Thread: Radiator performance and Thermostat operation

One problem.
Usual TD5 thermostat opening is 82C fully open 88C.

The temp out of the radiator on mine must be less than 82c as it usually sits at 82C to 85C at dry sensor on mounting bolt of wet ECU sensor. Heat soak yesterday in 31c to 85.5c after climbing steep hill.
Lets say 82C on a 28C day at 65KMH.85C idling in traffic.
about 85c towing caravan at 100Kmh on flat rising to 90s on significant hills.
No direct correlation to EGTs.
Regards PhilipA

Originally Posted by RRT

.....

In discussions, there was comparison with the TD5 Coolant sensor (which is dry mounted at the waterpump outlet) readings versus the DC Coolant Sensor (probe in the coolant itself just before radiator inlet). Having been closely monitoring the temps on the EWPC during driving as well as having Nanocom logs, there are noticeable discrepancies between the two. These can be with EWPC over 15C higher against the Nanocom Logs but mostly around 10C in normal driving. The logs correlated pretty well to what others have indicated, temps around hovering around the mid 80C up to low 90's but I also have temps going down to the 60C's as well .......

The coolant sensor for the TD5 is in the top housing(rad inlet) at the head. It is also 'passthru' in that the sensor is in the coolant stream(only just .. maybe a few millimeters .. can't remember exactly).
So technically the temperature reading of the TD5 coolant sensor should be the average of the temperature of the head.
If you use an IR temp gun around the head, you should see variable readings .. hot spots if you like .. some parts cooler, some parts hotter.
But the TD5s' std sensor should give you a fairly accurate reading of the temp of the head, as it's right there where the coolant flow comes out.

As I remember, your waterpump was fairly badly gummed/rusted up. If you haven't triple checked the condition of the outlet and sensor itself .. it could be affecting the readings of the sensor as it may also be gummed up.
I mention this only because you haven't strictly commented whether you have or haven't checked the outlet and or sensor for condition. There could be other reasons for coolant temp readings being 'out of whack' where wiring is a bit dodgy. IIRC the coolant sensor works in that as temps goes up, resistance goes down. This resistance is measured at the ECU.. so if you have lower than actual/normal temp readings could be higher resistance through the wiring.

You say you have temps going down into the 60's ... from the nanocom(ie. TD5s' sensor .. or the EWPC readings?
If the nanocom ... maybe time to check the temp sender and/or wiring.

If your temp sensor for the davies craig system is at the rad inlet .. which is not far from the TD5's sensor, I doubt you should see much, if any difference in temperatures between the two. Possibly some differences due to calibration or reaction times by each sensor, maybe.

Now that you have the DC sensor as another baseline, and if you had a thermometer to play with too(they really are handy tools to have, and 'geek around' with! ) ... under engine management TD5 in RAVE .. it has a table of resistance values for the ECT at various temp stages .. I'm sure with the stuff you've done recently, surely you have a multimeter to play with.

Also, once you've got to a happy place with it all, would be interesting to see results you get in various conditions .. whether highway driving, backroad driving, sandy tracks .. etc.

eg. I recently went for a short drive with the missus, for no reason, turned off the freeway and headed past Clonbinane, up into Disappointment forest area .. not for any wheeling, just meandering about.
Just prior to the top of the slowish climb up the backroad and into the state forest .. the windy climb(about 60-70 k/h type) .. where the Td5 I had would normally sit at 90°C in just about every flat road driving situation(A/C on or off) .. up this hill, it peaked just over 100 for a short slowish power demanding few minutes.
Personally I didn't like it.. I don't worry about 100°C temps too much, but so quickly and not towing ... I'd extrapolate that coolant temp experience into one where hard driving through the Big Desert/Murray Sunset NP sand dunes in warmer ambient situations .. AC may definitely need to be off .. dunno, never done it(yet) .. but I wasn't confident that it'd be fine.
I got a new thermostat .. just need the time to fit it one day.

So, take home from that short story is : would be interesting to read a specific condition/situation where it's run cool/hot .. and what road and what ambients .. and so forth.

Thanks gents, the question is, how much does the radiator cool the flow and if we are 85C at the head, then it is unlikely the thermostat would open at all if the radiator took off 10C or, is this too much to expect for a radiator?

As to the point made about the sensor and yes my Waterpump was a catastrophe, when first started up after changing the head and installing the system, had the EWC and Nanocom pictured together where the Nanocom was rising fast but the DC Sensor was not, they eventually met around 90C if I remember (if the pics are out of sequence its more the differentials anyway).


20211123_201845.jpg

20211123_200856.jpg
20211123_202813.jpg

In discussions with DC they confirmed that the head sensor are not an accurate coolant temp sensor and as you surmised is more of the head temp and therefore to expect a higher reading from the fluid sensor taking its reading from the internal flow which has been the case. The head sensor has occasionally not worked but by and large it does appear to match the readings of others.

