Can someone explain to me the "lower temp thermostat"?

In reply to Vigo (Forum Supporter) :

People also buy janky chinesium coilover suspension systems with no travel, no damping, and no isolation in the mounts, because they don't care at all about ride or handling as long as they can get the look just right.

I take a large grain of salt about any modifications people do without a clear goal for the outcome and/or an understanding of how the system they are changing works or is supposed to work.

Colder thermostats do help for some things, like drag racing your mostly stock 5.0 Mustang, because you're idling it down to a base level between 13 second bursts of wide open throttle, and the way their engines are laid out does reward colder coolant.  Just because it works that way for them doesn't necessarily translate to other situations.

I wouldn't see a colder thermostat being useful on a Porsche unless there was a definite issue in the engine that can be traced to higher coolant temps, and it was one of the last steps in a multi-angled attack on the cooling system that started with better ducting to the radiator, larger radiator cores, reprogrammed fan control strategy,  and/or water pumps with better flow characteristics at high RPM, because until you get the cooling system to the point where it is hovering around the thermostat opening temperature, the thermostat is going to open and stay open.  (What good is a 160 'stat if the fans don't come on until 230?)

In reply to Vigo (Forum Supporter) :

Ok.  I had more typed, but what's the point.

Pete. (l33t FS) said:

alfadriver (Forum Supporter) said:

In reply to Pete. (l33t FS) :

Uh, I don't think the thermostat has any effect on the coolant system pressure....  That's more about the rad cap.  

Unless the T-stat is totally sealed- which I've never seen one that does not have a small bleeder hole.

BLOCK pressure, not radiator pressure.  The radiator is on the low pressure side of the cooling circuit.

A rule of thumb I have seen for racing engines is to size the water pump pulleys/restrictor orifice (racing engine = no thermostat) for 50psi pressure in the block.

I assume that suction side thermostats work at all because they are only ever employed on aluminum headed engines, and aluminum is a wonderful conductor of heat compared to iron.

Still, T-stats have leaks designed in them, so that the pressure in the entire system is controlled by the rad cap.  Let alone the return line closes the entire system with a big passage.  I'm not aware of a pump that's sealed enough to generate significant pressure.  

I understand why one would want to run high pressures.  

Maybe I need to look into cooling systems more...

alfadriver (Forum Supporter) said:

Maybe I need to look into cooling systems more...

Will probably make you annoyed at the very least. I still don't know how you generate 50psi in the engine with a lower pressure rad cap... I always put my cap on a surge tank at the highest spot of the system, with the return of that tank into the lower rad junction just before the waterpump. Maybe you can generate a little excess pressure between the waterpump and the engine but I can't see it being this big, at least with a "standard" cooling layout. Offhand I would think it's lower at the inlet of the enging due to the lower temp but I would have to start mapping it out to proof this any more.

At higher RPM the water pump can move quite a bit of water.  As an example, a stock later year (magnum) Mopar Small Block pump is said to move 100 gpm at 5000 RPM.  At that kind of flow rate, even with a less than perfectly sealed pump impeller, the restrictions presented by the thermostat and radiator will lead to a significant pressure gradient in the system.  That pressure gradient is why the lower radiator hose is typically bigger than the upper and also why lower radiator hoses can in some cases collapse at high RPM despite the system being significantly pressurized. 

This pressure gradient is one big reason why crossflow radiators are better than downflow.  Downflow puts the cap on the high pressure side of the radiator, crossflow puts it on the low pressure side.  So overall system pressure is higher for a given cap pressure on a crossflow system, as you don't raise the pressure seen by the cap at high RPM.  Basically, the block and heads see the highest pressure, radiator inlet is moderate, then radiator outlet (return to pump) is the lowest pressure.  The gradient is also why some setups have issues with water pump cavitation at high RPM.  Increased pressure delays cavitation, but the water pump inlet is the lowest pressure part of the system. 

A similar gradient appears in the heater circuit, where you're feeding higher pressure coolant (from before the t-stat) to the core, then you have pressure drop across the core and a return to the lower pressure suction side of the pump.  This is why some engines have used restrictors in the heater feed line to avoid rupturing the heater core with excessive pressure at high RPM. 

