Radiator + fan thermodynamic model.

[Cold Fussion]

Hi there, I'm currently building a thermodynamic model for the engine cooling system of an FSAE car, and have come a little unstuck when dealing with modeling the fan (the fan in this case sits at the back of the radiator, sucking air through it). The end goal is to be able to specify a duct geometry with associated radiator, and be able to map the heat rejection as a function of engine temperature and vehicle temperature. At the moment I have modeled the heat rejection of the radiator as a function of the vehicle velocity using a NTU method, while utilising the methods of SAE paper 710208 for the convective properties of the radiator fins (though i am a little suspicious of the results, as the heat transfer coefficients are very high for the louvered sections), however now, I am a little bit stumped as to how to integrate the fan of the model.

My basic understanding of the fan is that it will increase the effective pressure drop across the radiator, which will increase the air speed through the radiator. My difficulty lies in when i consider the fan is working on moving air. If we consider an air free stream spining the fan, then there will be a vehicle velocity where eventually, the incoming air will want to spin the fan faster than the fans motor will spin the fan, and thus the fan will be providing a resistance to the flow. My question is how to I actually quantify this?



This is the fan curve for a similar fan to what we'll be using. If I were to work out the total pressure before the fan (based off of the vehicle velocity, duct geometry and radiator properties), would that mean the fan would add x amount of volumetric flow rate through the radiator? From this added flow rate, I would then be able to calculate the heat transfer from the radiator.

Secondary to this, is my question of calculating the drag induced by the radiator. In principal I would say the drag is the pressure drop multiple by the frontal fin area of the radiator. However I am not confident in doing this without a better understanding of the fan.

If someone could point me into the right direction that would be great, or point into some good reference material. Currently I've been relying on my Fox and Mcdonalds Fluid textbook, Incropera Fundamentals of heat and mass transfer, and Fundamentals of Heat Exchanger design. Thanks.

[marcush.]

I would be very very conservative with all this .
In automotive industry there are two paths how fans are integrated into the calculations -one is the volume-force method the other is the moving (rotating )mesh method in CFD .Both are less than really precise when it Comes to predicting temperature and flow effects.

you will always face losses due to blockage ,caused by the drive Motor of the fan ,insatallation ,shroud Details ,proximity of the fan to the Radiator mesh and of cause blockage in your Radiator ducting as well .
you can PM me ,i might have a few more details to share .

as for the restriction coming from the faan in the air duct you are right it will always have a negative effect as the Motor physically restricts the flow throu the mesh .Additionally there are Motordrive concepts (DC brushless ....)that will force the fan Motor to Keep its Speed no matter what resistance there is in the ducting ... so as a positive you could make this your friend at times using the fan to stop or impede flow through the ducting -wich could lead to lower drag numbers Overall .At least that´s what OEMs toyed around with recently to save a drop of fuel...

[Greenish]

For the fan performance, the typical method is to plot a system curve vs the fan curve, and that will determine the operating point of the fan. The system curve describes pressure drop required to push air through the radiator at a given flow rate, and will be approximately parabolic with increasing flow. You can determine this for your radiator by formulas or by modeling. You can then account for the extra flow due to ram air pressure, but you're right that this will effectively shift the curve. Incropera probably has something on this... but boy that book is not meant for reading by real humans. There are lots of web resources about fan curves and matching them to the system; a quick search reveals this, which isn't bad: http://www.greenheck.com/media/articles ... basics.pdf

Regarding drag, it's complicated and maybe someone here who knows about estimating internal flow can speak up. But in theory, yes, using the pressure drop x area is a good starting point... but you need to account for the pressure drop through the whole air passage back to free stream, not just the radiator. And then you need to add or subtract the pressure at the outlet (i.e. is it in a low- or high-pressure zone).

[Cold Fussion]

For the fan performance, the typical method is to plot a system curve vs the fan curve, and that will determine the operating point of the fan. The system curve describes pressure drop required to push air through the radiator at a given flow rate, and will be approximately parabolic with increasing flow. You can determine this for your radiator by formulas or by modeling. You can then account for the extra flow due to ram air pressure, but you're right that this will effectively shift the curve. Incropera probably has something on this... but boy that book is not meant for reading by real humans. There are lots of web resources about fan curves and matching them to the system; a quick search reveals this, which isn't bad: http://www.greenheck.com/media/articles ... basics.pdf

This is what I'm mostly confused with. Getting the system curve and thus the system to fan intersection curve should be fairly straight forward, however the variable flow through the radiator is what's confusing me. If I were to say find the curves intersection at 0.3 m^3/s volume flow, but then say the volume flow rate at 60 km/h is 0.6 m^3/s, which does this actually mean? It would be to the right of the fan curve, so does this mean it is effectively doing nothing? And similarly, does anything before this point mean the fan is adding to the flow rate?

[Smokes]

Basically if your intake flow rate goes over the BEP of the fan curve the fan blades stall. Typically the OEM fans are there to provide air flow through the Radiator when the vehicle is has very low air flow i,e stationary and low speeds. When there is high airflow the fans stall and wind mill. typically this does not matter to much in OEM applications as the air exiting the radiator fan hits the engines block,manifold etc. though it will matter more when fuel efficiency is required.
If you are ducting air through a rad with a clean exit you would be better off removing the fan out high air flow system and fit
a secondary intake with a ducted fan to push air in to the high air flow system to cool the vehicle when it needs the additional air flow at low speeds.
something like this

http://www.suzukaracing.com/1blower.html

[marcush.]

