MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGF Technical - Preventing thermal shocks- the best way to preserve the head gasket

Hi all,

I've had an interesting discussion with Simon Cuffham (scuffham@globalnet.co.uk) on the Elise forum regarding thermal shocks- the most probable cause of HGF in K-series engines (at least IMO!)

The problems associated with the standard cooling system become more apparent the more power enhancing modifications that are performed on the engine.

The issues involved can be summarised as the following:

1) The radiator is larger than the engine requires.
2) The radiator is located some pipe distance away from the engine.
3) The thermostat is stupidly placed on the intake to the pump and relies on the coolant by-pass to give it hot water to open.

These issues combine to cause the thermal cycling/ oscillations by the following mechanism:

Water returning from the radiator is so cold relative to the by-pass, it inappropriately closes the thermostat. The engine in the meantime is still working hard and getting hotter. However, the thermostat takes some time to open as the by-pass return coolant needs to warm the coolant at the thermostat, causing the engine to run too hot. The thermostat then opens completely, dumping a bolus of cold water into the block/head. Repeated cycles of hot/cold thermal cycling leads to the deterioration of the head gasket (at best!).

The solution is to remove the standard thermostat from its original location and remove (or block) the by-pass. Obtain from your nearest motor factor a remote thermostat with by-pass (used on some Rover SD1's (2600 models), BMW's etc.) and splice this into the outlet hose from the engine as the heater hose splits off and use the by-pass to feed the heater.

If you have an engine generating well in excess of 200 bhp and subject the car to track/race use then you should look at putting the thermostat at the front of the car with the by-pass connected back into the radiator return pipe. This solution gives a much larger volume of water at engine temp, the downside is that warm up time is much increased.

Using the latter solution, Simon says: "I use a 87C thermostat and no matter what I do the engine [temperature] will sit at 87C."

A more detailed explanation of the above can be found on http://www.users.globalnet.co.uk/~scuffham/Water System Mods.htm

Simon is scathing about the potential benefits of drilling the thermostat ring that we have discussed on this board before: "drilling the thermostat is NOT a solution, as it makes not a jot of difference to the problem, it just moves the start point." Thus the thermal cycling continues to occur, albeit from a cooler starting point, reducing the magnitude of the thermal fluctuations.

QED incidently already retail their own solution to this problem, bolting a replacement thermostat on the engine outlet. It consists of an alloy thermostat holder with bleed pipe that fits onto the hot outlet hose from the head with the bleed to the expansion bottle. The by-pass is then blocked off, thus the water HAS to go though the new thermostat. Nothing wrong with this approach except you loose the heater until the thermostat opens. See http://www.qednet.demon.co.uk/ for more details.

These are all very interesting solutions to the problem of HGF prevention. Anyone have any comments on the above?

Rob
Rob Bell


This does imply that having the heater on full from cold
will improve things by having a damping effect on the thermal
ups and downs during the warm up period.

Anything that adds to the quantity of water from cold will
have this effect to a greater or lesser extent.
So, drilling holes in the therostat should still help.

Having said that, neither of the above address the fundemental
problem as you have described.

Rover state that the thermostat was placed on the inlet rather
than the outlet to 'provide a more stable control of coolant
temperature in the engine.'
Interesting. They may have experimented and found what we
have to be the best compromise.

I further suspect that part of the design brief was to have
the car heat up inside quickly. Could these two factors
explain the use of a water as opposed to air valve to control
the heat inside? And why they didn't do as QED have done ?

Cheers,
Paul.
Paul Nothard

Hi,

had to read it a couple of times to understand it. 'My problem' *g*.
I'll try to work out a special 'Simons coolant system page' if nobody else volunteers. Credits to the author promised as ever. May be Carl or Jon like to modify in the above manner in their next long and cold winter.

Interesting stuff, though. But see also, he mentions *the oil temperatur will remain on the same high level*. Thats what I'm bothered in.

Regards
Dieter --> learned to open the heater at cold start all the time :))
Dieter Koennecke

Can't get anything from the link so don't know if this is covered. Has anyone actually measured temps of the coolant at various points to confirm that the thermal shock is as said?

