Sorry, another K-series cooling worry/paranoia - SORTED

[Smithy77]

Hi all, will try and keep this short and sweet, as far as possible.

I have an original Minister built R400 K-series engine which the previous owner dropped into my car a few years ago. He later got it DVA tuned to produce 230bhp@8000rpm. I bought the car in Feb 2016 and had the engine rebuilt by DVA in the summer. Before the rebuild, I noticed it overcooled quite badly with ambient temps <15C. This was with a totally standard cooling system: no heater; standard bypass; standard ali K-series rad, 82C stat with a large(ish) jiggle valve in (think it had a Ford stamped somewhere on it). Fitting a new QH 82C stat with 2mm bleed hole drilled made no difference.

Fast forward a few months after picking up the rebuilt engine. I decided to make a few cooling circuit mods to address the overcooling. The radiator was also weeping so I replaced this with a Coolex radiator, I'm told built to the same spec as the standard Caterham K-series rad, and I have re-used the old 9" Spal fan. I also decided to go down the Rover PRRT route: fitted a brand new genuine item (82C with soft pressure spring, as recommended) in the bottom hose between rad and submarine pipe; fitted a 32mm T in the top hose with 32mm bypass pipe to the PRRT; fitted a gutted stat/polo ring in the OEM stat housing to replace the removed OEM stat; and finally blanked off the old 16mm bypass between top water rail and stat housing. The results of this work on the road were fantastic! Now I have rock solid temps of 75-80C (indicated) in all conditions and all ambient temps.

Now here is the "issue". Went on my first Caterham trackday at Castle Combe in October. Bare in mind this was my first time on track (with my own car) and first time owning such a high powered K-series, so I have little experience to judge this on. I was finding that after a spirited 15-20min session on track, and after a cool-down lap cruising round in 6th gear, once I got into the paddock the rad fan would stay on at idle permanently. I fully expected it to stay on longer than normal as heat soaks out of the cylinder head, but I'm talking about it idling for 15-20mins without the fan switching off. In the end, I don't think the fan actually switch off before I switched the engine off. Does this sound like normal behavior? If not, could I be suffering from an inadequate rad fan or even poor rad flow at idle? It should be noted that whilst on track, temps seemed perfect with water and oil never going beyond 85C indicated.

I did have a minor gremlin (possible red herring) with the new cooling set up at first (before the trackday) whereby at idle, it showed the signs of little or no coolant flow through the rad. This was evident by coolant temp at idle rising up to 95-100C without the fan kicking in, but if you raised the revs to as little at 1300rpm, temps quickly began to drop and the fan kicked in. I put this problem down to two things: 1) the idle speed of 900rpm was too low for use with a PRRT, and I have since raised it to about 1050rpm, and 2) I noticed I was finding an airlock at the top of the rad, despite always bleeding properly when filling. When this appears I bleed it out by blowing coolant through the expansion bottle before tightening the bleed screw. With these two issues sorted, the system behaves exactly as expected at idle, with the fan switching on at 92C and switching off after a minute or 2 when the temp is down to 86C. No increase in revs is required.

I'm slightly concerned about the air lock issue, but not overly just yet because I have not kept the cooling system closed for a long enough period to monitor it properly yet. Replacing fiber washers to drain plugs, swapping a used PRRT for new, and generally little usage in between has meant the air could just be a result of residual air from filling the system finding it's way to the top of rad. I'll keep a close eye on it though as I'm aware this symptom can often be the result of a dreaded HGF. On the plus side, during the track day, coolant level remained very consistent and only rose by 10-15mm after a spirited session, which is normal. I know a few members have experienced coolant blowing out of the expansion cap (along with airlock in the rad) due to the onset of HGF.

Anyway, sorry for the long post. Any thoughts or advice greatly appreciated. 

[Jonathan Kay]

Have we read this before, or was that another one that was similar?

What was the indicated water and oil temperature when the fan seemed to have been on for too long?

It's going to be a bit hard to diagnose unless you can replicate the suspicious behaviour. When's the next track session? How about buying a new fan sensor/switch ready to swap?