Basically if your intake flow rate goes over the BEP of the fan curve the fan blades stall. Typically the OEM fans are there to provide air flow through the Radiator when the vehicle is has very low air flow i,e stationary and low speeds. When there is high airflow the fans stall and wind mill. typically this does not matter to much in OEM applications as the air exiting the radiator fan hits the engines block,manifold etc. though it will matter more when fuel efficiency is required.
If you are ducting air through a rad with a clean exit you would be better off removing the fan out high air flow system and fit
a secondary intake with a ducted fan to push air in to the high air flow system to cool the vehicle when it needs the additional air flow at low speeds.
something like this

http://www.suzukaracing.com/1blower.html

that toy with 280 cfm will not do much ...don´t you think so?

As I said before a modern electronics controlled brushless is set to a Speed no matter whta you do .so no windmilling you can even make it stand still at any airspeed hitting it.Easier to calculate though.
Most modern cars have vents in their shroud to allow air a clean path at higher airspeeds avoiding the fan blade...

it seems there are no good pics tobe found anywhere in the web.. the VW touareg has something like 16 or 20 flaps spluttered all over the shroud...



this is actually the Touareg/Cayenne item i was involved in development of this .

[PlatinumZealot]

Cold fusion, why reinvent the wheel?

There are so much factors that come into play here. Heat transfer, fluid dynamics... and even experimental data from the actual fan itself is needed. It's a bit too deep to go into all of this. You might even find that iterative methods will need to be apart of your model too.

My crazy advice is to forget about modeling the radiator and fan from scratch altogether and just focus on making a simple calculator that will use manufacturers given specs for the radiator(s) and the specs for the fan(s) along with inputs for your coolant temperature, car speed and ambient temperature to give you the actual heat transfer and the returning coolant temperature.

[Cold Fussion]

My crazy advice is to forget about modeling the radiator and fan from scratch altogether and just focus on making a simple calculator that will use manufacturers given specs for the radiator(s) and the specs for the fan(s) along with inputs for your coolant temperature, car speed and ambient temperature to give you the actual heat transfer and the returning coolant temperature.

This is what I'm doing.

[marcush.]

very helpful here is Mr Wolf -Heinrich Hucho publications .

[PlatinumZealot]

My difficulty lies in when i consider the fan is working on moving air. If we consider an air free stream spining the fan, then there will be a vehicle velocity where eventually, the incoming air will want to spin the fan faster than the fans motor will spin the fan, and thus the fan will be providing a resistance to the flow. My question is how to I actually quantify this?

Since your fan data for the pressure and flow at stagnant air speed is experimentally obtained, It would say you might have to follow though and conduct some tests to get data for the fan working at those high speeds.

You might even find the fan has choke point where no more air can flow through it ?? just saying... so I would experiment a bit.

[marcush.]

Sure has hell there is nót just calculations but a lot of testing going on.I remember Fan blade distance to the Radiator mesh having significant Impact on flow as has the shroud design and Details like Radiator carriers,airgaps at the sides etc etc ..all things not easy to model a´nd huge potential for non correlation of Simulation and actual behaviour.
The actual discrepancy between Elaborate Moving mesh cfd calcs and actual behavior maybe as much as 10% -as I have witnessed in Projects so in essence it is a far better bet to actually test the Hardware for Performance instead of trying to calculate...with dubious results.
You can always enter the real test results into your Simulation which is best practise in Automotive Simulation for a reason.

[Smokes]

Basically if your intake flow rate goes over the BEP of the fan curve the fan blades stall. Typically the OEM fans are there to provide air flow through the Radiator when the vehicle is has very low air flow i,e stationary and low speeds. When there is high airflow the fans stall and wind mill. typically this does not matter to much in OEM applications as the air exiting the radiator fan hits the engines block,manifold etc. though it will matter more when fuel efficiency is required.
If you are ducting air through a rad with a clean exit you would be better off removing the fan out high air flow system and fit
a secondary intake with a ducted fan to push air in to the high air flow system to cool the vehicle when it needs the additional air flow at low speeds.
something like this

http://www.suzukaracing.com/1blower.html

that toy with 280 cfm will not do much ...don´t you think so?



As I said before a modern electronics controlled brushless is set to a Speed no matter whta you do .so no windmilling you can even make it stand still at any airspeed hitting it.Easier to calculate though.
Most modern cars have vents in their shroud to allow air a clean path at higher airspeeds avoiding the fan blade...

it seems there are no good pics tobe found anywhere in the web.. the VW touareg has something like 16 or 20 flaps spluttered all over the shroud...

http://junge-autoersatzteile.de/auction ... G_5844.jpg

this is actually the Touareg/Cayenne item /I was involved in development of this .

Remember a good FSAE car will only weigh 300 kg and uses a water cooled bike engine, with a bike engine radiator designed to have restricted air flow to it i.e the front wheel of a bike. What I was suggesting was fitting a light weight bypass system. As a fan isn't needed for a FSAE car and adds unnecessary weight and drag.

How much energy did that motor consume when it was resisting the torque of the wind milling effect of the blades.

[marcush.]

I lost that bit of course .... but aiming at 300kgs you can afford to plug in 1 +kg worth of fan? wouldn´t it be more clever to carry some sort of chillers to work around those critical moments instead of carrying the extra weight for no real benefit in track Performance (apart from better reliability)?