Aero and radiators

I haven't cut the hole for my radiator duct yet and wondering the correct sizing. I know the opening should be smaller and open up toward the size of the radiator to slow down the air and create a high pressure area prior to entering radiator (more info see Carroll Smith). However since the face of the opening is more like 3/4 the height (not optimal but not horrible) and several inches wider, I'm not sure if I should open it up completely or leave some of the face remaining. Obviously this will create drag and disturb air but I'm not sure if that's better or worse then opening completely.

I tend to think opening all the way is the answer but looking for some feedback from those with more experience with aero. The exit is good.
PXL_20210326_213030805.MP.jpg
 

Scott

Lifetime Supporter
I’ve done extensive modeling of the radiator duct and outlet on the SL-C and here’s what I’ve learned. You want to maximize mass air flow through the radiator by maximizing the pressure differential across the radiator code. One way to achieve this is with a diverging-converging duct where the radiator inlet is smaller than the radiator core and the radiator outlet is slightly larger than the inlet.

As air travels from the inlet to the core it diverges which causes it to slow down. After the air passes through the radiator, it converges which causes it to speed up. Since air speed and pressure have an inverse relationship this creates a pressure differential which causes air to flow. This is the same principal that causes an airfoil to generate lift.

While all of the above is basic knowledge, I was unable to figure out what the ratios should be. I hired an x-F1 aerodynamicist and he’s done a bunch of CFD modeling on the SL-C. For our types of cars (i.e., very different than F1) he recommends that the radiator inlet is 40% the size of the radiator core and that the outlet is 14% larger than the inlet. The reason that the outlet is larger is that all of the air that goes into the inlet must exit the outlet and the radiator heated the air which caused it to expand. The 14% increase is an estimate based on how much heat he estimated the condenser and radiator reject.

In summary:
  • Duct all air that enters the radiator opening into the radiator
  • Duct all air that exits the radiator into a low-pressure zone
  • A Gurney Flap or molded flick can be used to decrease pressure at the outlet
  • Inlet should be ~40% smaller than the radiator core
  • Outlet should be ~14% larger than the inlet
Here are some images from the CFD modeling we've done:
1616811866760.png

1616811951876.png


1616812013294.png

1616812083626.png
 
I’ve done extensive modeling of the radiator duct and outlet on the SL-C and here’s what I’ve learned. You want to maximize mass air flow through the radiator by maximizing the pressure differential across the radiator code. One way to achieve this is with a diverging-converging duct where the radiator inlet is smaller than the radiator core and the radiator outlet is slightly larger than the inlet.

As air travels from the inlet to the core it diverges which causes it to slow down. After the air passes through the radiator, it converges which causes it to speed up. Since air speed and pressure have an inverse relationship this creates a pressure differential which causes air to flow. This is the same principal that causes an airfoil to generate lift.

While all of the above is basic knowledge, I was unable to figure out what the ratios should be. I hired an x-F1 aerodynamicist and he’s done a bunch of CFD modeling on the SL-C. For our types of cars (i.e., very different than F1) he recommends that the radiator inlet is 40% the size of the radiator core and that the outlet is 14% larger than the inlet. The reason that the outlet is larger is that all of the air that goes into the inlet must exit the outlet and the radiator heated the air which caused it to expand. The 14% increase is an estimate based on how much heat he estimated the condenser and radiator reject.

In summary:
  • Duct all air that enters the radiator opening into the radiator
  • Duct all air that exits the radiator into a low-pressure zone
  • A Gurney Flap or molded flick can be used to decrease pressure at the outlet
  • Inlet should be ~40% smaller than the radiator core
  • Outlet should be ~14% larger than the inlet
Here are some images from the CFD modeling we've done:
View attachment 114039
View attachment 114040

View attachment 114042
View attachment 114043
  • Inlet should be ~40% smaller than the radiator core
  • Outlet should be ~14% larger than the inlet
Do these inlet/outlet opening areas and ratios stay the same if the radiator core is tilted significantly from upright, and/or if the radiator core area is made up of a very thick core?

Jack
 

Scott

Lifetime Supporter
My radiator is tilted at 55 degrees which he indicated was fine. He had a rule-of-thumb angle at which he said you needed to be very careful and do CFD and wind tunnel testing to ensure good mass air flow. I don't recall the angle, but I remember thinking I've only seen those types of angles on race cars.

