K-series remote thermostat idea: am I being a plank?

[Mankee]

I think I've had too much coffeeeeeee this morning and have been perving on QED remote thermostats, reading the archives etc. etc. I'm keen to try the remote stat idea, as I'm getting a similar minor issue as Steve (Grubbster) of coolant being spat out of the header thank with sustained high revs and trackday use.

This thread has been extremely useful:

http://www.lotus7.club/forum/techtalk/k-series-overheating-good-solution-fitting-instructions

However, I've had a rummage around and was wondering why I can't just use a standard Rover thermostat housing (with stat of choice) plumbed into the top hose and the bypass T'd into the 16mm header tank to bottom hose submarine pipe, i.e. exactly the same as the QED instructions.

I've got an unused 74 degree stat to fit a spare housing, the flow direction arrows make sense (but is opposite to when the stat is fitted in the standard location???), a dummy stat and all I'd need are a handful of hose clamps, a reducer, a joiner and some stat and housing seals.

Is this a ridiculously cheap and simple alternative to a full "kit" or have I overlooked something and there is a vital flaw in my plan somewhere? Put me out of my misery, please!

PS. I've done the popular coolant circuit mod but will be returning it to stat and removing the stat (many racers don't run a stat) for the rest of the season's trackdaying anyway.

[Graham King]

Mankee, as no one else has replied then I'll give it a go. 

I do have the QED stat plumbed in exactly as you describe, I can't picture the standard k-series stat housing outside of the engine as I've only ever seen it connected to the block, I would have thought it was all doable but a couple of points I can think of.

1. You need to T it into the top 32mm hose, the input from the bottom hose is obviously 32mm but is the output side that is normally connected to the block?

2. The bypass outlet on the standard stat housing is 19mm (I believe) and the hose from stat to heater tapers down to 16mm. You would obviously need to replicate this somehow in your setup.

Apart from that I can't see any specific reason why it wouldn't work as ultimately it does exactly the same job. If the plumbing can be made to work (leak free) and I guess as you have a spare housing then you can see if it can be made to fit. 

Graham.

[Mankee]

Cheers for the reply Graham. Glad that there's nothing blindingly wrong with my theory at least!

Here is the standard stat housing not fitted to the block:
http://www.rimmerbros.co.uk/ItemImages/Large/PEM10025.jpg

As fitted to the engine, it would be the other way round, so the bottom-left end is shoved into the block behind the water pump and the upper-right end is connected to the J-hose, which is connected to the submarine and S-shaped rad outlet hose.

I'm proposing to fit the stat in the top hose as pictured, with flow from the coolant rail bolted to the front of the head entering bottom-left, leaving to go to the rad inlet upper-right and, when the stat is shut, bypassing via the 22mm port into a T-piece spliced into the header tank to submarine hose.

I've not measured them, but I'm pretty certain that the coolant in and out are 32mm, which is the same as the top hose I think. So it should be just like fitting the QED stat, i.e. chopping a section out of the top hose, bunging the stat housing in and crunching up some Jubilee clips.

So to downsize the 22mm bypass outlet to the 16mm header tank to submarine hose, it's a simple matter of a 22mm to 16mm silicone reducer, a 16mm joiner, some 16mm silicone hose and some clips. Too simple?

The only advantage of the QED stat I can see, apart from the shiny housing, is that you can spec a 16mm bypass port so that it is a simpler fit in a Caterham install, rather than the 22mm Elise/standard Rover bypass size.

(I'm clinging on to a glimmer of hope that my head gasket isn't leaking...)

[Graham King]

Mankee, it looks to me like it should work then, the bottom and top hoses are the same size so they will fit, the QED bypass is actually 8mm, the 16mm referred to on their website is the size of the hose it t's into, so the t-piece is 16-8-16 (if that makes sense). When I bought mine they only offered the 22mm t-piece, I rang them and pointed out it didn't actually fit a caterham and that it needed to be 16mm, so they had some of the right size made up sent me one and started offering the option on all new ones. They had assumed the installation in the caterham was the same as an Elise. 

