Cooling system doubts.

[wolf14]

Hello everyone,

Our team is working on a custom radiator for our KTM DUKE 390 engine, the first time. We had some doubts about the same.
1. How to select a thermostat? Our current thermostat being the stock one which opens at 88 degrees. We think it should open at a lower temperature since the stock one must have been designed considering antifreeze as coolant.
2. How to find engine heat rejection at various rpms? We needed that for radiator design calculations.
3. How to decide or find out the optimum temperature required by the engine to function efficiently?
4. What values of water pump flow rate have you'll got for KTM 390? (Just needed to verify!)

Eagerly waiting for your replies.

Team Octane Racing,
College of Engineering Pune.

[JSR]

Reading this might help a little

http://www.fsae.com/forums/showthrea...adiator-design

[DanielH]

Hello Wolf14,
My name is Daniel What is yours

With regards to your question about radiators, why are you building your own custom radiators. why not buy a slightly smaller radiator. or use the stock radiator. In my experience the cooling system is a place where you lose reliability and not gain performance.

1. The stock thermostat is usually pretty good. The engine tends to like to work around the 85-95C(just above the thermostat opening temp). Water has a better heat capacity than antifreeze water mix.
2. Test it. find the flow rate and difference of temp through the radiator. No one is going to give that info to you.
3. see 1.
4. We talked to a local race team( that uses our engine) or the manufacturer(which is local as well) and they gave us a chart of flow vs RPM. That beeing said the best is to get yourself a bucket and a stopwatch and test the thing.

Make the cooling reliable and legal and test it on a hot day. Then test a lot more. Then, once you have done a lot more testing, try putting a slightly smaller radiator.
That being said, calculations to back up the testing is definitely needed

[Kevin Hayward]

A couple of answers:

1) Why run a thermostat at all?
2) Test it (Thermodynamic calcs can get you close)
3) Test it
4) Test it

In assuming you have an ECU with the ability to control on and off temperatures for a fan (and a water pump if you go electric) then you can run your engine at a number of temperatures pretty easily. Also worth considering how your oil is cooled. If it is by a water/oil heat exchanger you are probably already aware that an engines performance may be more sensitive to changes in oil temperature than it is to water temperature. With that in mind you might be designing your cooling system with two temperatures in mind.

I haven't spent any time with KTM engines, but 88degC seems very low. I am aware of teams that run as high as 130degC continuously in endurance. 100-105degC seems to be a pretty good safe number, with a higher oil temp (wont tell you how much higher). At those sorts of temps your thermostat is just an open valve that still has the potential to fail.

Kev

[wolf14]

Hello everyone,

Thank you for the replies.

Daniel this is Sanath.

Thanks for your advice Kevin sir, but this being our first time study in cooling system, our scope is limited to just the cooling of the coolant. Too many things at one time will hamper our progress(step wise study).

I'm eager to listen from others too. If they have something to say on this topic. Also are there any other things to be considered or precautions regarding the same.

Sanath
Team Octane Racing.
College of Engineering Pune.

[Kevin Hayward]

Sanath,

What are you cooling with your radiator?

The point I was making was that your oil is likely cooled by your coolant. This either occurs through a water/oil heat exchanger or just in the fact that the coolant is generally cooling the engine. An alternative is to have separate cooling of the oil. I am not suggesting that you build a separate oil cooler.

You are asking the question of what temperature do you want to run the coolant at, and how do you decide what is best (Your point 3). Do you run 88deg as per stock or is there a better way to determine adequate cooling.

I propose some limits for coolant temperature:

1) Physical properties of water hoses and connections. There is a limit to how high you can run temperature as the system becomes highly pressurised in order to keep the water from boiling.

2) Physical damage to the engine. Components will undergo thermal stressing, some plastic parts will soften, parts weaken under temperature etc. This is dependent on design and material selection.

3) Limits from the behaviour of oil at various temperatures. Oil viscosity changes and certain additives are broken down at elevated temperatures. Different oils break down at different temperatures and have different viscosity vs. temperature behaviour. A good reason to understand a little bit about oils and the various bases used.

4) Cooling efficiency. Heat exchangers work on a delta T, if you can run higher water temps for other reasons it allows you to use a lower radiator as the delta T between the water and the air will increase improving efficiency.

My point is that it is likely that your engine's performance may be more closely linked to oil temperature than water temperature. You might find for instance that at a water temperature of 88deg your oil temp might be 160deg and too hot for good performance, or conversely that at a water temp of 88deg your oil might be 100deg, which may be too cold for ideal performance. You will not answer your third point without some understanding of how your water temp affects clearances, oil temp, and air temp. This can be done with testing, and try to make sure you are monitoring oil temperature.

...

On another note you may want to consider that the radiator is not the only way an engine cools. Surface area and airflow over the engine is not insubstantial, especially surface area and airflow over the sump or oil tank. By putting the engine in a car you will be affecting this cooling mechanism. Heat is also rejected by the exhaust, if this heat is retained around the engine due to structure around the exhaust pipe you could be dealing with a larger cooling issue than the original bike.

The easy "get it done" answer is to run a decent sized over the top radiator, but if you want to do it right you really do need to consider things such as oil temp, airflow over the engine, location in the car. Some of this can be hard to model, but a decent test program could get you a long way. There are no quick answers to questions 2 and 3, which are the ones that drive your requirements.

In the short term you are probably best to be heavy handed and make sure you have enough cooling capacity, and then do a bunch of testing for later years to help gain the small weight and performance advantages possible. With this in mind you probably will have more issues with radiator placement and getting clear air feeds and exits as well as appropriate fans and shrouding, than you will with radiator and water pump sizing.

Cheers,

Kev

[MCoach]

Don't bother with a thermostat, they like to break and are just extra weight at that point. They are also difficult to find ones intended for temps high enough for motorsport applications.

Electronic control of a fan or auxiliary water pump is a pretty reliable way to go about it if you want really fine control over temperatures all the time.
Tape over the radiator/duct is good for a set-it-and-forget-it adjustment. Good enough for Z, and world level karts, and cup cars.

Doing a temperature model should give you a decent estimate of radiator sizing. Temp in --> Rad thermodynamics --> temp out.
If you do it in excel, it can give you a nice static answer, enough to run with after a simple "hand calc". Something like Simulink can give you a decent a dynamic heating output.

If you get clever, you can couple that model with a car/track simulation file or recorded data and predict your temps to make decisions way ahead of time, knowing what size radiator you should need.

[Swiftus]

Thermostats like to break? I disagree with that point - they live for hundreds of thousands of miles in road cars with little problems.

And why the huge obsession with ultra-hot water temps? 130 C temps may make your engine more efficient, but you likely are so close to detonation all the time that your engine will just be looking for an excuse to have some unplanned disassembly. Increase the CR, decrease the temps, runs a thermostat so you don't have to think about how big is too big in your cooling system. We only get to use 93 octane fuels (or e85 which gets you a few more kPa boost or a little more CR). A smaller rad fits in the same space as a larger rad, but a larger one won't fit where a smaller one was.

[Housebaratheon]

I would run a FS car without a thermostat, for our application it won't matter much if the cooling system is cooling during engine 'warm up'. In my opinion it's not worth the risk, air easily gets trapped and there's always the risk of the thermostat not opening when you need it to.

[MCoach]

Thermostats like to break? I disagree with that point - they live for hundreds of thousands of miles in road cars with little problems.

I'm always replacing them on everyones' cars, one of the more common parts to fail. Anecdotal experience, though.
I'm sure that assembly and dis-assembly by amateurs for FSAE doesn't make that rate any better.

#Leanerismeaner
#ThisiswhyImhot