radiator design

[Colin H]

Can anyone recommend some reading material on heat exchanger design? So far I hav been relying on intercooler theory on the web and old fluids and heat transfer texts as reference for my rad design. What are other teams using?

Thanks for the help.

[BStoney]

I could use some insight also.

B.J. Stoney
University of Cincinnati
bjstoney13@yahoo.com

[Spike]

RAD Testing (http://sdsefi.com/air10.html)

[Leida Monterrosa]

Firstly excume for my English...

Colin... I`m Leida from Universidad Simón Bolìvar team in Venezuela. I would like to know if you found the information asked about radiators design. In that case, would you mind on telling me in what sources you found the information?

.................
How many universities use the original engine radiator?

Leida Monterrosa
F-SAE USB
www.formulasaeusb.com (http://www.formulasaeusb.com)

[Frank]

We had an excelent cooling system last year.
radiator 350*370*25mm, standard water pump, 8 amp fan. The radiator hang up the back of the car(ugly, but effective)

This year we had a terrible cooling system.
radiator 240*295*40mm, electric water pump, 6.5 amp fan (way to small).

Both units were well shrouded.

After the engine started overheating, we tested the new system experimentally.

We got an urn (you now the big things for heating water for coffee that they have at functions)..

We heated the water to 80 deg C, and ran the electric fan and water pump, using thermometers either side of the radiator.

we found the system was good for about 6kw.

we then varied the water flow rate (pump), and air flow rate (fan), outputs by varying the supplied voltage. We measured the air flow rate with an airspeed meter (little fan with a counter). Obviously you can simply wiegh the water after a 1 min run to find mass flow rate.

We found that both airflow and waterflow rates increased cooling capacity quite linearly (for given flow rate of other). BUT it was proportional with factors quite a bit less than one.

Both functions also had a quite large scalar constant. ie.. the zero flow rate condition caused cooling (as you would expect)

example:
cooling= 0.2*wfr + 1.5kw (for given afr)
cooling= 0.6*afr + 1.2kw (for given wfr)

Now, for our system, the waterflow rate had quite a low linear constant, the airflow constant was nearer one.

i cant remember the exact figures, but for example (from the original condition) doubling the waterflow rate gave us 20% more cooling

doubling the airflow rate gave us 40% more cooling.

we were in a rush, so more academic statistical analysis wasn't done,we simply said "right put a bigger fan on quick!"

Now like I said earlier the system initially shifted 6kw.

We then put the engine on the dyno. When the engine reached about 6kw the cooling system failed (the temperature started rising)

Several text books suggest that the shaft horsepower from an engine is roughly the same as the heat leaving the (water) cooling system. So this made sense....

Now we considered the use of a formula SAE at competition. 32mins racing, 3 mins between heats=35mins

fuel consumption was 4 ltrs
calorific value of fuel = 48Mj/kg
SG fuel 0.75
20% of heat into engine leaves via radiator
= 14kw (average)

from talking to engineers at a radiator manufacturer, it is obvious that it ALL the data is derived experimentally, and without prior data you're stuffed. and even when the system is installed it might still not work.

the most important thing is understanding what a thermostat does. You should use one. and the cooling system temp should be running constant temperature ALL DAY (retrieve the data from Motec computer). ONLY when this is the case do you have a cooling system that is "working".

my personal advice is a 300*350*25mm radiator, with a "ballsy" fan (8amp) and a good shroud. The unit should be experimentally tested before the car is run. and you should preferably do the testing before the dyno, and use the unit on the dyno.

keep in mind we're australian, so 35 deg C days are NOT uncommon, so its pretty important here.

(sorry about the essay)

http://fsae.com/groupee_common/emoticons/icon_eek.gif

[PJL]

Any updated on radiator designs now that the Detroit competition is over?

The radiator (Al, 10"x12"x1.625", 18 FPI) and fan (pulls air @ 26 KPH) combo we used worked fine. The only problem we had was with packaging. This year we plan on using a radiator approximately 14"�x7"�x2"�. They are similar volumes just different frontal areas.

How did everyone size there radiator? Heat transfer data manipulation, experimental methods, or monkeys throwing darts at a wall?


Any suggestions, or comments?

[Sam]

I would advise against using the thicker core to give a similar volume, you will probably find that your cooling capacity will decrease.

due to the low speed these cars run at, even with fan assistance, cooling will most probably be acheived only in the first 1" of core, the rest of the volume will act as a heatsink, slowing heat transfer from the fluid. a thicker radiator wil only benefit from serious airflow.

PJL.. you may find that your 14X7 radiator will perform better with a 1 - 1.5" core

I found this out the hard way. see frank's message above...

Sam Graham
Engine Group Leader 2003
UQ Racing

[Charlie]

Sam-

What kind of airflow rates are you talking about (that thicker radiators are not a benefit)?

-Charlie Ping
Auburn University FSAE (http://eng.auburn.edu/organizations/SAE/AUFSAE)
5th Overall Detroit 2003
? Overall Aussie 2003. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[BeaverGuy]

For good reading I recomend Compact Heat Exchangers by W.M. Kays and A.L. London.

I was a bit lost until I found this book. I haven't had any fluid or heat transfer classes yet but this book has everything you need to know if you have had the classes or are willing to think a little bit in order to figure out how they are defining their values.
I'll hopefully be writing a matlab program to calculate radiator sizes this week.

The reason for thicker cores being less desirable is hotter air travels across the thicker core decreasing effectiveness. So for cores of equal volume a thinner core is more efficient.

PJL, are you sure about the velocity that your fan pulls air? if so then it would be a a fairly stout fan at nearly 1200 cfm.
I'm designing the cooling system for this years car using the NTU method with some experimental fluid flow values, though in previous years I believe monkeys threw darts to determine it.

The radiator on the past 3 or 4 cars has been 8 x 12 x 2 1/2. It works ok at speed but the fan isn't large enough and it doesn't have a real shroud. Prelimnary calculations say a radiator of the same frontal area and about 3/4" thick will be fine.

With a gas-liquid heat exchanger, increasing gas flow, in this case air, will almost allways be helpful because the heat capacity rate(Heat capacity times mass rate) of the gas is generally less than a quarter of that for the liquid. The closer the rate ratio is to 1 the more efficient the system will be and the smaller it can be in size.

I haven't found much information on shroud design, just some recomendations from the American Petroleum Institute. Where have you guys found your info?

[Rhoades]

Can anyone give me the dimensions on the stock CBR F4i radiator?
This info would be greatly appreciated.

[Mechanicaldan]

Well, I just did a little thinking. Our engine is 600cc, and is raced pretty hard at FSAE. Being around motorcycles for a long time, I've noticed that the factory motorcycle race teams always put on larger radiators. So, it made sense to not use a 600cc radiator, but pick one from a larger displacement sportbike. Most sportbike radiators are curved, but not all are. The '93-'94 Honda CBR900RR and '98-'00 Honda CBR1000XX radiators are flat. We used a XX radiator with 2 smaller puller style fans without any problems, and I believe we could go to 1 slightly larger fan. I just purchased another XX radiator off E-bay for $50. I love E-bay. I have not seen a RR radiator on E-bay, so good luck. The extreme sportbike guys crash those bikes A LOT making videos.

Yes, I do realize that there should be some analysis going into the cooling system design, but the testing proved that what we have works and holds any temperature we set in the ECU. More testing will prove if we can get rid of the weight of one fan. Cheap, aluminum and light, and it works. Well, only cheap if you get them used. The XX radiator is $500 from a dealer, and the RR radiator is $600.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Project Director

[BeaverGuy]

I just finished writing a program for radiator sizing. It is a javascript app and is currently available hereRadiator Program (http://www.geocities.com/mustangdude.geo/radiator.html).

I would like all the feedback I can get. For those of you that have been able to do a fair amount of testing on your cooling system if you could give the program a try and tell me if it seems to give you the same numbers that you have gotten experimentally or by hand it would be very appreciated.

[BeaverGuy]

I have fixed all the errors in the program that I have found. I have also added a little bit of an explanation for some of parameters.

I know a few of you have at least looked at the program if you haven't used it. Any feedback I can get would be very helpful.

Thanks

[ATOhoosier]

I've read a couple of places (mostly webistes and postings) that the heat output from the engine can be approximated by using the shaft power. Is this just a rule of thumb or can someone direct me to book or something? Some of the team members are designing the cooling system for our senior design project so we need something more to go on than forum postings. Thanks a lot. Just reading these posting have been incredibly helpful. Hope to see you all in Detriot in 2004!

- Jonathan Strunk
Mercer FSAE
Macon, GA

[BeaverGuy]

One way that you could measure the heat load of the engine is to measure temperature before and after the radiator and also the flow rate of the water. By doing this you know the change in temperature and the amount of water that is flowing now multiplying temperature change times mass flow rate times specific heat will give you the heat load, Q.

The method above using the specific fuel consumption of the engine will give you an aproximate value if you have a good idea of how much of the heat is rejected by the engine into the cooling system.

[gug]

i would guess that the shaft power thing assumes every engine has a constant inefficiency. ie. rule of thumb. dont bother with it, it cant get you accurate enough for a senior design project.

- if it isnt coming, you need a bigger tool.

[ATOhoosier]

Thanks for the help everyone!

Jonathan Strunk
Cooling System Team Leader
Mercer FSAE
Mercer University
Macon, GA

[Scott Wordley]

With regards to beaver guys comments above check out these pictures

http://www-personal.monash.edu.au/~fsae/03build.htm

Scroll down to where is says James' Radiator Test rig. Logging temps on the car will be used for comparison.

Regards,

Scott Wordley

http://www-personal.monash.edu.au/~fsae

[LukeT]

Last year our car used 2 7x9x1.5 radiators from Ron Davis Racing, about 16 fpi. It stayed nice and cool, as long as both had fans on (not even shrouded, at that) when it idled. Right now we're testing a 12x9x1.4 18 fpi radiator off the Honda Rubicon ATV, using a mathmodel that a (now)grad student developed it should be of adequate size...


-Luke Thompson
Vandals Racing 2003-04
University of Idaho

[David Kieke]

Last year we used a stock F4i radiator (~9x17x1) along with a small oil cooler. Both had shrouded fans and we have never had any issues. We have reduced the radiator size this year while keeping the same size oil cooler (though changing from steel to aluminum). The F4i (and earlier Hondas as I recall) have oil jets directed at the back of the piston so cooling the oil (with air, not just the stock liquid-liquid) can have a more dramatic effect than increasing the size of the water radiator.

[LukeT]

What sort of oil cooler setup are you running? We're looking into removing the stock F3 cooler, replacing it with a remote filter mount, oil thermostat, and ATV-sized external oil cooler. It's looking like we need to machine 4 parts to do that, plus buying steel braided hose and fittings, did you guys do something similar?


-Luke Thompson
Vandals Racing 2003-04
University of Idaho

[David Kieke]

That's basically what we did, but only machined a threaded extension for the oil filter. We ran an oil cooler from Pegasus (~4" x 10") along with a sandwich plate from Pegasus (removed the stock oil cooler, threaded the sandwich plate on, and threaded the oil filter onto the plate). Then we just needed braided hose to go from the sandwich plate to the cooler.

We ran an F4i, and I've never worked with an F3. If it has the larger rectangular oil cooler (like the F2) there may be some fabrication tasks to convert it.

David

[PJL]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by LukeT:
....... Right now we're testing a 12x9x1.4 18 fpi radiator off the Honda Rubicon ATV, using a mathmodel that a (now)grad student developed it _should_ be of adequate size...


-Luke Thompson
Vandals Racing 2003-04
University of Idaho<HR></BLOCKQUOTE>

same size as we used last year and we found it to bee a bit too big. This year i think we will go with Approx 7*15*1.5...... at least thats what the calcs and tests are saying.

[Hihanexal]

I am a new member of the formula racing team at Mayaguez. I have to design a radiator and I need the basics, since this is the first time I'm doing this. I hope someone can help me.

sincerily, Nexal Flores

[BeaverGuy]

It is my first time with the team and my first time designing a radiator. But I have everything down pretty good right now. The best thing to do is read Compact Heat Exchangers by W.M. Kays and A.L. London. It will give you the theoretical background neccesary to design the radiator once you have the real world data like heat load and fluid flow rates.

[ATOhoosier]

We're a first year team experiencing all sorts of problems so by advide may not be the best. However, we've found "Compact Heat Exchangers" by Kays and London to be extrememly helpful. The only problem we're running into is that there isn't really much in the book in terms of core cofigurations for small cars and motorcycles. Another book we're trying to find is "Fundamentals of Heat Exchanger Design" by Shah. Appearently Shah worked for GM and his books have more of an automotice slant on the information contained.

On that note, how much design are the more experienced teams doing? Are you able to find core configurations/ tube pass configurations for common bike radiators? If so where from?

Thanks.

Jonathan Strunk
Cooling System Team Leader
Mercer FSAE
Mercer University
Macon, GA

[BeaverGuy]

you don't need much info on core configurations as that is what you are designing. You do however need tube and fin size information , for which a current radiator would be a suitable source. As for the number of passes to make that is also your decision as is the number of tubes deep to make the core.

we got our radiator from Griffon radiator last year and the dimensions for the tubes are .085" high by 1.25" wide and a common wall thickness for alluminum tubes is .010" and fin thickness of .006"

These are the numbers I'm going to use for tube height, wall thickness and fin thickness while adjusting tube width, during my design.

I'm not doing much with the cooling design this year other than sizing the radiator, however I woudl like to get an electric water pump and a have the ECEs build a water pump and fan controller next year that way it might be possible to go to an even smaller radiator.

[PJL]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Frank:
.........................
Both functions also had a quite large scalar constant. ie.. the zero flow rate condition caused cooling (as you would expect)

example:
cooling= 0.2*wfr + 1.5kw (for given afr)
cooling= 0.6*afr + 1.2kw (for given wfr)

Now, for our system, the waterflow rate had quite a low linear constant, the airflow constant was nearer one.

.......

http://fsae.com/groupee_common/emoticons/icon_eek.gif<HR></BLOCKQUOTE>
Did you find any correlations between the overall heat transfer coefficient water flow, airflow, water temp differences?

Just wondering because i received heat transfer data from our radiator MFG and i've been trying to determine an experimental heat transfer coefficient and match it to values I calculated.

[BeaverGuy]

PJL,
depending on the flow configurations you can get analytical relationships between heat transfer coeeficient, and flow rates and temperatures. However, for an unmixed crossflow arangement there isn't a closed analytical solution relating the number of transfer units to the efficiciency and heat capacity rates.
Using the following relations you can find a heat transfer coeeficient from your test data for flow rates and temperatures. The terminology is that from Compact Heat Exchangers by Kays and London.
Ntu=A*U/Cmin
Eff=[Cc*(Tc,out-Tc,in)]/[Cmin*(Th,in-Tc,in)]
where R=Cmin/Cmax
Eff=f(R,Ntu), if you can find a relation for unmixed crossflow and solve for Ntu, you should then be able to solve for the heat transfer cooeficients


I just found an SAE paper relating temperature and flowrate changes to Specific Disipation. There may be some helpful equations in there for you.here is the link (http://www.delphi.com/pdf/techpapers/2000-01-0579.pdf)

[waz]

We have just tested our 2004 radiator 275*300*40mm and a 10 inch davies craig fan.

Calculations have been made following the test day and we estimated the following. At air flow of 250L/s we achieved 10.4 kW of cooling. The pressure drop accross the core is 55 Pa. This was confired by water side calculations. From the data recorded we would estimate that the out car requires 30 - 50 % more capacity. It is interesting that people are saying not to use greater than a 25mm core though.

We measured an air off temperature of 65 degrees and 30 degree ambient at incoiming water temperature of 105 degrees.

[rjwoods77]

Anyone ever use an air cooled motor?

