Enhancing the Cooling System

[eec564]

Alright, I'm transporting a hijacked thread here to try and make things right. Here's what we've got thus far.

BTW, has anyone considered removing the thermostat and installing a variable speed control for an electric water pump? It could come on at preset speeds... maybe 10,25,50,75 and 100 percent depending on water temp. It could come on at 10 percent at 140 degrees, 25 percent at 145 and increase every 5 degrees until you reach 100 percent speed at 160 degrees.

I don't think that would work very well, as you'd need a closed-loop system to measure flow. 50% voltage on an electric pump won't equal 50% flow, and a lot of 12v pump motors won't run under load at less then 9 volts, and 70% flow or so, depending on pressure.

I don't think that would work very well, as you'd need a closed-loop system to measure flow. 50% voltage on an electric pump won't equal 50% flow, and a lot of 12v pump motors won't run under load at less then 9 volts, and 70% flow or so, depending on pressure.

Thanks for the information eec564. I didn't know this. I wonder if designing a pump and control system would be worth the bother? It's just a thought to improve maximum flow by eliminating restrictions through the thermostat.

 

Maybe a new thermostat design is the simpler way to go... one through which water passes freely through the (((center))) when fully open rather than having to traverse )))around((( the thermostatic valve. Or perhaps an electronically controlled valve could replace the themostat? I'm just talking efficiencies here.. and it's probably useless dribble.

Well' date=' the point of the cooling system is two fold, keep the engine cool, and keep the engine at a stable temperature under all (ideally) driving conditions.

 

That's why removing the thermostat causes issues. Newer cars rely on switched electric fans in addition to the thermostat. The main reasons old style thermostats are still around is their simple, reliable, and cheap to manufacture.

 

For race engines, a thermostat that opens at a lower temperature is used, to give the cooling system a head-start. A more efficent radiator makes the engine cool better, requiring less coolent flow. The entire point of eliminating the thermostat would be to allow better coolent flow, resulting in a higher maximum colling effect, the same result as a better radiator, better airflow through the radiator, or any number of cooling system upgrades. Everything I've mentioned is simpler, reliable, tested, and most likely cheaper.

 

That said, I really like the idea. Especially if you're already using a custom fuel injection controller that could handle the computing needed to adjust the extra cooling capacity to prevent overheating by precooling the engine and radiator slightly under high engine loads . I'm thinking along the lines of extra code in megasquirt, or a small pic microcontroller, with MAP and temp sensors. Of course, you could always use a small on dash controller to adjust the base cooling level and leave the temp vs cooling effect ratio stationary and not adaptive. I just like to see base designs where extra features possibal to go ahead and use them.

 

The big question I see is how to modulate the amount of cooling effectively, accurately, and FAR MOST IMPORTANTLY reliabally. A mechanical valve might be nice, such as a stepper motor controlled 2-inch ball valve to limit flow predictabally but still allowing full flow. Or a modulator control for an electric fuel pump, which would require a flow sensor as a certain voltage/current/modulation provided wouldn't result in a consistant flow rate (experience with electric motors, loads, stall points, starting voltages, etc tells me this, I forget which textbook it was in).

 

So there's a few ideas, got any of your own? And do you think we'll be chased by people trying to catch us with their hijacked thread?[/quote']

I really like your idea of modifying the Megasquirt code. Is it proprietary? For the vast majority not using Megasquirt (I'm using carbs) a stand-alone controller is in order as well. Like you said' date=' a manual switch for a blast of precooling might work as a simple and inexpensive solution for that little extra edge. Could one tie into the proper existing sensory data without affecting the engine controller? If it's a matter of digital noise affecting the disparate controllers, UHF filters might eleiminate that issue. Perhaps the controller could also accept TPS data responding immediately to WOT as well as quickly rising temperatures. A step motor sounds like the way to go for valve control. How many steps, do you think... a half dozen? A flow sensor would certainly work, but my concern is sensor failure in the harsh, grungy environment of engine coolant. Also, like you mentioned, simple is good. Given the variables associated with electric water pumps, I agree with your initial ascertation that it's best to let them run full speed. A stepped motor controlling a large ball valve is the most reliable solution.