Plus I would add 4 to your list Rob. This is that there is less than 2 litres of coolant volume within the engine, which means that there is a designed capacity for rapid and greater temperature changes of the coolant.

Rog
Roger Parker

So what will happen it the thermostat is removed altogether? There'll be no need to do any thermostat drilling, no need to search for lower temp thermostat, no danger of the thermostat failing/stuck.

Any side effects this will have other than the engine will take a longer time to warm up?
Jason

If it ever warms up at all..
Growler

Growler is quite right Jason- the engine will never warm up to the detriment of engine efficiency, not to mention mechanical wear...

The thermostat drilling was in Simon's opinion something of a bodge. His experience comes from endurance racing an Elise (equipped with a VHPD + spec engine)- wherein he suffered from 4 successive HGFs before arriving at the solution detailed above. Since then the problem has been solved. I have asked for any available testing data- but I suspect that the solution was arrived at empirically. Re-located thermostat = no more HGF.

Rog, the link doesn't work because of spaces in the URL- try cut 'n' pasting >> www.users.globalnet.co.uk/~scuffham/Water System Mods.htm <<

Good point regarding the probable 4th point, and I'd have to agree that this is a major contributer to the thermal cycling issue.

Why hasn't Rover re-located the thermostat as desribed? Good question to which I have no answer. The system works well in low output FWD cars (as described in the Water System Mods.htm link above), and probably works reasonably well in the MGF under constant load conditions. Add in some heavy right foot conditions, and it appears that things go rapidly pear-shaped...

Rob
Rob Bell

All,

I did not just arrive at this solution, it took several months and a few HG's!

The way I finnaly proved it was to borrow a dozen thermocouples, place them in all the pies arround the car allong with a flow gauge in the radiator cct, then a test session at Bruntingthorpe (Bloody great long runway!)

the data I got from this was that at 100mph with the engine at 95+C (water exit temp from the head), the radiator return was less than 40C. Ambient temp on the day was 22-25C

What then happens is as I describe it in the post above.

Simon (S)
Simon

As mentioned before on this board HGF failures are attributable to an unfavourable temperature gradient either side of the thermostat, this is due in the most part to overcoooling via the long radiator run, this is not a new phnenomenon but has been experienced in just about every mid-engined car.

The remote thermostat solution was implemented in the 1980s on the Ford RS200 rally car which had a mid mounted Cosworth BDT turbo producing 400BHP+, the cycling effect in the cooling system destroyed a few engines before Hart moved the thermostat to a remote location. YThis has been common practice on many mid engined cars since.

Drilling the thermostat does lessen the cycling problem to an extent and we have data captured from thermo-couples in the cooling system to show its effectiveness. It is not a complete solution however.

The coolant temperature cycling is also quite hard on the front liner.....

It's a pity that MG/Lotus havent seen fit to follow the examples of the past.

Dave
David Andrews

Does this imply that cars in colder climates are more prone to suffer from HGF? Are HGFs more common in the winter?

Being in Cyprus where we have hot weather most of the year and real high temperatures (40C+ in the shade)during the summer I am always impressed with the capacity of the MGF cooling system to keep the coolant temperature at a constant level all the time - even with the aircon at full.

Spyros
Spyros Papageorghiou

Rog,
FYI, copy and paste works for Simons link. Another engineer with different skills, though. :)))

Anyway, added a mirror to the mgf-net.de and a link also to the original document at mine.
http://www.lame-delegation.de/mgf-net.de/thermostat/

Mirror Link with credits for the author :))

http://www.lame-delegation.de/mgf-net.de/thermostat/Water_System_Mods.htm
(just hit it)
Simon, let me know if you disagree. I collect each and everything in this way :)

Dave,
>It's a pity that MG/Lotus havent seen fit to follow the examples of the past.

What do you think is the reason ? Do they bother the costs for a remote thermostat ?

I had a visit at my mechanic this afternoon cause the coolant expansion tank cap at my MGF got broken after 12k miles.(high level in the expansion tank).

I asked him about the remote thermostat installation.
He got a *grin* and mentioned. 'My godness, that crap part got just damaged at my BMW...' ... but may be worth a try at the MGF.

