hot air can be moved on.

The thermoswitches are available with a range of different temperature set points. I have a complete list of the part numbers and their corresponding temperatures if you want, although the standard part sounds about right and I don't think there is any need to tinker around with the target switching points. So one possibility is that you might want to replace the thermoswitch as new ones are not expensive, however ...

Just imagine that all of you temperature readings could be lowered by 4°C throughout. You would get 82°C in the cruise, the temperature would then rise to 92°C at which point the fan would kick in and bring it down to 88°C. Do those numbers look familiar? Yep, EXACTLY what you would probably expect from the thermostat and thermoswitch set points.

So is the most likely explanation that everything is working just fine but your temperature gauge is overreading by 4°C? To be honest that's not much of an error at all. You could very easily check the calibration of the temperature sender in hot water against a thermometer and you could also check the calibration of the thermoswitch in the same way, although both of those would mean a limited loss of coolant and the need to refill and bleed again.

If I were you, I would buy a non-contact infra-red digital thermometer from eBay (for example £8 here http://www.ebay.co.uk/itm/Handheld-Non-Contact-IR-Laser-Infrared-Digital-Temperature-Gun-Thermometer-UK-/381315287841?hash=item58c8298b21) and take some readings on the head, the water rail near the temperature sender and the radiator and see what it is really running at, and compare those with the gauge readings. If your gauge is indeed running 4°C hot then you've got nothing to worry about and it is all exactly as you expect. To be honest, even if it isn't, I don't think you have too much to worry about as those temperatures are not far off and not excessive or dangerous. If you wanted to trim your gauge a bit, a 5p resistor in line with one of those wires we extended could be used to trim it to read correctly at and around the mornal range of temperatures.

[Jonathan Kay]

I personally would like to try to understand why some K Series cars seem have thermal issues at idle and some don't.

Me too.

I'm tempted to start a survey just to see if there is any pattern as to which cars have overheating problems and which don't.

Please do.

Jonathan