Assessing AJ16 Thermostat Efficiency

[ptjs1]

For some time now, I've been concerned that my thermostat may be jammed partially or fully open. As we know the minor gauges on facelift cars are not known for absolute accuracy and my temp gauge does seem to have been reading consistently low over the last year or so.

So, today I've been trying to measure the temp during warm-up using two k-couplers hooked up to a temp reader. However, I'm not sure of the validity of the results I'm getting.

Therefore, I wonder if someone who understands the precise component locations of the AJ16 engine could give me some opinions on the following:

- Is it realistic to tape the k-couplers on the thermostat housing, one on the engine side ie right by the sensor, and one on the curved surface just on the rad side of the housing? Or will I get thermal dissipation across the alloy of the housing such that the readings appear to be similar even if the stat was shut during warm-up?

- Would I expect the top rad hose to warm-up in the first 5 mins of idling if the stat was fully shut, i.e. is there some inevitable thermal dissipation into the top hose when the stat is shut?

- How different would I expect the readings to be if taken on the outside surface of a rubber rad hose compared to the external surface of the alloy stat housing?

Tks

Paul

[BrianG]

Paul, I really wouldn't bother with trying to assess the validity of the thermocouple measurement placements. Thermostats do not last forever and they are not really that expensive, so just change it! On both my X308 and X350s, when the temperature gauge started reading low a change of thermostat brought an instant solution to the problem. They only seem to last reliably for about 5 years in my experience, so apart from the fag of changing one (depending on car model), if its age is uncertain I would always advocate a new one to eliminate uncertainty.

[Tintin]

Hi Paul

I replaced my thermostat last year and will gladly fire up my engine and take a few readings with a portable thermocouple and meter if that would assist - comparison with another engine may help resolve your doubts. I have been meaning to invest in an infrared thermometer for situations like this - rather easier than taping sensors to components - but haven't got around to it yet. My gadget is pretty basic, but I guess it is the temperature difference you are looking at, not the total value.

Robbie

[ptjs1]

Brian & Robbie,

Tks for your feedback. I agree that it's just easier to replace the 'stat although I think I replaced it about 5 years ago anyway. Still, I've now ordered a new one and will fit it tomorrow.

I ran some more tests today but still not sure of what it's telling me. I started off by wiring up 4 k-couplers to both sides of the 'stat housing, the top hose and the cylinder head. I only have a 2-channel reader and wanted to monitor the 2 sides of the 'stat housing but also wanted to occasionally x-reference the top hose and the cylinder head:

I then fired up the engine, ran the aircon to speed up the process (Annoyingly for the 4th year running, my aircon isn't very cold!) then connected my OBDII plugin so I could monitor the ecu-read coolant temp also.

The two readings on the meter represent the engine side of the 'stat housing followed by the top hose side of the 'stat housing. My phone shows the actual coolant temp as read by the ecu. This is also from the engine side of the 'stat housing. There was a differential between the two halves of the 'stat housing but the top hose side did seem to be heating up immediately from start-up. don't know if that's from a partially-open 'stat or should I expect that from the bypass of coolant through the "jiggle-hole"? As the engine temp increased, the differential remained broadly proportional and the coolant then reached 82c.

The top hose side of the 'stat housing was starting to get a little closer in temp to the engine side of the housing and I expected them to suddenly equalise as the 'stat fully opened but that didn't happen! The coolant temp didn't go any higher than 84c after 20 mins of running the engine and the temp gauge didn't go much higher than about one quarter up the scale.

I suspect there may be two problems: a 'stat that is partly open all the time and which never fully opens but still stops the engine from reaching full operating temp. And also a gauge that should be reading a bit higher even at only 84c. And of course, there's that pesky aircon which probably needs regassing and a leaking seal replaced!

So, I'll replace the 'stat at 9am tomorrow morning, head off to watch Chelsea beat Tottenham at Wembley and then see if the car runs properly on the Drive-It Day run on Sunday.

Robbie, Thanks for your offer of monitoring your own engine. If you do have 2 couplers that you could place in the same places as I did on the 'stat housing, it would be interesting if you got the same results. But please don't feel obliged to do it.

