The Function and Malfunction of the Diesel Thermostat - Page 20

T-Cut · 23rd March 2013, 15:15

Another update on results from a failed thermostat. I was sent a stat from a 2004 diesel that had been replaced because of low running temperature (thanks Matthew). Visually it looks perfect, almost brand new. I assume this is the latest type fitted in production. It has a bright metal main valve with a thin rubber edge seal. The small air/jiggle vent is a small ball bearing loosely held in a pressed metal cage. This and the main valve are completely airtight/water tight from the outside. I decided the best way to detect opening is to see when it fails to hold water.

The stat was immersed in a pan of water and positioned so that the main outlet socket was vertical and this was filled to the brim with water. Only the plastic flanges of the housing were in contact with the metal pan. The wax capsule was completely submerged. A digital thermometer probe was placed in contact with the wax capsule.

The pan was heated slowly and the water gently stirred. The temperature rise was about 1-2°C/min. The water meniscus in the socket reservoir was watched and when it first started to drop the temperature was noted. I then repeated the test from scratch so I could take photos just before and just after this point.

Initial drop in level was at 55-56°C. By 57-58°C the water had all escaped.

So this clearly illustrates the problem. Cold coolant was bleeding into the engine from around 56°C.

I'll remove and dissect the capsule to see if the wax behaves like the first sample.

EDIT: The thermometer used is a brand new Silverline digital with an accuracy spec of plus/minus 1°C for the range 47-150°C.

EDIT2: I'd like to repeat this test on a brand new stat. If anyone has one they're willing to lend me, I'll run the test and return it to you.

TC

Mike Noc · 25th March 2013, 06:59

Good work TC. Will be interesting to see what you find when it is dismantled.

Jules · 25th March 2013, 07:38

Morning TC
Send me your address and I'll forward a new stat for experiments.
Will any cruelty be involved?

HarryM1BYT · 25th March 2013, 08:12

Quote:

Originally Posted by T-Cut

Notice the little lump.

I've more closeups to add when I get them all uploaded, but will have to wait for now.

I also have some accurate melting data which I'll add as I get time.

TC

I can see what looks to me like an air pocket as the wax has set, did you mean that rather than a lump?

The pocket appear in the photo to have a rough black + sign in it.

Jules · 25th March 2013, 08:15

Black + sign.
Must be the Devil

T-Cut · 25th March 2013, 10:44

Quote:

Originally Posted by Jules

Morning TC
Send me your address and I'll forward a new stat for experiments.
Will any cruelty be involved?

Cheers Jules. PM sent.

TC

T-Cut · 25th March 2013, 10:53

Quote:

Originally Posted by HarryM1BYT

I can see what looks to me like an air pocket as the wax has set, did you mean that rather than a lump?

The shape of the cast is lumpy looking as well as pocketed. The photos can be interpreted in various ways due to the lighting. The wax itself is also odd. I'll post more soon.

TC

T-Cut · 25th March 2013, 12:02

Reading back through this, there's not been a clear explanation of the secondary valve found in this stat. It's also present in the V6 stat and in the later (PRT) stat fitted to the 1.8 engines. The PRT operates on both temperature and pump pressure of course.

Anyway, this extract and photo are from Wikipedia and explain the basis of inlet flow temperature control.

Engines which require a tighter control of temperature, as they are sensitive to "Thermal shock" caused by surges of coolant, may use a "constant inlet temperature" system. In this arrangement the inlet cooling to the engine is controlled by double-valve thermostat which mixes a re-circulating sensing flow with the radiator cooling flow. These employ a single capsule, but have two valve discs.

Thus a very compact, and simple but effective, control function is achieved.
The wax used within the thermostat is specially manufactured for the purpose. Unlike a standard paraffin wax which has a relatively wide range of carbon chain lengths, a wax used in the thermostat application has a very narrow range of carbon molecule chains. The extent of the chains is usually determined by the melting characteristics demanded by the specific end application. To manufacture a product in this manner requires very precise levels of distillation.

The smaller spring behind the secondary valve (which floats on the piston/spindle) is to accomodate any overthrow by the piston. A properly functioning stat (like the early MikeNoc example) shows a pressure/wear ring on the closing face of the disc.

