Miata cooling system thread
#242
A car that runs at the right temp w/o a tstat by definition, does not have enough cooling capacity. If it only runs cool when it is un-regulated, then the system is maxed out and can't handle any more heat. In Oldguys example, 110* ambient temps, drafting our a quicker driver would likely have it overheating.
We have found that deleting the tstat leads to overcooling. Every car we take to the track has a programmable ECU and is tuned to be happy within a prescribed temp window. The difference might be that our cars, in our heat, with our drivers, will overheat unless everything is done; Ducting / Reroute / Big radiator / 98% distilled water / 19psi cap / E85. Thus we have excess cooling capacity and it needs to be regulated. In the winter, our cars won't get over 140* without a tstat. We run tstats in our oil coolers for the same reason. 140* oil isn't good.
In short, if deleting the tstat saves you the cost of a reroute and works for you, go. Just know that you might not have enough spare cooling to handle more load than you have been putting in the system.
We have found that deleting the tstat leads to overcooling. Every car we take to the track has a programmable ECU and is tuned to be happy within a prescribed temp window. The difference might be that our cars, in our heat, with our drivers, will overheat unless everything is done; Ducting / Reroute / Big radiator / 98% distilled water / 19psi cap / E85. Thus we have excess cooling capacity and it needs to be regulated. In the winter, our cars won't get over 140* without a tstat. We run tstats in our oil coolers for the same reason. 140* oil isn't good.
In short, if deleting the tstat saves you the cost of a reroute and works for you, go. Just know that you might not have enough spare cooling to handle more load than you have been putting in the system.
__________________
#243
It might be that my coolant system is balanced by luck. The higher flow matches the smaller radiator not that flow affects much. I have run the car on 36 degree C days with no overheating issues and on cooler days it just needs more warmup time. I have always thought most coolant systems self balance. The hotter the radiator the more efficently it removes heat.... so you need to do something drastic like halve the cooling or double the heat load to raise the temp by a significant amount (ie 10-15%)
Emilio, do you think there is much benefit to running a faster flow of coolant through the engine to keep the temp differential lower? It would be very interesting to run two coolant sensors, one at entry, one at exit to see the temp difference.
Also, anyone here run an electric water pump?
Emilio, do you think there is much benefit to running a faster flow of coolant through the engine to keep the temp differential lower? It would be very interesting to run two coolant sensors, one at entry, one at exit to see the temp difference.
Also, anyone here run an electric water pump?
#245
sorry people, my intent was, certainly, not to " start " such intensity
my statements are all accurate
you people don't know me and I don't know all of you
this is the only forum I've been on that had a minimum posting requirement to place an ad
I started out only wanting to place an ad for my car
although happy to answer any questions I'll certainly defend myself
enjoy your forum
my statements are all accurate
you people don't know me and I don't know all of you
this is the only forum I've been on that had a minimum posting requirement to place an ad
I started out only wanting to place an ad for my car
although happy to answer any questions I'll certainly defend myself
enjoy your forum
#248
It might be that my coolant system is balanced by luck. The higher flow matches the smaller radiator not that flow affects much. I have run the car on 36 degree C days with no overheating issues and on cooler days it just needs more warmup time. I have always thought most coolant systems self balance. The hotter the radiator the more efficently it removes heat.... so you need to do something drastic like halve the cooling or double the heat load to raise the temp by a significant amount (ie 10-15%)
Emilio, do you think there is much benefit to running a faster flow of coolant through the engine to keep the temp differential lower? It would be very interesting to run two coolant sensors, one at entry, one at exit to see the temp difference.
Also, anyone here run an electric water pump?
Emilio, do you think there is much benefit to running a faster flow of coolant through the engine to keep the temp differential lower? It would be very interesting to run two coolant sensors, one at entry, one at exit to see the temp difference.
Also, anyone here run an electric water pump?
FM and few here have been playing with the Davies-Craig EWP electric pumps. It is worth note that a properly configured tstat based PWM table or control algorithm will run whatever coolant flow is required to meet target temps. That might not necessarily be max flow.
