Process Control in a Car Cooling System

I thought the electric fan was made necessary by the introduction of the transverse engine, which took the belt driven fan away. I do not know of an electric coolant pump but it is possible. most are direct drive by the engine.

And BTW the conventional system does have a thermostat, typically in the upper rad hose connection at the block. the pump is part of the motor and so it does change flow rate with engine speed.

That thermostat opens and closes to try to maintain the best operating temperature for the engine. The pump circulates coolant within the block until the operating temp is reached and then the thermostat opens to begin to regulate the temperature to that set point.

The task of the radiator is to remove the excess heat passed to it by the opening of the thermostat, so yes, it will vary there with the energy output of the engine.

Any electric fan / pump / electronic thermostatic control system will not be a conventional cooling system and will introduce many more failure points than the original conventional designs of old. Every time I see a new fangled electric / electronic system to replace a mechanical one, I get the same reliability issues in mind. It would take an extremely robust and well executed design to reliably replace the "conventional" automotive cooling system. I have had to fix many more electric cooling system failures than "conventional" ones. Some ideas are not better or more worthy just because they are new.

I think you need to study up on how and why the cooling systems on vehicle engines are designed.

What you say here is no where near being correct.

Of course this will not reach a stable temperature. This is a car with constantly changing heat production as mechanical loads change from acceleration, hills, etc. Also as the vehicle's velocity changes the air mass moving across the radiator will change, too. The fan will only add to this amount of air mass not control it.

I would guess that this would still work well enough to not roast the engine. This is not likely to thermally run away.

You are not describing a, "car with a conventional cooling system".

The answer depends on the cooling capacity of the system in the current environment.

What's wrong with a mechanical water pump and a thermostat?

"What's wrong with a mechanical water pump and a thermostat?"

It's too simple and reliable therefore it must be re engineered!

Nothing wrong with a mechanical fan, a mechanical pump and a thermostat, that's just not the system to be analyzed in this discussion.

Your question for this discussion is not going to be answered toward a working unit. The problem as others have began to state is the work or heat produce by any internal combustion engine varies to the degree of load placed or required of it. This is why we control the heat at a constant rate and cool the excess by over sizing the radiator to the worst case possible (or near as can be figured mathematically).

Also by controlling it in this manner, the engine oils are keep in a state that allows for moisture removal thus increasing the service life of same.

The problem I envisage is - if the water pump runs constantly there's a danger of overcooling when on the open road in cold weather, even though the fan is switched on low temperature. It might be possible to stop/start the pump, but the engine might not like it, specially stopping flow through the cylinder head.

The pump must circulate coolant through the engine all the time. The radiator with its thermostat will bypass hot coolant through the rad as needed. The thermostat will control the temperature of the cooling fluid. The thermostat on top of the radiator will not restrict the flow of coolant through the engine...just the flow through the radiator. It will only let water into the radiator if it needs to cool the water. I cannot see any way the water would run too cool if you have an internal thermostat on the radiator.

Why would you want to run it without a thermostat in the first place? It won't give you any more performance....

That confirms my point - need for continuous coolant flow through the engine.

I wouldn't want to run without without a thermostat, but that's what the OP stipulated.

I considered modulating the speed of the electric water pump, but it only does about 5 gpm now. When I built the car there weren't many choices and I used a marine pump which is still working after 20 years! I've driven the car on highway in freezing weather and the temp will stabilize around 200F without fan.

I'm surprised temp stabilises that high with pump running and no thermostat, but the proof of the pudding........

Given all the particular circumstances for this setup the stabilization temperature will most likely depend upon the size of the radiator; the larger the cooling capacity the cooler the fluid returning to the engine, all other things being equal.

It should stabilize between T1 and T2. However, if you have an air conditioner (AC), the temp. may run lower than T1, especially in cooler weather, since the AC overrides the sensor and you don't have a thermostat. I'm familiar with this company and they are legit. Why do you want to do this - is this a street car?

Forget the bit about the coolant pump, most are driven by the engine and therefore are inherently load sensitive, plus this controller doesn't have any output for an electric pump anyway.

It's a nice box, but why would anyone undo the factory setup?

To answer the question of stabilization, if the load on the engine is stable (as in constant speed highway driving), I would expect the coolant temperature to be relatively stable. Stop and go city traffic would probably be just the opposite, with the fan speed always chasing the change in load.

Why? Because the controller doesn't measure or respond to the ΔT across the radiator, it just knows the temperature of the coolant "somewhere" in the system. The ΔT is representative of all the forces acting on the cooling system; engine speed and loading, thermostat position, plus the cooling effect of airflow and ambient temperature.

Read the instructions; it's not waterproof (it has two holes for adjusting the potentiometers), be gentle when you adjust the pots, and don't put it where it can get too hot (sorta precludes under the hood of any car stuck in traffic).