One thing to be aware of is the QED solution isn't without it's own compromises. Warm up is slightly compromised, in that engine temps take a while to stabilise. It's ok if stationary as there is no airflow through the rad, but if driving on a cold day then during warm up there is cold airflow through the rad which further cools the water in the rad so when the stat does open it then dumps a rad and bottom hose full of cold water back into the system, which causes that stat to close again and so the cycle starts again. Will a łarger bypass exacerbate this or make it less likely? 

Graham.

[Sallen43]

I am considering the same, but using a later freelander thermostat housing. It looks exactly the same as the QED system, but not as shiny.

i'm interested to hear how you get on as the freelander thermostat and housing is £25

cheers,

sam

[Jonathan Kay]

Please could someone do (or point us to) a schematic sketch of the before and after circuits?

Thanks

Jonathan

[Graham King]

Jonathan, here it is http://files.qedmotorsport.co.uk/k-series_qed_remote_thermostat_fitting.pdf

Diagram shown is for an Elise but it's basically the same for a Caterham, essentially then housing inserts into the top hose and the bypass goes to the bottom hose (so bypassing the rad) via the hose from the header tank to the submarine.

Graham.

[Jonathan Kay]

Thanks, Graham.

I need to read that few times again. I'm not convinced by their logic about positioning in terms of temperature control alone.

Jonathan

[Dorset7]

Imagine  I can feel a waterless coolant episode about to start - I'll pull up a chair and grab a beer.  

[Mankee]

This is my amazing sketch on the back of an envelope:

https://www.dropbox.com/s/n4ccx8tisl0mjt6/file%2030-07-2015%2018%2049%2032.jpeg?dl=0

I can't remember exactly, but do the QED instructions mention removing/gutting the standard thermostat?

[Dorset7]

[Jonathan Kay]

Is the position of the temperature sensor moved with this modification, please?

One thing to be aware of is the QED solution isn't without it's own compromises. Warm up is slightly compromised, in that engine temps take a while to stabilise. It's ok if stationary as there is no airflow through the rad, but if driving on a cold day then during warm up there is cold airflow through the rad which further cools the water in the rad so when the stat does open it then dumps a rad and bottom hose full of cold water back into the system, which causes that stat to close again and so the cycle starts again. Will a łarger bypass exacerbate this or make it less likely?

Very interesting. The lower the thermal inertia of the water in the bypass (thermostat closed) circuit the less oscillation, other things being equal(!), I think... 

Jonathan

 

[Dorset7]

No. 

[Graham King]

Jonathan, my experience is that once up to temperature then the temps are controlled very well, other than whilst sitting stationary in traffic mine never goes over 80. It is prior to reaching a permanently open stat position that it is in my view compromised. I should also mention that I have a Radtec rad (stabdard not extreme version) so this may exacerbate the cooling effect whilst the stat is closed and the water is stationary in the rad.

Graham.

[Jonathan Kay]

Once the thermostats are open what's the difference between the standard and modified system?

Thanks

Jonathan

[Graham King]

Jonathan, once all up to temp then there's no flow difference, however the stat will be reacting to water that has just left the engine not to the temp after it has been cooled via the rad as in the standard setup.

During warm-up however, in the standard set up the rad is part of the flow, whereas in the modified setup the rad is bypassed, so warm-up is quicker but once the stat is opened the rad comes back into the circuit and all the cold water held in it returns to the system, on cold days this changes the temp enough to close the stat again starting the cycle again. 

If the car is stationary the effect is minimal but if car is moving then the effect is noticeable. I reckon the gauge drops by about 10-15 degrees to start with, each subsequent cycle is smaller until it stabilises fully. I'm not sure how much effect this has on the engine but I know some say the k-series is sensitive to thermal shock. 

Hope that all makes sense.

graham.