The DC sensor is the one located closer to the radiator inlet, the other sensor is low fluid

20211204_121332.jpg


I am getting away from the question initially asked though and do not want it be about the EWP and EG necessarily. Sure it reason I am asking as to try and figure if the system is genuinely viable. The fact that in nearly all the forum posts as to temps they have taken the TD5 as gospel as to temps and I assert that this is a falsehood due to my experiences, then them being in the mid 80's where they sit on the Nanocom, the thermostat would not be operational for almost 90% of the time due.

If as I have found with what I believe the real temps to be at the radiator inlet in normal running being actually mid 90C's and this would at least open the thermostat but tends to indicate the drop in only around 10C, not nearly as much as I would have thought.

So as to my original question, what is the temp drop across the radiator expected to be?

I was thinking about your issue with wanting to know the temperature loss across a radiator.
I don't think it would be possible with the equipment I have ( laser temp gauge) as you would have to remove the fan shroud to be able to get a reading on each end, and that would change the reading. Also The viscous fan would most probably have to be removed, again nullifying the reading.
To me you would need a dry sensor at the fins near the inlet and a sensor on the fins near the outlet, and even that could be affected by the conductance of the aluminium.
Ideally a wet sensor in both places but who is going to ruin a radiator for that?
maybe the nearest could be a laser on the inlet hose and the outlet hose.
And the temperature drop would depend on the ambient temperature.
Regards PhilipA

If you use an IR temp gun around the head, you should see variable readings .. hot spots if you like .. some parts cooler, some parts hotter.

Yes I recall several years ago I read the temperature at the rear of the head , as someone had a gauge sensor there vs Nanocom and it was 5C hotter at the back than the Nanocom reading. I don't know the accuracy or his gauge however as I was helping someone out with ABS codes.

Using a thermistor attached using thermal glue to the radiator metal adjacent to the radiator outlet on both a TD5 and P38 I can confirm that on a hot humid Queensland day idling in the traffic the outlet value is 63-65 degrees - both with new radiators and genuine good viscous fans. Its always about 30 - 35 degrees below the coolant leaving the engine. On cold days or cruising the margin is higher with the lowest outlet temperature I have seen being 55-60 degrees.
Regarding the TD5 I also have a thermistor fitted into the lower crossover pipe immediately after the thermostat - this sits at 75 - 78 as it is a mix - see attached.
This value is the most critical as it determines the amount of vapour in solution as the coolant enters the very low pressure region of the water pump inlet.
The water temperature inside the oil cooler is the critical one as it gets very hot with the new radiator mod - up to 115 degrees - see attached images.
Pressure and flow are the most critical of values - you need to check your coolant overflow bottle cap - all the copies I have tested let go at 7psi which is below the correct spec of 140kPA or 20 psi. The only caps that passed spec are those with Reuter stamped on the rim next to the 140kPA stamping. My system sits at 17psi with a Reuter cap when towing my tinnie(750kg) north.
Boiling point with 50% coolant and 7 psi is only about 120 degrees and with 20 psi it is 135 degrees. ARE Cooling
Coolant flow rate is very important - can you alter flow rates on the new electric pump? IT would be worthwhile to collect outlet temperature date with variable rates of coolant flow.
Attached is also some data on cooling performance and mods when Jeep changed the coolant pump when trying to cool their diesel model - pump inlet diameter/speed and flow rate were critical to make it work in tropical environments.
Hope this helps - I have more info if you want to PM.

Originally Posted by peter51

Using a thermistor attached using thermal glue to the radiator metal adjacent to the radiator outlet on both a TD5 and P38 I can confirm that on a hot humid Queensland day idling in the traffic the outlet value is 63-65 degrees - both with new radiators and genuine good viscous fans. Its always about 30 - 35 degrees below the coolant leaving the engine. On cold days or cruising the margin is higher with the lowest outlet temperature I have seen being 55-60 degrees.
Regarding the TD5 I also have a thermistor fitted into the lower crossover pipe immediately after the thermostat - this sits at 75 - 78 as it is a mix - see attached.
This value is the most critical as it determines the amount of vapour in solution as the coolant enters the very low pressure region of the water pump inlet.
The water temperature inside the oil cooler is the critical one as it gets very hot with the new radiator mod - up to 115 degrees - see attached images.
Pressure and flow are the most critical of values - you need to check your coolant overflow bottle cap - all the copies I have tested let go at 7psi which is below the correct spec of 140kPA or 20 psi. The only caps that passed spec are those with Reuter stamped on the rim next to the 140kPA stamping. My system sits at 17psi with a Reuter cap when towing my tinnie(750kg) north.
Boiling point with 50% coolant and 7 psi is only about 120 degrees and with 20 psi it is 135 degrees. ARE Cooling
Coolant flow rate is very important - can you alter flow rates on the new electric pump? IT would be worthwhile to collect outlet temperature date with variable rates of coolant flow.
Attached is also some data on cooling performance and mods when Jeep changed the coolant pump when trying to cool their diesel model - pump inlet diameter/speed and flow rate were critical to make it work in tropical environments.
Hope this helps - I have more info if you want to PM.

Thanks Peter I applaud for you have done a lot of work and this can only take us forward!