[jonno]

Rob,
I've seen pictures of what I beleive is an american team at one of the early Formula Student events. They used an air cooled honda v-twin, longitudinally mounted with a shaft output. I've sent you the pic (is there a way to post pictures without them already being on a website?)
I'd be concerned that sufficient airflow was behind the driver to cool the head, especially at the low fsae speeds

[David H.]

Hey Jonathan

I'm starting a FSAE program here at Virginia Commonwealth University (VCU) and i was talking w/ my friend about having our car air cooled. My friend said through his experience with FSAE in Cali their biggest problem was cooling and where to put the radiator so it didn't create much drag. Could you send me a few pics on the air cooled FSAE car you saw. Thanks

~David
formulavcu@gmail.com

[BeaverGuy]

If I was to use an aircooled engine I would probably do it the same way as water cooled engine. I would build ducting to blow air over the engine instead of a radiator. In addition I would probably see how a fan could help the cooling.

[jonno]

David
Hope you got the pic i emailed you of the air cooled car. I'd have concerns about getting sufficient cooling to the cylinder head with it being sat directly behind the driver. you'd more than likely need some impressive ducting to get it working well. If drag is your biggest concern, i'd have thought a water cooled system would minimise this. You don't need sidepods: you can eliminate them reducing frontal area. Brunel have run without sidepods for at least 4 years now
http://www.brunel.ac.uk/faculty/tech/Racing/Images/galleryFS2003/4.JPG
The rad is ground mounted infront of the engine: no sidepods, therefore reduced frontal area & reduced drag. The fan is quite essential however: this sucks air from underneath the car exploiting a nice loophole in the FSAE rules where fans may be used for cooling. Although not to the extent of Gordon Murray's Brabham BT46B (http://www.f1rejects.com/hall/beta/index.html#2), a side effect of this is that you are removing air from underneath the car, creating some downforce.

[alfordda]

Jonathan-

We ran our radiator in the same location for Detroit 04, but we had some cooling issues. Our radiator shroud was pretty poor, it only sucked air through about half of our radiator. Have you guys had any cooling problems?

[jonno]

none whatsoever, hottest the coolant has ever got is 84 degrees C so far: fine for the UK and Detroit. The pic is of BR-4, which is an inferior system to BR-5 - but there are no pictures of BR-5 on the internet yet. BR-5 uses the Davies Craig Booster Pump (http://www.daviescraig.com.au/main/display.asp?pid=27), not the much heavier - but more powerful - water pump, therefore our flowrates are real slow. However the newer system has minimal coolant side pressure drop, no bends in any of the hoses, cross flow heat exchanger etc. The key is the 200mm x 200mm x 19mm heat exchanger: 98fpd louvered fins (0.076mm gauge material) and similar gauge reformed tubes.
In terms of shrouding, i was particularly anal about getting the shroud airtight (critical) - i can't beleive how many teams "attach" a heat exchanger somewhere in an airstream and expect air to flow straight through it. Our shroud was manufactured from fiberglass. However, I filled the unit along all seams with silicone sealant (including the circle between the fan and the shroud). Also, the fins are extended to the headers: all the air that fan is sucking through has to pass through the fins.
Finally, you want to place your fan as far as you can from the heat exchanger, as there will be a "blind spot" behind the hub of the fan. This is well depicted on this cpu fan site (http://www.coolermaster.com/index.php?LT=english&Language_s=2&url_place=product&p_serial=ISB-V73&other_title=ISB-V73Aero%204)
Hope that's all usefull

[ATOhoosier]

Hey David,
I wish I could help you out but unforunately, funding was pulled for our FSAE team and the equipment purchased or constructed was used for various senior design projects. Sorry i couldn't be of more help the only suggestion I have is try to keep it simple at first. We were going to place it behind the drive and have a fan either pushing or pulling air over the radiator coils. Good luck with the team, I hope you guys have more success than we did.

[David H.]

Wow that sucks that you got ur funding pulled... I'm really hoping that doesn't happen to my team. Thats my biggest fear right now, not finding enough sponsors, and not having VCU back me up. My second fear is aquiring all the technical info needed to build one of these cars. I've been reading the forums for a while and i get lost easily on all the terms u guys use. But thanks for the tips on radiators. I'm thinking of just going for either one or two bike radiators and putting them under the car w/ fans, that seems like the best and simplest idea to me.

[Barnard1425]

As I look at that picture, a few things come to mind. First, with the radiator in that position the exhaust header and muffler are moved to a higher location, thus raising the center of gravity of the car. It also looks like the radiator placement is forcing the team to choose between either a really long car or an unnecessarily upright driver position (again, raising the cg).

Is drag on sidepods really that big of an issue on the tight Detroit courses? Has anybody come across any data regarding drag over sidepods in slower speed racing conditions?

[jonno]

as i said before, the picture is of br-4 (car from 2003 formula student comp), the reason for the exhaust abruptly coming up is because the spaceframe is in the way (bear in mind that br-4 was made by a team of 6 people in 1 year) therefore the exhaust was a bit of an afterthought: hence it's interference with the spaceframe.
The car is not "unnecessarily" long, sure it's a 64 inch wheelbase: but 4 inches over minimum is not such a bad thing... neither is the fractional increase in c of g by raising the 4kg of exhaust pipe slightly. The competition is about evaluating where compromises have to be made, and the raising of that exhaust was evaluated as the correct thing to do to fit in line with the marketing angle of servicability pushed with BR-4.
In terms of lengthening the car / sitting upright the answer is neither. With the CBR600 engine and the chain drive, the rear portion of the car cannot be any shorter: the head is not far off touching the firewall. The driver lies down in BR-4: the fuel tank fills the space under his back, and the radiator then fills the gap between the fuel tank and the crankshaft.
Whilst the cars may not actually get up to speeds where drag from sidepods is considerable, the way the brief markets the car is for a weekend racer who desires performance: therefore drag minimisation can be argued in in the design judging. The car would be no shorter with the cooling system in sidepods, the driver's driving position would be unchanged by moving the radiator to sidepods, therefore the reason that the radiator is there is that it not comprimising anything, and we get design points for it: in fact there was a rumour going round that had we run at formula student 2004, the cooling system was in contention for the cosworth engine prize (BR-5's including coolant weighs less than 5kg). So the judges must like something about it http://fsae.com/groupee_common/emoticons/icon_wink.gif

[alfordda]

Jonathan-

I've been looking for a pic of our setup, but haven't found a good one. I'm not an engine guy by any means, so forgive me for asking a stupid question here. Is that booster pump in addition to the stock water pump or a replacement for it? I don''t know the exact dims of our radiator, but it is about 6x12x1. I can put my hand under the car and only feel air being pulled through about half of the radiator. So, I know our shourd is the problem, well at least its the first step.

[jonno]

The electrical booster pump replaces the stock mechanical pump: the mechanical drive is used for the oil dry sump scavenge pump: hence why we use the electric pump.

Shroud sealing and distance as i said before is critical. Glass fiber is cheap, so make yourself a mould out of expanded polystyrene, wrap release film around that, lay glass fiber on that & you have a shroud! Seal it like i said with silicone, get the fan as far away from the heat exchanger as possible, and you should have even airflow.

[danzim]

what's the best setup for testing radiator/fan systems? did someone say something about using a tea urn? are people finding that testing supports theoretical calcs.?

dan

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by waz:
We have just tested our 2004 radiator 275*300*40mm and a 10 inch davies craig fan.

Calculations have been made following the test day and we estimated the following. At air flow of 250L/s we achieved 10.4 kW of cooling. The pressure drop accross the core is 55 Pa. This was confired by water side calculations. From the data recorded we would estimate that the out car requires 30 - 50 % more capacity. It is interesting that people are saying not to use greater than a 25mm core though.

We measured an air off temperature of 65 degrees and 30 degree ambient at incoiming water temperature of 105 degrees. <HR></BLOCKQUOTE>

[jonno]

A performance calorimiter if you can find a company that will let you use one! Visteon UK are one of our sponsors so did it for us for free!
The tea urn method would be a poor 2nd place, fiddly to set up, pretty inaccurate too I'd guess at...

[jonno]

btw danzim, I found that the testing didn't really support my calcs (could have made a few errors on the colburn j factor and friction f factor) I designed mine with visteon, so had access to the library data, which we ran through SBD-HX (a lumped heat exchanger program). Since that uses test data as a base it is very accurate, and my core was predicted to within 5% of actual performance: max heat rejection 16.8kW

[Daves]

Through our worthless testing, we saw a max heat dissipation of 72 kW through a single heat exchanger on a CBR600 F4i engine. This is useless because the engine was under no load.

[jonno]

72kW???
Our fan is 455cfm and coolant flow rate is about 9L/min (around 2.3 gal / min). Ambient temperature in the Visteon Calorimiter was about 30 deg C. You sure you got the numbers right? How did you measure the heat rejection?

[danzim]

dave.
how did you run your experiments? i need to get testing once i've found a suitable radiator.

anyone know of a suitable bike radiator for FSAE? like about 200 x 300 mm? GSF bandit 400 is 400 x 200 x 25- too big!

surely the reason why dave's heat rejection is so high is because the stock CBR radiator is much bigger than the radiators used for FSAE (300 x 400 or so?). what are the dims of your's jonno?

[jonno]

200 x 200 x 19 but size alone doesn't really mean anything. what if a 200 x 200 x 19 core had unlouvered fins at 42 fpd (fins per decimeter). would that have the same performance than my core that has 98 fpd and are louvered? no, because the air side surface area is over twice the size on my core. Therefore purely saying that his core is bigger is not a valid argument. My reason for concern is that my heat exchanger is extremely efficient: it is all prototype materials and concepts that visteon plan to bring online in a few years (it took several attempts to make the core since the material is so fine: the core kept melting in the CAB furnaces). It is possible that dave has a heat exchanger that can reject that much energy, however it is highly unlikely, and also unnecessary. We run a CBR at around 70BHP. This is approximately 50kW. Engine heat rejection is about 1/3 max engine power (an approximate rule of thumb). 1/3 of 50 is around 16.6kW. Therefore (unless dave is running a 290BHP engine (quite unlikely) then either his method of working out heat rejection is wrong, or he has a hugely overspecced system and will get massive fluctuations in head temperature. I think his problem is the former....

[BeaverGuy]

With the numbers that he gave there is no way the heat rejection given is correct. The flow rate of air he gave cannot remove that much heat.

400 ft^3/min=0.19 m^3/s

0.19 m^3/s * 1.2 kg/m^3(density)=0.228 kg/s

.228 kg/s * 1.01 kJ/(kg*K (specific heat)=0.2308 kW/K

0.2308 kW/K * 80 K (normal temperature difference)=18.4 kW

So for his numbers to be correct he would have to have nearly a 320 degree airside temperature rise.

[Daves]

Our testing used thermocouples in the water flow right before the exchanger and directly after the exchanger. We tested a set of dual exchangers from White Oak Radiator, and we tested a dual-pass Visteon/C&R radiator from 2 years ago.

Both of our cooling system setups used the engine jacket water to cool the oil as well. We used the equation qdot=mdot*c*deltaT to calculate the heat rejection.

[drivetrainUW-Platt]

this is slightly off the topic, but what teams run solid metal radiator tubing. I know Madison did it on the 04 car, looks like it would weight less than all that rubber, and also added a little more bling since they could chrome it. I would think that it might radiate more heat than rubber, acting as an inline radiator???

[Denny Trimble]

We do aluminum cooling lines, they're the blingin'est.

[Daves]

Aluminum tubing saves about 1 pound of weight per foot compared to rubber hose.

[jonno]

Dave,
I don't like to be blunt, but you're fundamentaly wrong somewhere: like i said, your engine needs to be in the region of 300bhp to necessitate that kind of heat rejection, and as beaver guy said is your airside delta T 320 degrees (i.e. fierce hot)? unlikely. especially with a few year old visteon model. and even more especially if it's U flow (dual pass), since coolant side delta P drop is large: killing your coolant flow rate for no major thermal benefits.

[danzim]

So what are people saying about bling-bling chrome cooling lines?
what about safety?
anyone know what is lightest and cheapest?

Also, has anyone done the tea urn testing to support their theoretical calcs? had any success?

[Daves]

Jonno, you're right. We did our testing with no load on the engine, and our flow meter had too high of a pressure head loss to be accurate.

[Chris Boyden]

Jonno and Dave,

it seems like the big difference is the flow rates of the pumps. Jonno, you said your running about 2.3 gal/min and Dave about 8 gal/min. ~3.5 times different. Dave's 70 kW
would drop to about 20 kW at the lower flow rate.

[BeaverGuy]

Mathematically there is nothing wrong with your calculation. However, the fan flow rate that you gave can not support that amount of heat rejection unless there is an airside temperature rise of 320 degrees. To do that you would need water temperatures at least that much warmer than the inlet air.

I think your error is in one of the values you gave. Either the flow rate of water, water temperature difference, or fan flow rate. The water and fan flow rates are both reasonable numbers. However, the water temperature difference is incredibly large. There should be no need to cool the water 36 degrees.

If you would be willing to give your water and air inlet temperatures and exchanger parameters; Core size, Tube width, height, and thickness, fin thickness, and fin spacing. I will do a theoretical heat rejection calculation with the fan and water flow rates that you gave.

[jonno]

dave, chris

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> your engine needs to be in the region of 300bhp to necessitate that kind of heat rejection <HR></BLOCKQUOTE>

Your calculation appears to be correct in that your numbers add up, but is the equation correct. Unfortunately i don't have my project and my books here so can't give you a definative answer, but basically what i am trying to say is 72.8kW is SERIOUSLY wrong for a FSAE car. That 1/3 max engine power is a good approximation: so as my quote says, are you
a) planning on obliterating the competition with a top secret 300BHP, 600cc restricted engine,
b) calculating something wrong
c) got a heat exchanger that is totally the wrong size for the application

again, my personal gut feeling is b). Sorry to criticise your work, dave, as soon as I get my books and project, I'll let you know the heat rejection formula's or whether i am completely wrong.

Regards,

[Garlic]

Mr. Gray is right. Be vary wary of living in a calculation-based world. Make sure you compare your numbers to known values. If you look around at reputable sources, heat rejection of a heat exchanger in automotive applications is generally less than half of horsepower output.

[Daves]

Our water pump flow rate needs to be tested again. Our team did not test it right 2 years ago, so they did not get accurate results. I believe they tested it with no thermostat in the engine. Can anyone recommend an inexpensive and small venturi flowmeter that we could use to measure our flow rate?

[Grover]

Aluminium tube for coolant pipes is the only way to go. Very light. Use it wherever possible.

[jonno]

dave, just thought where did you get the value of 8 gal/min from? Possibly it is the max rated of your water pump... if the radiator has a large coolant pressure drop (being a U-flow it will) this will reduce the flow rate a lot (at visteon we used to see around 40% going from an identical x-flow to u-flow. U-flows are basically for packaging improvements). our pump will go up to 15 l/min, but once it is matched with the system and all the pressure drops taken into account, it actually runs at 9. This may be one source of error in your calcs...

[danzim]

peops, how do you measure waterside flow rate?
F4i water pump is mechanically driven and pump speed is proportional to engine speed.

