cooling tower

[Dave W]

Dave,

If you get the engine up to 200* to start a pull I'm assuming the water in the tower is the same temp. Is this water supplied by a cool supply like from the hose bib or does the tower act as a radiator and if so how would it keep the water cool for multiple pulls.

[86rocco]

On the dyno I've been using, the cooling tower is a tank about 15" in diameter and about 30" tall. The engine water pump draws water from the tank through the engine and back into the tank. The tank has a heating element inside to preheat the system, there's an overflow and there's a thermostatic valve which allows cold water from the mains to enter the bottom of the tank, the hot water exits through the overflow into a drain then out to the sewer.

[DaveMcLain]

The cooling tower is a very simple setup that works automatically to keep the engine at a stable temp. I run at about 190 degrees most of the time. How the cooling system works:

I have a circulation pump which is a 12 volt unit from CSI, it's mounted to the dyno frame and it pumps water to the engine. Most of the time I use a stock water pump bolted to the engine as a manifold, it doesn't have a belt or anything driving it, it just routes the water to both sides of the engine. I have a friend who uses a 110volt swimming pool pump for circulation. Plug it in at the start of testing, unplug it when you're done...

The cooling tower is made from a piece of 8 inch .125 wall aluminum tubing about 36 inches tall with a cap welded onto each end. It has a inlet and outlet fitting near the top and near the bottom. Inside it has a stand pipe which is a piece of 1-1/2 aluminum tubing. It extends up inside of the tower to within about 6 inches of the top and about 4 inches above the hot water return fitting. I would say a larger tower would tend to be more stable temperature wise but a smaller one would warm up faster, I think mine's about the right size.

I also have a piece of clear plastic hose hooked between two fittings on the side to let me see the coolant level in the tower. When I fill the engine for the first time generally I just fill it till water runs out of the drain letting me know it's full.

On the bottom of the cooling tower there are a few fittings. One's a drain for the system, another is a place to fill the tower. One fitting is for the sensing bulb of the thermostatic valve and the last one is where cool water from the thermostatic valve enters the tank.

The thermostatic valve has an adjustment to set the opening temp. and that's all it does, when the temp in the tower gets warm enough it opens and allows cool water from the water supply to enter. This forces the water level to go higher than the stand pipe and it spills over and runs out of a hose. I have a large piece of agricultural hose, that yellow stuff that serves to drain the hot water away from the testing stand.

If you look at the video of a dyno pull on this site you'll see the cooling system "dump" right near the end of the pull. I just let the circulator pump run between pulls most of the time to keep the engine at a nice even temp.

Having a non pressureized system like this is quite safe because it can't really blow hot water all over the place if something would happen. It does keep you from being able to test at coolant temperatures higher than about 210 degrees which isn't all that big of a deal but it might be to some folks who would prefer a more complicated closed system.

Hope this helps.

[Tony]

One problem I can see with this is that the engine cooling system is never going to be pressurized.

In a race engine running at full power, the cylinder heads MUST be highly pressurized to prevent steam bubbles from forming around the combustion chambers and exhaust ports. This water pressure increase arises from two separate causes.

First there is the normal static water pressure in the whole closed cooling system controlled by the radiator cap release pressure. But far more important than that, is the additional dynamic pressure created by the engine water pump running at high Rpm. Many people are not aware of this, but the water pressure in the cylinder heads should be running at something like 40 to 50 psi at full power engine Rpm.

This additional water pressure is accomplished by the water pump flow being restricted at the thermostat which is always located at the cylinder head outlet. This is designed to always offer a significant restriction to flow, even when fully wide open at full operating temperature. Have a look at your thermostat, and be surprised how small the water opening actually is. It is made that small for a specific purpose.

It is also why many racers have had problems when removing the normal factory thermostat. There is an urban myth that removing the thermostat causes the water to circulate too fast through the radiator for it to cool down. That is plainly nonsense. The real reason is that removing the thermostat vastly reduces the cylinder head high rpm internal water pressure. Flash steam bubbles form, the cooling system pressure spikes, and water is driven out out past the radiator cap. Within seconds the whole cooling system can be boiling and almost dry.

On an engine dyno, I would much prefer to use a marine heat exchanger, and a fully closed engine cooling loop exactly as it will be in the vehicle. And drive the water pump directly off the engine in the normal way. A pressure gauge indicating internal cylinder head water pressure should indicate if the water pump is working properly at maximum engine Rpm. If the pump cavitates, slow it down ! A cylinder head water pressure gauge should be an addition to any serious race engine dyno. If water pressure falls off at the top end, find out why and fix it.

The factory water pump and operating speed may need some modification, but it is not difficult, and well worth the trouble. Keep an eye on the cylinder head water pressure, you may be quite surprised.

All this is much less important for a very short burst of power than for a long pull flat out. A horrible cooling system might be quite o/k for a drag car. In something like a boat it can be one possible cause of detonation, or cracked cylinder heads in a really high power engine.

Developing a reliable and properly functioning cooling system is all part of developing reliable high power. I would much prefer to see the dyno cooling system be identical to the installed vehicle cooling system in every respect. The only difference being that the radiator is replaced with a water to water marine heat exchanger.

On the secondary side of the marine heat exchanger, you can let your imagination run wild as to how best to provide sufficient cooling capacity. But the engine cooling loop itself should remain sacred.