Regards
Dieter
Dieter Koennecke

In reference to Dieter's comment about starting up with the heater on:

Surely it would be better to start up with the heater off, in order to get a small amount of coolant up to temperature as fast as possible. Then and only then allow the 'cold' water still in the heater matrix into the general cooling cycle (and don't turn the heater from cold to hot unless the car is idling - ie not putting much heat into the water, as the thermostat will close when you turn the heater.)

Thoughts? Or have I got the wrong end of the stick?

N837 OGF
Hugh

If you have the heater on from cold, you'll have more water to heat and a slower rise in temperature.
That'll open the thermostat more slowly.
Also when the thermostat is opening there is more water for the newly entering cold water to cool down.
Thus, it smoothes the curve somewhat.

That's my take on it, but I may be wrong.

Comments ?

This may be a cheap (nae free) way of reducing the incidence of HGF.
Or maybe not... but I gotta try something. :-)

P.
Paul Nothard


Hi all,
indeed very interesting info from Simon ! Some time ago when I wrote about the diff. in heater system between Elise and "F" the point was that the Elise is slightly better off in HGF -terms due to the fact that the heater is ALWAYS on and the air temp. is regulated by flaps regulating the amount of hot air entering the cocpit.
There are a few good housed thermostats on the market (BMW (!) , Mercedes etc.) that could be turned into a workin system on the "F" ! This is surely something to consider when the season comes to an end and my car is on stands again.... That the oil fills a big role in cooling the engine is something that I guess all of us have experienced at one time or another, especially during hot times as present ! The big question is what system to use to cool down the oil ?
Personally I am no big fan of long oil-feed tubes and radiator with pressure drop and yet another thermostat in the system ! Maybee the more direct water / oil cooler is a good thing once the water cooling is sorted out ?

Regards from hot and sunny Sweden ,
Carl.
Carl

Simon has very kindly sent me an image of the oil cooler installation on his car. Unlike the "usual" installations on the MGF whereby the cooler is typically found at the nose of the car, Simon has his installed on the bell housing without thermostatic control. This system appears to work perfectly, controlling oil temperature peaks and without over cooling the oil.

My immediate thought that on the MGF this may not be possible because of packaging restraints and a lack of airflow. But I've been thinking about it and I think that a solution is possible through the removal of the resonance box from the air intake system and/or using the MR-2 type air scoop that Roger has fixed to his car at the moment.

As regard to remote thermostats, has anyone got part numbers to hand?
Rob Bell

I should mention that the oil cooler on Simon's car is a oil/water heat exchanger. Neat.
Rob Bell

http://www.lame-delegation.de/mgf-net.de/2001Bayern/dk_DCP_6019.jpg

shows the oil cooler position as done by Philip Class.

Yes, a part number set of any electric remote thermostat assy would be nice. Needs not be be from BMW, cause I heard already of 'not so reliable'.
Dieter Koennecke

Hi,
I just have visited the mechanic and asked him for a picture of the romote valve he mentioned at his Beemer.

I have really noo idea whether it's this one which Simon means.
So I added a very sensitiv text far below to the document.
http://www.lame-delegation.de/mgf-net.de/thermostat/index.htm
Thje mecanic told me that it's the heater valve. It gets controlled by a potentiometer.
So it's an analogue input which needs some more electronics to convert a temperature signal for controlling the valve.

It's the view to the electric heater valve of a BMW 7, 3Liter. (looks a little older that thing)
http://www.lame-delegation.de/mgf-net.de/thermostat/beemer_7cp_6168.jpg

May be Jon F. can suggest something else from a DB.

Please correct me if this is all complete rubbish.
Regards
Dieter
Dieter Koennecke


Hi Dieter,

It must be the old type of thermostats with one inlet ,one outlet to rad and possibly one to the heater. Will source one this afternoon and try to take a pict. As they are old eq. they are also not that hysterical high on temp. - more modest 80-86 deg. C.
Regards , Carl.
Carl

Dieter, I think that Carl is right. Simon mentioned the remote thermostat from the Rover SD1 (2600 6-cylinder model) in the text of his explanation page- so I suspect we are looking at a "traditional" thermostat and housing...

Could the existing thermostat housing be suitably modified and relocated? This needs further investigation...
Rob Bell

A fascinating thread.
1) The cooling system has excess capacity (is this not a requirement ?)
2) The engine block contains only a small quantity of coolant
3) The coolant temperature gives absolutely no indication of status (once thermostat opens completely).