-anyone got any data for acutally flow rates?
-what causes the pressure drops? does this result in a much lower flow rate through the radiator tubes?
-am i correct in thinking the waterside isn't too important because the heat exchange is so airside critical (anyone else got
Cp.mdot_water/Cp.mdot_air = about 5?)


air side.
i've got data for fans from merlinmotorsport website etc. are the values for flow rate they give realistic?
does it not depend on geometry of duct?
what materials should you make it out of?
where can you get really cheap (2nd hand?) fans from?
anyone got any tips?

danzim

[jonno]

mechanical pumps are tricky to work with, like you say they are directly proportional to crank speed. Electrical pumps are real simple, they will only work at one speed. Pressure drops are caused by loads of factors. Flow shears in hoses, therefore slowing the flow. Sharp turns in flow also create pressure drops. constrictions also cause pressure drops. Waterside is not too critical if you can get a pump to overcome it, the same statement is also true for airside, you can get a fan big enough to overcome it. obviously if there is more to overcome though, the fan and the water pump need to be bigger: adding to total vehicle weight.
Pump curves are pretty easy to understand, really, well described here (http://www.engineeringtoolbox.com/34_635.html) this page shows you the system pressure drop, the pump head and where the two lines cross is the flow rate you end up with

[Chris Boyden]

Jon,
I agree with most of what you said, but you can design or buy a controller that would lower the water pump speed(flow) at cold temps or increase pump speed at high temps all with the goal of maintaining a set temp. So they're not always constant. But they are much easier to characterize, because all you need is a variable power supply, a gate valve, and a pressure gauge, a timer and a bucket to generate a pump (flow vs.pressure) curve for the entire system.

[Matthew Robinson]

The way I am doing cooling is a fairly simple approach

1. Determine heat rejection needs at full load at varying RPM
2. Determine Average mass flow of fuel during endurance lap
3. Determine radiator size needed to reject this number.
4. Determine fan size needed to reject maximum number

[Daves]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>4. Determine fan size needed to reject maximum number - Matthew Robinson <HR></BLOCKQUOTE>

How do you determine the fan size? I have been assuming a convection coefficient of 100 W/(m^2*K). Do I just fiddle with the convection coefficient until it matches our experimental results? Can the convection coefficient only be determined experimentally? What is the relationship of convection to fan speed?

[Matthew Robinson]

get pressure change due to fan (from fan data or experiment)
relate pressure change to pressure change on radiator data (or by making radiator data)

[jonno]

Determining a fan is all to do with matching again. Once a coolant flow rate is determined, and providing you have an airside pressure drop curve, you match the fan to that.
In writing my last few posts, I'm beginning to realise that without a performance calorimiter or a full complement of data associated with the core, you should either use the core that came with the engine you plan to use or go for overkill (which is a bad idea, cos your engine will work inefficiently, although this is better than the other end of the scale: a warped head due to extreme heat). Radiator system matching is quite delicate and is much more critical than a lot of people realise...

[Chris Boyden]

In my experience, just going with the core the bike engine came with isn't enough. The duty cycle that the engine sees taxes the stock system, and voila....you overheat, even though you're making less power than the bike.

[jonno]

I can't see why it wouldn't chris... stock cbr600 f3 approx 110bhp, brunel racing restricted cbr600 f3 approx 70bhp, less power and therefore definately less heat to reject. On the bike i beleive the rad is mounted in the airstream without a fan. On an fsae car, most teams use a fan, ensuring flow over the core. The only problem i can see with it is that the core is actually too efficient at rejecting heat, so you will run cold. Why do you think the duty cycles are any more on the fsae cars than on the standard bikes these engines were intended for?

[Daves]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Engine heat rejection is about 1/3 max engine power (an approximate rule of thumb). -- Jonno <HR></BLOCKQUOTE>

I agree with this statement, but can anyone tell me where it is written? I need to cite it.

[rjwoods77]

Air cooling rocks. No cooling weight. Forced fan is so nice.

[MikeWaggoner at UW]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Dave @ L.U.:
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Engine heat rejection is about 1/3 max engine power (an approximate rule of thumb). -- Jonno <HR></BLOCKQUOTE>

I agree with this statement, but can anyone tell me where it is written? I need to cite it. <HR></BLOCKQUOTE>

Citation from Harley site:
http://www.peganet.net/personal_pages/ironman1/harley.html

You can also look up efficiencies of IC engines, and do a simple calculation from that.

[Dave M]

so how about some pump flow numbers?....

[Chris Boyden]

We have an 11" Spal pull Fan on a R6 stock radiator instead of the Honda F4i radiator. The system overheats with this setup at hot track conditions. Every thing else is stock except for the hose and the fan w/ a carbon shroud. I do have a bend that is pretty tight, which could be killing the flow, but I don't know how much less flow. Most of the time it's pretty good, but it shouldn't overheat at all. I used to think that it would work as well, due to that same logic, less horsepower = less heat. After boiling coolant, I started questioning that logic in it's entirety. Yea, I think it still applies, but after getting burned, a better radiator is definitely in order.

[Jarrod]

on a bike, the only time you are making big power, and so more heat, is when you are riding hard, ie fast, so high air speed. On an FSAE car, you are working them very hard pulling out of corners and so on, with low air speed. Fairly fundamental difference.

[jonno]

jarrod,

almost all manufacturers design their systems to reject all heat at w.o.t. This figure is the max heat rejection. An engine will not exceed it in standard form. The rad heat rejection figures are then set up, established and the rad designed around that. I will refer you back also to the picture of br4 (the url has been changed so i have reposted it) http://www.brunelracing.co.uk/media/formulastudent2003/4.jpg
As you can see the rad is flat mounted. If the fan wasn't working / not there, the radiator would receive practically zero airflow through it (similar to not moving whatsoever) all airflow through the core is via the fan and shroud. Therefore, speed of vehicle has nothing to do with cooling.
So, Chris as to your concern that you need a better rad: i seriously doubt it unless something is blocked within the rad. It rejects the heat of the bike running at over 110BHP, it will reject enough heat for your application. your quote is correct, less horsepower = less heat. For proof of this check my previous quote on the site, Engine heat rejection is about 1/3 max engine power . Your problem will lie in coolant flow or air flow. I'd try and eliminate the kinks in the pipe, and rethink your water pump (if you are using an electrical). If you are using the bike mechanical water pump, it should do the job... In terms of air flow across the rad, get the seal sorted so all air passes between the fins, not around the gap between header tanks and shrouds etc: that is the key. Hope this is of help...
Regards,

[GTmule]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Fairly fundamental difference. <HR></BLOCKQUOTE>

Fan + presumably better supply air quality, would negate this, IMO.

[Joel Miller]

Jonno,

Looks like very cheeky exploitation of the "no power ground effects" clause - fan forced diffuser tunnel. Did the rulesmen have any reservations about this? Did it provide a measurable benefit?

[Daves]

_

[Daves]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>A project that can be started before the engine decisions are finalized is the cooling system: This is VERY important. The racecar that can't cool can't do anything else. The final specs of the system will depend on the engine decisions, but there is plenty of groundwork to be done. Cooling system calculations are no mystery. There are probably a hundred million vehicles on the road each day that don't overheat. However someone on the team has to learn how it is done. Someone has to build or borrow a test rig to measure all the relevant temperatures, pressures, and flow-rates. If you are using a liquid cooled engine, you have to source the heat exchanger or radiator from somewhere. Someone has to make sure it is mounted in the chassis in a way that it won't be damaged, has sufficient cool air coming into it and has somewhere for hot air to escape. Don't make any assumptions that a stock bike radiator will be sufficient. Countless races have been lost because of this. -- Alan Gruner @ http://students.sae.org/competitions/formulaseries/fsae/reference/orgteam02.htm <HR></BLOCKQUOTE>

A motorcycle radiator may or may not be suitable for your application. There are too many variables to compare an FSAE vehicle to a motorcycle. The only way to really know if it will work is through testing. I recall reading that NASCAR teams use a different radiator for each course. That is how fine tuned their cooling system is for different variables. This makes it sound reasonable that the stock bike radiator may not be ideal for FSAE, but then it still depends on the FSAE conditions.

[Daves]

Shroud or no shroud?

Perma-Cool states,"All Perma-Cool electric fans are designed to be installed directly to the radiator core, close enough to not need a shroud." They also write, "A general rule of thumb is that you cover at least 70% of the radiator core area with the blades of the electric fan."

We will use a 13" x 12" radiator, which is 156 sq. in. A 12" fan will cover about 113 sq. in., which is 72% of the radiator area. Should this be sufficient without a shroud?

Perma-Cool statements from:
http://www.perma-cool.com/faq/efans.html

[jonno]

dave i will quote you directly in answering your question of should it be sufficient without a shroud<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> Someone has to build or borrow a test rig to measure all the relevant temperatures, pressures, and flow-rates <HR></BLOCKQUOTE> you will only know it is sufficient when you have all that data, i listed it all earlier in this thread. The NASCAR quote is a bit extreme, i doubt many / any fsae teams have the resources and funds to make a formula student rad, an fsae rad and an fsae-a rad. when you repeatedly finish top few then you need to think about optimising your setup to that extent....

Those perma cool quotes are ok, but i can guarantee that the further away from the core a fan is, the better the heat rejection will be, purely because the air will flow evenly across the whole core. With a 70% coverage, what's the remaining 30% of the core doing?

Joel, with the power ground effects i've mentioned that previously on this thread, there's pretty minimal benefit in terms of power ground effects, but some is better than none! (the brabham fan car used titanium blades, and span so fast that on early versions the blades sheared off, our little plastic 8" fan ain't quite up to that!!!)

ps dave how do you resize the pictures? or have you hosted them elsewhere?

[syoung]

Dave,

Have you invented a fan with square blade area? Surely your 12" fan only covers (pi)r^2= 113 sq. in. (= 72% of radiator area), not 144. And remember that your fan isn't just made of blades - you have to subtract the area of the hub, which won't be pulling any air through, if you're installing the fan right next to the radiator.

Not that I pretend to be an expert - I reckon I'm learning more from this thread than anyone else - but it often takes a newbie to ask a stupid question.

Cheers,

Simon
Warwick Formula Student

[Daves]

Save the image to your computer, resize it, and upload it to your website. Next, edit your post and replace the URL with the new ones.

[CMURacing - Prometheus]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> Of all the heat produced from combustion, approximately 1/3 goes out the exhaust, 1/3 is dissipated through cooling and the remaining 1/3 is used to overcome friction and produce mechanical power. These percentages make it very clear that there is plenty of room for improved thermal efficiency. <HR></BLOCKQUOTE>

http://www.peganet.net/personal_pages/ironman1/harley.html

this seems to suggest that the portion of work rejected as heat and usable power are equal, i.e. not heat is 1/3 of max horsepower. or am i still reading this wrong? obviously jonno's car didn't overheat, so he's doing something right...

[Brendan]

Question - How does the inclination of the radiator affect its cooling effectiveness? I've seen other teams inclining rad's towards the front of their car and Ferrari use double inclined rad's in F1. So is it worth investigating? Obvious advantages are that sidepod size and therefore weight and drag resistance is reduced, but does an angled arrangement result in better cooling?

I'm new at this, so any help would be appreciated.

Cheers,

Brendan
Cardiff Uni FS - Cooling System Guy. http://fsae.com/groupee_common/emoticons/icon_cool.gif

[BeaverGuy]

Inclining a radiator, well if it is perfectly perpendicular to the flow then the velocity of the air through the radiator should be that of the flow. If the radiator is at an angle to the flow the air velocity through the radiator would be the componoent of the flow perpendicular to the radiator. That is the simple way. If you have scoops ducting the air then it would depend on the duct orientation and bends. So it really depends on the situation.

Example... last year our car had the radiator over the rear differential angled towards the front of the car. This was an extremely bad orientation and when they put it that way I essentially said, "You're a bunch of morons and I take no responsiblity for the cooling system from this point forward." To remedy the sitation they built a scoop that turned the airflow around 100 degrees and directed it into the radiator.

http://oregonstate.edu/~gilletjo/images/fsaegroup_small.jpg

[prelude]

I thought I'd chime in with a few things I've learned:

1. This is probably most important: Air speed through the radiator is not equal to the ground speed of the car. In all likelyhood it's not even close, especially as the speeds go above 30mph or so. A better way to look at it is the pressure difference across the radiator. You're looking to maximize this. Some of you may have gotten the performance curves from visteon, that's the easiest way to determine your air speed through the radiator. Determining the air speed through the radiator experimentally is something of a pain I imagine.

2. Shrouds aren't really about ducting air over the radiator. If the radiator didn't offer any resistance, then all the air would just flow through it by virtue of the fact that you're flinging it through the air. Shrouds are about evenly distributing pressure across the face of the radiator. Looking at the front, if there were no shroud, then you get a roughly parabolic pressure curve at the face of the radiator with nearly zero pressure at the edges (since the air is simply slipping around the edge of the radiator), and maximum pressure at about the center. This is inefficient use of the radiator, since you're essentially only making use of the center of the radiator.

Now with the shroud, pressure down the length of it becomes roughly uniform. So now you have a roughly uniform pressure at the face of the radiator. Due to the exponential nature of the radiator-air-speed vs. pressure differential, the result is to bring up the airspeed at the edges more than you bring down the center (the pressure at the face of the radiator is roughly the average pressure at the inlet of the shroud), making the radiator more efficient. You're doing basically the same thing with the shroud on the rear.

[GTmule]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> Some of you may have gotten the performance curves from visteon, that's the easiest way to determine your air speed through the radiator. <HR></BLOCKQUOTE>

Can someone send me this? I've ben looking for this kind of thing for a while.

boulware<at>gmail.com

[Daves]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>this seems to suggest that the portion of work rejected as heat and usable power are equal, i.e. not heat is 1/3 of max horsepower. or am i still reading this wrong? obviously jonno's car didn't overheat, so he's doing something right... <HR></BLOCKQUOTE>

Both of these are fairly good assumptions. About 1/3 of the engine energy does go to cooling system heat and oil. Also, the radiator needs to be sized to dissipate about 1/3 the heat of the power output of the engine. The difference is that the engine is not always running at maximum power. Even under full acceleration, the rpm varies, so the power output varies. According to Visteon, the 1/3 power ouput rule of thumb is conservative for normal operation of a vehicle.

[Big D]

jonno,

does your car have any sort of lower shroud or scoop under the car? I am just thinking of the under-car air-flow, which is probably more or less parallel with the rad, not perpendicular, like what would be ideal. Do you just use strong enough fans that this doesn't matter?

obviously you know what's up with cooling, and your car works well. I am very interested in this layout, mainly due to C.G. considerations. Feel free to divulge as much or as little as you want, several of us on my team have independantly come to the conclusion that this is where our rad needs to go, so it would be nice to know.

[jonno]

Guys,

sorry i've not posted for a while on this topic, several questions have been fired my way, unfortunately I don't have any books here so I've been hesitant to reply from memory...
anyway Big D: the cooling system on BR-5 was initially designed with our undertray partially acting as a scoop. The plan was to have a channel around 10mm high, but the width of the car, tapering down to the width of the radiator as air progressed under the car. In testing, the system works better without the duct though, but this may be down to the fact that the duct on the undertray had next to no design consideration go into it. I'm sure a well designed duct would help though.
We're trying to keep the layout for the 2005 car for FS (BR-5 will be at FSAE), but at the moment, our new wishbone geometry and new engine position are hindering our packaging space. we will possibly have to get a rectangular shape core.
Our fan isn't anything special, just a kenlowe off the shelf fan (8" dia if memory serves correctly). The parallel flow doesn't create an issue, but as i've said many times before on this thread the ducting of the shroud (and the sealing of that ducting more importantly) is key, to ensure that ALL air that the fan pulls has come through the fins of the radiator, and not round edges where pipes go in / out etc. Once our core was in position, shrouded up, hosed up ready to roll, i got a silicone sealant gun and filled all the small holes to ensure that the flow was through the fins.
Your point of exit is also quite beneficial to pulling air through and helping the fan do it's work... if you get your air to exit the shroud in a low pressure recirculation zone on the car, it will pull more air through.

As regards inclined cooling, this becomes a tricky art, because you get an uneven pressure distribution along the core. Simply holding a core in a region of airflow will not work, higher f.p.d. cores act more and more like a solid wall as speed increases. I did some CFD studies looking at inclination of our groundmount core (the trailing edge of the core closer to the ground than the leading edge), and what ended up happening was the rearward side of the core ended up with an area of huge pressure infront of it. Therefore you end up with a nasty distribution of flow across your core, with next to none on the low pressure leading end, and loads on the high pressure trailing end. This is not to say it won't work, it will although you must remember it is not at the max efficiency. Having said that, with a good fan behind the core, it will pull through a more even distribution, provided it has good shrouding!!! prelude is pretty spot on with his 2nd point here.

