Kirby
How does the thermostat actually work?? Does it actually adjust the flow orifice size based on the coolant temperature…or does it just open fully at a temperature. On my other cars…a sign of a dud thermostat was the fact that the temperature gauge wouldn’t shift far off the “cold” mark. Replace the thermostat and the temperature gauge sat at half way…so these thermostats were actually temperature control valves.
On my thermostats there is a small spring loaded disc on the engine side attached to a spring loaded plunger that has to plunge to allow flow thru the thermostat.
I presume the spring loaded disc sits above the bypass passage when cold and then closes the bypass passage when the thermostat has plunged a bit. But…this spring loaded disc can then seal off the bypass flow and then allow the thermostat to plunge (open) more until the spring on the flap is fully compressed. So it looks like there is flow adjustment movement in the thermostat AFTER the bypass is closed off.
So it looks as if the thermostat is designed to be a temperature control valve…but I don’t know the temperature range over which the thermostat first opens until it is fully open…that would be an interesting test. If it opens fully within…say…5 degrees…then your Jag is essentially “overheating” above that 5 degree range…and is only temperature controlled because heat transfer increases with the increased temperature…so the nature of the system is self correcting…
I might get some time this weekend to measure this…
Matt
Ah HAAAH …if all else fails read the manual…the Haynes manual,states …Thermostat starts to open- early cars-79-83 deg C, Later cars-88 deg C. Fully open 93-96 deg C.
So the early cars have roughly 14 deg range and later cars a 5 deg range…interesting. I must watch my temp measurement to see if it rises above 96 deg C when in hard use…
My answer to that is “yes”, because you are now out of control, although as there is a time lag between the temperature seen by the thermostat and the journey through the radiator to deliver cooler water (which goes through the engine on its way) so what the thermostat wants to do and what the engine may be generating by way of heat are always a bit out of sync, so I’d not be overtly worried about a 1:1 correlation at all times. What would be worrying is if the temperature had already risen and the demand for cooling had subsequently also risen whilst the revs had dropped. I’d worry in much the same way as I’d worry about aquaplaning in a straight line but causing no damage to the car - you ARE out of control, but the “bad” consequences often never materialise because there is a wide latitude of results which cause no material damage and still let you get back on track easily. I’d agree there is mostly no cause for alarm, but I’d rather it were controlled by design rather be be put in a position where one day I run out of luck.
“Load” with cars is typically a measurement which dictates how much fuel to pump into the car and typically is set up as a table one of whose axes is either manifold pressure (“MAP”) and throttle plate opening. Atmospheric pressure at sea level is 100kPa and total vacuum is 0kPa. A car without wild cams will typically idle at around 40kPa so this can be considered “not under load” or “lightly loaded”. When coating downhill, cars typically are even less loaded, so much so that the fuel is often cut by the ECU and the MAP may be at a level of 30kPa - totally unloaded. Moderate acceleration or going uphill is going to see a MAP above idle levels and maybe as high as 70kPa. Above that is what I’d define as “heavily loaded” and the throttle plate is typically opened up to allow more air into the engine. Doing more work demands more fuel. Increasing rpm is more work in a shorter time frame and the percentage split that the injectors are firing versus not increases.
Superimposed onto this is the desired air:fuel ratio that the car is set to. Downhill, this may be super lean, at partial throttle, lean and at the upper edges of the graph it’ll be rich.
Acceleration typically involves pumping in more fuel but we don’t really spend much time doing that otherwise we’d lose our licenses to drive. Speed on its own isn’t an issue as you may cruising be at partial throttle.
Kirby:- if you read Roger Bywater’s site, you’ll see that rather this being a cause for alarm, these tests were simply more rigorous than the ones previously considered for the engine and that Jaguar simply fitted more powerful fans to the car to pass them. It’s unhelpful to imply to owners that their cars are sure to overheat “out of the box” unless they panic now and take some sort of action. All v12s are not equal - the underhood space for aircooling is different on the different models, they have different fans fitted and the radiator design on the etype is different from that of the XJS. They’ll also be running different spec thermostats (for emissions purposes) so their default minimum temperature will be different.