Paul

[oldtimer]

Paul,
I have read through this thread a couple of times and cannot resist putting my penneth in.
As I understand, your concern is that the water temperature gauge is not reading what you believe it should. So your dilemma is whether to believe the gauge or suspect the thermostat - this I can understand.
Just to remind folks, the gauge contains a bi metal strip with a heater wire wrapped around it. By passing a current through the pizzo wire (as in an electric fire) a rise in temperature takes place and causes movement of the bi metal strip and with certain internal arrangements a needle deflection takes place. The current flow through the heater wire wrapped around the bi metal strip in the gauge is governed by the so called sender ( in fact it is not a sender) it is a thermistor which is a temperature sensitive resistance. By carefully cutting the tip off a redundant thermistor will reveal a small carbon looking disc about .005 thou thick. This disc is the secret to the whole system. The matching of the thermistor to the gauge is crucial. It becomes fairly obvious that if the resistance alters then this will effect the current flow to earth and thereby alter the gauge reading. In your case it could be something as simple as a dirty connection.

Each thermostat contains a wax with copper dust mixed in to improve conductivity. The Wax remains totally inert until the temperature stamped on the end is reached. The purpose of the thermostat is to govern engine water temperature at the level as stamped on it. So, on engine start up the stat will be fully closed with the water flowing through the by pass arrangement. No movement will take place until water reaches the stat operating temperature, the stat will only open sufficiently to maintain spec temp. Depending on ambient temp and loads imposed will govern the opening response. At this point I should point out that vehicles are / were tested in hot climates so the margin for overheat should not be reached in our temperate climate. The sole purpose of the stat is to maintain water temperature at the specified level , should the cooling system have a secondary problem the stat will continue to open to a total lift of 10 mm . Beyond that point the stat is saying you are on your own now mate I cannot do any more. So , that is when it is followed by a boil up.

From what I remember Warranty faults were very low and in most cases claims were rejected. Excellent test facilities were in place and without going into to much detail , silicon fluid was used in the test bath because operating temperatures got higher and higher and a DTI is required to determine the exact start to open point.

If one is determined to test it simply have a pan of cold water and one of boiling water, tie a piece of string on to the stat and transfer from hot to cold several times. Ensure the door lifts by 10 mm and hold up to the light to ensure fully shut.

If I needed to remove the stat I would bin it and buy a new one

I think that is enough rambling on for one night - past my bed time.

I hope that helps just a little in your reasoning

Oldtimer

PS nearly forgot, Get yourself an infra red thermometer and check the temperature at the bottom hose

[Zennan101]

Thanks Oldtimer,
Very clear explanation. I have learned something about how the temp gauge works. Trick is remembering it!

Rod

[ptjs1]

Oldtimer,

Thanks for your input. Useful info and of interest to many members I'm sure.


Ok, new stat fitted by 9am and definitely necessary! I'll post pics and details later. Off to Wembley - come on Chelsea!

Paul

[J44EAG]

The thermostat maintains a minimum operating engine coolant temperature, NOT the maximum.

The maximum temperature is governed by the amount of "work" produced by load placed on the engine, the ambient local temperature of the day and the efficiency/size of the coolant to air radiator. Too small a radiator and the engine will run hot or boil up. Too large and the engine will not reach the usual useful running temperature of a typical 91-93 degree figure.

"My engine is getting hot and so am I. What do you want me to do? Switch off, boil up or blow up?"..... Michael Caine in the Battle of Britain film, sat at dispersal, waiting for permission to take off on a hot sunny day. A Mk2 Spitfire was never blessed with a decent sized radiator.

Mike

[oldtimer]

Mike ,
Sorry , you really are totally wrong -- try reading my thread above again

[Brakebuster]

i think we are talking at cross purposes here

the engine thermostat opens at a set temperature for efficient engine running ( a relatively cold engine is not efficient )

once open it plays no further part in maintaining engine temperature for that cycle

a poorly performing or non working thermostat will cause issues with engine temperature once it goes past the set point of opening , i am not talking extreme temperature environments here though...

poor preforming stat produces irregular warmup cycle

non working stuck open stat results in relative cold engine until whole system gets to ideal temp ( it if ever gets there )

non working stuck shut thermostat will eventually result in engine overheating

thats how i see it

BB

ps , this is not talking about how a stat is built or operates, and or how big or efficient the rest of the cooling system is on any particular vehicle



Paul, the only real place to take an accurate temp reading is on the water rail to the pump , the thermostat housing has no flow other than from the jiggle pin until the stat is open ( which you suspect faulty ) , and as such is not reflective of ' engine temp '
i would also allow a small amount of time once the stat opens so the slug of ' cold ' water in the rad has had time to warm up to system temp ,

the method of temp taking is personal to you , i personally have a IR reader ,

[Tintin]

This is fascinating stuff and very informative but I don't think it answers the original post. I ran up my engine this morning from overnight cold and allowed it to idle for about 20 minutes. I took temperature readings every 2 minutes. The readings I got from the water pump housing and cylinder head side of the thermostat housing were almost identical. I also took readings from the top hose. The results were as follows.