 

This is moving coolant flow control away from the dumb pump to a smart flow control valve. How much advantage is realized I don't know. An important factor though is allowing for huge instantaneous volume on demand. This might realize a few net HP while helping to protect ones' engine from excess heat.

 

I'm old school having owned a few GM Gen1's ranging from 283 to 455 CID. I think the old carbureted engines are affected more by temperature variation than todays' computer controlled offerings. In my experience, running at the correct temperature is paramount. Even five or ten degrees can make a difference.

 

Yes... we're probably in trouble for hijacking this thread. You have the option though of pointing an accusing finger at me:smile:[/quote']

[eec564]

5 or 10 degrees F different in block and head temp dosen't make much of a difference for a daily driver, but it does make a good sized difference in how well the engine runs. Power, fuel economy, emissions, the first and second ones go down and the last one goes up if the engine isn't at the optimum temp it's tuned for. Granted, an engine operating at 500* F would be great if it were designed for that, go play with ideal heat pumps and the like, I'll stick with currently in use and easy to manage designs.

 

A stepper motor dosen't sound bad, but you'd want a fairly powerful one to move such a large valve, and a backup to fully open the valve if the stepper were to fail. The megasquirt code wouldn't be hard to modify to contorl this, you can easily monitor engine load by looking at the manifold pressure, or use a manumatic system, one that ran off a PIC or similar microcontroller for normal operation and had the option for enhanced cooling under all circumstances.

 

Perhaps placing the valve under the hood in the middle of the hose right after the water neck would work, after taking out the thermostat?

 

I can think of one major feature of this system. After a hard drive, you could use a turbo timer, and cool the engine FAR more by leaving it running for 1min then just by letting the normal cooling system slowly bring it back to 180. Because we all know engines continue to get hot after you turn them off, this could be great in the hot summers we're having ALL over the country. Also, the quicker cooling with less after-run time would result in better engine life and less gas wasted trying to keep your engine happy. I know I live up a VERY steep hill and have to idle down my car at 1-2K for one min after a spirited drive up on a hot day.

 

Okay, that's it for me tonight, I've got an early morning, and I'll sleep on it.

[Guest Mike]

5 or 10 degrees F different in block and head temp dosen't make much of a difference for a daily driver' date=' but it does make a good sized difference in how well the engine runs. Power, fuel economy, emissions, the first and second ones go down and the last one goes up if the engine isn't at the optimum temp it's tuned for. Granted, an engine operating at 500* F would be great if it were designed for that, go play with ideal heat pumps and the like, I'll stick with currently in use and easy to manage designs.

 

A stepper motor dosen't sound bad, but you'd want a fairly powerful one to move such a large valve, and a backup to fully open the valve if the stepper were to fail. The megasquirt code wouldn't be hard to modify to contorl this, you can easily monitor engine load by looking at the manifold pressure, or use a manumatic system, one that ran off a PIC or similar microcontroller for normal operation and had the option for enhanced cooling under all circumstances.

 

Perhaps placing the valve under the hood in the middle of the hose right after the water neck would work, after taking out the thermostat?

 

I can think of one major feature of this system. After a hard drive, you could use a turbo timer, and cool the engine FAR more by leaving it running for 1min then just by letting the normal cooling system slowly bring it back to 180. Because we all know engines continue to get hot after you turn them off, this could be great in the hot summers we're having ALL over the country. Also, the quicker cooling with less after-run time would result in better engine life and less gas wasted trying to keep your engine happy. I know I live up a VERY steep hill and have to idle down my car at 1-2K for one min after a spirited drive up on a hot day.