__________________
#249
Thermal transfer efficiency is increased with delta but those gains don't overcome the maximum thermal rejection of the system. You also need to consider that our medium is water just below boiling point. We don't want to purposely have 245° coolant at the head just so it will transfer faster when it hits the rad. I've never done any testing on flow rate vs rejection capacity. I do recall pulling a tstat or two over the years on someones else's car that was overheating as a band-aid to finish the day. I'm not an engineer so I can't comment on the contact time vs heat transfer of 20-30psi, 230° coolant through the head or through the rad. I'm going to guess that rate of flow vs BTU transfer is a bit of a bell curve. Slow rate way down and coolant is quickly maxed out to near boiling and the part being cooled gets too hot. Speed it up to fast and not enough heat is absorbed by the coolant, although the rate would be at it's highest possible due to the delta between the part and coolant. So somewhere in the middle is a flow rate which takes advantage of the greater temp delta where the transfer rate is highest while coolant is circulated fast enough to avoid over heating the part without flowing so fast that the coolant doesn't have enough time to absorb heat off the part.
FM and few here have been playing with the Davies-Craig EWP electric pumps. It is worth note that a properly configured tstat based PWM table or control algorithm will run whatever coolant flow is required to meet target temps. That might not necessarily be max flow.
FM and few here have been playing with the Davies-Craig EWP electric pumps. It is worth note that a properly configured tstat based PWM table or control algorithm will run whatever coolant flow is required to meet target temps. That might not necessarily be max flow.
#251
No it will not. Flow too fast and you reduce the time for transfer to take place. There most definitely is a point of diminishing returns WRT flow rate. I'll agree that ensuring your system is capable of high flow rate is important and that proper functioning is reliant on it. What I would like to see, is some data on total BTU rejection vs flow rate. Maybe Keith or Jeremy can chime in as I think they have some data on this.
__________________
#252
No it will not. Flow too fast and you reduce the time for transfer to take place. There most definitely is a point of diminishing returns WRT flow rate. I'll agree that ensuring your system is capable of high flow rate is important and that proper functioning is reliant on it. What I would like to see, is some data on total BTU rejection vs flow rate. Maybe Keith or Jeremy can chime in as I think they have some data on this.
The math is simple, it's m *Cp*detlaT. is how much heat transfer takes place.
Improving m linearly improves heat transfer with contstant Cp and detal T. In reality if you double the m the detla T will go down some, but it won't be cut in half, and you'll get more heat transfer.
I agree you should test this and see what happens. Here's an easy test, go remove the thermostat (maximize m dot) and see what temps do. They will go down, not up.
#254
Reduce the time for heat transfer to take place?
The math is simple, it's m *Cp*detlaT. is how much heat transfer takes place.
Improving m linearly improves heat transfer with contstant Cp and detal T. In reality if you double the m the detla T will go down some, but it won't be cut in half, and you'll get more heat transfer.
I agree you should test this and see what happens. Here's an easy test, go remove the thermostat (maximize m dot) and see what temps do. They will go down, not up.
The math is simple, it's m *Cp*detlaT. is how much heat transfer takes place.
Improving m linearly improves heat transfer with contstant Cp and detal T. In reality if you double the m the detla T will go down some, but it won't be cut in half, and you'll get more heat transfer.
I agree you should test this and see what happens. Here's an easy test, go remove the thermostat (maximize m dot) and see what temps do. They will go down, not up.
__________________
#255
If you quadruple the flow through the engine/radiator, total heat rejection would go up. The higher flow rate results in the average tube temperature of the radiator to go up, and thus the delta-T between the tubes/fins and the air goes up, and heat transfer goes up.
#256
Now it's obvious. FWIW I've tested this on my car, thermostat vs no thermostat and it overcooled without a thermostat, even under loads that would typically result int the engine running 200+, it wouldn't go over 180 without a thermostat.
If you quadruple the flow through the engine/radiator, total heat rejection would go up. The higher flow rate results in the average tube temperature of the radiator to go up, and thus the delta-T between the tubes/fins and the air goes up, and heat transfer goes up.
If you quadruple the flow through the engine/radiator, total heat rejection would go up. The higher flow rate results in the average tube temperature of the radiator to go up, and thus the delta-T between the tubes/fins and the air goes up, and heat transfer goes up.