Sorry it's a bit longwinded, but that is the airside design to the best of my knowledge, if i've said anything wrong please correct me, and if you'd like me to detail anything further please ask, I am by no means an expert, but I do have a pretty good cooling system knowledge.

Regards,

Jonno

[Big D]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> ....I am by no means an expert, .... <HR></BLOCKQUOTE>

http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

well, if you say so.... Thanks for the informative reply.

[Daves]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>I can't see why it wouldn't chris... stock cbr600 f3 approx 110bhp, brunel racing restricted cbr600 f3 approx 70bhp, less power and therefore definately less heat to reject. On the bike i beleive the rad is mounted in the airstream without a fan. On an fsae car, most teams use a fan, ensuring flow over the core. The only problem i can see with it is that the core is actually too efficient at rejecting heat, so you will run cold. Why do you think the duty cycles are any more on the fsae cars than on the standard bikes these engines were intended for? -- Jonathan Gray <HR></BLOCKQUOTE>

The FSAE cars weigh more than the bikes, so more load may be placed on the engine.

[jonno]

true, but the bike is geared a lot taller, and the bike comes across steep hills, the formula student track (and i assume the fsae and fsae-a tracks) are pretty flat

[HaroonSheikh]

I have calculated the dimensions of the radiator as 14x7x2 for a 600cc engine ....and aluminium as material for radiator is it possible...that i can reduce these figure because i am having problems with packaging ...urgent help required ..plz

[HaroonSheikh]

these figures are in INCHES ...14X7X2 cubic inch

[prelude]

As it just so happens, that was the exact same dimension we had for our radiator last year.

Didn't cool for worth a damn.

But it wasn't really because it wasn't the right size. The real deal is ducting. You've gotta force as much air as you can through your radiator.

I presume you've calculated your heat rejection and the expected air temps and coolant temps? You'll also need your coolant flow rate... not to offend, I'm assuming you already have those.

The problem is, you're pretty much gonna need roughly the same size area on your radiator.

Hmmmm... I'm rambling a bit. Try to clear up your question a bit more than just "I need a smaller radiator", and I'll do what I can to help.

[BeaverGuy]

Depending on what your packaging requirements are and who will be making the core you could go to a thicker core, make it square, relocate it, use two smaller radiators, try a copper/brass radiator, or probably a myriad of other things.

[Daves]

Switching to a copper/brass radiator may be an option, but extensive testing would have to be done to show that it could cool the same amount as the aluminum radiator. For a good article about a go-kart guy using a copper radiator, see http://www.ozsuperkart.net/TechArticles/radiators.htm

[Daves]

Perhaps a copper radiator would not be much better. According to some websites, typically the tubes are made of brass, but the fins are made of copper. To join the two together, a lead solder is usually used. This solder lowers the conductivity between the tubes and the fins, so an aluminum radiator may actually be the same or more effective for the size.

A rather new technology that will most likely replace the lead solder is Cuprobraze. This method brazes the copper and brass together (at a much lower temperature than the soldering process), so better conductivity and lighter weight are possible. However, Cuprobraze is not very common yet, and I am not sure what, if any, radiator shops are currently using it.

So for the time being, aluminum radiators in general have more cooling per unit mass because more time and money has been spent developing them over the past 30 years than has been spent on developing copper/brass radiators.

Therefore, the simplest way to cool your radiator better while not requiring much more space would probably be a fan that flows more air.

http://www.copper.org/environment/trends/auto_radiators.html
http://www.sae.org/automag/metals/13.htm

[Daves]

We got our Perma-Cool 1650 CFM fan in today. The surprising part is that it covers 80% of our radiator with no shroud. Too bad it weighs 4 lbs.

[Nitesh]

Could anyone suggest me the core dimensions for an R6?
Also how effective are twin radiators? Are they equivalent to doubling the core?
What should be the fan specs?

[jonno]

Nitesh,

as regards core size please read the many posts on this thread already: to summarise i could give you a core size but it would mean nothing as it subsequently depends on fin geometry, fin material gauge, fin pack, louvers, tube profile, tube wall gauge, tube dimple pattern aka turbulence promoters (if any), core layout, and many other factors. size alone means next to nothing.
Twin radiators have the benefit of being two smaller rather than one larger unit: you can use two smaller sidepods then. You may think that two half size cores would be equivalent to one full size one, but they usually aren't, as the flow enters the first core, gets cooled and enters the 2nd colder obviously. what this means is that the delta t between coolant and ambient air isn't as great, therefore less heat is taken from core 2. A taller single core would have a slower flow rate across more tubes, rejecting more heat to ambient.
the fan should be specced so that the fan's pressure curve crosses the rad's pressure drop curve. the point where they meet is the rate that the fan wil pull air through. If this is sufficient from your heat rejection 3d plots, then you will effectively cool the coolant when the car is stationary (assuming all your ducting is good and airtight)

[NovaCat2005]

I was just wondering, why not run the radiators in parallel? This would make the inlet temperatures the same at each rad. It would be a little bit more piping, and you would need a way to distribute the flow equally to both rads.

[Marshall Grice]

I can tell you from experience that parallel radiators that aren't exactly flow balanced perform poorly.

on another note, has anyone calculated and/or measured the heat rejection for gsxr 600 motor? We've got some estimates but haven't had a chance to measure it yet. just curious if anyone else has some numbers to corelate to.

[danzim]

is everyone designing their cooling system for heat rejection equal to 20% max power? is that anything like what is produced? surely the low speeds of FS mean likely to get less heat? what about you guys in the US, what speeds do you get up to?

Does anyone know what the biz is regarding oil cooling? i've heard it's well important for the single cylinder engines.


dan

[BeaverGuy]

Our cooling system is designed to reject about 60,000 BTU/Hour which comes out to around 30% of max power.

[Daves]

what about you guys in the US, what speeds do you get up to?

Our rules state that about 66 mph is the top speed seen in the endurance event.

[Jesus C]

Hey guys....i'm trying to figure out what wall thickness to use for our aluminum tubing. I really want to run this instead of hoses. Can anyone recommend a size? Also....does the pressure cap determine the overall pressure through out the radiator? any help would be appreciated thanks

[fade]

Jesus, Smith reccomends .049 wall. And the cap would determine overall pressure unless your hose clamps are loose. http://fsae.com/groupee_common/emoticons/icon_frown.gif

[Daves]

Can anyone confirm the following radiator temperature drop data? We saw a 9.6?F (5.3?C) average drop when coolant was flowing through the radiator. The peak temperature drop while coolant was flowing was about 15?F (8.3?C).

http://www.geocities.com/fif4183/images/09.txt

[jonno]

If you start running pressures much in excess of 2 bar then the welds start to pop open on the tubes of the radiator, something to bear in mind but i doubt you'd get anywhere near that high.
Max temp drop our heat exchanger does is 9.39 deg c, this is with an airside flowrate of 440 litres / second and a coolant massflow of 748 kg / hr. This equates to a 6.8kW heat rejection. Highest heat rejection is 10kW at the same air flow but a higher coolant mass flow rate of 4473 kg/hr. The temperature drop here is only 2.1 deg c.
Your figures of temp rejected seem reasonable, but as I have just explained may mean a low flowrate. What you need to watch over a prolonged period of time is your head out temp (I'm assuming that is the blue plot on your graph) and make sure it doesn't skyrocket whilst in use.

[kreuk]

Hi guys, this is my first year doing engine cooling design and I have a question for you guys...how to design a core radiator? my philosophy is the T(water,in) and T(air,in) are known, then use NTU formula Qmax=Cmin*dTmax...then the question is how do i determine the Cmin since the Cmin is influenced by mass flow rate that is known from either fan and water pump, on the other hand fan and water pump is selected based on the core radiator and T(out)! Designing radiator size then selecting fan and motor pump then other components....is my procedure right?? or designing radiator and selecting fan and water pump are done simultaneously so then we do iteration??

[danzim]

i'm just coming to the end of my first year of cooling. the only experimental knowledge i have (wasn't given any resources to do testing) is that our CBR 600 F4i doesn't overheat with 2 radiators 210 x 210 core size, normal aspiration, poorly ducted.

this year i've tried to come up with a single radiator system based on theory. i've based things on wanting to lose 30% of max engine power.

some people have been talking on this thread about engine power and coolant load both being a third of heat added (exhaust enthalpy being the other third). i've seen figures that agree with this in a book but they were for a family car at steady load (not max power). anyway, i think you're gonna be massively overspecced if you do that for forumla student! remember that max power is only achieved at max engine speed. the rest of the speed range is much much less.

i used the NTU method to find out radiator effectiveness, the big problem (apart from whether theory is anything like reality!) is the mass flow rates, particularly airside, ie. see jonno, prelude, dave etc.'s previous posts about shrouding and flow.

airside flow: you guys all seem to use fans so you know what flow you can get. this is gonna drain a bit (given you need flow rate >0.3 kg/s?) of power and weigh a ton isn't it? well specced 2 rad system should weigh less than 10 kgs. presumably you don't see much benefit from using the motion of the car to force air through your system since fan reduces the flow area?


was gonna pipe in a flow meter to see what flow rates i get from standard mechanically driven water pump (at max engine speed/power) but was told not to (only 3rd year!). anyway, has anyone got any tips for how to measure

what mass flow you can achieve?
rad stuff- temp drops/flows (cheaply!)?
what you need to lose

i think the airside mass flow is the main factor. would you find this out from average car speed to find average heat loss? is 1/3 max engine power then a good average heat you need to lose?

anyway, gonna go testing and see whether my one underspecced rad somehow does the job. if not gonna add one more little one!

anyone got any advice on sizing an oil cooler? what about thermostats and temp gauges?

have i done it wrong?!!
hope that vaguely helps the few of you that are less experienced than me!

dan

[danzim]

forgot to say that it is definitely all about ...

Compact Heat Exchangers by Kays & London.

Bit of headfuck at first but got all the theoretical info you need to build your models.

[jonno]

Danzim,

core size means next to nothing. My core for our '05 car is 190mm x 160mm and there is only 1.
The 1/3 thing is for standard road cars. It's for tests like the ones Aston Martin have been doing recently on the V8 Vantage ie going to the middle east and then drilling it through the desert in 40 deg heat at 150mph for hours on end. It's a worst case scenario but likely to happen since some oil sheikh will undoubtably pummel it through the desert at 150mph, and want his aston to work the next day. therefore you can go below 1/3 (I think i've mentioned either somewhere on this or another - possibly single engine arguement thread that the course only has WOT condition for 15% of a lap). Most load is produced at the WOT condition, hence most heat is created at WOT. Since you are only at WOT for 15% of the time, you can drop below the 1/3 max power bit.

Our fan pulls 4A which is fine with our CBR and a red top 8 across 4 coils. The WR has a poxy alternator (120W) and barely creates enough juice to run the coil, let alone anything else. 4A isn't to severe a draw though in the scheme of things, and i'd say a fan is essential. If nothing else it simplifies your airside calculation - if you know your core pressure drop you know exactly what your fan will pull through it (and exactly how many Amps it will draw in doing so). The fan (http://www.kenlowe.com/) weighs 778g, and the sh http://fsae.com/groupee_common/emoticons/icon_wink.gifroud weighs less (it's just a thin fiberglass thing) so weight is no real issue there. By the way a well specced cooling system weighs under 6kg (including coolant and fan and shroud) http://fsae.com/groupee_common/emoticons/icon_wink.gif
In terms of measuring bits and pieces on your dyno, a rotameter is easy enough to install to get a flow rate. Thermistors and pressure transducers are cheap as chips from RS, whack a few of those at various places (I'd strongly reccomend either side of your rad right at the inlet and outlet , and either side of the engine, again at inlet and outlet) and that is your coolant side pretty much sorted.
Most bike engines reject oil heat to the coolant,the CBR has a water jacket around the oil cooler on the front of the engine I assume most other formula student engines are a similar setup, unless you are running a dry sump I doubt any oil hoses have come into the equation, it will all be internal. Therefore if you are on a wet sump, the easiest way to monitor oil temp is to drill a thermistor into the sump, so it's sitting in the oil.

Hope that is of interest, gotta get back to studying for finals now.

[Rob Davies]

Jonno,

What fan are you running that weighs 778Kg? What is the quoted volume flow rate on that?

We have a 225 mm (largest we could fit in) SPAL suction fan that has 1100 m^3/hour air flow at 0mm H20 so it will be much less when pulling through the radiator. This fan weighs 1Kg and we reckon it was the best we could find, so just wondering if we were wrong - lol.

In my opinion fan flow rate is almost as important as size of radiator. From my rough calculations on our radiator, doubling air face velocity increases heat dissipation by 1.6.

We have the same battery problems as you guys so we went for a 240mm x 240mm x 32mm rad, actually bigger than we could fit as its angled 35 degrees. Angleing (upto a point....) has no noticable effect from our testing. Hopefully the fan wont be used much when we are moving.

See you at Bruntingthorpe....

Rob

[jonno]

If you look carefully you will see fan (http://www.kenlowe.com)was underlined in my previous post, as it is in this (it's linked to the manufacturers website: Kenlowe. We use the 8" model, which will give just over 1000 m^3/hr, although I imagine our pressure drops reduce that. No idea what it actually is off the top of my head though. Your fan sounds pretty similar to ours though, 200g is next to no difference really I wouldn't loose any sleep over it.

Increasing air flow massively increases heat rejection - validated in several physical tests I've done at Visteon over several cores. The change in heat rejection from a 2 l/sec to 10 l/sec airflow with a constant coolant flow rate is of the order of 5 times hence why I have always stressed sealing your fan shroud is critical. It really is one of the easiest ways to make your system more efficient. Switching the fan will save you loads of electricity, 4A is quite a drain on the singles! We're stuck with pretty much permanently running it, cos we get next to no ram air with the ground mount core.

[Daves]

For you metrically challenged (like me), that's about 588 CFM (1000 m^3/h).

[Rob Davies]

200g id say is quite a lot if its on an item of the car that can be reduced in weight without reducing factor of safety (too low) or performance. I would trade the 100m^3/hour for the 200 grammes but I wont lose sleep over it as I know there will be bits of our car much further from optimal than the fan (we also got a good deal on it) and it will still be an improvement over last years which is a success in its own right.

I guess im gonna sound stupid saying that as we will probably weigh in twice as much as delft but we have to start somewhere...

Catch can is such an example. We had a 150g hiking water bottle, I looked all over Iowa, USA when I was on holiday for an aluminium beer can but apparently they cant legally sell beer cans that big so I bought a steel 1 litre beer can at sainsburys weighing in at 100g then a teamate brings in a 50g shampoo bottle that gets soft at 120degrees C but still has some strength so our supervisor says to go with that. 200 grammes saved over the two catch cans there.

One of the main areas we tried to save weight on the cooling system was to reduced the tubing lengths. This decreased the volume of water in the tubes and hence weight. Could well have reduced the tube diameters somewhat but didnt want to mess with the flowrate. The fact that Jonnos rad is right next to the engine will really help this....

Here is a pic of our radiator shorud and fan:

http://pg.photos.yahoo.com/ph/ssdav9/detail?.dir=9d34&.dnm=a999.jpg

Can you tell im bored of exams and just want to get into building the car ( I cant get it out of my mind even though I should be concentrating on exams).

Jonno, are you going to the IMechE driver training on 16th June?? We are taking ten drivers (nearly all our team) but unfortunately we will be taking last years car as we will never be ready in time.

Laters, Rob

[kreuk]

DOes someone has any idea how to find the max. power produced from engine? will it be the same with motorcycle specs?

[Daves]

You can assume it is higher than the actual work you are outputting to the crankshaft because some parts rob power (alternator & water pump). Therefore, you could probably estimate that the stock motorcycle max power is near your maximum engine power.