In reply to rslifkin :

wow... how much HP does it take to push 100 GPM of water? easily 6 inches of lift as well, that's a lot of work!!

Robbie (Forum Supporter) said:

In reply to rslifkin :

wow... how much HP does it take to push 100 GPM of water? easily 6 inches of lift as well, that's a lot of work!!

I'm not sure, but with the prevalance of underdrive pulleys and electric water pumps in racing applications, it's gotta be significant.  I'm betting the pump sizing was based on adequate flow at low RPM on the street. 

I'm enjoying this discussion.  The part that makes me happiest is the evidence that suggests German engineers didn't get it all right, which fits my world view quite nicely.

You need a cooler thermostat to compensate for a few different design failures.  If the cooling system is below required capacity, the low temp gives you a few more seconds, or minutes before you get too hot.  If the combustion chamber design lends itself to detonation at higher temps, or the programming is iffy, or the castings are not able to retain stability, or ..., a lower temp thermostat might compensate enough to keep you together.  Otherwise, 220 is a lovely temperature to run. 

I raced the E36 M3 out a twincam Neon, and it made no difference at all what the thermostat design temp was, it ran 220.  Didn't seem to matter if it was 60 or 100 out, it ran 220.  Car ran great, never missed a beat, never cooked an engine, except for that one time that the fuel pump died in the middle of a race, and I coasted into the pits with absolutely no cool down time at all.  No idea what temp the head got to with no circulation, but it warped the head enough that my next race weekend, the car would get really loose in turn one after about six laps.  That was when it started spitting coolant under the right rear tire.

Anyway, thats my story about how a Neon is better than a Porsche.

rslifkin said:

Robbie (Forum Supporter) said:

In reply to rslifkin :

wow... how much HP does it take to push 100 GPM of water? easily 6 inches of lift as well, that's a lot of work!!

I'm not sure, but with the prevalance of underdrive pulleys and electric water pumps in racing applications, it's gotta be significant.  I'm betting the pump sizing was based on adequate flow at low RPM on the street. 

There is a number here https://www.c1pulleys.com/pages/cooling-system-basics but it's almost too big to believe. Another calculator here http://irrigation.wsu.edu/Content/Calculators/General/Required-Water-Pump-HP.php even with some goosed up numbers of 50psi and 100gpm and low efficency it's double digits hp at max. If I use that second calculator to do the math on a 20gpm Meizere electric pump with 30psi and 90% efficiency 13.2v electrical system it's 21A load which seems about right (and only .4hp of load x the inefficiency of the electrical system so maybe sucking up 1hp vs overpumping with a mechanical pump).

rslifkin said:

At higher RPM the water pump can move quite a bit of water.  As an example, a stock later year (magnum) Mopar Small Block pump is said to move 100 gpm at 5000 RPM.  At that kind of flow rate, even with a less than perfectly sealed pump impeller, the restrictions presented by the thermostat and radiator will lead to a significant pressure gradient in the system. 

I will politely disagree here. You're confusing GPM with PSI.  A water pump is not technically a pump at all.  It's a fan that moves water.  If you were to take the outlet hose off that pump at 5000 rpm and lay your hand over the outlet, you might get a couple psi until your hand sealed over the aluminum at which point you would get effectively zero psi.  It's not a positive displacement pump.  It doesn't have to lift a single millimeter of water because it is circulating in a closed loop.  For every inch of water is has to lift, there is an equal pressure of water pushing down by what is already up there.  It doesn't matter if the engine is 10' higher than the pump, it requires no pressure to circulate the water because the head pressure seen at the outlet is exactly the same as what is seen at the inlet.

What you're suggesting is that the pressure of circulating water increases when the thermostat is closed. This is not possible.  A) because the pump doesn't generate any real pressure, and B) the block is a completely open circuit with minimal restriction.  The pump doesn't care if the stat is open or not.  It blows water.  It doesn't care if somewhere down the line the water is diverted through a radiator, a heater core, or an engine block.  Even if you closed all of those off, you won't have much if any pressure gradient. 

