I thought I would post my original flow rate calculations that I did for the old KVRC. (Back in 2015). You can see that I calculated what I considered to be realistic values, but we (Julien and I) agreed to reduce them to make the challenge a little easier whilst we introduced this extra part of the challenge: we always planned on increasing them when everyone had got to grips with them:
Machin wrote:Just as a bit of background about myself; I currently work for an engine manufacturer that supplies engines to military applications (land and marine); not exactly race engines but they are still high performance sequentially turbocharged internal combustion engines to which I have access to all of the testing data. My job is helping our customers design their vehicles to accept our engines, including specifying their cooling, exhaust and intake systems.
From our test bed data I know a good modern engine has an energy break-down roughly as follows: Mechanical output = 40%, Energy dissipated through cooling system = 21%, Radiated heat = 2%, Wasted in exhaust = 37%. Total =100%.
After application of our power curve we typically see average power outputs of about 300kW across the typical KVRC lap. So using the figures above you would need about 160kW of cooling (21/40 x 300 = 160kW) on average over the lap. We run the CFD at 100mph (44.7m/s), which isn't too far off the average lap times of these cars, so we're looking for a cooling flow in the test that is sufficient to dissipate the full 160kW cooling requirement. With air at 1.2 kg/m^3, specific heat of 1.005kJ/kg.K and a 25degC delta T we need 5.3m^3/sec (160/(25*1.2*1.005) = 5.3)... We rounded that figure down to 5 to account for the fact that the average lap speed is normally a little more than 100mph.
For this year, to ease ourselves and competitors into this we're looking for a further reduced figure of 3m^3/sec (that's total for the whole car - both side pods, so 1.5m^3/sec through each side pod).
Also this year we have mandated a minimum radiator face area, again just to ease ourselves and competitors into this extra dimension to the competition. The minimum area of each radiator is 0.2m^2, so we're looking for a "face velocity" through each radiator of about 7.5m/s (1.5/0.2), which is about 16% of the free-stream velocity of 44.7m/s
Let me know if you think there are any errors in that, but I think the numbers all seem reasonable. The proof also seems to be proven in the results: we had one car which was submitted without a front wing and he achieved the cooling flow figures easily; proving we can't be too wrong. I think the reality is that Variante's huge front wing assembly is simply unrealistic. It will be interesting to see what he cuts it down to in an effort to get sufficient cooling, but CAEDevice, JJR and TF have all achieved the cooling requirements whilst still having pretty good front downforce....
As I say, let me know if you think there are any errors you can see in the maths, but I'm confident on the 160kW cooling requirement for a 300kW mechanical output as I "stole" that ratio off some actual engine test bed results!
Pressure loss through Radiators:
Machin wrote:I did a quick calculation based on the method used in "Flow of fluids through valves,fittings and pipes" by Crane (
http://www.amazon.com/Fluids-Through-Va ... B003152YTG). Taking one sidepod on its own and a figure of 2.65m^3/sec flowing through it I found that the pressure difference required to generate that flow through a radiator in the duct ("modeled" as a mesh with a flow coefficient according to the book referenced above) was 400 Pa.
Engine intakes and Exhausts:
Machin wrote:In these cases the pressure requirement was simply derived from typical values from real engines; above which the engines start to lose power: On the inlet side pressure can be anything down to 0Pa (Infact, typical engines can pull against a slight negative pressure, but assuming there is some pressure loss in the intake duct I thought 0Pa was a good enough value). On the exhaust side a typical engine can "overcome" anything up to a back-pressure of +980 Pa without derate. The velocity was also based on typical "real engine" values: 20m/s on the inlet side and 30m/s on the exhaust side, requiring a total combustion inlet area of 0.015m^2 (0.3m^3/sec/20 = 0.015) and a total exhaust outlet area of 0.01m^2 (0.3m^3/sec/30m/s = 0.01m^2).