[kreuk]

Compact Heat Exchangers by Kays and London.
Ntu=A*U/Cmin
Eff=[Cc*(Tc,out-Tc,in)]/[Cmin*(Th,in-Tc,in)]
where R=Cmin/Cmax
Eff=f(R,Ntu), if you can find a relation for unmixed crossflow and solve for Ntu, you should then be able to solve for the heat transfer cooeficients


I just found an SAE paper relating temperature and flowrate changes to Specific Disipation. There may be some helpful equations in there for you.here is the link

Ya..but how do u deal with the area since that what we're designing for?

[BeaverGuy]

Originally posted by kreuk:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Compact Heat Exchangers by Kays and London.
Ntu=A*U/Cmin
Eff=[Cc*(Tc,out-Tc,in)]/[Cmin*(Th,in-Tc,in)]
where R=Cmin/Cmax
Eff=f(R,Ntu), if you can find a relation for unmixed crossflow and solve for Ntu, you should then be able to solve for the heat transfer cooeficients


I just found an SAE paper relating temperature and flowrate changes to Specific Disipation. There may be some helpful equations in there for you.here is the link

Ya..but how do u deal with the area since that what we're designing for? </div></BLOCKQUOTE>

That looks like a post I originally made so I'm going to respond. The reference I was making was for a situation in which the the area and heat transfer data was allready given. They simply wanted the heat transfer coefficients.

However, if you want to find the transfer area, you start with the same equations and work backwards. A good reference for this would again be Kays and London's Compact Heat Exchangers. In the examples they make some assumptions about the overall heat transfer coefficient and log mean temperature difference. This means that to achieve good final value you may need to iterate. This makes a computer a very useful tool for this method. I did this innitially two years ago. However, I haven't been able to verify the programs I wrote with physical testing. If you would like more information feel free to PM me.

[Engine Dude]

Hey guys,

I am seeking some help on the cooling system design.

Firstly, my objective is to have a radiator which has a suitable geometry and physical configuration to maintain a suitable engine temperature. The temp must be sustained for the events and for the time between the events.

Now, if you know the temperatures,
1) temp coolant in to radiator
2) temp coolant out of radiator
3) air temp onto radiator
4) air temp from radiator,

I assume you can calculate R and P to get F of the correction factor chart.

As for these temperature valves, is everone just running a stock setup and measuring the temperatures at various locations? Could anyone share some rough values to compare with?

Also, when calculating Cr, what is Cp,c and Cp,h and where do you get them from? Unfortunatly our uni is quite poor and we have a limited selection of books. Wish I could get a copy of that Compact heat exchangers book

Cheers

Sam.

[BeaverGuy]

If you have the temperatures and core configuration you can calculate R,P, and F. Cp,c is the specific heat of the cold fluid and Cp,h is the specific heat of the hot fluid.

However, I don't see the LMTD method as very practical. Usually you only have one hot side and one cold side temperature when you are designing the exchanger. You can set an ideal outlet temperature for the coolant but that outlet temperature should be based on the flow rate of coolant and the amount of heat that the radiator needs to reject. Both of those are allready involved in the effectiveness-NTU method though. If you can't get a hold of Compact Heat Exchangers, your heat transfer text should be sufficient. Though, you will need to make some of the connections yourself because they don't tend to connect the exchanger analysis with the basic diffusion and convection heat transfer correlations.

[Charlie]

Recently I was looking at the Ansible website (makers of Aerolap) and they have a Cooling system simulation program that is free for download. No experience with it but it might be worth playing with.

http://ansibledesign.com/ad/products.htm

It uses wind tunnel data, which is not applicable in FSAE because most cars use a fan. But I'm sure there would be ways to approximate and it might be a good learning tool..

[Travis Garrison]

Engine Dude,

you should definetly setup your own tests as soon as you can...but in the mean time I'll post some rough values, and maybe they will be far enough of as to anger those who have done more thourough testing and then they will in turn post some better data http://fsae.com/groupee_common/emoticons/icon_smile.gif

coolant temp in: ~210
coolant temp out ~150-160

air temp in ~70-75
air temp out ~190

This was taken with thermocouples on our current setup.

The coolant temp in comes from the engine sensor, the out was taken from the OUTSIDE of an aluminum coolant line so grab a big 'ol grain 'o salt for that one (measured temp was 150, when the inlet pipe said 197 and our temp gauge was floating around 200-210)...

The air temps were taken with thermocouples as well, placed in a couple locations about an inch back from the core.

The Fan was a 1000cfm by manufactures specs but I don't have actual flow numbers.

In any case, that is all VERY rough data but maybe it will work as a starting point or at least a reality check...

-Travis Garrison

[Engine Dude]

Too all,

Thanks for your suggestions dudes. I come from A team in Australia consisting of 7 and I am in charge of engine so I have a few tasks to complete. So im just trying to maximise the performance without doing a full study on the system, but I am doing all I can.

Beverguy,

Assuming I have the temperatures (which I dont yet), do I then calculate R and P from the Correction Factor graph? If my mc.Cp,c < mhCp,h do I multiply effectivness by P, and if
mc.Cp,c >= mhCp,h multiply by RP?

Also, before you get effectivness, can you......
-get Uo from resistance model from specific setup
-plug in values of Ao
-get mc (massflow coolant) from pump data (electric pump flow rate)
-get cp,c from somwhere,

to get the NTU?

I have this lecture from some uni on the web. but I cant remember where i got it from. I will post it tomorrow. Anyway it has a table outlining some relationships, which is basically where I am getting all my ideas from.

I realise I may be a little far off here but Ive only been doin it for a few weeks. Thanks heaps for your posts dudes. Thanks for you values Travis.

Regards,

Sam

[kreuk]

Hi guys, does anyone know how much normally pressure drops and inlet pressure on both water and air side for fsae car? our team is using Yamaha R6....thanks

[BeaverGuy]

Engine Dude,

Sorry for the late reply, if you haven't found the answer I hope this will give you some more direction.

The heat capacity ratio, m*Cp, of the warm fluid, water, will always be higher than the heat capacity ratio of the cold fluid, air, for a radiator.

For correction factors, Incropera and DeWitt don't have any correction factors involved with the effectiveness-NTU method only with the LMTD method. Also, you use R and P to find F which is multiplied times the LMTD of a counter flow exchanger to get the LMTD of the cross flow exchanger.

Cp,c is simply the specific heat of the cold fluid, i.e. air, you should be able to look it up or better would be to find a table or equation for its temperature dependence then pick the value at the LMT and use that for your calculations. This is obviously based on assumptions because you won't have the outlet temperature.

The mass flows will come from the water pump flow rate and either the fan flow rate or a flow determined by the car velocity and duct geometry, I always use the fan flow rate as it is likely to be a worse case scenario.

Uo can be preliminarily determined from the cooling system geometry and fluid properties. However some of the fluid properties that are relevant are highly temperature dependent and will affect Uo. That combined with Ao and Cmin, smaller heat capacity rate, should be enough to calculate NTU. Then effectiveness can be calculated and you can see if the radiator rejects enough heat and see what the output temps would be then iterate if your assumptions weren't close. That is pretty much the method outlined in Compact Heat Exchangers and the method I use to evaluate radiator geometries to see if they are sufficient to reject the necessary amount of heat.

Alternatively, if you know how much heat you need to reject and what the maximum possible heat rejection is, q,max=Cmin(Th,i-Tc,i), Then you can calculate the effectiveness and use this to get the NTU which combined with Cmin and an assumed value for U will give a value for A which if you take the tube and fin geometry available can give you a radiator core size. Then you go through the first method to make sure that it will reject enough heat. It probably will have a different U than was assumed and iterate with the new value of U until it converges. This is the method I use to design a radiator core geometry.

[Scott Sinclair]

Uo can be preliminarily determined from the cooling system geometry and fluid properties. However some of the fluid properties that are relevant are highly temperature dependent and will affect Uo. That combined with Ao and Cmin, smaller heat capacity rate, should be enough to calculate NTU.

BeaverGuy,

Firstly, is Uo the airside conductance or is it the total conductance (or neither)? and even after reading Kays and London (and obviously not understanding it very well!) I don't get how you determine Uo? can you please shed some light on this.

I've got all the inlet and outlet data and am having difficulty getting my head around the model to evaluate it.

Thanks

Scott Sinclair

[BeaverGuy]

I believe that Kays and London use Uo as the airside heat transfer coeefficient, however it doesn't matter as long as you use the appropriate heat transfer area. If you use the air side coeeficient use the air side transfer area.

The heat transfer coeeficient for an object without extended surfaces and of the same transfer area on both sides will be the inverse of the sum of the transfer resitances.
i.e 1/Uo=1/h1+1/h2+...+1/hn. However, because of the extended surfaces and difference in area it becomes 1/Ua=1/(hw*Aw/Aa)+1/ha*na. This is ignoring the resitance of the wall as is done in Compact Heat Exchangers but not in Incorpera and DeWitt's book.

Where Ua is the airside heat transfer coeeficient, hw is the water side convection coeeficient, Aw is the water side transfer area, Aa is the airside transfer area, ha is the airside convection coeeficient, and na is the airside extended surface temperature effectiveness.

If you have Incropera and DeWitt's book these concepts are also explained in there and probably in most other Heat Transfer texts as well.

[kreuk]

Scott,

Once you find the way how to do it, Kays and London is a book where you can find the required graph(s) and table(s).

Btw, the method of calculating water side is a slightly different on my way compared with the air side,that I havent justified yet. Does any of you guys who has already justified the method for water side since it has flat tubes?

Thanks for your help guys,
nb: thanks for your message to Beaver Guys

[Scott Sinclair]

Thanks guys, I think i'm starting to get my head around it now.

On another note, I remember earlier in this thread some people asking for the flow rates of mechanical pumps. I found a paper which gives the flow rates of a mechanical pump on a "common 600cc motorbike engine". It doesn't specifically state the type, but a good indication can be gained.

I'd post the paper but I've only got it in hard copy.

Basically:
At 2500rpm, flowrate = 9.5 L/min
6000rpm, flowrate = 30 L/min
10,000rpm, flowrate = 59 L/min

So if you plot those points and put a straight line through it, you get an indication of the flowrates at different engine revs for a "common 600cc engine."�


Scott Sinclair

[Neil]

I wanted to get your opinions on double inclined radiators and it's possible applications to FSAE cars. The interesting thing I noticed is that the airflow does not seem to fully go into the radiators. This is as close as I got to be to the F2005 so I couldn't take a better look and was so giddy I couldn't think to zoom in.

Here is the super high res link: http://img137.imageshack.us/my.php?image=f2005_big.jpg

http://img137.imageshack.us/img137/2699/f2005_small.jpg

[Scott Sinclair]

Anyone have any opinions or experiences with double core radiators? any known pros and cons?

Any idea on the performance difference between a 40mm thick single core or 2 rows of 20mm thick tubes?

Cheers,

[scooter2131]

An obvious con is that your setup will be more complex and weigh more due to extra material required for the radiators and more water in your system. Unless you're having major cooling issues, you'd be better off optimizing one radiator than trying to mess with two. That's just my opinion.

[Mike Claffey]

What I think he is talking about is the difference between a single core 40mm thick radiator and 40mm thick dual core radiator (2 cores each of ~20mm but sharing some endtanks). I get the feeling the dual core would be harder to pump through, any thoughts?

[Scott Sinclair]

Yeah sorry Scott, I meant it the way Mike explained it...so its still one unit but with effectively 2 skinny cores sitting flush against eachother.

[BeaverGuy]

In general, a single core of the same thickness and basic design as a dual core will perform better, though marginally. However, if the tubes in the multi-core design are offset things could change because the airflow has been dramatically altered.

It doesn't really matter to me whether dual or single core. As it all comes down to what is available from the manufacturer. If one manufacturer has cores with 2 1 inch tubes and another has a single 2 inch tube, but the dual core manufacturer can get it to me cheaper I will take the dual core because the performance difference isn't that large.

[jonno]

Ford UK used to make a Transit van with a two row core (common tanks but effectively two cores) and for a given core frontal area, the transit 2x19mm core performed better than the aston martin 42mm thick single core (produced in the same plant in the same manner from the same materials) ie an effective 38mm core vs a 42mm core. I think the improvement was down to the fact that more coolant was nearer the outside of the tubes therefore improved heat transfer took place. Not too sure about pressure drops, the approximations i was told were based on common flow rates from performance calorimiters. The reason that few 2-rows are made are the manufacturing complexity of trying to balance two sets of tubes and fins in place prior to braze. If you can get hold of a correctly sized 2 row core for formula cars the fine, but I doubt many 2 rows will go down to the size required in formula student: the thicker core and higher performance will result in a core with miniscule frontal area - and therefore low airside flowrates.

[kreuk]

Originally posted by jonno:
the thicker core and higher performance will result in a core with miniscule frontal area - and therefore low airside flowrates.

would that be high airside flowrates with low air core speed, since turbulent flow is what we want.

what if the core reduced to become thinner with larger frontal area (significant difference in water side length); eg. 350*330*40 becomes 270*320*20 (350 to 270 - tubes length)??

Is there any rule of thumb for airside flow (core width) must be less than waterside flow(core height) especially for those using single fan though it is dual phase??

[kreuk]

Do you guys have any idea which one's better? either let the cooling fan run continously or put some sort of controller and let the cooling fan run at the particular temp.??

[The X Factor UCV]

I personally beleive that leting the fan runing at a particular temp, triggered by a sensor is better than always running.

We made a little test, and having it running all the way consumed the battery pretty fast.
Our fan has a great cooling capability, it only remains on for 10 to 15 sec, and the temperature never goes over 87ºC (about 170 ºF).

You can have an alternate swicth in the control panel, so the pilot can turn it on and leave it so for as long as he wants, this of course needs more concentration from the pilot in order to monitor the temp. I suggest this because if you make the circuit by switching on the fan when you start the engine, you'll hace the fan running unnecesarily for a long time mean while the engine reaches its operational temp, thus yo are wasting your battery with no reason.

I think thats pretty much it.

I Salute You

[Andycostin]

I think that you'll find the major issue to deal with is current available from the alternator, as X Factor mentioned. Our car is running with barely enough current with a thermo operated fan, and this is even worse if you have a starting prob during the endurance stop - you may not have enough power in the battery to keep trying to start the engine.

Just my 2cents

[kreuk]

I've seen the posted message pretty much says : cooling fan area must be 75% of the radiator ...how does it deal with the airflow produced by the fan? not too offend but it must reconsider other factors, doesn't it? I'll try to change by 1" below the supposedly '75%area' fans with higher flowrate for sure, so what do you think guys?

I have no practical experience background before in engine cooling, just refer to everybody's favourite book Kays and London,the way it should be done is:
know the surface selection charcateristics+problem specs.+op.conditions,etc
1.then determine fluids output temp.
2.get C, effectiveness then Ntu
3. get thermophysical properties
such as stanton number, core mass velocity!,Re. numbers for both fluids
then obtain h values, U value, and further will obtain specific length of radiator height, wide and thickness.
Any comments guys? Oya on the other hand I tried to combine gas-air to intercooler design concepts to generate result, however still nil.

therefore, could someone help me with your way to calculate the area? My input basic input are:
thi=110 C
tai=40 C
mw=0.15 kg/s
ma=0.45 kg/s
inlet pressure 110 kPa,
pressure drop: water: 20kPa
air: 0.32kPa (both those values are gross estimate?)
the rest input is on your own in doing that and what source available with you.

I assumed you all guys knows how important pressure drop in prior calculations since I've never read the many postings about it?so then if you guys know the most approximate estimate values for both pressure drop, I would appreciate it very much. the output shows giant dimensions 12"x14"x1.5"

Any new fresh ideas are needed to continue the last breath!!! Pls someone can calculate the input of mine and tell me what's wrong??