Jonathan

[Smithy77]

JK, thanks for your reply. I may have posted previously about the overcooling issue separately  You will have also read many other similar sounding K cooling issues on track, but I've not seen anything like this specific issue myself.

I think the indicated water temp was about 95C whilst idling back in the paddock, not sure about oil. Also measured temps of the coolant rail and top of rad using a laser thermometer and both were reading in excess of 100C but <105/110C from memory.

I've since done some extensive testing on the thermo switch using the laser thermometer and it behaves perfectly, so we can rule that out.

Next time on track will be end of Feb earliest.

[Smithy77]

I am starting to think perhaps the PRRT is not the ideal set-up for a track car. There is a pressure relief valve in it which controls the bypass flow, and this is the only thing which controls bypass circuit and it is modulate by pressure alone, not temp. It also has 4 small holes in it so even when closed there will still be bypass flow. As pressure is exerted through the bypass, it pushes open the bypass valve. The thermostat part of the PRRT controls only the rad circuit. The theory is, as the stat moves from closed to open, resistance of the rad circuit is reduced and flow increases, thus relieving pressure and closing off the bypass to an extent and regulating the pressure between the two circuits. 

That's my understanding of it anyway. However, the "problem" I can see with this is that the rad, by its design and construction, is inherently high resistance is it not? Compared to the resistance of a 32mm bypass and a relatively weak pressure relief valve I would have thought. So even when the stat is fully open, the rad might provide so much resistance still that the bypass valve is still wide open.

Would love anyone with a greater understanding than me to share their thoughts on this....

[Smithy77]

So I've been told that this behavior after a track session is NOT normal, so there is a problem to resolve, and I'm thinking it boils down (excuse the pun) to poor rad flow at idle. Increasing the idle speed, as alluded to above, seemed to help it but maybe not enough. 

Before ditching the PRRT, I'm going to try reducing the bypass diameter from 32mm to 16mm in an attempt to increase rad flow. To test the theory without buying smaller hoses/reducers, and without going on track, I will first knock the idle speed back down to 900rpm. This will be in an attempt to reproduce the previous and noticeable issue of poor rad flow at idle which needed a blip of throttle to get the fan to kick in. If I can get this happening again, I will simply try clamping the bypass hose to restrict it quite a bit. If this has the effect of boosting rad flow and have the fan kicking in again, I think I'll be on to something.

If anyone can reaffirm this theory as having reasonable logic, it would be appreciated.

[Jonathan Kay]

I like changing only one thing at a time, and think the proposal might be logical.

Can you add a sketch of the circuit?

Jonathan

[Smithy77]

Thanks Jonathan,

Link to diagram

Another idea... if reducing the bypass down to 16mm should, in theory, increase bypass resistance and boost rad flow, if I was to increase the length of the 16mm bypass, would this increase bypass resistance even further?

[Jonathan Kay]

It's really complicated, but for this purpose resistance to flow might be roughly proportional to length and inversely proportional to fourth power of radius.

Jonathan

[Smithy77]

So just to be clear, longer length = higher resistance to flow, and smaller radius = higher resistance to flow?

[Jonathan Kay]

Yes x2.

Jonathan

[Smithy77]

Excellent! Thank you. 

[Titanium7]

Just an observation, having fitted two of the PRRT systems, on both I didn't blank off the water rail to stat 16mm hose. On one I had a Heater but the second car just left the pipe in place and both systems worked as they should. 

[Smithy77]

Thanks, did you do any trackdays with the PRRT?

[Titanium7]

Oh yes!

[Smithy77]

And you had a 32mm bypass without the problem I experienced? That would suggest something else might be at play. Think I will still experiment with reducing bypass diameter as I tried clamping mine today to reduce it somewhat and it did seem to improve rad flow

[Smithy77]

I can't see how deleting the OEM bypass between coolant rail and OEM stat housing could cause any issues...... or could it?? (just thinking aloud)

[Smithy77]

Time for an update.