I guess this product is for the guy who wants to do away with a mechanical fan and go electric. It can't be for energy conservation, it will never duplicate a factory setup, and it may even screw up the computers and pollution controls. Nice toy, no joy; kinda like those folks who put their own fan controllers into their PCs rather than use the ones built in to the motherboard.

It will depend on the ambient air temp, relative humidity, movement of vehicle, engine load and other conditions.....Since the heat load is constantly changing, the fan will vary between off, 1 and 2 speeds according to the variables....as to whether the temperature will stabilize between speeds, it depends on if the heat load is constant....

It will only be as stable as

• the configuration of the hardware
• the variations in the load on the control system
• the tuning in the controller that is varying the output signal

will allow, noe of which can be seen from here.

What is the point of maintaining a stable temperature, anyway?

Let me answer some of the questions. The car is my street rod, a '23 T-bucket, built with a remote radiator, electric water pump and electric fan. I experiment with different ways to control coolant temp. Started with simple thermostatic switch, then used a Dakota Digital unit that switches two fans with individual off and on temps. Got the Painless unit now and will install it this winter. I'm wondering what will happen. Will the temp stabilize or will the system always be chasing? We can simplify the analysis by assuming the car is just sitting at idle, removing some variables.

This is a very basic process control loop, without any self-tuning or sophisticated algorithms. Adjusting T1 and T2 is the only "setpoint" control.

The temp sensor is presently in the motor and the fans of course are on the remote radiator. Would it be better to have the sensor at the radiator?

any analog system controlled by electronic servo system will chase, thats what they do when inputs change. The amount of chase and the response time varies with the rate of change, the location(s) of sensors and the program of the controller. You are apparently not concerned with keeping a stable operating temperature of the engine because you do not have the most simple thermostat installed on the block, which is the best way to do it. The rest of the external, remote cooling system should be dealing with only the surplus heat form the engine beyond normal operating temperature - which is what your product is designed for, not for regulating engine core temperature. The electric pump is fine, if it is used to circulate the excess heat from the engine to the remote radiator, but that is the only real reason for its use here. the engine should have its own built in pump and thermostat and the high tech stuff should deal with the rest ( external, remote )of it. If you attempt to regulate the core temperature by regulating the entire large remote system you will find that the response time to keep the core stable will be too long and you will have many overheating events. T1 and T2 ( fan ) are for the external coolant system not the internal one. For this, yes, put the sensor at the rad because thats what you are regulating with the fan. Putting the sensor at the engine looks at heat that you have no direct ( electronic ) regulation for. What is the engine make and size you are using, its more important information than the type of vehicle.

I would install a thermostat, and an electric fan. Don't worry about it chasing temp...as long as there is air through the rad, the internal thermostat will handle it properly.

A constant speed electric fan will suck less energy when you are traveling because the incoming will windmill the fan, taking the load off the electrical system. So it will be self correcting and more efficient than people seem to think it will be.

With an electric water pump, you will have a fairly heavy load on your electrics, but it should give you an extra half second on the quarter mile. Just have an alternator which can handle the charging. You may will need a bigger battery...which will add, what, two or three pounds to sprung weight. You need that because there will sudden spike loads due to that electric water pump. Make sure you have the fluid bypass in place.

Good project. I have seen it done on the track here at Luskville. People seem to think going all electric works a treat, but they usually just replace all that complex electronics with a cockpit switch. An off the shelf timer to shut down the fan after a few minutes seems to be as complicated as they get.

Good luck with this project. Nice to see people working on cars.

Why is there a need to maintain a stable coolant temperature when the job of the coolant is to remove as much of the waste heat from the engine as practicable once the engine is inside its normal operating temperature range?

It's a 350 SBC crate motor. The electric water pump was needed because the motor has a Potvin blower which mounts to the front of the motor and there is no place for a mechanical pump, or a thermostat. Here's the setup:

Radiator is in the back, with long hoses in the frame.

The Dakota Digital unit worked fine for temp control, but it was constantly switching the fans off and on. I think the Painless unit, with it's fan speed modulation, should be easier on the fans. But they have lasted for 20 years already, so it's not a big deal apparently.

Guees then go with it , its all by guess and by golly from here. Cyclinging fans constantly proves the analysis of large system hysterisis. Have to find an algorithm to predict delta T to reduce the resopnse time and frequency of fan cycling. Placement of sensor will be important, trial and error for best location, not necessarily at either end of system. Nice looking rig, do you show it anywhere

Yes, car is 20 yrs old and has had many changes. Show it locally in MSP and big events. Took it to CA twice, once for the LA Roadster show. Here's a bunch of pics:

https://plus.google.com/photos/104971212334307682869/albums/5552767776876190417

Wish I had seen that picture before I wasted half an hour on my replies.