Sorry for the long story, hope to hear any comments from you all guys!! Thanks in advace

[Alfonso Tan]

Diggin' up an old thread...

I'm totally new to this cooling system thing, but I've been designated as U.Washington's cooling guy for this year. I'm conveniently taking my first heat transfer class this quarter, so I'm getting a slight idea on how to determine radiator and fan dimensions/requirements, but am still a little foggy on a few things:

-Visteon is hooking us up with a radiator this year, so the radiator we use will be one of the units they offer. With each radiator, I was given a few graphs: heat rejection vs. avg. core face air velocity, air side pressure drop vs. " ", and coolant side pressure drop vs. coolant flowrate. Being that I've got this stuff, determining radiator size seems pretty simple, which kinda bugs me. What I'm thinking is I can get an average core face air velocity based on a fan (the worst case scenario is that the fan is the only thing providing moving air), and with that information, along with the amount of heat rejected from the engine, decide on a radiator using the Visteon graph that shows the proper heat rejection rate for a given average air velocity. What's missing from this?
-To find radiator dimensions using the NTU method (as recommended a few pages back on this thread), after determining the NTU, does the area of the front face of the radiator follow? I've got this textbook that states that NTU = UA / Cmin where U is the overall heat transfer coefficient, A is the area, and Cmin = Cp*(mass flow rate). This makes it seem like area can be found through this calculation. How is total volume takin into consideration?
-How are the fins accounted for? Fin area doesn't seem to be part of the equation anywhere.
-A teammate was able to locate the 'Compact Heat Exchangers' book by Kays and London. Like I said though, I'm new to this stuff (so the info doesn't register with me easily) and unfortunately don't have time to sit down and read 300+ pages. Anybody able to point me to the most useful sections of the book?

Any and all input is appreciated http://fsae.com/groupee_common/emoticons/icon_smile.gif

[BeaverGuy]

The area in the Ntu=U*A/Cmin refers to the total transfer area, including fins, if U is calculated on the Airside or the interior tube area if U is calculated on the waterside.

Your heat transfer text will have the most important parts of the design methodology. However, "Compact Heat Exchangers" goes through acquiring U from the geometry of the exchanger and has tables of fin geometries that are helpful. When I first looked at the book, before I had taken Thermo or Heat Transfer, I found the examples in the back to be helpful and that may be a place to start. You'll have to refer back to other areas in the book

As far as the tables that Visteon is providing you, that should be all that is neccesary if you have the coolant, air, and temperature data.

[Alfonso Tan]

^^ ...Thanks for the advice. I'm running into a new problem unfortunately. I'm able to make fair assumptions about temperatures and flow rates, but to use any of these calculations, I think I need total area (fins included). Visteon doesn't specify a total area, or very many properties about their fins other than fin height. This is where I'm getting caught up.

Also, I've got a text book that states that the overall heat transfer coefficient, "U" is in the 25-50 W/m^2*K range. Does that sound reasonable? With the Cmin and NTU values I have, this results in a huge area (which might be reasonable depending on total area including fins). I really need to figure this stuff out... small report due monday http://fsae.com/groupee_common/emoticons/icon_eek.gif

and lastly, for anyone that's gotten a sponsorship from Visteon or has their pdf files with info regarding their radiators, they've got fin density in a unit "fpd," what's that mean?

[BeaverGuy]

fpd=fins per decimeter.

And yes I would say that 25-50 W/m^2*K is reasonable though you might want to stick to the low half of the range.

Calculating the total transfer area can be messy and you will have to calculate the fin effectivness also. And these in turn have an impact on U. That is why I think the task is best for a computer because of the iterative nature.

[Alfonso Tan]

The total transfer thing turned out to be pretty far off from reasonable. I'm not exactly sure why, but using some reasonable estimations on temperature and the NTU method, an expected area of ~22 m^2 was found, when it really should be around 3.5 m^2.

A team mate that was in charge of the cooling system last year came up with an idea of reading pressures off our current system with MAP sensors and using Visteon's performance curves to determine radiator sizing. This sounds like a good idea, but might not be a good representation of this year's radiator setup. Is there another way to use to pressure curves experimental data? How are they used in calculations?

[harini]

hi,
i dont know much about designing a shroud so i was hoping someone could give me some info on this.

[Scott Sinclair]

This has been discussed previously on this thread, but the most important thing is to make sure that all of the air that passes through the shroud is forced to travel through the radiator core. This means sealing around the edges of the radiator so no leakage occurs.

Another thing is to make sure that the fan isn't mounted directly onto the back of the radiator. A gap of at least 25mm (between fan and radiator) allows the fan to pull air from a greater area of the core, hence increasing its efficiency.

[harini]

thanks for the info....one other thing..
i was wonderin if not mounting the radiator perpendicular to the air flow affected the efficiency a lot.i have read that the pressure distribution behind tha radiator will not be uniform.how much does this affect the performamce.

[andrewd]

Originally posted by Scott Sinclair:
This has been discussed previously on this thread, but the most important thing is to make sure that all of the air that passes through the shroud is forced to travel through the radiator core. This means sealing around the edges of the radiator so no leakage occurs.

Another thing is to make sure that the fan isn't mounted directly onto the back of the radiator. A gap of at least 25mm (between fan and radiator) allows the fan to pull air from a greater area of the core, hence increasing its efficiency.

so you believe that increasing the gap between the fan and the core will increase the efficiency of the system?

the fan should not be the sole provider of air in fact it should be as small as possible to only provide such cooling during zero/low speed low load operations. if you are using you fan at all speeds you have incorrectly sized your cooling system. PERIOD

[Scott Sinclair]

so you believe that increasing the gap between the fan and the core will increase the efficiency of the system?

If you mount the fan directly onto the back of the radiator, it is only pulling air through the core area which is exposed to the fan blades, which when you subract the area of the motor, isn't much. Having a gap between the radiator and fan allows the fan to pull air through a greater portion of core, increasing its efficency.


the fan should not be the sole provider of air in fact it should be as small as possible to only provide such cooling during zero/low speed low load operations. if you are using you fan at all speeds you have incorrectly sized your cooling system. PERIOD

I'm not fully understanding the logic here. Are you suggesting that if you are using your fan lots, your radiator is too small? If so, to fix this, a larger radiator is required, which weighs more, and then more water is required to fill it, which weighs more again.

[andrewd]

the core volume of water is minimal compared to the volume of the hearder tank and trasnfer piping. the weight of a 10" fan is slightly more than a similar sized radiator (with water)

so i believe you are wrong there. things like fin density and pressure drop play an important role in deciding the overall efficiency of the system. improving on these can also lower your system weight

[Scott Sinclair]

the core volume of water is minimal compared to the volume of the hearder tank and trasnfer piping

Fair point


things like fin density and pressure drop play an important role in deciding the overall efficiency of the system.

Fin density and pressure drop are two of the many variables which play an important role in determining system efficiency.

With regards to the original question about designing shrouds...the amount of air leaking around the core and the distance between fan and core are two variables which should be considered when designing the shroud.

[Maverik]

I agree that designing to have your fan run all of the time is a badly designed cooling system. You need a fan for times when the car is :
1) not moving but the engine is running
2) moving, but at very low speeds (hence no airflow through the radiator)
3) under high engine loads where temperatures soar in the cooling system/engine
The idea is to utilize an available resource, the moving air around the car. Make a duct to your radiator that will cram "clean" air to your radiator when driving. If you do it right, the fan will hardly need to come on while driving. If you can't make a tricky duct, then position that radiator perpendicular to the air flow. Ideally, you should think of your fan as a backup system. That way you save current draw, and have one less thing to worry about if your fan happens to fail during endurance.

[andrewd]

Originally posted by Maverik:
I agree that designing to have your fan run all of the time is a badly designed cooling system. You need a fan for times when the car is :
1) not moving but the engine is running
2) moving, but at very low speeds (hence no airflow through the radiator)
3) under high engine loads where temperatures soar in the cooling system/engine
The idea is to utilize an available resource, the moving air around the car. Make a duct to your radiator that will cram "clean" air to your radiator when driving. If you do it right, the fan will hardly need to come on while driving. If you can't make a tricky duct, then position that radiator perpendicular to the air flow. Ideally, you should think of your fan as a backup system. That way you save current draw, and have one less thing to worry about if your fan happens to fail during endurance.

well put.

so who here believes the inlet duct is of more importance than the rear one?

would make more sense wouldnt it now

[Maverik]

Im gonna have to say both are equally important depending on what you are going to do with your design. The most efficient design will decrease the velocity through the inlet duct, through the radiator, then converge in the outlet shroud to increase the velocity back to free air stream velocity. Ahhh, I love fluids and aerodynamics!

[SnailRacer]

Back to the fan question... I don't think its unreasonable to design a cooling system with a fan running all the time. If you're going to have a fan on there anyway, may as well use it. Just make sure you weigh the cost (current draw, reliability) vs the gain (mass, packaging).

[Maverik]

But you are missing the whole design challenge. The fan is a backup system, period. Trust me, you do not want your engine to fail because your cooling system crappped out. If it's electronic, it can fail any time. By all means, definitely design with the fan how you want, but a few examples of the benefits to my way are:
- Less current draw, so if your driver spins out, that much more juice is available to restart the car.
- The incoming air through the duct can be redirected to the radiator and parts of the engine or electronics.
- As long as the duct is somewhat air tight, you can do much better calculations on efficiency because you know the air stream flowing through the radiator is "laminar like" in that it is a uniform flow.
- When the fan is on, it doesn't have to work as hard to pull the same amount of air.
- And as stated before, your designing for a little redundancy so that if your fan fails during enduro it isn't going to be what brings your team to it's knees!
The cooling design is all about sizing and testing testing testing. If you notice, I haven't really touched on core geometry. Even though it makes a difference what size your core is, you can make any core dissipate the heat you want given high enough water and air flow rates. When you want efficiency though (less fan and water pump work, hence less power and current draw) every aspect of the system should utilize what is already available. Do some flow rate calculations with air at a velocity of 35 mph and compare it to what your fan pulls through the same area. You might be surprised. Then, depending on your radiator orientation in the frame, compare the frontal projected area parallel to the centerline of the frame. These are a few of the things you should consider in the cooling design. Make everything as efficient as you can!

[Scott Sinclair]

Maverik, or anyone else,

Have you done any modelling to determine the effects of the chassis, wheels spining/turning and the wishbones on the airflow reaching the radiator?

[Maverik]

Alas, my university has the licenses to ansys icem cfd program locked away for "research oriented programs" only and they wont spend to get another... But you can get a fairly good idea by thinking about where the duct/radiator is in relation to the control arms and their geometry. Our body is fairly aerodynamic and my duct placement is tight to the body directly between upper and lower control arms and away from the tire. So the opening is as far away from the vorticies created from the tire spin and control arms. Keeping the location there I can assume without too much error that the airflow is "undisturbed." I can't wait to get my hands on some cfd to really know... or a real wind tunnel!!

[Igor]

I find the idea of designing for a non functional radiator a bit odd. None of the other parts on the vehicle are designed to not work, why is this part different? Wouldn't energy be spent more wisely in making sure the fan doesn't fail? Nobody designs their cars with six wheels so they can keep driving when a wheel breaks off.

On another note. I have never done any calculations on it, but would it be possible to use a small and thick core coupled to a centrifugal fan without putting the core in the airstream? Centrifugal fans are much more efficient than the propellor style and you'd be able to mount it right next to the engine. It'd save all the hoses, intake ducting and some packaging issues.

Igor

[Scott Sinclair]

Check out page 5 of this thread for a picture of the Brunel car with the radiator mounted out of the airstream.

[Igor]

Hmm, ok, should have read the entire thread first... http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

The part about the centrifugal fan remains though.

Igor

[Marshall Grice]

i'd say that running the radiator with no fan/very small fan and relying on free stream air and ducting is exactly like running wings at fsae. the air speeds are typically slow enough that to make it effective the component weighs way more then it could. It's obvious that it works, it's just a compromise. I'd also imagine that putting a well ducted fan on the back of a good free stream ducted setup would pretty much kill any air flow through the radiator with the fan off. so your options are to design a well ducted radiator with no fan (or put a fan in there and mess things up), or just design the fan to do most of the work and free up all the other constraints associated with free stream ducting.

[jonno]

Not really looked at this thread since graduating but since i've been away there are some interesting points being raised again that i think have been discussed earlier but anyway just to reply with my opinions on cooling systems...
firstly, maverik, I am against any backup systems, they are wasted weight - the original system should be designed to do it's job.
secondly, on fans themselves, a lot of people use the spal fan which while for the same size (say 8") pulls slightly more air through the core uses over twice the power when compared to the kenlowe 8"fan (from memory they are 4A vs 10A). so by that you can run the kenlowe 2 1/2 times longer using the same ammount of current.

Radiators by their very nature inhibit airflow through the core, this is a byproduct of extended surface areas with louvres and tight fin packs of radiator cores. Whilst nice long ducting ahea of the core helps, it does not solve the problem. Formula one and other formula style cars get away with no fans because their average speed is high (i'd hazzard a guess at 100mph but please feel free to correct this). Talking to the organisers of fsae, the average speed of the enduro track is 30mph. I personally don't feel it's enough to run with no fan, the speed is just too low coupled with the resistance to flow of a radiator. Therefore with accurate matching of fans, a well built fan shroud downstream of the core, a well sealed fan shroud using tape or silicone sealant to fill in the gaps, and also carefully placed blobs of silicone sealant on the gaps between the ends of the fins and the headers, a fan can be made to work very efficiently, giving you a guaranteed rate of cooling. The picture of the brunel car on page 5 relied solely on it's fan to generate all it's cooling airflow, yet it runs with many restarts on a red top 8 (8 amp hour battery) easily lasting longer than enduro. the fan was switched by an ecu controlled relay running off the engine (coolant) temp sensor, switched on at 100 deg c, then off again at 95 deg c (from memory). The coolant won't boil cos it's a pressurised system.

[Maverik]

I never meant to champion designing for failure of your primary system, so let me correct that if I gave that impression. In my defense though, free airstream works on fsae cars. It is also hard to really compare formula 1 to our cars in terms of cooling since they make 10-15 times the amount of heat! As far as the fan being an obstruction, when you are talking about formula 1 speeds then the drop off in efficiency is quite high by putting a fan in there. In our lower speeds, it really comes down to a judgement call. Yes it will be less efficient than if no fan was there, but then you lose cooling when not moving faster than 20 mph (roughly.) Airstream velocity at 30 mph (13.4 m/s) equals about 1000cfm through a radiator with .038m^2 free flow area. Now, that's ideal, so lets say the ducting is 50% efficient through the fan blockage and pressure drop across the core and effectively only 500 cfm gets "used" for lack of a better expression. That's almost as much as what the fans we use can do. Again the downside is no cooling when the car is not moving fast or not moving. We do very strenuous testing on our vehicle with the car being on for long periods of time so we need that fan. Now when it comes time to endurance will the free airstream work, I believe it will. Hopefully the increased overall efficiency, disregarding the extra pound, of the system will turn out such that the fan won't need to be on all the time and during driver change the car will fire right back up. Trust me, our system needs to be reliable. Plus I have about six months to experiment. I would always rather take off if I dont need something then to put more on. That starts to delve into the electrical side so I wont get to much farther on that. I had actually not heard of Kenlowe (lol welcome to florida), can you give me a little more info about the fans? I looked at the website but it was very limited. It's a bit refreshing to see new conversation in this thread... the ducting was briefly talked about, but most of the thread has been devoted to NTU and cooling calculations. Merry Christmas!

[Marshall Grice]

I don't really understand why you're considering the car not moving as a critical part of the radiator design. if it gets hot and you're not moving, turn it off. Not like we're going to be stuck in traffic in these things.