[Jonathan Kay]

1. In answer to your Subject... "No", but I'm learning a lot.
2. How to decrease the thermal downshock from that slug of cold water from the radiator... in the absence of major changes you could either decrease the volume or increase the temperature of the water in the radiator (and that bit of the circuit). A small bypass or hole in the valve would do this by keeping the radiator warmer but would slow the whole system warming up.
3. Why does that lttle coolant circuit to the electrical temperature sender exist rather than the sender being moved? Is it just easier to do the conversion that way?
4. ... once all up to temp then there's no flow difference, however the stat will be reacting to water that has just left the engine not to the temp after it has been cooled via the rad as in the standard setup. At a given heat output from the engine it doesn't matter which side you sense it in control terms although the cut-off value would need to be different.

But 4 suddenly made me realise why this problem is harder than I thought. In most control systems the sensor(s) that controls the valve(s) and the valve(s) can easily be separated. In car thermostats they're inherently together. Are there any modern engines or racing engines that separate the sensor(s) and actuator(s)? Are they available off the shelf? That's going to be more expensive and potentially less reliable, but it would make the whole thing much easier to control.

Jonathan

[Mechanical Moz]

Some modern engines use an electric water pump and control the flow rate instead of using a thermostat. Davies Craig make aftermarket kits but they cost a lot more than a thermostat.

In the 90's BMW used electronically variable thermostats, they were ECU controlled and used to raise the coolant temperature at steady cruise.

[Jonathan Kay]

Thanks... something like that would make this a lot easier to solve.

Jonathan

[Jonathan Kay]

Similar thoughts on a K in an MG F.

Jonathan

[Smithy77]

Sorry to resurrect an old thread.

I will soon be getting back my 230bhp R400 K engine after a fresh rebuild. Before the rebuild it suffered quite badly from the unexplainable K-series overcooling issue - any cruising above 40-50mph and with ambient temps below 15C, the temp gauge settled at a constant 60-65C. This was on a standard coolant circuit, without heater, and fully functioning 82C drilled stat. It would seem nobody can explain how this phenomenon occurs; if the coolant is 60C as measured as it exits the top of the head, it can be no warmer when it reaches the standard stat housing, so in theory the stat should close, and therefore coolant temps should rise, but they don't. The symptoms are exactly what you would expect from a stat stuck open. Blanking the radiator helped a lot to reduce the problem, which must mean that while the coolant is measured at 60C, that the stat must be staying open (after initial warm up) if blanking the radiator helps to control the temps.

Anyway, the reason for resurrecting this particular thread is because I plan on adopting Mankee's idea (and indeed using his parts) to make a DIY version of the remote stat. However, I plan on deviating slightly from his original idea which, as I understand, was based more closely on the QED circuit whereby the original "heater bypass" remains in situ, but the bypass out of the remote stat is plumbed into the the expansion bottle bypass circuit. I'm not sure I like the idea of having 2 bypasses bypassing the radiator as you are making it easier (by reducing the resistance) for the coolant to flow through the 2 bypasses rather then through the radiator. So my idea involves deleting the heater bypass circuit completely and introducing a new bypass from the remote stat directly to the original stat housing. 

Link to coolant diagram

It should be noted that the bypass inlet/outlet from the stat housings are 20mm and I plan to reduce the bypass hose to 16mm for the majority of it's length to help maintain some of its resistance. The QED kit has an 8mm bypass outlet I believe which Tee's into the 16mm hose between expansion bottle and submarine pipe, but I think you are also supposed to retain the original heater bypass circuit too.

I'm hoping this might go some way to resolving the overcooling and general temp stabilisation. Will this solution be any better or worse than a) the standard circuit, or b) the QED kit? Thoughts/problems/ideas welcome before I go experimenting on my new engine!

[Graham King]

Smithy, since this thread I drilled a small hole in the stat in the QED system, as Jonathan mentioned in his previous post. This has completely removed the cycling I referred to initially, but as suggested, it does now  take longer to warm up. Your assumptions around the QED fitting are correct, 8mm bypass and leave standard bypass in situ. To be honest I've never really thought about why or what effect that has or what would happen if it was removed. 

I'm happier with the overall system now and aside from whilst stationary the water temp never rises above 80, even during pretty high speed runs during the Classic last weekend where it was hot. 

Graham.

[Smithy77]

Graham, thanks for the update. Gives me the confidence to crack on with it as planned