So just as a quick re-cap, with the OEM cooling set-up I was suffering with chronic over-cooling:

I then tried running with the Rover PRRT set-up as follows:

Over-cooling successfully eradicated. However, while on a track day, the above set-up gave me the problem/symptoms of poor rad flow at idle. When I came off the track, the rad seemed to really struggle to get rid of the heat from the stinking hot engine and the rad fan stayed on constantly for 15+ mins while idling in the paddock.

As was discussed earlier in the thread, I then tried reducing the 32mm bypass to 16mm in an attempt to increase rad flow. This did seem to have the effect of increasing rad flow at idle as the average time the rad fan stayed on during a heat cycle seemed to drop by 10 seconds or so (very similar ambient temps). Now, I've not tested this thinner by-pass on track yet (Snetterton booked 25/02), but I took the car out for a road test on the weekend and experience yet another issue - an increase/exacerbated thermal cycling. Having let the engine come up to temp (20mins driving), I'd be cruising along at 40-50mph with the temp nice and stable at an indicated 80C. As soon as I boot it and give it higher revs, the temp suddenly goes from 80C to 70C - not good for the head gasket!

So, I'm now thinking of ditching the PRRT and going for a remote stat in the top hose. Not an expensive QED stat, but using a 2nd OEM stat housing in the top hose with a drilled 82C stat I have spare:

The are 2 main difference I can see between my version and the QED set-up:
1) My bypass from remote stat T's into the 32mm bottom hose, where as the QED kit has you T it into the 16mm hose between expansion bottle and submarine pipe. (This is mainly because I already have the hose cuts and hardware for this, but I also think it could have the advantage of a) reducing the volume of the rad circuit, and b) increasing the distance for hot and cold coolant to "blend" before entering the engine.
2) My OEM bypass is deleted completely, but I think the QED kit might assume there is a secondary bypass for the heater.

Before I go ahead and cut up my cooling system, AGAIN, I just wanted a sanity check to make sure I'm not doing anything stupid. Any other views/observations also welcome.

Thanks!

[Mechanical Moz]

I have that thermostat arrangement, albeit with an aftermarket stat housing.  Warms up quickly, never had overheating problems and never had an airlock or problems refilling the system.

1) Radiator bypass goes from stat housing to submarine, the submarine has two 16mm outlets.

2) I do have a heater and an uprated ali rad.

[Superwhite R283]

Hi Pete

Hope the new plan works. Unfortunately I cant offer any advice.

My set up (still the same as when I bought the car) is as your OEM diagram (but now with no jiggle valve as im running jenveys) but with no stat. Works well, doesn't over heat, when it gets to 80 it stays there (road or track), only draw back is on cold days it takes a while for the water to reach 80, especially on the road. Just made up some grill blanks to see if it speeds it up :o)

 

Cheers Darren

[Ian B]

Why was the heater loop removed?

Would that not help stabilise the head temps?

[Smithy77]

Cheers Darren. I don't think I'd get on with going statless, what with the worry of messing about with radiator blanking on cold days. I know it works for some people though.

Ian B, the heater loop was "removed" for two reasons - 1) I don't have a heater and 2) with the Titan DS system, the close proximity of oil pipes crushes/kinks the bypass hose (reducing x-sectional area by about 50%) where it connects to the bypass inlet on the stat housing. I believe this is a standard Caterham "feature" on the DS K-series cars. I say "removed" because I technically I still have it, it's just moved from behind the engine to in front of the engine. 

This K cooling is doing my nut in, and the more research I do, the more my nut hurts! I just can't fathom the behaviour of my PRT. It's designed to equalise pressure, but I have poor rad flow at idle. It's supposed to prevent thermal cycling/shock, but not on my car. I am definitely ditching it.

Regarding the remote stat idea, the more reading I have done the more I am sceptical about it. By all accounts it's does little to help and it removes a key feature that the OEM set-up was designed to do, and the majority of members I have spoken to seem to agree. Taken from a very well reference link in the BC archives:

"The X remote thermostat

Under this description fall some devices who shares the same principles, although different in shape and plumbing. They are originated from a completely inverted approach: If the original cooling system is no good, let use a cooling system which has proven to be very effective on other engines. This system follows the lines of the remote thermostat as fitted on CIB engines. Unfortunately the problems on the K are of different nature (as explained above).