As you note the danger temp of the system when it starts to vapour is a lot higher than a lot of people realize. The ingrained 100C at sea level being boiling point and therefore as being a danger, well it would be in a non pressurized system, but still leaves 20% headroom. This is where the non water systems seem so interesting as their vapour points are a lot higher again. I note that my cap has the 140kpa rating.

Again without making it about the EWP, it works by pulsing in the initial instance and as the temps approach the set thermostat temp, it pulses faster until it gets around -5C below set temp where it starts flowing and as temps rise, increasing the flow in relation to temps, until full flow. This flow sits between the two amounts on the Jeep info as I have the largest available 150LPM which was over and above DC specified one for this engine size, they have simplified this down and it is now the recommended one.

1626064259.EWPEBPSelectionTable12-July-2021.jpg

02bc14f29ca10f1ca8f26950bb8457ecd1107a25.png
There are some interesting points to see in the Jeep chart by the way, the coolant temps shown would be to most on this forum, HORRIFIC and CATASTROPHIC if seen on a TD5. I would suggest that it goes with my surmising that the TD5 is a lot better engineered than many give credit for and the Jeep is designed for climates that match and in Death Valley exceed the demands of greater Australia and not 'just Europe' heard oh oh oh oh so often. Taking out the ridiculous 145kph speeds and a more applicable 88kph up a 6% grade the inlet temps are 104C (matching my 'worse' test yesterday) and 117C which I needed a fire under the sensor to achieve LOL

The other point I would make is that the revs being at least 3700rpm on this are not indicative of TD5 working speeds as to waterpump speeds. Take a third off and get it to 2500rpm and the numbers for the 88kph speed would have proportional flows if it is linear of 78lpm and 133lpm, hmmm, hmmm.



I am not sure though that your lower temps are an accurate representation of the actual outlet 'coolant' temps somehow.

I say this for two reason,

• First the thermostat would never ever ever open as it would take an upper temp at over 110C
• Second anyone who has had a stuck thermostat knows, it boils very very quickly as many could attest to once heat gets in the engine.

Something you didn't qualify was how the upper temps were measured and if it was via the standard sensors, you are therefore seeing the upper range of temps that many others have indicated on many forum post is almost the danger zone. Would say though that in my findings and it has to be noted as anecdotal for fairness as I do not have these figures recorded from the EWPC, that at lower speeds the Nanocom and actual fluid temps do meet somewhat, separating again with air flow through the engine bay

Even more interesting figure is to how would the the oil cooler temp be so high as this would be 20C higher than the radiator inlet temps.

Don't think this is out of school, went back through the correspondence with DC product engineer and this was post first drive home on the 12th December, will include the Nanocom log, where I had set the temp at only 80C. No tune except the basic start it up tune Shack provided for the changeover to the NNN, but all other 'parts' in place, this was sent to DC by the way. IT was a 26C day.

HOMEZ (2).CSV

"...Looking at the results, It looks like the effort to re-locate the pump and redo the Fan mounting have fixed the issues.

In hot conditions I would totally expect the temperature to run a bit hotter when in traffic as the A/c will lower the effective ness of the cooling system.
I would recommend setting the controller hotter as running in the high 70’s is almost too cold as the factory thermostat for the TD5 is 88C.
This would also explain why the engine would get up to 87 when in traffic on a hot day, because this is where is designed to run at.

NOTE: we recommend setting the controller to 5c Above the rating of the thermostat so 93C in your case, if you want to run it a bit cooler I recommend an 88C setting.
I would expect the normal operating temperature range to be 90C-100C for the factory system.

Regarding the result of your Nanocom,
You will not get a good comparison between he EWP controller and the Nanocom because the sensor is NOT measuring the coolant temp.
It is measuring the temperature of the cylinder head/engine block the temperature of these parts is not going to move relative to the coolant temperature
This is because the heat coming from the exhaust, combustion chamber, intake air and exhaust gases. This heat is then being removed by the coolant.
So unless you can log the actual coolant temp is will always be hard to get an accurate comparison."

he was the bloke that stated categorically to mount the EF flush to the radiator by the way.

Suppose for discussion

• even mounted to the fins of the radiator there is still going to be discrepancies due to airflow over the sensors as the sensor will have about 60% not in contact anyway.
• Suppose similar to the exhaust manifold, the manifold metal itself is not rising to 700C and then back to 500C in seconds and neither does the block
  

Sensors mounted both top and bottom should give a good approximation I suppose now thinking about it as the Delta is what I am after all said and done. Going to see whether I can find a cheap twin set of sensors / probes out there to stick into the radiator. But then again, a temp gun (as I have no fan or shrouding to get in the way) I can shoot the top and bottom corners on a stabilized system. Get back to you gents!

The TD5 stock coolant sensor measures coolant, the only time it measures head temperature only, is when you have run out of coolant, and then it will do it very poorly indeed.

The head temp will be different to block temp in almost all situations, and pretty much always significantly hotter than the block.

The coolant going past the stock sensor will be the hottest coolant in the system(at operating temp of course).

A quick test with an uncalibrated IR temp gun shows about 10°temp difference across the top of the radiator.... YMMV.