[Maverik]

To each his own, I have good reasons for all of my decisions which may or may not apply to other teams. The bottom line for my team is we need a fan that wont be constantly on, yet still be able to cool. The main difference we are talking about here is whether to put a forward duct on the radiator or use a fan right? Aside from an extra pound from the duct, what is the harm in adding it? That is the main thing I was trying to get at. The duct can't fail, it can't suddenly stop channeling airflow to the radiator, it helps diffuse the air before the radiator. I don't consider the duct a critical part of the design, but an important one.

[CowtownBallin]

Originally posted by Marshall Grice:
I don't really understand why you're considering the car not moving as a critical part of the radiator design. if it gets hot and you're not moving, turn it off. Not like we're going to be stuck in traffic in these things.

So, what happens when you have a hot engine, with hot coolant flowing through it, and you shut the engine down? Coolant stops flowing, nothing there to take the heat away quickly.

[Chas]

We get very hot days in Colorado and the air is fairly thin here. Last year we used a 12"x12"*x2" single pass radiator with a 13" fan.

This year we're trying a double pass radiator to limit fluid bypass in our radiator. Has anybody here used double pass?

[JuicedH22]

We used a double pass radiator for the 04 and 05 seasons, both with terrible results (we overheated last year in endurance, after finishing 10th in autocross...)
Over the summer, we did some testing, in one test, we took the dual pass radiator, and modified it to a single pass. Comparative testing showed MUCH better results for the radiator after the change to single pass.

[ad]

Originally posted by JuicedH22:
We used a double pass radiator for the 04 and 05 seasons, both with terrible results (we overheated last year in endurance, after finishing 10th in autocross...)
Over the summer, we did some testing, in one test, we took the dual pass radiator, and modified it to a single pass. Comparative testing showed MUCH better results for the radiator after the change to single pass.

what sort of water pump arrangement were you using for this double pass radiator? was it unchanged for the single pass test?

[JuicedH22]

The water pump was the stock unit, we used the same pump for both apps. the same pipe diameters as well, the only thing changed was the configuration from dual pass to single pass

[Maverik]

Likewise... I think it's a misconception that a two pass performas just as well as a single pass. Just because the water "stay's exposed longer" doesn't mean you are going to transfer any more heat. An ideal radiator is with the core infinitely short and an infinite number of rows. The reason has everything to do with the gradient of the water temperature through the radiator... that is why we are NOT running a dual pass anymore. (We overheated the past two years as well with a dual pass.) Great for packaging but only if you are willing to give up on the efficiency!

[ad]

Originally posted by Maverik:
Likewise... I think it's a misconception that a two pass performas just as well as a single pass. Just because the water "stay's exposed longer" doesn't mean you are going to transfer any more heat. An ideal radiator is with the core infinitely short and an infinite number of rows. The reason has everything to do with the gradient of the water temperature through the radiator... that is why we are NOT running a dual pass anymore. (We overheated the past two years as well with a dual pass.) Great for packaging but only if you are willing to give up on the efficiency!

More often than not it is the air thats the limiting factor. But maybe your headers were too restrictive or provided an inefficient flow through the radiator, and going to a single pass fixed that.

[jjusb]

Single pass radiators are not always better, but neither are single pass.

It depends on a number of factors.
If you use double pass, you would limit the number of rows of water thus increasing the reynols number for a given mass flow. Also you would get a higher water pressure drop than using a single pass.

Using a double pass would double the total lenght of the exchanger but decrease the trasfer area for each pass.

As most engineering problems there are compromises for each design, you must test which is best for you.

[Chas]

Yeah we've used single pass on previous cars. We're hoping for the best. Our radiator company thinks two pass will be more efficient. He may be right due to the amount of fluid bypass in our past radiators, but we'll see.

[skeltz]

Having just done alot of testing into this myself, i might be able to shed some light on parts of it for you. i can only really speak from an airflow perspective but firstly, if your not already using one, use a thin cored radiator. having tested one an inch thick i found that at the lower speeds we seem to overheat at, the air flow out the back is only 50% that at the front so a thick core would be aweful. A large fan to pull air through is critical, as regards to fan size, im still looking into that more closely, but volumetric flow rate seems to be the key. The spall fans are nice and light and pull air at a fair speed but generally have big hubs, thus the flow rate is reduced. Also, if your not running a shroud at the front of your radiator i suggest you do. Not only does it maintain an even pressure distribution across the core, ensuring more efficient cooling, but also increases the air speed out of the rear of the radiator. As regards to the rear shroud, i would lean towrads keeping it straight but packaging on these cars being as it is doesnt always allow that. However, i would look into how far away from your radiator your fan is. the effect of a shrouded fan behind a raditor is optimum at a certain distance behind the rad, meaning that its pulling effect is most uniform across the core at a certain distance away from the rad. Exercise some caution tho as the further back you move the fan, the more compromised its effect will be on the raditor core.

Hope this helps mate, still learning myself so any suggestions any one can make go ahead.

[Jersey Tom]

For what its worth, we've driven with Chas's double pass radiator / shroud / fan setup, and it works great. Cold as ice!

[illiterate]

has anyone tried putting it under the car?

[Chas]

Like Tom said, I'm happy to report that we may have too much cooling this year (better than not enough). We have a smaller core than last year's single pass with more cooling. 800-900cfm fan and a decent shroud.

Go double pass!

[harini]

hi...
can anyone give me the water flow rates of the mechanical pump in a 600 cc honda F4i engine or any other 600 cc engine.

[Scott Sinclair]

The first post on page 8 of this thread gives this info for a F4 mechanical pump

[Freedom Fries]

Scott,

Did the paper you have with the flow rates inlcude any information about the pressures that developed those flow rates? I would really like to get an idea of the flow rate of this pump at different pressures.

Thanks

[BeaverGuy]

The paper is SAE 2004-32-0084 or JSAE 200444371 and is titled Small Engine Cooling and the Electric Water Pump. It does include some pressure data for pumps but the pressure data is circumspect. And I would say the same for a good portion of the rest of the data in the paper. That said there is some interesting and relevant information in the paper but I don't really think it is worth buying a copy from SAE.

[MacV]

Jonathan:

I am an alumnus of the Mercer FSAE team from 1997 and would like to know the status of your current team. Please e-mail me at

Mac
cmqv@hotmail.com


Originally posted by ATOhoosier:
We're a first year team experiencing all sorts of problems so by advide may not be the best. However, we've found "Compact Heat Exchangers" by Kays and London to be extrememly helpful. The only problem we're running into is that there isn't really much in the book in terms of core cofigurations for small cars and motorcycles. Another book we're trying to find is "Fundamentals of Heat Exchanger Design" by Shah. Appearently Shah worked for GM and his books have more of an automotice slant on the information contained.

On that note, how much design are the more experienced teams doing? Are you able to find core configurations/ tube pass configurations for common bike radiators? If so where from?

Thanks.

Jonathan Strunk
Cooling System Team Leader
Mercer FSAE
Mercer University
Macon, GA

[MacV]

UOF tried this a few years back. We (Thor and Mac from Mercer 1997)ran with them at 103rd street in Jacksonville and the car had cooling problems. The problem that I would note is that there are a lot of stones to deal with.

Mac


Originally posted by illiterate:
has anyone tried putting it under the car?

[JuicedH22]

Originally posted by illiterate:
has anyone tried putting it under the car?

Yes, our 04 car has the radiator laying flat, behind the drivers seat, it works relatively well. (single pass, 18x6x1.5) The under tray has a hole with a screen cutout in order to allow air to the radiator, and dual fans w/ hig CFM to get enough air through...

[Jenner Collins]

Hi All,

Well, finally read all ten pages of posts! Some good info...

Last year we had our radiator on the back of the car behind the engine, near the oil tank. It ran with a 12" Spal fan(I think...) with a short Carbon shroud on the back - the rad was about a 12"x12"x1.

My gut feeling is that the air flow to the front of the rad would not have been spectacular, but the reasonsto move were that the engine was over cooled and/or that you hit cones with a rad on the side.

This year we have decided to have a duct fore and aft which will cure the cone-hitting rad problem... As for moving the rad to due to over cooling, I would have thought make the rad cross section smaller...

...so...

We tested a 10"x12"x1" with a 75%(of rad frontal area) duct and it worked fine. Although we have not gone too crazy as it was on a cold ambient temp day. So we may go a touch smaller on rad core, but we'll see.

I think Jonothan Gray has posted this a few times, we concur, it is very important to seal the duct. The air does not want to go through the rad, not if it can find an easier path. It is like electrons in the respect - always wanting the easy life - buggers!

Deffinately makes sense to run the fan a little further down (or up I guess...) away from the rad too.

We are looking at angling the rad, again like said before, if you angle the rad AND angle your duct so the stream of air comes to the rad face at 90?, this would be equivalent to having no angle at all.

I feel it is more important to get the front duct right, however as the air is a fluid, it 'knows' what it is about to hit. By this I mean, if you have a high pressure or some sort of blockage past your fan downstream of your rad core, the air will not want to pass through it. This is why in F1 you see so many chimneys and the like, the aim is to create a low pressure past the rad to 'suck' the air across the rad core.

Jonno, thanks for the post about the Kenloww rads, will check them out.


Thanks to all the post before me, I hope this helps some of you.

Regards,

[Nitesh]

At what rpm do you take the water flow rates for your calculations, as the flow varies with the rpm. Do you take the rpm for the max torque/power or perform your calculations at various conditions before arriving at a conclusion?

[Jenner Collins]

Hi Nitesh,

We did our testing by actually making a test duct and running the car on track on a very simple circuit where driver influence would be minimal.

You can see the testing on youtube.com if you search for 'Brookes Winter Testing'. Or go to our website: www.brookesracing.co.uk (http://www.brookesracing.co.uk) and there is the same video and also some photos. It is a big aluminium thing hanging off the side of the car http://fsae.com/groupee_common/emoticons/icon_smile.gif


Jenner

[Marwan]

Hello,

Thought this may be the best place to raise this...

What are people's thoughts on using either the pusher or puller fan set up? Does anyone have experience with both and if so which do you prefer?

I've been reading some material that suggests benefits to both but am yet to see the variations explored in detail on any of our previous cars, so I don't have much to work with other than my reading!

Cheers

[Maverik]

I haven't used both, mainly because of packaging issues, but here is my opinion...
The only thing you need to get your radiator to cool is a flow rate across the core, for that to happen you need a pressure drop. Assuming the same loss for pull or push fan the only difference is going to be your deltaP will be in a higher range than for a puller (though the drop will be the same.) I would imagine there could be something there in regards to specific heat values, but in all likely hood unless your pressure range goes from 100kPa to 1MPa the effect is likely negligible.
The next point has to do with where you position your radiator. Most teams have them on the side in some fashion so they are open to the free stream flow associated with the moving car. With a pusher, the restriction in front of the car to free stream air seems like it would overcome any benefit (at least at the speeds we operate at) to even thinking about free stream air. With a puller behind the core, the effect of a restriction should be less as there is a longer distance for disturbances to propogate to your duct opening and the air has to go through the core anyway. But again depending on the size of your fan, you might only see marginal gains in this area.
So in the end, I think it really comes down to packaging and what size you think your fan and core should be.... unless of course you are thinking of a sucker car.

[neeraj shah]

hello we are team orion from india competing for the first time in fsae.

can any one suggest when to turn the fan on and off. should it depend on the temperature of the coolant or the engine rpm.
also we are using the radiator that came with the hondacbr 600 f4i engine will it work efficiently or do we have to look for other options.

[Maverik]

Read through this entire post and some of what you are asking might be answered, otherwise the best thing would be to get a heats book and learn how compact heat exchangers are designed (ie figuring out heat transfer coefficients for a given radiator design, etc...) once you have that knowledge a good design guide is Compact Heat Exchangers.

[Marwan]

Thanks for the info Maverik. My thoughts were actually similar to yours in terms of pusher flow restriction.

[egg-uon]

With regards to using Kays... I am currently trying to size our radiator for a 80bhp engine. I've been reading the kays guide to producing the size of a radiator, but it asks for a lot of variables that I can't seem to work out. Values such as change in pressure (do I just assume this?) and Ntu (which is found from the earlier equation... Area*conductance / Cmin. this seems to mean that I have to assume a size at some stage for the initial calculations?

Has anybody gone about finding their rad sizes this way, what sort of values should I assume? Any help would be fantastic.

[BeaverGuy]

There are a couple ways to go about it. The way I would reccomend is to take all your known and design values which should be the amount of heat the radiator has to remove from the coolant, Water temperature at the radiator inlet, Air temperature at the radiator inlet, air flow rate, and water flow rate.

Use these to find your effectivness and the heat capacity ratio. Then take these values and an assumed value for the NTU and itteratively solve for the actual NTU. You can't directly solve for the NTU because an automotive style heat exchanger is a cross flow heat exchanger that is best modeled as both fluids being unmixed.

Here I would then assume a value for the conductance based on the air side transfer area. Then using the relationship between NTU, Conductance, Heat Transfer Area, and Cmin you can get a value for the heat transfer area. Now you need some data about the radiator core, with fin and tube sizes and spacing being the bare minimum. Using the core data, your input flows, and desired heat transfer area you will now, depending on how detailed the radiator core data was, either be finished with a value for the conductance or you will need to use some heat transfer correlations to ultimately calculate the conductance. EIther way chances are it won't be the same as what you assumed so you will need to iterate with a new value for the conductance.

[Destruction]

Hi people,

Whats the deal with louvered fins ? Have not been able to find any theory to make some calculations. Are there any books that someone could recommend detailing louvered fins ?

Thanks

[BeaverGuy]

in The EFFect of Changes in Ambient and Coolant Radiator Inlet Temperatures and Coolant Flowrate on Specific Dissipation which can be found here along with every other SAE paper DELPHI has published since 2000Delphi Technical Papers 2000 (http://www.delphi.com/news/techpapers/2000/)
there are some references mentioned about the methodology to calculate the Nusselt number for louvered fins.

[jejuu]

what is exactly the overheating temperature of an engine?how do we know that our radiator design will not cause overheating to the engine?

[BeaverGuy]

The overheating point of the engine can be defined as one of two points. The first is when the coolant begins to boil, for a 16psi cap this is ~230 oF if I remember correctly. The second is the point at which the engine can no longer run properly or damage begins to occur.

The way to design such that you don't reach the point of "overheating" is to design your cooling system to be energy balanced at a temperature below this. This is really just a quick and easy explanation as the reality of it is far more complicated.

[Maverik]

If you want to get tricky, you can play a dangerous game with "overheating." The hotter an engine is, the more efficient it gets... until mechanical failure (thermal expansion can cause bearings, gaskets, seals, etc... to fail) or detonation. So if you want to, try figuring out a couple of design parameters such as what kind of film thickness you need to support your main/rod bearings and what pressure/viscosity is needed of the oil to supply that film thickness readily... then figure out the temperature of your oil at that point and boom, overheating point. I wouldn't suggest trying to run close to it though...

[Mechanicaldan]

Just wanted to add a little info regarding pusher and puller fans. We ran a Spal pusher fan on a CBR1100XX radiator, and it moved so much air it actually bent all the cooling fins closed. We then had overheating problems that we couldn't find until we finally removed the fan from the radiator. It was very obvious what our problem was. 2 hours of twisting the fins straight, and mounting a puller fan corrected our problem. We then used Spal puller fans from that point forward, with custom carbon fiber ducts to seal the airflow between the radiator and the fan.

[Schultz]

Radiator with ducting dyno
Does anyone have any useful or previous radiator dyno experience? We have some great ideas that we can discuss but looking for more from someone that has built one.

[James_C]

Originally posted by Jenner Collins:

We did our testing by actually making a test duct and running the car on track on a very simple circuit where driver influence would be minimal.


Jenner

Jenner,

Care to summarize the results of this test? Did the inlet duct significantly reduce system temps or the duty cycling of the cooling fan?

Maybe the weight of the inlet duct is not warranted by its impact on cooling, but I have yet to conduct testing or some kind of analysis. Although the final decision might boil down to aesthetics.