The starting ambient temperature was 15C.
After 4 mins the housing was at 36C and the hose at 26C.
After 8 mins the housing was at 50C and the hose at 36C.
After12 mins the housing was at 58C and the hose at 38C.
After 16 mins the housing was at 70C and the hose at 50C.
After 20 mins the housing was at 65C and the hose at 46C.

The hose readings were taken from the top surface - the under surface was several degrees cooler suggesting a bleed effect from the joggle pin for the first 16 mins. At 16/18 mins there was a pronounced change - the whole hose became hot and the temperature both sides of the stat fell as the coolant in the engine primary circuit and radiator secondary circuit mixed. A correctly operating stat (mine was new last year) causes a sudden temperature change after about 18 minutes and 70C (subject to ambient temperature and other variables) as the coolant mixes on stat opening. A stat stuck partly or fully open or closed will show appropriately different temperature differences.

Robbie

[Brakebuster]

i think we are trying to apply laboratory temp readings in the real world , and comparing apples with oranges

the starting ambient temps will be different from car to car ( cold is not cold, its relative) , the coolant types and mixtures are different with different thermal dynamics , the engine revs wont be exactly the same, even 50rmp on idle will alter results as the engines will reach temp at different times, and so on...

at the end of the day , the stat opens at a set temp , ( written on the stat with a degree of built in tolerance +/- ) if it doesn't do this , it needs replacing ,

BB

[ptjs1]

Robbie,

Thanks for doing that. It was very good of you. As you have noticed there is approximately a 16 degree difference when the outside alloy of the stat houses reads 70c compared to the coolant temp inside.

Your readings seem to indicate two things:

- That your 'stat started in a closed position and opened close to 88 degree coolant temp as expected.
- That the natural heat transference and bleed valve mean that the top hose does start to warm straight from cold although mine was obviously warming more than yours, implying a partially open 'stat.

As I mentioned, I fitted a new 'stat yesterday morning. Thanks to Mark at Guy Salmon who opened up at 8am so that I could collect my parts and get them fitted before I had to leave home at 9.30am

With the stat housing cleaned up, the new 'stat and a new seal fitted, I started the engine. Within 7 minutes it warmed up straight to an ecu-read 93c, dropped to 86c as the new 'stat opened and then sat and cycled continuously between 86c and 91c. Very different to before! The temp gauge now sat much closer to the mid-point. I checked for leaks then left to watch Chelsea beat Tottenham, Oh yes!!

Today, I went out on a 100-mile drive with 120 Triumph TRs, great drive! At 50mph the temp cycled continuously between 87c and 91c. At 75+mph, it cycled continuously between 90c and 93c. That's about what I would expect. The temp gauge also sat just fractionally under 1/2-way which is where it used to be a year ago. However, at idle, the gauge drops a bit as my minor gauges do under-read at idle so I probably have the usual slight earthing problem either at the bulkhead or at the instrument cluster. One more thing to look at.

So that should be the problem solved with a partially-stuck open stat replaced. But....this afternoon I looked at the old 'stat. It was definitely closed after I extracted it from the car. Obviously, during the job, I removed and discarded it quickly to avoid too much coolant loss. But it was definitely closed when I picked it up off the floor. So, I put it in some 90c water and it opened! I tried it 4 times and it worked every time. If you look closely at the second pic you can see it starting to open:

So now I'm confused. Did it get stuck open in the car, freed itself as I removed it and is now working perfectly? Do I have a working spare to keep or should I bin it? Who knows?!

As regards the thoughts about using an IR thermometer. Yes, I do have one but I used the k couplers for two reasons. Firstly, I wanted to measure at least two places simultaneously and was also potentially going to do the measurements whilst driving the car. Therefore I really needed to do it with k couplers taped in place. Also, I was going to take a lot of readings and I've read a lot about how IR readings can vary significantly unless you take them from exactly the same distance every time. The "focal length" can give rise to false readings unless you constantly get the distance right. So, horses for courses.

Just to conclude with my naive view on the role of the 'stat. As we know, it can't actually absolutely control the temp. I view it as trying to hold the temp at an exact temp by allowing the car to warm up to that temp by staying closed, opening itself to then let the coolant flow so that the coolant around the engine doesn't get too hot, and closing itself again when the temperature of the coolant drops below that ideal temp. In that respect, it does a pretty reasonable job.

Tks to everyone for their comments and especially to Robbie for his practical help. Good debate guys! Another one that I'll probably get around to writing up for the magazine.

Paul