Even if you DID have some pressure gradient, a thermostat doesn't seal.  Then yet, on top of that, we're talking about a hydraulic system in which the fluids don't compress.  If you get 10 psi higher in the block, it only takes a few tablespoons of water to leak past the stat to equalize that pressure on the radiator side.

Think about it.... if your thermostat sticks closed and the vehicle overheats... what goes first?  The radiator cap... meaning that enough of the mass of coolant is still able to get by the stat and vent at the weakest point.

In reply to Curtis73 (Forum Supporter) :

You're correct that a centrifugal water pump is not positive displacement.  However, that doesn't mean it can't generate significant pressure.  Because it's not positive displacement, flow rate will decrease as output restriction increases. 

Here's a datasheet from a centrifugal pump as an example: https://www.defender.com/pdf/506420_Jabsco.pdf

At open flow, that pump moves about 21 gpm.  However, looking at the curve, against 12 psi, it'll still move 10.6 gpm.  At 16 psi, it's down to 5.3 gpm.  At around 18.5 psi, the output falls to zero. 

Streetwiseguy said:

You need a cooler thermostat to compensate for a few different design failures.  If the cooling system is below required capacity, the low temp gives you a few more seconds, or minutes before you get too hot.  If the combustion chamber design lends itself to detonation at higher temps, or the programming is iffy, or the castings are not able to retain stability, or ..., a lower temp thermostat might compensate enough to keep you together.  Otherwise, 220 is a lovely temperature to run. 

Agree and disagree :)

First sentence, yes.  Second sentence (detonation) disagree.  Temperature has almost nothing to do with detonation tolerance.  That combustion event lasts milliseconds and burns at about 2500 degrees.  It doesn't care if the water temp is 100 or 1000 degrees.  Detonation threshhold comes from HEAT exchange, not temperature.  If you have a combustion event (using made-up numbers) that converts a total of 100 joules from chemical to heat energy, 15 of those joules escape through the metal into the water, 10 of those joules soak into the piston and get absorbed by the oil, some are lost to sound and light, others will be used to overcome friction.... etc.

As long as your head chambers and the water behind it continue to absorb their responsibility of 15 joules, things will continue just fine.  It doesn't matter what the temperature of the coolant is.

When your aqueous coolant reaches 235 and you notice it is starting to knock, it has nothing to do with the fact that the water is 235 degrees.  It has to do with the fact that the chambers are the hottest part of the whole water jacket and they are the ones that start nucleate boiling first.  Everywhere there is a steam bubble stuck to the water jacket, heat can't transfer as efficiently and your coolant can no longer absorb the 15 joules of heat energy it is supposed to.

Now do that same exact experiment with a higher pressure cap.  You won't hear that knock until 245 degrees because the increased pressure has increased the boiling point of the coolant.  I have run non-aqueous coolant to 290 degrees and never heard a single ping.  It has nothing to do with the temperature, it has to do with ability to transfer heat.  Once aqueous coolant starts nucleate boiling, it loses it's ability to transfer the same amount of heat and you get detonation.

pre-ignition has nothing to do with the temperature of the coolant.  It has to do with the net heat being transferred out/being kept in.  If this were not true, then aluminum heads wouldn't be known for their ability to run higher compression.  It's not that they're able to have higher compression, its that the need higher compression to compensate for the fact that they might be facilitating 20 joules of energy transfer compared to iron.

rslifkin said:

In reply to Curtis73 (Forum Supporter) :

You're correct that a centrifugal water pump is not positive displacement.  However, that doesn't mean it can't generate significant pressure.  Because it's not positive displacement, flow rate will decrease as output restriction increases. 

Here's a datasheet from a centrifugal pump as an example: https://www.defender.com/pdf/506420_Jabsco.pdf

At open flow, that pump moves about 21 gpm.  However, looking at the curve, against 12 psi, it'll still move 10.6 gpm.  At 16 psi, it's down to 5.3 gpm.  At around 18.5 psi, the output falls to zero. 