The principle behind the thermostat is that the volume of wax cold (and solid) is different form its volume when hot (and liquid). Look in the archives for my r&d on thermostats.
What is upsetting is that the thermostat is misnamed and thus misunderstood. It is really just a water divider. It splits off the percentage flow of water through the bypass and to the radiator. It does not control the temperature - it reacts to the temperature. The control function is a feedback loop because the cooling system is designed to always overcool the water it is given. The result is that the water splitter tries to choke off the radiator’s supply of water and the two are perpetually fighting each other in a ping-pong battle which the radiator is designed to win.
When the total amount of water flowing is low - i.e. at low rpm, the thermostat doesn’t know any better, it’s just a water splitter. The only active element in the system is the thermo switch at the bottom of the radiator. If, in the designer’s judgement, the water that has already exited the radiator is still too hot, then the fans are cut-in.
I’d don’t think I’d describe the thermostat as a temperature control valve. Because the radiator overcools, the thermostat acts to choke off its water supply, i.e. “it’s a minimiser of water flow to the radiator”. In doing so it ensures that the average temperature of the smaller volume of water that the radiator path returns will be at a higher temperature because more went through the bypass than before. It certainly never lowers the temperature - once the split of water is 100%radiator : 0% bypass, it has no influence on either temperature or water flow.
kind regards
Marek
[quote="MarekH, post:23, topic:356539, full:true
I’d don’t think I’d describe the thermostat as a temperature control valve. Because the radiator overcools, the thermostat acts to choke off its water supply, i.e. “it’s a minimiser of water flow to the radiator”. In doing so it ensures that the average temperature of the smaller volume of water that the radiator path returns will be at a higher temperature because more went through the bypass than before. It certainly never lowers the temperature - once the split of water is 100%radiator : 0% bypass, it has no influence on either temperature or water flow.
kind regards
Marek
[/quote]
Hi Marek
The fact that the thermostat adjusts the flow based on the temperature of the water passing it means it is a temperature control valve doesn’t it?? The stuff it lets passed is going to be cooled …all other things being equal…heh heh…an economist phrase…
Dammit…sent early…but the flow in the engine system can also change…which should be factored into the characteristics of the pump and the thermostat discharge coefficient…to attempt to minimise the effect of the variation in flow on the variation in heat to be discarded. It is an old fashioned control system.
A modern system would use an electric pump and various sensors to achieve tighter control…but this option was not available when out machines were designed…so we have what we have…but the range of the thermostat temperatures is quite large and the ability of the system to reflect heat increases as it gets hotter…so it naturally corrects in the right direction…which is al least a bonus provided by the physics of the situation!!!
Regards
Matt
It is a control valve, but it reacts to temperature and is designed to maintain a minimum temperature by shutting off the supply of water to the radiator.
It has no control on how hot anything can get as it doesn’t control how much heat was generated. It’s just a water splitter which is a bit out of time with what is currently going on. It systematically dumps the radiator’s excess cooling capacity by choking its supply of water.
The temperature yo-yo’s just above the rated working temperature because the radiator always has excess cooling capacity which is dumped on the next pass.
When people’s cars get hot and stabilise at an elevated level (as reported earlier), that just happens to be the temperature at which the average heat generated equals the average heat dissipated - the thermostat is almost certainly fully open long before then. If at that point, people load their engines more without a commensurate increasing in radiator capacity then bad things will happen.
The thermostat only wants your engine to stay at or above the minimum temperature. It has no tools to control anything once it has done that.
The correct operating temperature is thus at or just above the thermostat temperature.
kind regards
Marek
Nicely composed explanation Marek. Thank you for that.