TC

Frank Incensed · 25th March 2013, 13:34

Quote:

Originally Posted by T-Cut

.............................Engines which require a tighter control of temperature, as they are sensitive to "Thermal shock" caused by surges of coolant, may use a "constant inlet temperature" system. In this arrangement the inlet cooling to the engine is controlled by double-valve thermostat which mixes a re-circulating sensing flow with the radiator cooling flow. These employ a single capsule, but have two valve discs....................................

The effect of my inline stat opening sees the OBD coolant temperature reading drop rapidly from 99 to 84 / 85. When I originally got the car, the temperature rarely got to 88 / 89 and the drop to its "normal" running temperature in the mid 80's was much more gradual. Were MGR just playing safe and not risking any further HGF debacle, or do you think we are in dangerous territory with this constant slow rise and rapid fall of coolant temperature, TC?

Cheers
Robert

T-Cut · 25th March 2013, 19:51

It's an interesting point. The in-line approach doesn't compensate for the loss of pump inlet flow mixing that the standard stat is designed for. Whether that's as important with the diesel as the 1.8, I very much doubt. Whether it's in dangerous territory I also doubt, but only time will tell on the mod. I suspect the last thing any diesel engine is likely to experience is significant thermal shock because the temperature gradients aren't that high. What you describe isn't alarming (IMO). High gradients don't seem to be the norm, but I'm sure someone will describe one. Anyone who feels their gradient is too steep could try a lower rated inline stat.

EDIT: Just thought a bit more on this. It seems to me that the inline unit is acting as a simple throttle for the main stat. The main stat is still operating as an inlet flow mixer. It simply starts doing it at too low a temperature (i.e. 'failed'). Looking at the opening range experiment, the OEM unit will probably be pretty wide open by the time the throttle stat is opened. This will increase flow of both hot and cold coolant into the engine. The result seems to me a hybrid of the inlet flow control and the older style outlet stats.

TC

23rd March 2013, 15:15	#191
T-Cut This is my second home Rover75 and Mreg Corsa. Join Date: Nov 2006 Location: Sumweer onat mote o'dust (Sagin) Posts: 21,753 Thanks: 341 Thanked 3,660 Times in 2,924 Posts	Another update on results from a failed thermostat. I was sent a stat from a 2004 diesel that had been replaced because of low running temperature (thanks Matthew). Visually it looks perfect, almost brand new. I assume this is the latest type fitted in production. It has a bright metal main valve with a thin rubber edge seal. The small air/jiggle vent is a small ball bearing loosely held in a pressed metal cage. This and the main valve are completely airtight/water tight from the outside. I decided the best way to detect opening is to see when it fails to hold water. The stat was immersed in a pan of water and positioned so that the main outlet socket was vertical and this was filled to the brim with water. Only the plastic flanges of the housing were in contact with the metal pan. The wax capsule was completely submerged. A digital thermometer probe was placed in contact with the wax capsule. The pan was heated slowly and the water gently stirred. The temperature rise was about 1-2°C/min. The water meniscus in the socket reservoir was watched and when it first started to drop the temperature was noted. I then repeated the test from scratch so I could take photos just before and just after this point. Initial drop in level was at 55-56°C. By 57-58°C the water had all escaped. So this clearly illustrates the problem. Cold coolant was bleeding into the engine from around 56°C. I'll remove and dissect the capsule to see if the wax behaves like the first sample. EDIT: The thermometer used is a brand new Silverline digital with an accuracy spec of plus/minus 1°C for the range 47-150°C. EDIT2: I'd like to repeat this test on a brand new stat. If anyone has one they're willing to lend me, I'll run the test and return it to you. TC Last edited by T-Cut; 23rd March 2013 at 16:07..