Fitting the thermostat on the engine outlet could solve the thermostat dynamic issue (because of the blending/dampening effect of the coolant inside the engine and the engine itself) and the static issue (as the coolant temperature at the engine outlet is higher compared to the inlet). The thermostat will work beautifully and also the temperature measured by the senders at the engine outlet should be stable. Sadly all the control over the flow temperature at the engine inlet is gone and any level of gradient and temperature fluctuation is allowed inside the engine, as long as the final temp is constant. Although this setup can't possibly prevent HGF in the long run, the risk of temporary overheating as described in c. is sensibly reduced."

So where does this leave me then? Well, going back to the beginning when I had the OEM circuit but I was suffering with the bad over-cooling, I think the problem was associated with the fact that the stat which I removed (when converting to the PRT) had a very large jiggle valve in a 4-5mm hole. I'm hoping the stat was the problem causing the over-cooling. 

I think I'm going to try and go back to an OEM set-up but with a minor variation (don't worry, there is method to my madness ), and that is to move the stat a little way upstream of the coolant inlet, but still on the cold side of the radiator, like so:

"Why would you not just go totally OEM?", I hear you cry. Well, unless I am missing something, the above set-up should function in exactly the same way as the OEM set-up. The plumbing, in theory, is exactly the same, except for the fact the bypass and stat are now in front of the engine instead of behind. There are also some advantages to this over standard OEM:

1) The bypass is not kinked.
2) Access to the stat is improved by about a million%
3) There is a longer path for hot/cold coolant to "blend" before hitting the pump (possibly irrelevant?).
4) I have all the pipes/hardware to do this with a neat instal, so it will cost nothing and I can leave the inaccessible original stat housing well alone.

I'll be using an 82c stat with 2mm hole drilled. The new stat housing will have to be fitted in the VERTICAL section of the S-shaped bottom hose - will this have any adverse effects on operation?

Am I mad, or is this ingenious?

[Smithy77]

Well that works! 

Tonight I have been busy changing the cooling system as per my latest diagram above. Ran it up to temp in the garage and all behaved perfectly as expected. Then went out for a 30min test drive. No over cooling, virtually no thermal cycling. Works pretty OEM to be honest. Did some cruising at 40mph, cruising at 80mph, and then some higher revving at the end and the needle stayed between 78-82c the whole time. And when it did move it was very gradual, not the near instant 10c drop I had last time out with the PRRT. 

Acid test will be Snetterton on Saturday to see if my rad flow issue is sorted when I come in from some hot laps.

Some pics:

So to conclude, I'm back to OEM(ish) set-up with OEM behaviour, but now have the advantage of a very accessible thermostat. The differences between now and my original set-up, where I suffered badly with over-cooling, are: 1) stat moved up-stream, 2) new 82c stat with 2mm hole, 3) bypass which is not kinked. We can only assume really that the old stat, with a 4-5mm jiggle valve, was the culprit. The PRRT issues remains a mystery, but I would advise that they are avoided from my experience.

Will report back after Snetterton. Thanks for the help

[elie boone]

With the remote thermostat i would feed the bypass from that thermostat into the expansion bottle rather than bypass the radiator. But as your system work fine now there will be no need to change.

[Smithy77]

Elie - Not that it matters now, but out of interest, what difference would that make? I did allude to the fact that your suggestion was as per the QED install/instructions, but I couldn't see how the different approaches would function any differently - they both bypass the radiator.

I am assuming your comment relates to this diagram for the DIY "REMOTE" stat:

[elie boone]

The bypass just before the radiator offers the least resistance especially with a low flow and no reduction in diameter. if you feed the bypass into the expansion bottle you create a loop that is separate to the radiator, the diameter of that bypass will always be significant smaller and once the the engine is fully up to temperature the main flow will go trough the radiator. Basically you created the same now but with a complex plumbing.