I would like to see the coolant flow from the engine to :- 1) heater (2) bypass (3) radiator (no change there). The thermostat would then mix the return from (2) and (3), with (1) returning to the engine uncontrolled by the thermostat. At the moment the bypass is mixed with the heater return. The radiator fan should be controlled by a sensor on the radiator (!)
Where should the sensor for the temperature gauge go ? engine outlet might be the most useful ?
Is a thermostatic valve available to mix bypass and radiator return ?

Steve
Steve

Picture of Simon's water/oil cooler installation: http://www.users.globalnet.co.uk/~scuffham/PIC00011.JPG

Steve, I guess that a temperature sender on the engine outlet would give the most representative indication of engine coolant temperature. But isn't this where the existing temp sensor resides?
Rob Bell

Hi,

Yes, very interesting stuff. We seem to leap from one theory to the next, keep heater valve open, knock out jiggle valves, drill thermostat, etc. but this one has empirical evidence from Dave and Simon, albeit proven at the very high-end of the performance envelope of the engine. I guess what we have to decide is what type of driving we do, what amount of modifications we are willing to make, what risks we are willing to take with any modifications, what value we put on the car (a car with a radically modified cooling system would be virtually unsaleable), and how paranoid we are about being hit with an HGF.

I keep the heater valve open on starting, in fact I keep it on all the time, unless the heat is actually unbearable. (Please Mike, make a conversion to duct cold air to the face vents!) As the bypass hose is a very short er, bypass directly into the heater return pipe I did wonder how much use keeping the heater valve open was, but as the heater heats up then it must be circulating more coolant, mustn't it?. Strangely, front-engined installations don't have the bypass hose (not up to 1.6, anyway, I haven't looked at a 1.8). How does the coolant circulate at startup when the thermostat is closed? How does hot coolant get to the thermostat so that it opens? I don't know, but it seems to work.

I don't know if I've interpreted everything correctly (you're not the only one to have to read this stuff twice, Dieter!) but some of the proposed changes to the cooling circuits - especially the QED as described by Rob - imply that the bypass hose is removed or blocked and no circulation can take place until the thermostat is open. This seems like one of the worst things you could do to the system, surely the low-capacity coolant in the engine must circulate at all times?

As for drilling the thermostat, this seems like a sensible modification. It might not be enough for power users, but it might act as a small element of relief for more modest drivers, a sort of mini bypass on the main coolant circuit. The only problem is that the system has to be drained to do it, and many of us fear that almost as much as an HGF. At least this modification doesn't seem to bring any great possibility of damaging the engine.

Regards, Kes.
Kes

Kes,

I don't have a FWD car to hand, but I would place good money on the FWD 'K' does have the same by-pass pipe, otherwise the stat would never see hot water - this could be the heater cct though, on a FWD it would not be very far away.

Anybody got a Rover 200/400/600 (25/45/75) they can take a pic of?

Simon (S)

PS. drilling the stat is not really going to help as it is just going to cause the stat to be more closed that it should be at any given time, therefore, negating the hole, (unless the hole is so big that the stat never opens at all!)

Simon S

Simon, about drilling the stat.
The point isn't that it'll stay closed longer - which you are correct in saying. That's a side effect (problem?) of the fix to allow more circulation at a wider range of temperatures.

cold:
Drilling holes will allow a more even rise in temperature up to operating temperature. The implication being that there is more temperature fluctuation while warming up.

hot:
Are we also saying that the overcooling by the radiator will allow a cycle to be set up with slugs of cold water closing the thermostat? Thus drilling holes will stop this cycle by allowing water to flow at all temperatures.

It's not a perfect fix. In fact, it's not really a fix. As I see it it's a cheap way of reducing the problem.

I'm trying to work out if the free fix of having the heater on will reduce the temperature fluctuation in the same way. It should help from cold as there is more thermal inertia in the system with the heater on. When the stat opens and we get the fluctuations discussed above will the extra flow round the heater help or hinder the temperature cycling ? I currently think that this free fix can do no harm and probably does some good, so I'll keep doing it. :-) But I'd be interested to see if people agree with it or not.