 

Okay, that's it for me tonight, I've got an early morning, and I'll sleep on it.[/quote']

 

Good points, all. I'm glad one of us had the forethought and courtesy to move this to a new thread;)

 

I see removing the water neck/thermostat and replacing it with a universal valve. This is mounted via adapter plates to the intake and adapter water outlets to the hose. I don't think a valve will be much larger than the average water neck... maybe an 1.5 inch wider? The new neck is extended couple inches to accomodate mounting of the step motor. The motor doesn't need to be large. Since a tiny bit of flow-by around the valve is harmless, the ball valve can be fairly loose... it doesn't need to "seal". This eliminates the need for a high torque motor if slick Teflon ball and bushings are used. The motor need only be strong enought to hold the valve in place. The ratcheting system or pressure detents at each step will serve to stabilize motor/valve position. The key is to keep the ball loose while avoiding chatter between the ball and bushing. I fully agree that a backup (emergency solenoid?) must be incorporated.

 

I'm surprised that auto manufacturers haven't already done this. Electric fans continue to run after the engine is shut off but this only cools the radiator. If an electric water pump continues to run and the thermostat stays wide open, this will cool the entire power-plant.

[Twizted]

One problem with high-flow radiators and removing the T-Stat is TOO much flow. The watermix will not be in the radiator long enough to cool so then you send the hot water back into the engine to be super-heated. Increased flow is good and does help but just remember that there is a point where it becuase less effective and then harmful.

[Guest Mike]

One problem with high-flow radiators and removing the T-Stat is TOO much flow. The watermix will not be in the radiator long enough to cool so then you send the hot water back into the engine to be super-heated. Increased flow is good and does help but just remember that there is a point where it becuase less effective and then harmful.

 

Hi Twizted. Thanks for your insights.

 

The electronically actuated ball valve with dedicated controller replaces the thermostat and performs the same function... just more accurately/efficiently and with less restricted flow. It still controls coolant flow. I've heard/read many times that too fast coolant flow precludes efficient convective heat loss. I sincerely believe this isn't right. It's one of those ideas that's started, isn't proven, and is pertetuated. Of course, I could be wrong but it just doesn't seem logical to me.

[Phantom]

Several things.

1) The idea of an electric water pump has it's pro's & cons. They would include:

a) More complex and potentially less reliable system.

Less efficient theoretically because of mechanical losses from

generating the elecricity and then converting it back to mechanical

pumping vs. just straight to pumping.

c) Does allow to not run the pump at initial start-up as engine is

warming up thus having less overall drag on the engine/fuel

consumption.

d) Does allow for circulating water after engine shutdown to cool the

engine down.

2) For optimal cooling you want turbulent flow in your piping. Laminar flow reduces the heat transfer rate so you want the water to move at a brisk rate.

3) For those concerned about residence time in the radiator and cooling effectiveness that is controlled by:

a) the face size of the radiator - height x width

the number of rows - the more rows the slower the water moves

and the more effective the air cooling.

c) The the capacity of the fans.

There is a minimum and maximum water flow velocity based on tube size to get the most effective cooling out of a radiator.

4) A variable speed water pump controlled by a temperature sensor with an output range based on temperature would allow the water flow to be varied based on maintaining a set temperature. That is esentially what your thermostat does now but it varies the water flow based on variable back pressure determined by the valve position. The electric pump would vary the flow based on the temperature feedback. In low load conditions this would result in less energy consumption but probably slightly higher in extreme heat condtions.

[Guest Mike]

Several things.

1) The idea of an electric water pump has it's pro's & cons. They would include:

a) More complex and potentially less reliable system.

Reply: More reliable pumps can be engineered

Less efficient theoretically because of mechanical losses from

generating the elecricity and then converting it back to mechanical

pumping vs. just straight to pumping.

Reply: More efficient pumps can be engineered

c) Does allow to not run the pump at initial start-up as engine is

warming up thus having less overall drag on the engine/fuel

consumption.

Reply: True

d) Does allow for circulating water after engine shutdown to cool the

engine down.

Reply: Yes indeed

2) For optimal cooling you want turbulent flow in your piping. Laminar flow reduces the heat transfer rate so you want the water to move at a brisk rate.

Reply: Yes' date=' but how does converting to a control valve affect turbulance through the block and radiator?[/b']

3) For those concerned about residence time in the radiator and cooling effectiveness that is controlled by:

a) the face size of the radiator - height x width

Reply... EMPHATICALLY YES.

the number of rows - the more rows the slower the water moves

and the more effective the air cooling.