The problem we are trying to avoid is small pin head sized steam bubbles forming on the hottest parts of the head casting. The steam then drives away the cooling water, and the steam bubble rapidly explodes in size. The result is often a cracked cylinder head. Very high static water pressure inside the cylinder head helps, as does sufficient water flow that will naturally occur with an engine water pump running at many thousands of Rpm.

Some food for thought, but an engine dyno is the ideal environment for working up a good high pressure cylinder head cooling system.

[DaveMcLain]

I don't think it would be all that difficult to make a heat exchanger like you're talking about especially if you're going to be water to water. I would be willing to bet a simple loop inside of a cooling tower like mine would do the job. I guess you could build it that way to give you options. You could run a loop through the tower and when you want to run closed loop hook it up and plug the other holes. Unpressurized unplug and use the holes in the tower and leave the loop open.

Sometime I would like to test with a closed system but I've not had any problems with my setup, the pulls are just not that long for the type of testing I do, 15 seconds is about the max. I used to run with an engine driven pump and it was a hassle with different engines, pulleys etc. Not to mention having a belt to deal with all the time.

I've even tested fans on my dyno so really if you build a versatile system you can test just about anything or any combination of things.

[Tony]

Yes, I agree. All this cooling stuff can be pretty well ignored for very short acceleration sweeps on a dyno, and for drag racing, or even most high power street engines.

But a high powered race engine that has to hold full throttle for maybe a full minute (or more), might be in some serious trouble without a properly pressurized cooling system.

Also surprisingly, a standard 150Hp car pulling a horse float up a very long hill will be in greater danger from boiling than a light weight 800 Hp turbo car.

It all has to do mostly with how long you need to hold full throttle to get the job done. Quite often, the more power you have, the less time you can keep the throttle fully open. That is the main reason most people get away with huge power increase and still run a stock radiator and cooling system.

A cooling loop in the sump of a cooling tower should be ideal. Water to water heat exchangers require remarkably little surface area to transfer a lot of heat.

[86rocco]

[color=#000000]Based partly on the discussions in this thread, I've been thinking about re-doing our cooling tower. To that end, I've been doing a few calculations but I appear to be doing something wrong, the numbers I'm getting aren't making sense. So does anyone have any rough esimates how much cooling an engine requires?

For example let's say I've got 10

[Tony]

My understanding is that for many typical gasoline engines, roughly half the heat of combustion goes down the exhaust, one quarter turns the crank, and the other quarter ends up in the cooling system. That is after all the heat flows have become stable.

So as a guess, the cooling water will gain roughly about 200 Hp worth of heat if the engine is delivering a continuous steady 200 Hp. That estimate may not be exact, but it is at least something to start from.

One horsepower equals 10,693 Calories per minute. And one Calorie heats one cc of water one degree Celsius per minute.

So one horsepower will heat 10.693 litres of water one degree per minute.

Forty horsepower will heat the same water flow by 40C. So something around 53.5 litres per minute may be a reasonable starting point.

[86rocco]

Thanks Tony. I'd been working on the math a bit more and when I plug 25% heat loss through the cooling system into my work sheet, I come up with 48.6 L/min so we're in the same basepark so after seeing your numbers, I feel a lot more confident about my calculations.

It's a little rough but in case you're interested, here's my spreadsheet, it starts with the energy content of gasoline and the brake specific fuel comsumption of the engine and works the calculations from there.

[Thomas Vaught]

As some of you know, I work for Ford Research.

We do a lot of engine dyno testing for long periods of time, (like
6 hours at max rpm plus 10 percent). We use a cooling tower and
a LARGE heat exchanger (with city water at 40 degrees) to control
the engine temp.

The engine is set up just like an engine in a vehicle (except no radiator
as the cooling tower/ heat exchanger takes its place), thermostat, etc.

Hope this helps.

Tom V.

ps We also use a oil cooling cart with a large volume of oil that is cooled to maintain a proper temp as well as remove any air from the oil.

[86rocco]

That's the sort of thing I'm thinking of doing although, I don't plan on putting our engines through nearly the torture tests that Ford does. There's a GM engine plant not too far from where I live, I visited there a few times when I was a kid, I can vividly remember seeing engines in their dyno cells, screaming at high rpm and glowing hot, when you're a 12 year old obsessed with cars, it doesn't get any better than that!

Anyway, now that I think I've got a handle on the cooling loads required, I've got a starting point for the design of the heat exchanger for the cooling tower and I know a few people that are well equipped to help me with that. BTW, The engines I'll be testing are mostly 4 cylinder normal aspirated engines putting out between 180 and 250hp

[Tony]

The heat rejection figures are a bit elastic because ignition timing and air fuel ratio are going to effect heat rejection. But something roughly around 50 Lpm for 200 Hp sounds about right.

A water to water heat exchanger may not need to be as physically large as you might expect. Marine heat exchangers are a good example. The engine running temperature may be 85C and the river/sea cooling water may only be 15C. Heat exchanger capacity increases directly with the temperature difference and surface area. Marine exchangers, though astonishingly small, are still conservatively sized for continuous flat out full power operation.

But it is definitely the correct way to do it. Keep the engine pressurised cooling circuit IDENTICAL to in the vehicle, and replace the radiator with a water to water heat exchanger.