The centrifugal pump you linked generates pressure.  The centrifugal pump on your engine generates next to none.  It of course has to generate some otherwise it wouldn't pump, but you can't just link to a random industrial pump that makes pressure and assume that all centrifugal pumps make the same.

alfadriver (Forum Supporter) said:

Pete. (l33t FS) said:

alfadriver (Forum Supporter) said:

In reply to Pete. (l33t FS) :

Uh, I don't think the thermostat has any effect on the coolant system pressure....  That's more about the rad cap.  

Unless the T-stat is totally sealed- which I've never seen one that does not have a small bleeder hole.

BLOCK pressure, not radiator pressure.  The radiator is on the low pressure side of the cooling circuit.

A rule of thumb I have seen for racing engines is to size the water pump pulleys/restrictor orifice (racing engine = no thermostat) for 50psi pressure in the block.

I assume that suction side thermostats work at all because they are only ever employed on aluminum headed engines, and aluminum is a wonderful conductor of heat compared to iron.

Still, T-stats have leaks designed in them, so that the pressure in the entire system is controlled by the rad cap.  Let alone the return line closes the entire system with a big passage.  I'm not aware of a pump that's sealed enough to generate significant pressure.  

I understand why one would want to run high pressures.  

Maybe I need to look into cooling systems more...

Well, I stuck a pressure gauge on an engine and played around with it... 

45psi was in a Vulcan engined Taurus at 3000rpm.  

Pete. (l33t FS) said:

alfadriver (Forum Supporter) said:

Pete. (l33t FS) said:

alfadriver (Forum Supporter) said:

In reply to Pete. (l33t FS) :

Uh, I don't think the thermostat has any effect on the coolant system pressure....  That's more about the rad cap.  

Unless the T-stat is totally sealed- which I've never seen one that does not have a small bleeder hole.

BLOCK pressure, not radiator pressure.  The radiator is on the low pressure side of the cooling circuit.

A rule of thumb I have seen for racing engines is to size the water pump pulleys/restrictor orifice (racing engine = no thermostat) for 50psi pressure in the block.

I assume that suction side thermostats work at all because they are only ever employed on aluminum headed engines, and aluminum is a wonderful conductor of heat compared to iron.

Still, T-stats have leaks designed in them, so that the pressure in the entire system is controlled by the rad cap.  Let alone the return line closes the entire system with a big passage.  I'm not aware of a pump that's sealed enough to generate significant pressure.  

I understand why one would want to run high pressures.  

Maybe I need to look into cooling systems more...

Well, I stuck a pressure gauge on an engine and played around with it... 

45psi was in a Vulcan engined Taurus at 3000rpm.  

You measured 45 psig ... and the radiator caps are usually rated below 1.4 bar ... shouldn't water be coming out of the reservoir as the pressure would overcome the radiator cap spring?

Curtis73 (Forum Supporter) said:

The centrifugal pump you linked generates pressure.  The centrifugal pump on your engine generates next to none.  It of course has to generate some otherwise it wouldn't pump, but you can't just link to a random industrial pump that makes pressure and assume that all centrifugal pumps make the same.

Put a pressure gauge on an engine and play around with it.  It's really fascinating.

I had a 385 Ford that was making 120psi by midrange RPM and was exploding 200psi burst rated heater hoses at high RPM.   You can generate some serious pressure with a water pump!

Slippery (Forum Supporter) said:

Pete. (l33t FS) said:

alfadriver (Forum Supporter) said:

Pete. (l33t FS) said:

alfadriver (Forum Supporter) said:

In reply to Pete. (l33t FS) :

Uh, I don't think the thermostat has any effect on the coolant system pressure....  That's more about the rad cap.  

Unless the T-stat is totally sealed- which I've never seen one that does not have a small bleeder hole.

BLOCK pressure, not radiator pressure.  The radiator is on the low pressure side of the cooling circuit.

A rule of thumb I have seen for racing engines is to size the water pump pulleys/restrictor orifice (racing engine = no thermostat) for 50psi pressure in the block.

I assume that suction side thermostats work at all because they are only ever employed on aluminum headed engines, and aluminum is a wonderful conductor of heat compared to iron.

Still, T-stats have leaks designed in them, so that the pressure in the entire system is controlled by the rad cap.  Let alone the return line closes the entire system with a big passage.  I'm not aware of a pump that's sealed enough to generate significant pressure.  