Reply: Yes, yes YES!!!

c) The the capacity of the fans.

Reply: Absolutely!!

There is a minimum and maximum water flow velocity based on tube size to get the most effective cooling out of a radiator.

Reply: So VERY true!!

4) A variable speed water pump controlled by a temperature sensor with an output range based on temperature would allow the water flow to be varied based on maintaining a set temperature. That is esentially what your thermostat does now but it varies the water flow based on variable back pressure determined by the valve position. The electric pump would vary the flow based on the temperature feedback. In low load conditions this would result in less energy consumption but probably slightly higher in extreme heat condtions.

Reply: The idea is to AVOID those extreme heat conditions with a EXTREMLY HIGH VOLUME SMART control device and a RAPID RESPONSE valve in lieu of the very SLOW thermostat.

[eec564]

One problem with high-flow radiators and removing the T-Stat is TOO much flow. The watermix will not be in the radiator long enough to cool so then you send the hot water back into the engine to be super-heated. Increased flow is good and does help but just remember that there is a point where it becuase less effective and then harmful.

 

That's why a high-efficency (most likely aluminium) radiator would also most likely be used, the custom control would primarily provide adaptive cooling, not simply reactive cooling.

 

In reguards to Phantom's last post:

Point B, the mechanical pump could easily be used with flow controlled by a valve but this would eliminate point D, which I REALLY like, especially with the speed it could happen at. But that could also happen quickly and efficently with the engine running briefly and the valve fully open.

 

Point 4, controlling full range non-stepper 12V motors is a pain.

 

Everything else, Mike said it all.

[Guest Mike]

That's why a high-efficency (most likely aluminium) radiator would also most likely be used' date=' the custom control would primarily provide adaptive cooling, not simply reactive cooling.

 

In reguards to Phantom's last post:

Point B, the mechanical pump could easily be used with flow controlled by a valve but this would eliminate point D, which I REALLY like, especially with the speed it could happen at. But that could also happen quickly and efficently with the engine running briefly and the valve fully open.

 

Point 4, controlling full range non-stepper 12V motors is a pain.

 

Everything else, Mike said it all.[/quote']

 

eec564... Controlling a non-stepping motor...

 

What about robbing a motor speed control from a 14V cordless hand drill or similar device? Could one of those be controlled by a variable resistance circuit, which accepts data from your control unit (or by a Megasquirt) and also receives feedback from an RPM sensor on the pump? Could the drill circuit be beefed up to handle the higher load? I'm clearly no expert but it's a thought... sort of.

[veritech-z]

This stuff is pretty interesting, but the particulars of the design are a bit over my head. I can say that having 7 years in the auto parts industry at both the aftermarket and dealer level, I have personally witnessed cars that overheated in the absence of a thermostat that ran better with it in. In fact, I've even seen (and subsequently owned after it overheated and would no longer run) an S30 that exhibited this behavior.

[eec564]

Mike: Robbing a motor speed control from a cordless drill wouldn't work so well as they're not rated for continuous duty. A dc motor controlled by PWM (pulse width modulation) would work just fine, as long as it had a feedback sensor for rpm, or better yet coolent flow. That would eliminate the need for a control valve in the sytem too.

 

I see the two main choices being the standard mechanical water pump, mechanical fan (I don't see much use running the fan without coolent flowing) upgraded radiator, no thermostat, valve of some sort, lots of sensors, and control circuitry as one route. The other route, which I like better, is electric pump, no thermostat, lots of sensors, electric fan, upgraded radiator, and controll circuitry. When using megasquirt, most of the sensors are already built in, so all that's needed for the control side is a high current PWM motor controller and either a rpm sensor or preferabally a flowmeter.

 

veritech-z: We wouldn't be simply running the cooling system at max flow, which would overheat the engine, without the restriction of the thermostat. When coupled with a more efficent radiator and sensors to tell it the temp of the radiator, oil, coolent, transmission, engine load, etc...an adaptive cooling system could cool far better under greatly varried driving conditions. Such as go ahead and cool the radiator down to 160 keeping the engine at 180 with lower coolent flow to give a little extra head room and don't have the system playing catch-up when you really get your foot down or turn on the AC.