The 93-96 fully open range, coincidentally, is exactly the temperature our OP was concerned about.
Cheers
DD
Excellent 
Cheers
DD
Dear Doug,
I think that the temperature range observed matching the thermostat spec range is coincidental as the upper bound of temperature is a function of the amount of cooling the radiator does to it (good!) plus the amount of work done by the engine before the water gets back to the thermostat (bad!).
I think 5-8 degree range might still be a reasonable expectation and perhaps the text could be referring to the upper bound of the 88’c thermostat only.
Here is a datalog to demonstrate my car warming up and sitting on the driveway, with 82’c thermostats. The coolant temperature is the blue line on the middle graph (65’c -100’c) so 82’c is the mid line of the graph. The temperature yo-yo’s from ~80 to 86’c, not over the extended range.
At point A and before, the car is still warming up and the temperature is going up steeply and at a constant rate. I’d deduce that the thermostat is closed and water is only circulating through the bypass.
At some point between A and B, the thermostat must be opening and at B the rate of cold water coming into the engine is equal to the rate of hot water passing the thermostat - no further rise in temperature occurs. It peaks at 82’c
Between B and C, the temperature is plummeting at a rate even faster than it was going up. Previously, the thermostat was closed and opening it for the first time allows cold water from the radiator (this is a cold start remember) to replace the hot water in the heads. This is a one-off and the temperature goes down to about 75’c.
As the temperature now does an about turn, I’d deduce that the thermostat must be closing or closed sometime before point C. Most interestingly, the temperature now climbs from C to E, initially at a rate equal to that the that which it achieved before it got to B. That tells me that the thermostat is likely closed until the kink at D, when the temperature climbs, but at a reduced rate. I’d deduce that D is where we see the effect of the thermostat opening.
Since the water coming into the engine has now already been heated before its hop to the radiator, E is higher than B, at around 86’c. It then cycles between about 80’c and 86’c until the car is driven away.
The temperature is now moving about (see H and I) but not immediately in direct proportion of time in relation to load or rpm. The thermostat is barely opening here and the temperature is quickly taken back up but there isn’t a 1:1 correlation with what has happened and the cooling delivered as there is a time lag between the two. Essentially the thermostat is always “a bit wrong” and we know from E-F-G that the temperature could be anywhere in a 6’c range for the same given load depending on how much water has already been sent down the radiator route.
The temperature figures in this example are of a car running lpg (propane). The vaporiser which converts liquid propane to gaseous propane sits between the temperature sensor and the thermostat and probably takes about 5’c out of the water which is why the temperatures at H and I probably reflect a minimal amount of sporadic flow to the radiator as the temperature varies between 78-80’c. A car runnig petrol will probably be 5’c hotter at the thermostat and send a higher proportion of water to the radiator as a result. Under steady state conditions, there is no reason not to expect the E-F-G pattern to not occur. The rate at which it climbs will be a function of load and rpm and the rate at which it drops will be a measure of how strong the radiator and fans are and the flow to and over them.
Given the above logic, if an 88’c thermostat is seeing extended elevated temperatures, I’d speculate that the bypass route is not being fully shut off when the thermostat is opening to the radiator. Conversely I’d also not be worried too much or expect it to be cycling in a pretty sine-wave like manner once the car is moving and the load and rpm are changing.
kind regards
Marek
Hi Marek
The Haynes manual has three separate lines with the thermostat data on them…so I think the fully open info is for all the thermostats.
But…nevertheless…the trace of the water temperature looks like a slightly under damped control system response to me.
From the info in the manual I am inferring that the Jaguar Engineers were happy to have the temperature fluctuate about 14 degrees on the early cars. And to have set a maximum running temperature at about 95-96 deg C. This is comfortably below the boiling point of the coolant at 15psi of about 120 deg C…but I wonder what the effect of the 120 deg C would have on my valve seats in the relatively highly expanded heads. To me that is the “Do Not Approach” temperature…but I don’t know if anyone has a feel for this magic number.