[kapps]

How many teams use a diffuser before the radiator? Is there an optimum velocity through the radiator that I should be shooting for?

[Maverik]

Just like everything else in engineering, you need to determine the velocity required... which is directly related to the area and the external convection heat transfer coefficient. In my experience, our competition speeds are slow enough where a diffuser before the radiator won't help significantly and makes packaging that much harder... unless you are an aero car and care about the drag

[kapps]

Thanks for the info. I'll have to run some calculations. We probably will be an "aero car" so at the very least, it'll help with the drag.

[Conrad]

This link was posted in this forum about copper-brass vs. aluminum radiators:

http://www.ozsuperkart.net/TechArticles/radiators.htm

The author of the article mentions that copper is stronger than aluminum, but copper has an elastic modulus roughly 110-128GPa, a shear modulus roughly 48GPa and a yield strength roughly 70MPa; whereas, aluminum has an elastic modulus roughly 70GPa, a shear modulus roughly 26GPa and a yield strength roughly 400MPa (that is, copper will deform more at the same stress levels, and yield at a lower stress = not nearly as strong)...

Am i missing something here? To the defence of the article, there is a good point about the solder between the fins and the tubes of the copper-brass radiators creating a high contact resistance. From what I have read, it sounds like copper-brass brazed radiators can be designed to be more effective per unit mass, but at a much higher cost (since these products are not highly available, yet, and therefore likely need to be designed for specific purposes at higher manufacturing cost).

Sorry, to summarize, it sounds like the article is wrong; although, copper-brass radiators can be designed to have a higher heat transfer per unit mass than aluminum, but at a higher cost.

This "daubery" guy seems to think so anyways...
http://www.eng-tips.com/viewthread.cfm?qid=61247&page=9

[Mike Hart]

What are peoples opinions on using kart radiators? There are a few out there for ICAs that are pretty big by kart standards (370x200x~1in flat and 415x200x~1in curved) and could suit an FSAE application in my opinion. The only thing I'm a little worried about is not knowing what the fpi is on them. They look fairy dense but not really sure. The average speeds have got to be fairly similar and the rads are totally unducted. The 4 Stroke Aixro Wankel engined kart generates 50bhp and gets away with this rad:

http://www.zercustoms.com/news/images/Mercedes/RENNtech-Aixro-XR50-kart-engine-3.jpg

[fart can]

Originally posted by Daves:
Aluminum tubing saves about 1 pound of weight per foot compared to rubber hose.

Because we care, we just weighed exactly one foot of stock rubber radiator hose. It weighed about 150 g. This is roughly 1/3 of a lb. This means that using aluminum lines adds negative weight.

Revolutionary.

[Biggy72]

Sounds like all the more reason to run aluminum tubing. Actually I'm thinking I'll just run lines everywhere since we'll probably be a little on the heavy side.

[screwdriver]

Sounds funny. But don't overdo, you'll loose grip http://fsae.com/groupee_common/emoticons/icon_wink.gif. Maybe if you install them upside-down, they'll generate downforce.

@Mike:
Regarding the Kart radiators, TU Graz's radiators looked strikingly similar this season to the one in the picture above, if my size reference is correct. What you have to pay attention to, though is what kind of engine the cart runs on. If it has similar power and capacity, you could propably use it. If not, you're better off with something else.
On a side-note: we recently found out that our water-pump was the culprit in our cooling problems. It proved way less efficient when under back-pressure.

[Mike Hart]

The Graz radiators were off a KTM apparently, but it really wouldn't surprise me if they are fundamentally kart radiators. To me they look very very similar in size to a Rotax radiator which is 135x290x35mm.

http://www.spellfame.co.uk/acatalog/PICT0026.JPG

I was surprised they cooled ok with them, particularly due to a complete lack of ducting and tiny little fans stuck right against the back of the core. A Rotax generates about 30bhp though from 125cc and they have to put slats on the completely unducted core to stop it overcooling, so they must be fairly efficient.

I got a nice complicated heat exchanger book out the library yesterday and have Kays on reserve so will do some calcs and see what I come up with http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Steve O]

Im impressed that this post has gone on fairly continuously for 5 years...

On a more related note, we have been doing a lot of dyno testing with the stock radiator and fan of our yfz and a small fan to keep air flowing across the engine while doing loaded testing without overheating at all. Honestly my opinion is that if you can design it into the space you have, use a stock rad or an aftermarket rad meant for your particular engine... it will probably be the most efficient as far as size and weight you can get and still have a good margin of safety. It's also cheaper and easier than custom designing your own. Why reinvent the wheel? Are teams really seeing a big benefit ,other than packaging, with a custom designed radiator? (that question was not rhetorical, I am honestly interested in knowing!)

On the topic of aluminum tubing, I wonder if that increases heat dissipation through the lines themselves? If so I would say location of plumbing would be a lot bigger deal, keeping it away from air intake and fuel lines and keeping it somewhere that the added dissipation can be toward your advantage. Interesting to think about but I think I will stick with the rubber for the added downforce! :P

[ad]

On the topic of aluminum tubing, I wonder if that increases heat dissipation through the lines themselves? If so I would say location of plumbing would be a lot bigger deal, keeping it away from air intake and fuel lines and keeping it somewhere that the added dissipation can be toward your advantage. Interesting to think about but I think I will stick with the rubber for the added downforce! :P

The only real benefit of aluminium tubing is the reduced weight per unit length over rubber tubing. However if you have too many connections (i.e. not one piece tubes bent to required shape) these benefits deminish at best pretty quickly. Aside from this the freedom of bending to many different bend centre line radii and angle is of great convenience.

I would say that the heat dissipation is minimal given the low surface areas, however it would be better than rubber given the differences in the through thickness resistance component.

[MPetti]

I have been tasked with designing the cooling system for our '10 car. I've read through this thread a few times now, and I'm still having trouble understanding exactly how to figure out the heat rejection of the engine. I hear a lot of people throw around the 1/3 of max engine power idea, but what exactly is the max enigne power. Power at the shaft? Total engine enthalpy? I'll try reading it again, but I can't find a clear answer anywhere I look, and any help would be greatly appreciated.

[Pennyman]

We basically used the stock engine (f4i) setup as a baseline, and took water temps before and after the radiator with no forced convection (no fan) as a worst case scenario. Sized the radiator based on the NTU-effectiveness method from there.

In terms of weight savings, just for fun we chopped our main rubber coolant line and replaced it with about 18in of aluminum tube with small rubber ends. We saved 1/3lb over the rubber alone.

[kapps]

Mark, are you guys a new team? If not, was your cooling system not up to snuff last year? Many on here will say that modeling and calculating a cooling systems efficiency won't get you very close to actual. I tend to agree. I've been doing our cooling system for the past 2 years. I started with no knowledge and had no help from previous teams. If you are like I was, I recommend looking at the stock radiator. You can get rough sizing requirements there. If you don't have or want to use the stock radiator, ATV radiators work well. If the ATV has comparable displacement and the radiator has similar frontal area as the stock one, it should work pretty well on your engine.

If you aren't a new team and have previous work to go on, where were the shortcomings last year? Increasing cooling efficiency can be done by increasing radiator size or simply better ducting and higher flow fans. You could use two smaller radiators instead of on larger one if space is an issue (this is what I did this year).

[MPetti]

We are by no means a new team here at Rutgers(though this is only my second year on the team). We just had serious cooling issues last year preventing us from finishing endurance at Virginia and Michigan last year (though it was a little more involved than bad cooling at MIS).

I know its not very accurate to try to calculate everything, but seeing as we just switched over to the honda CBR and had big heat issues the first year we used it, I need something to base my setup off of at least initially. All I have so far is a system that didn't work last year. I will have to take a look at the stock radiator size though as a rough base point.

[Wesley]

I'd recommend as much data gathering as possible to determine what kind of heat loadings you're going to see.

A couple of thermocouples before and after the engine and radiator(s) will give you heat rejection numbers. Really, even that isn't necessary. We did some really long dyno pulls on a 1500HP superflow chassis dyno, and basically we could look at logged coolant temperature data and watch the increase over time - knowing system fluid mass and a little conservation of energy tells us the heat rejection by the motor. It actually worked out pretty close to Heisler's 1/3 engine horsepower suggestion. After that, do a couple hand calcs just to see what kind of size range you're going to need a radiator in, include an appropriate safety factor, and go.

[radi]

Hey i just went through all the post in this article and i really cant understand how to get the area of heat transfer of a radiator core while calculating the NTU( NTU=UA/Cmin)............
can anyone pls help me out with this????

[Wesley]

Well, the air-side transfer area is the total area of all the fins, (both sides) plus the tubes, (assume negligible losses other than the core)

The water-side transfer area is the inner wetted area of the tubes.

[Ani_V]

Iam new to the cooling systems in my team. Was going thru the posts. I came across Heisler's one third suggestion over heat rejection. Does it mean ideally if the engine generates X BHP the heat absorbed by the cooling water should X/3 for optimum performance?

[VFR750R]

It's a generality, but yes.
In Heywood the following are listed for SI engines at max power.
Power = 25-28% of fuel energy
Cooling system = 17-26%
exhaust 34-45%
Misc = 3-10%

Compression ratio and total engine piston area are the largest factors in where a specific engine lies in these ranges.

The listed values assume steady state, so if you base your cooling system requirements over MAX power(not average), it will be close to 1/3rd of that max.

[Kevin Dunn]

We used the same assumptions that VFR750R did from the Heywood book and came out with a really good cooling system (hottest it got in Cali Enduro was 190F with a 90F+ ambient temp).

You can look up the lower heating value of the fuel your using and compute how much of the energy is going to be dissipated into the cooling system (17-26%). This has to be done using the mass flow rate of the fuel @ the maximum power RPM (or atleast that would give you worse case). Then you can size your radiator accordingly as far as how many kW of heat dissipation you'll require. I think we even did some calculations on fans and such, but I think we might have went overkill on the fan we chose.

[Ani_V]

@ above

So that means knowing the fuel flow rate @ MAX POWER RPM, its Lower heat value I get to know how much heat is to be rejected. And then knowing the pump capacity i will get to know the water flow rate in the engine block.

Now the thermostat full valve opening temp of our engine is 95 deg celsius. I just have a doubt.
Should I take the temp at radiator inlet as 95 deg? And to what temp approx should the water be cooled by the radiator??

[Wesley]

In my opinion if you design your radiator to reject the same energy as the engine rejects to the coolant at max power, your system is oversized.

If you can, get an old car and do some testing and answer a few questions.

Things to measure:
- Mass flow H2O of your system per RPM.
- Inlet and outlet temperatures of the engine
- Average mass flow rate of air through radiator (fan or otherwise)
- Track ambient temperature during tests

Things to calculate:
- How often do you use max power while on an endurance course?
- How much heat do you reject from the engine per lap? In other words, what's your average power dissipation over the whole lap?
- During the longest period of full acceleration you expect to see, what will your coolant temperature rise to?
- What temperature ranges do you need to be able to handle? What's the average track temp at each event you're attending?


Design considerations:
- How can you affect heat rejection and capacity of your system?
- Can you run a smaller radiator while increasing system volume via a swirl pot or header tank to allow less maximum heat transfer but increase the amount of heat your system can absorb per degree temperature increase? This will allow your system to accommodate periods of acceleration and "catch up" during periods of part throttle or deceleration.
- Or should you decrease coolant mass and run a larger radiator, maintaining a more constant temperature via thermostatic control?
- Is natural airflow sufficient for your needed air mass flow rate? If wings or other elements obstruct your radiator, how large a fan is sufficient to generate required flow?

And another important consideration: what are your oil temperatures? Are they too high? Are they too low? Increasing oil temps not only reduces the amount of heat the oil can carry away from engine components (specifically pistons to prevent detonation) but it also reduces its viscosity, leading to increased friction levels and MORE heat. Treat your oil temperature with just as much importance as your coolant temperature.

[Ani_V]

@ wesley

I am sorry to have forgotten to mention that mine is a rookie team. We are participating for the first time and that to on a very strict budget. So obviously we do not have the liberty of such detailed testing. Could you suggest a basic approach that could help me explaining the design of my radiators? Even approximations would do.

[Wesley]

Ani_V, every team should have access to two thermocouples. Put them on the cooling lines, put them on the oil pan, you can get useful numbers even from cheap setups. Start with a stock radiator, available cheaply, then determine what your heat rejection is to see if you can use a smaller or more efficient one. Or at the very least, have numbers to prove to the judges that your radiator is sufficient and is cheap enough to fit in your price range.

[Adambomb]

+1 for cheap testing.

With a bit of creativity, work, and a good plan, it is possible to get good data with almost no financial investment. You don't always need a fancy full data aquisition system.

[theloyal]

Hi Guys,
I am designing the cooling system, all i have done is that at constant Flow rate of Water and Air, i find out the heat rejection then after reading Fundamentals of Heat Exchanger by Shah i design a system capable of cooling that number..
Then after that i come to the real case where my flow rate of water is varying with Different RPM which i have also calculated but finding air flow rate which is also varying is problem faced by me. Can someone suggest me that how can i found air flow rate, remember i am using electric fan which is operated by thermostatic switch..
Regards

[Wesley]

You should be designing your ducts to give you appropriate airspeed for cooling.

As far as the fan is concerned, you can guesstimate with the manufacturers free-flow specs, but remember you have the radiator pressure drop which will reduce flow. You can always get a hand-held anemometer and measure airspeed in a matrix across the radiator inlet with the fan running to average out total airflow, but that isn't the most accurate way.

[heat exchanger]

Can any one give me link for get free books for automobile radiator in louver fin aluminum brazed radiator thermal design calculation.

[Adambomb]

www.google.com (http://www.google.com)

[Vino]

I got an oppurtunity to work for the cooling team.I'm toltally new to this. I want to size the radiator, could any one tell me what are all the basic inputs required to size the core? what is the minimum and maximum distance we have to maintain between Radiator and Fan(Suction).

[Wesley]

I highly recommend obtaining a thermodynamics textbook and going through, at the very least, the heat exchanger segment. Any first year engineering thermodynamics class will cover these in much greater detail than possible on a forum.

But, in a nutshell, you need to know the mass flow rates of the air and the coolant (trickier for coolant, as it varies with RPM,) the specific heats of both, temperatures of both, heat transfer coefficients (depending on the method you want to use)

The calculation of the actual heat rejection isn't the difficult part, a couple thermocouples can tell you that. The part where the real decisions come in:

What do you want your system volume to be? System volume provides a "buffer" increasing the total heat capacity of the cooling system. This buffer can be sized to a specific amount of running time - say, an autocross lap - before the system reaches an overheat state. Water is also heavy.

Are you getting good airflow? You may rely on a fan for idle cooling only, or also utilize one for augmenting at-speed cooling. Do you use one or two or 12 radiators, and where do you put them to ensure you get the airflow your calculations say you need? What kind of pressure ratio do you need to get proper airflow through your radiator?

What impact does fan distance or a fan shroud have on heat rejection at low speed? At high speed?

What measures will you implement for air/steam control? A swirl pot? Vent lines? Header tank?

I would start with Carroll Smith - I think it's Design to Win that covers basic cooling system design.

Then hit the engineering texts. Once you understand the quantities and effects at work in a heat exchanger (aka "radiator," which isn't technically correct) you can start looking at OEM radiators to get an idea for scale.

What size radiators do OEM's use? Are these still acceptable solutions in a race environment? What compromises do FSAE teams make that will limit the effectiveness of a stock cooling system (race conditions, poor airflow, high idle/restricted conditions.)

This is the stage where you make yourself a list of questions - what do you NEED to make it work. After that, you can start answering the questions and making design decisions.

I encourage new teams to start with a stock system, then, money and time permitting, go to a more specialized system. The OEM components are available very cheaply compared to custom components - and as long as you can show that they are sufficient for your needs and design intent and understand the data that supports your claim, the judges will back you up.