I understand why one would want to run high pressures.  

Maybe I need to look into cooling systems more...

Well, I stuck a pressure gauge on an engine and played around with it... 

45psi was in a Vulcan engined Taurus at 3000rpm.  

You measured 45 psig ... and the radiator caps are usually rated below 1.4 bar ... shouldn't water be coming out of the reservoir as the pressure would overcome the radiator cap spring?

The cap was OFF!  

Measuring pressure in the block.  Before the thermostat.  Radiator has nothing to do with it.

It's like measuring fuel pressure at the fuel injectors vs. at the gas tank.

Pete. (l33t FS) said:

Well, I stuck a pressure gauge on an engine and played around with it... 

45psi was in a Vulcan engined Taurus at 3000rpm.  

Good to have some actual data.  I'd be curious to see pressure at a few different points in a cooling system: before t-stat, upper rad hose, lower rad hose.  Maybe if I feel like draining the cooling system on the Jeep in the spring and sacrificing a pair of radiator hoses at some point I can do a test on it.  I've got t-fittings that'll fit both rad hoses, just have to find a port to use on the intake coolant crossover that doesn't already have a temp sensor in it. 

Pete. (l33t FS) said:

Slippery (Forum Supporter) said:

Pete. (l33t FS) said:

alfadriver (Forum Supporter) said:

Pete. (l33t FS) said:

alfadriver (Forum Supporter) said:

In reply to Pete. (l33t FS) :

Uh, I don't think the thermostat has any effect on the coolant system pressure....  That's more about the rad cap.  

Unless the T-stat is totally sealed- which I've never seen one that does not have a small bleeder hole.

BLOCK pressure, not radiator pressure.  The radiator is on the low pressure side of the cooling circuit.

A rule of thumb I have seen for racing engines is to size the water pump pulleys/restrictor orifice (racing engine = no thermostat) for 50psi pressure in the block.

I assume that suction side thermostats work at all because they are only ever employed on aluminum headed engines, and aluminum is a wonderful conductor of heat compared to iron.

Still, T-stats have leaks designed in them, so that the pressure in the entire system is controlled by the rad cap.  Let alone the return line closes the entire system with a big passage.  I'm not aware of a pump that's sealed enough to generate significant pressure.  

I understand why one would want to run high pressures.  

Maybe I need to look into cooling systems more...

Well, I stuck a pressure gauge on an engine and played around with it... 

45psi was in a Vulcan engined Taurus at 3000rpm.  

You measured 45 psig ... and the radiator caps are usually rated below 1.4 bar ... shouldn't water be coming out of the reservoir as the pressure would overcome the radiator cap spring?

The cap was OFF!  

Measuring pressure in the block.  Before the thermostat.  Radiator has nothing to do with it.

It's like measuring fuel pressure at the fuel injectors vs. at the gas tank.

Freaking quote is huge, but I am on my phone and cannot edit easily. 

Any chance that your sensor was in an air bubble and you were measuring the pressure of the air? 

I'd like to believe that over water pressure ...

Slippery (Forum Supporter) said:

Freaking quote is huge, but I am on my phone and cannot edit easily. 

Any chance that your sensor was in an air bubble and you were measuring the pressure of the air? 

I'd like to believe that over water pressure ...

Even if there was an air bubble, it should be at the same pressure as the surrounding water. 

There can be no pressure without restriction.

iceracer said:

There can be no pressure without restriction.

Other than the whole getting hot part, yes.

In reply to Robbie (Forum Supporter) :

Coolant flows too fast so it cannot absorb heat and cool the engine.

iceracer said:

In reply to Robbie (Forum Supporter) :

Coolant flows too fast so it cannot absorb heat and cool the engine.

That's an old wives tale.  If it overheats without a t-stat and doesn't with a restrictor, either the thermostat is important to controlling bypass or the reduced block pressure with it wide open is causing coolant to boil in the heads. 

Here's the OBD output from my car with the normal Porsche coolant system and stock 185F thermostat.  Outside air temp is around 95F.

I doubt I'd see any benefit from a cooler thermostat.