[Guest Mike]

This stuff is pretty interesting, but the particulars of the design are a bit over my head. I can say that having 7 years in the auto parts industry at both the aftermarket and dealer level, I have personally witnessed cars that overheated in the absence of a thermostat that ran better with it in. In fact, I've even seen (and subsequently owned after it overheated and would no longer run) an S30 that exhibited this behavior.

 

This is over my head too. I'm just interested in possibilities... new and better ways of accomplishing tasks. The difficult work, i.e. analysis, design, development, etc., is left to those better suited to succeed... not me:confused2

[Guest Mike]

Mike: Robbing a motor speed control from a cordless drill wouldn't work so well as they're not rated for continuous duty. A dc motor controlled by PWM (pulse width modulation) would work just fine' date=' as long as it had a feedback sensor for rpm, or better yet coolent flow. That would eliminate the need for a control valve in the sytem too.

 

I see the two main choices being the standard mechanical water pump, mechanical fan (I don't see much use running the fan without coolent flowing) upgraded radiator, no thermostat, valve of some sort, lots of sensors, and control circuitry as one route. The other route, which I like better, is electric pump, no thermostat, lots of sensors, electric fan, upgraded radiator, and controll circuitry. When using megasquirt, most of the sensors are already built in, so all that's needed for the control side is a high current PWM motor controller and either a rpm sensor or preferabally a flowmeter.

 

veritech-z: We wouldn't be simply running the cooling system at max flow, which would overheat the engine, without the restriction of the thermostat. When coupled with a more efficent radiator and sensors to tell it the temp of the radiator, oil, coolent, transmission, engine load, etc...an adaptive cooling system could cool far better under greatly varried driving conditions. Such as go ahead and cool the radiator down to 160 keeping the engine at 180 with lower coolent flow to give a little extra head room and don't have the system playing catch-up when you really get your foot down or turn on the AC.[/quote']

 

It sounds like you have a theoretical prototype in mind. If you build it... they will steal it... unless you patent it before they do. This actually sounds fairly easy for someone like yourself with the proper background experience.

 

If you decide to give this a try, do you mind sharing your findings? I'll build one if implementation is within my limited abilities. Perhaps use of an existing OEM PWM motor control, a magnetic P/U to read pump RPM or an in-line low-restriction flow meter, and common OEM or aftermarket sensors for those without computer-controlled engines or Megasquirt systems?

[veritech-z]

I'm following the discussion, I realize what you are doing would not be an on/off flow switch, but a progressive restriction. Somebody made a comment earlier in the thread that they didn't believe that a car with no thermostat would overheat simply because there wasn't a thermostat in place, and I was just chiming in with some real world experience that yes, it is in fact possible for this to happen as I have seen it several times. That being said, does anybody know how Formula 1 cars handle their cooling situation? I would imagine that if they are using solenoid operated intake and exhaust valves, surely if this were a possiblity worth exploring it would have been done?

[Guest Mike]

I'm following the discussion, I realize what you are doing would not be an on/off flow switch, but a progressive restriction. Somebody made a comment earlier in the thread that they didn't believe that a car with no thermostat would overheat simply because there wasn't a thermostat in place, and I was just chiming in with some real world experience that yes, it is in fact possible for this to happen as I have seen it several times. That being said, does anybody know how Formula 1 cars handle their cooling situation? I would imagine that if they are using solenoid operated intake and exhaust valves, surely if this were a possiblity worth exploring it would have been done?

 

If Formula 1 cars do use such a system, then perhaps this can be adapted for street? If this hasn't been developed, perhaps it's only because we've been happy enough with the old technology. I think it's worth trying since this may actually be a useful improvement.

 

A proactive instantaneous approach is always better than reactive with lag time with regard to any solution to a problem. It seems to me that this also applies to automotive cooling systems.

 

Is it appropriate to drop ones' paradygms in pubic?