Cor, this is too much for a hung over Sunday morning. Think I'll just go back to bed now....

P.
Paul Nothard

All FWD cars share the same configuration where the bypass from the head outlet feeds the heater and the heater return attaches direct to the thermostat housing. However there is a control valve in this circuit that provides for 100% flow shut off through the heater core.

The full bypass system operates by allowing coolant to circultate via the heater take off pipe from the head and if the hose by the heater valve are examined a 'squared' connector exists between what is the heater feed and the return to the thermo housing. What this does it to provide the bypass route for coolant during the warm up cycle.

Overall in the MGF the whole issue is about the thermostat being unable to control coolant temps within as narrow a band as will be found with the same engine in a FWD application. There will be a number of routes that can be followed to achieve this better control. Perhaps a step back from the details of the individual parts will provide for some lateral thinking??

Rog
Roger Parker


Hi all,

the radiator return rail bleeder screw could fairly easy be used as a measuring point for those thermal oscillations. A cheap digital thermometer where itīs thermistor is situated inside a slightly longer screw will make fast temp. changes noticible.
In this way each proposed change ,( holed thermostate , open heater etc ) could be recorded.
The tempsensing thermistors of those cheap thermometers are only about 2,5 mm in diam. and would fit nicely into an drilled screw..... Dieter - what is the dimension, M 4 or M5 ?

Regards , Carl.
Carl

>better control. Perhaps a step back from the details of the individual parts will provide for some lateral thinking??

Of course, I really have problems to follow up. :) 'My language problem', don't bother.

What is special at an MGF engine cooling system ?
It's basically nothing else than a system of a 'heater' (the engine) and a 'cooler' which should reach a constant temperature in a short time with a reasonable gradient of temperature rais... and reverse..
Nothing else than some physical terms. Each engineer for analogue control sytems has the same problem each day, but lots of support for closed loop systems. :)

What I'm talking about are signal forms of input and output quantities of an analogue system. Controlled continuous with regard to value and time.

I should be a profesional for this technics cause it was the main content of my studies on getting an engineer 30 years ago.
Unfortunately I had never had to use all that mathematics stuff on PID controllers, closed loop etc..

I wish I were a student again :)
Dieter Koennecke

Carl,

that's not going to help you understand the problem.

If you want to measure it you need to put a temp sensor in the head outlet and another in the thermostat housing.

You also need so way of measuring the water flow rate in the return pipe.

I am not sure what you hope to achive, as this has already been done.

Simon (S)
Simon S

>>I guess that a temperature sender on the engine outlet would give the most representative indication of engine coolant temperature. But isn't this where the existing temp sensor resides? <<
This sensor controls the radiator fan. The sensors for the ECU and temp. instrument are by the thermostat. The damping on the instrument renders it useless - it only shows a problem when the coolant has drained away.

I would like to use the sensor on the engine outlet to control flow through the radiator and to drive the temp. indicator.

Steve
p.s. I was looking over a VW 2 litre the other day and noticed the oil/water heat exchanger by the filter and thought it was neat.
Steve

Simon,
if I am able to read the coolest point in the system surely any alteration to the cooling system layout can be recorded .This as any step towards a temp-diff. between inlet /outlet water should be in the 15-25 deg C band and not a temp. shock around 40-60 deg.C.

By placing a sensor at that point IMO any step in the right direction will be noticed , let it be just holed thermostat or more advanced relocated thermostat etc.

Regards , Carl.
Carl

Hi,

Steve, there are two temperature sensors at the engine outlet manifold, actually in the heater branch part of the casting. The forward pointing sensor feeds the ECU, which then controls the radiator fan, and the downwards pointing sensor feeds the temperature gauge. The is no sensor of any kind down by the thermostat.

My '95 214, a wet liner I guess, has (apart from the radiator circuit) the heater takeoff pipe running round the engine and through the bulkhead into the heater matrix. It returns out of the bulkhead, through the control valve, and back to the thermostat housing. There is no bypass hose anywhere. I looked at a new 1.6 a year ago, and another equally new 1.6 a few days ago, and I could see no bypass hose. The engine compartments are mighty crowded and it's not easy to follow all the pipework. The 214 has vanes controlling the heater so you can switch from hot to cold immediately: I started it from cold on Sunday with the heater valve off and coolant started circulating around the heater circuit - and also along the top hose to the radiator. This is just confusing. I'll look at the 1.6 again tomorrow if it isn't raining.