What I question, being a idiot at math, is the math you stated that supports an infinite increase in a Miata cooling system (not hypothetical model) efficiency with increase flow rate. If that's so, we should be spending more time with auxiliary or standalone water pumps.
What you're stating, if I understand it correctly, is that time the coolant spends in contact with a given area of metal doesn't matter. It's the volume of coolant that contacts a given area of metal that all system design should be based on. As that idea pertains to air flow volume on a given area of the rad, more is better so that supports the idea (for us not mathy folks).
__________________
#257
I mean it's obvious that removing the tstat can lower coolant temps, but also leave the system unregulated. My take is that if the system only does the job running at max capacity, it probably has no reserve in case the load increases. That's fine for F1 where cooling aero drag is pared down to the nth. For a typical HPDE Miata though, some spare capacity is a good idea.
What I question, being a idiot at math, is the math you stated that supports an infinite increase in a Miata cooling system (not hypothetical model) efficiency with increase flow rate. If that's so, we should be spending more time with auxiliary or standalone water pumps.
What you're stating, if I understand it correctly, is that time the coolant spends in contact with a given area of metal doesn't matter. It's the volume of coolant that contacts a given area of metal that all system design should be based on. As that idea pertains to air flow volume on a given area of the rad, more is better so that supports the idea (for us not mathy folks).
What I question, being a idiot at math, is the math you stated that supports an infinite increase in a Miata cooling system (not hypothetical model) efficiency with increase flow rate. If that's so, we should be spending more time with auxiliary or standalone water pumps.
What you're stating, if I understand it correctly, is that time the coolant spends in contact with a given area of metal doesn't matter. It's the volume of coolant that contacts a given area of metal that all system design should be based on. As that idea pertains to air flow volume on a given area of the rad, more is better so that supports the idea (for us not mathy folks).
I agree removing the thermostat is a bad option, only mentioned it to "prove" that increasing m dot over the factory setup does in fact improve cooling, even at load. I also agree that having a system with overhead is desirable, I've spent a lot of time trying to achieve that on my own miata.
The factory cooling system dumps 20-30% of the hot water that exits the block right back into the mixing manifold. If you want to get a 20-30% improvement in mass flow rate to see if that "helps" improve the cooling, build a thermostat that switches the water exiting the heater core to go to the mixing manifold when the water is say below 180, and exit to the radiator when above 180.
Beyond that, yes a better water pump would help. If you had the "perfect" water pump, it would pump so much coolant that the engine block, and the tubes of the radiator, would be the same temperature.
#258
It's continuous systems with deltas everywhere.
Once you reach turbulent flow the heat transfer profile thickness is constant and small.
Infinite water flow will make all the water the same temp making the hotspots in the head just a fraction hotter than the radiator fins (since the heat transfer profile thickness never will be zero).
Remember that the water pump spins quite fast and probably reach a peak flow before redline (I don't think I've seen flow data for a Miata pump, and not for one installed with a reroute).
Too slow though...
Hmm, one way to make the head temp even more even would be to increase the pump and increase the heater circuit (or other path bypassing the Tstat+rad). Even temps enables reduced marginals which can enable more power.
Once you reach turbulent flow the heat transfer profile thickness is constant and small.
Infinite water flow will make all the water the same temp making the hotspots in the head just a fraction hotter than the radiator fins (since the heat transfer profile thickness never will be zero).
Remember that the water pump spins quite fast and probably reach a peak flow before redline (I don't think I've seen flow data for a Miata pump, and not for one installed with a reroute).
Too slow though...
Hmm, one way to make the head temp even more even would be to increase the pump and increase the heater circuit (or other path bypassing the Tstat+rad). Even temps enables reduced marginals which can enable more power.
#259
Moderator
iTrader: (12)
Join Date: Nov 2008
Location: Tampa, Florida
Posts: 21,026
Total Cats: 3,123
I disagree. If the radiator tubes were the same temperature then the radiator would not be performing it's only function. Coolant must be cooler leaving than entering or it would not be exchanging any heat.