25th March 2013, 07:38	#193
Jules This is my second home ZT 260 SE Twilight and 10 other 75 ZT's :O Join Date: Jan 2007 Location: Conwy NORTH WALES Posts: 11,094 Thanks: 512 Thanked 1,116 Times in 763 Posts	Morning TC Send me your address and I'll forward a new stat for experiments. Will any cruelty be involved? __________________ [SIGPIC][/SIGPIC] Newbies do now!! 1. Plenum drains..all 3 or 4 year dependent 2. Cooling fan..All speeds functioning 3. Bonnet cable divider block

25th March 2013, 08:15	#195
Jules This is my second home ZT 260 SE Twilight and 10 other 75 ZT's :O Join Date: Jan 2007 Location: Conwy NORTH WALES Posts: 11,094 Thanks: 512 Thanked 1,116 Times in 763 Posts	Black + sign. Must be the Devil __________________ [SIGPIC][/SIGPIC] Newbies do now!! 1. Plenum drains..all 3 or 4 year dependent 2. Cooling fan..All speeds functioning 3. Bonnet cable divider block

25th March 2013, 12:02	#198
T-Cut This is my second home Rover75 and Mreg Corsa. Join Date: Nov 2006 Location: Sumweer onat mote o'dust (Sagin) Posts: 21,753 Thanks: 341 Thanked 3,660 Times in 2,924 Posts	Reading back through this, there's not been a clear explanation of the secondary valve found in this stat. It's also present in the V6 stat and in the later (PRT) stat fitted to the 1.8 engines. The PRT operates on both temperature and pump pressure of course. Anyway, this extract and photo are from Wikipedia and explain the basis of inlet flow temperature control. Engines which require a tighter control of temperature, as they are sensitive to "Thermal shock" caused by surges of coolant, may use a "constant inlet temperature" system. In this arrangement the inlet cooling to the engine is controlled by double-valve thermostat which mixes a re-circulating sensing flow with the radiator cooling flow. These employ a single capsule, but have two valve discs. Thus a very compact, and simple but effective, control function is achieved. The wax used within the thermostat is specially manufactured for the purpose. Unlike a standard paraffin wax which has a relatively wide range of carbon chain lengths, a wax used in the thermostat application has a very narrow range of carbon molecule chains. The extent of the chains is usually determined by the melting characteristics demanded by the specific end application. To manufacture a product in this manner requires very precise levels of distillation. The smaller spring behind the secondary valve (which floats on the piston/spindle) is to accomodate any overthrow by the piston. A properly functioning stat (like the early MikeNoc example) shows a pressure/wear ring on the closing face of the disc. TC Last edited by T-Cut; 25th March 2013 at 12:23..

25th March 2013, 19:51	#200
T-Cut This is my second home Rover75 and Mreg Corsa. Join Date: Nov 2006 Location: Sumweer onat mote o'dust (Sagin) Posts: 21,753 Thanks: 341 Thanked 3,660 Times in 2,924 Posts	It's an interesting point. The in-line approach doesn't compensate for the loss of pump inlet flow mixing that the standard stat is designed for. Whether that's as important with the diesel as the 1.8, I very much doubt. Whether it's in dangerous territory I also doubt, but only time will tell on the mod. I suspect the last thing any diesel engine is likely to experience is significant thermal shock because the temperature gradients aren't that high. What you describe isn't alarming (IMO). High gradients don't seem to be the norm, but I'm sure someone will describe one. Anyone who feels their gradient is too steep could try a lower rated inline stat. EDIT: Just thought a bit more on this. It seems to me that the inline unit is acting as a simple throttle for the main stat. The main stat is still operating as an inlet flow mixer. It simply starts doing it at too low a temperature (i.e. 'failed'). Looking at the opening range experiment, the OEM unit will probably be pretty wide open by the time the throttle stat is opened. This will increase flow of both hot and cold coolant into the engine. The result seems to me a hybrid of the inlet flow control and the older style outlet stats. TC Last edited by T-Cut; 25th March 2013 at 20:12..

25th March 2013, 06:59	#192
Mike Noc This is my second home Rover 75 CDT Manual Connoisseur SE, Rover 75 CDT Automatic Connoisseur SE & a Freelander Td4. Join Date: Jul 2009 Location: Hampshire Posts: 11,553 Thanks: 3,470 Thanked 3,119 Times in 2,247 Posts	Good work TC. Will be interesting to see what you find when it is dismantled.