My first car was a 948cc MG midget, '62 or thereabouts. How could you have so much fun in a car with around 50 bhp? I thought it was crude at the time that you had to open the bonnet and turn off a valve to shut off the heater (heater on in September and off in May). I just didn't believe some 35 years later that the F would have the same crude, insensitive, unmanageable system, albeit controlled from inside the car. My second car, an MGB of the same year, had fresh-air face vents. I wonder how long we'll have to wait for that feature to appear in the F?

Am I interpreting the QED system correctly in that NO circulation takes place outside the engine until the thermostat opens?

Regards, Kes.
Kes


"Am I interpreting the QED system correctly in that NO circulation takes place outside the engine until the thermostat opens? "

Almost, it uses a small blead pipe to return back to the expansion bottle and has a very small blead hole in the stat plate, enough to make sure that it "see's" the water from the head even when fully closed.

Simon (S)


Simon S

What we need I think is a cooling circuit diagram similar to that found on Simon's web page that shows where the proposed remote stat fits. That should help us all visualise/ conceptualise what is proposed.

Moveover, I think that a repeat of Simon's studies with thermocouples etc would be useful as there does appear to be some significant differences in the way that the Elise coolant system works. I am happy to participate in such tests, but would need the necessary monitoring equipment...
Rob Bell

OK,

I have updated the drawing to mark the places.

It should be noted this is for an Elise, and as you all know, the MGF heater cct is not always on, so option 1 is not going to work in your case.

Simon (S)
Simon S

Interesting thread. Normally i keep well out of HGF threads (a lot of it goes over my head - if you excuse the pun!), but in line with Roger P's call for a bit of lateral thinking and after reading the excellent piece on Simon's page i had an idea...

Simon wrote...

>> As the engine starts to heat up (being on full throttle) the hot water from the by-pass (this is injected into the stat housing directly onto the stat wax bulb) opens the stat as far as it can to let more water flow round the radiator. The water it lets into the stat housing (and on then into the pump & engine) is way colder than the bypass water (that’s at engine temp), this hits the wax bulb of the stat and cools it down dramatically, shutting the stat (this takes about 2 sec’s), the engine, in the mean time, gets a sudden injection of cold water, (at 200Liters/min pump rate), replacing the entire water content of the engine. So all of a sudden, the cylinder head water jacket goes from 95+C to 40C, then, as the stat is now shut, starts to heat up all over again. <<

So, if i read this correctly, the engine gets cyclic "pulses" of cold water due to an overlarge radiator, exposure of the underbody pipe runs and the way in which the thermostat operates. Knowing nothing about cooling systems and being basically experienced in removing bodywork and interior trim it occured to me that fitting the Cheetah spotlights in the front bumper would reduce air flow to the rad and "boxing in" the underbody pipe runs would prevent them being over cooled. This would work in a similar way to the radiator shields seen on MGBs and Minis in cold weather (they use them to keep the heat in the engine).

Surely this would be easier to do than draining the cooling system and cutting it about to install new components?

If i'm talking crap let me know.

SF
Scarlet Fever


Hi all,
in the VERY old days I (and probably Ted !)used a small marvel to ease out any coolant fluctations !Remember the manual operated radiator curtain with itīs thin chain that had to be used in the wintertime in order to get any heat in the cabin at all? It was only on the hottest summer days the radiator was exposed to the full airflow...

regards , Carl.
Carl

Scarlet Fever,

Yes, you are right to a point, but what you need to remember is that at slow speeds when it's Very hot outside, you need the radiator!

Ypu need to bear in mind this only happens at prelonged high speed with you right foot nailed.

Other probelm with this is that with the Elise, if you restrict the airflow thought the radiator, it screws up the aerodynamics at the front real bad!

Simon (S)

Simon S

This thread was discussed between 04/07/2001 and 11/07/2001

MG MGF Technical index

This thread is from the archive. The Live MG MGF Technical BBS is active now.