Regards
Matt
Mine has only ever got to 95-96’c once and that was when the fans failed to come on because of a faulty inline fuse in the engine bay whilst in stop-go traffic. Just because the thermostat is fully open at that temperature, it doesn’t really say anything about the maximum expected running temperature, What the graph shows is that the temperature will move about but always try to return to its floor temperature - in this case 82’c.
Even applying a full 14’c range, you should only expect to see an 88’c stat’d car go to 102’c and you’d expect the fans to be running full pelt, so getting above that sort of temperature beggars belief that the coolant system isn’t under strain.
kind regards
Marek
So Marek, can we then assume that if you were to fit electric fans to an 82C car, the ideal would be that at least one of them starts when reaching 82, and all fans at high speed should engage at 96?
Without overthinking it, when below 82’c, the thermostat is shut because we wanted to be at least at 82’c. Once we are at 82’c the thermostat starts opening and that’s because we don’t really want to be hotter than 82’c. The further away from target we are, the further it opens and if the switch at the bottom of the radiator says we didn’t cool the water enough, it cuts the fans in on top of that.
Given that the thermostat is slightly out of sync with the returning water and the engine load, logic says you want lots of cooling as soon as you are above target.
Given that there is a time lag involved in seeing how well the radiator does its job, I’d say you’d simply cut both fans in at by about 88’c. I see no logic in allowing the temperature to go to 96’c for any reason - it wants to run at 82’c not meander about all over the place, so the closer you make it run to 82’c the better.
The datalog above has the fans cut in at 87’c and turn off again at 83’c, which they will do if the radiator hasn’t already dealt with things. Very occasionally in stop-go traffic, the temperature creeps up past that level, but it is a rare event. The cooling system is entirely stock.
kind regards
Marek
you guys can sure overthink things!!
here is a small excerpt from TWR racing of XJS V12, i do believe they had some knowledge of the overheat subject.
Ron,
That’s a racing car - it has different cylinder heads, a different cooling system, it runs at 5000-7000rpm exclusively, it probably doesn’t idle well and probably only idles when it’s sitting broken in a gravel trap. It might not even have any thermostats fitted.
It also has the benefit of being rebuilt between races if necessary.
The answer to the title of this thread is that the operating temperature is going to be yo-yo-ing just above the thermostat temperature. The more efficient the cooling system, the tighter that range will be.
kind regards
Marek
Makes sense to me. That´s a good range:
Start to kick in at 83, fully kick in at 87.
Dear Carlos,
I think your fans will simply run all of the time if you do that. The radiator cooling on its own should happily deal with initial (and most) cooling requirements - the fan is only needed if the returning water is estimated to be still too hot, i.e. if the radiator on its own isn’t doing enough. (Some cars with aircon may well have fans that do run most of the time, but that’s not because the cooling system needs it -it’s because the aircon system needs it.)
In my example, the fans turn on if it rises to 87’c but once they do, they keep going until the temperature has been dropped back down to 83’c.
kind regards
Marek
Again, again, again: Operation of the electric fans should NOT be governed
by temp at the thermostat housing. It should be governed by temp at
radiator outlet, which should be perhaps 15 degrees cooler. If you wait until
82C there, you’ve waited too long.
– Kirbert
Kirby
Maybe maybe maybe…I think the most accurate strategy is to measure the temperature of the water just after it has passed thru the hottest part of the engine…this is closest to the action and is the most accurate bit of data to the temperature that’s needs to be rejected from the cooling system. The temp after the radiator will vary depending on engine speed and fan on or off and air temp…and you can’t determine how much cooling to apply because you don’t know how much work the engine is doing…but measuring the output temp you KNOW how much heat you need to get rid of…and in the 80’s you could only switch the little fan on or off…but that would move air thru a new part of the radiator…and was close enuf for guvmint work back then…
Matt