A couple questions on fin geometry and mechanical water pumps

[Chadwick Huntington]

Hi everyone,

I'm a sophomore at the University of Houston (second year team). This is my first year at FSAE and I personally have little background on car design and I had a couple of questions about radiators.

I've read pretty much every post on radiator/heat exchange design and I couldn't find much info on fin geometry design. One post said that "compact heat exchangers" by kays and london would answer my questions on fin design. I went ahead and ordered the book and Im still waiting for it to come in. Any thoughts on fin geometry preferences? Is it worth the time to do a full investigation into the subject? Our team is relatively small (6-7) people so man-hours are limited.

My other question is about heat rejection on mechanical water pumps. Since the flow rates vary, I was wondering if anyone could give me better estimates on the proportional increase of flow rate vs required heat rejection. The numbers I've seen floating around are 25k btu/hr at rest and 60k btu/hr max.

So...currently, our setup would be a 12x12x1" radiator with a perma-cool 12" fan that blows about 1650 cfm. Basic calculations indicate that this setup will reject roughly 32k btu/hr assuming 150gph water flow at rest, and 66k Btu/hr assuming 1000gph at 10k rpm. Are my calculations far off?

[Chadwick Huntington]

Hi everyone,

I'm a sophomore at the University of Houston (second year team). This is my first year at FSAE and I personally have little background on car design and I had a couple of questions about radiators.

I've read pretty much every post on radiator/heat exchange design and I couldn't find much info on fin geometry design. One post said that "compact heat exchangers" by kays and london would answer my questions on fin design. I went ahead and ordered the book and Im still waiting for it to come in. Any thoughts on fin geometry preferences? Is it worth the time to do a full investigation into the subject? Our team is relatively small (6-7) people so man-hours are limited.

My other question is about heat rejection on mechanical water pumps. Since the flow rates vary, I was wondering if anyone could give me better estimates on the proportional increase of flow rate vs required heat rejection. The numbers I've seen floating around are 25k btu/hr at rest and 60k btu/hr max.

So...currently, our setup would be a 12x12x1" radiator with a perma-cool 12" fan that blows about 1650 cfm. Basic calculations indicate that this setup will reject roughly 32k btu/hr assuming 150gph water flow at rest, and 66k Btu/hr assuming 1000gph at 10k rpm. Are my calculations far off?

[VFR750R]

I think you should talk to the radiator manufacturer you would like to use on fin design. Fin design is much more complicated then it first appears and in the end physical testing and experience of the manufacturer will help alot in coming to a final solution.

Mechanical water pumps are quite nice in that they provide an almost linear increase in water flow with RPM. And engines happen to need heat rejection proportional to RPM (horsepower, fuel lb/hr burned, ect) assuming WOT.

So what you'll find on a dyno is that at any RPM the temperature delta across the engine stays the same. Unfortunetly radiators do not remove heat linearly with water flow but are actually impressively insensitive to it. They are sensitive to air flow. Mechanical fans make a crazy amount of sense on engines run steady state. In FSAE though, there is alot of off throttle time, so you need to look at average heat rejection needed over time (an endurance lap for instance), and average air flow during that lap. And of course the more air you can get to the radiator just from the car moving will be less you have to force through it with a fan that draws current.

[Chadwick Huntington]

Thanks for the reply. I fired off a couple emails to see what I could gather from the radiator manufactures.

So, on the second part, you're saying I could run a dyno, find the rpm to hp ratio, and determine the amount of heat I need to reject at different RPMs?

Also, could I somehow validate my design choices through a simulation program like Ansys or CAITA?

[VFR750R]

Well if you are running on the dyno you could measure it directly at different RPM points, so you won't have to calculate the rpm/hp ratio. But in the end it depends on your average power over time while you are racing, which will be much harder to measure.

Although you could validate your design choices through a heat transfer simulation program (which I believe Ansys will do), but it won't get you much closer then a thorough hand calculation and you'll spend a ridiculous amount of time on it when in the end only real testing will show you if you are right or not. You'll just be guessing 50% of the variables for your model anyway.

[Wesley]

We're just using the dyno and a couple thermocouples to check our system capacity. Heat rejected by the engine to the coolant must be less than the capability of the radiator to reject heat on average.

Fin density and shape really aren't something most FSAE teams mess with, at least at the actual design level - all you need to know if heat rejection per area, so you can spec out your core size from a manufacturer.

You don't have to design a system to reject your max power heat load - it would work, but it would be gross overkill. What VFR said is right about coolant flow rates not impacting the heat rejection as much - and if you look at the theory behind it, you'll understand why - with a specific heat four times greater, a density 800 times greater, and a heat transfer coefficient on the order of 5-10 times higher, the water temperature isn't as greatly affected by the heat transfer as the air - hence, air flow rate has to be huge compared to that of water.

Just grab a couple of thermocouples and get some datapoints on the dyno - it will tell you exactly how much heat you need to reject to keep the engine cool. Figure out how much power you use on an average lap and under different conditions and test a few rad configurations. Same thing with fans.

[Chadwick Huntington]

Thanks for the insight VFR750R.