The following equation is helpful:

1616892932644.png

I can't figure out how to enter Greek letters, so this will need to suffice:
  • A: Area of radiator core
  • DT: Delta T, the temperature difference between the coolant and the air flowing through the radiator
  • K: Radiator heat conduction coefficient
  • M: mass flow rate of air through the radiator
  • Q: The amount of rejected heat
The outlet increase vs. the inlet is driven by how much the radiator, condenser, etc. increase the temperature of the air (i.e., DT). If you know that number, it's simple to calculate (with another simple formula) how much the air has expanded and what the optional exit size would be. However, that will only be optimal at one DT and engine and ambient temp are anything but constant. So long as the angle of the radiator doesn't have a big effect on mass airflow, the angle doesn't change the ratio (it will affect drag and perhaps downforce).

Before I engaged the aero guy, I bought an expensive custom radiator from C&R. It's slightly larger than the stock one (bigger A) and it has a thick high-end core so its K is significantly higher. So everything should be great right? Well, when he got the pressure drop across the core data below from C&R for the thick core + condenser + fan shroud and did a CFD analysis, he didn't like the mass airflow number (M). He suggested a thinner core, but I'm not going to do that unless I have a problem. C&R is top notch and I took their recommendation not to mention that CFD isn't a perfect science -- there's a reason engineers still use expensive wind tunnels! He spent some time designing a molded "flick" for the leading edge of the outlet (does the same thing as a Gurney Flap, but looks better), optimizing the location of the outlet, etc. Time will tell, but I'm pretty sure I'll be OK with the thick core.

55mm core
Inlet​
Inlet Pressure​
Pressure​
Mean Core​
Ambient​
Air​
Temperature​
At Core Face​
Drop Across Core​
Face Velocity​
Mass Flow Rate​
Density​
(°C)​
(mbar)​
(Pa)​
(m/s)​
(kg/s)​
(kg/m³)​
35.0​
1.10​
110​
2.00​
0.473​
1.15​
35.0​
2.89​
289​
4.00​
0.945​
1.15​
35.0​
5.35​
535​
6.00​
1.418​
1.15​
35.0​
8.45​
845​
8.00​
1.891​
1.15​
35.0​
12.22​
1222​
10.00​
2.364​
1.15​


42mm core
Inlet​
Inlet Pressure​
Pressure​
Mean Core​
Ambient​
Air​
Temperature​
At Core Face​
Drop Across Core​
Face Velocity​
Mass Flow Rate​
Density​
(°C)​
(mbar)​
(Pa)​
(m/s)​
(kg/s)​
(kg/m³)​
35.0​
0.91​
91​
2.00​
0.473​
1.15​
35.0​
2.39​
239​
4.00​
0.945​
1.15​
35.0​
4.41​
441​
6.00​
1.418​
1.15​
35.0​
6.97​
697​
8.00​
1.891​
1.15​
35.0​
10.07​
1007​
10.00​
2.364​
1.15​

So to answer your question regarding core thickness... if all things are equal, a thicker core will increase K and decrease M. You'd only want the thicker core if the increase in K was greater than the decrease in M (i.e., Q is larger) and you're OK with the increased drag. Since Q is larger the air will expand more and you'd ideally have a slightly larger outlet.

The F1 guys spend the time and money to test and determine what all of the numbers are. C&R sent data above which helped with the mass air flow modeling, but they were unable to provide radiator heat conduction coefficient (K), You'd probably need to test your exact radiator to get that number. I couldn't find anyone at Vintage Air that even understood the same questions regarding their condenser. Worst yet, I wasn't able to provide any data on how much heat the engine generates which means that I couldn't specify the problem I was trying to optimize. This means that most of us are going to have to rely on rule-of-thumb calculations. My aero guy did his thesis on F1 radiator ducts and then spent 18 years doing aero design and wind tunnel testing for McLaren, Honda and Mercedes. He had a list of questions that I couldn't answer so despite spending the money on CFD analysis there is still a fair amount of rule-of-thumb.

Interesting a stock SL-C has a radiator core that's 45.7% smaller than the inlet which is pretty close to the 40% rule of thumb. My custom radiator is 41.7%, but that was luck. At the time I ordered it I didn't know about the 40% rule-of-thumb and was just trying to maximize A. The stock radiator "outlet" is a disaster, but there are a lot of street SL-C's with the stock setup and no issues. Net-net you can look at similar cars and see what's working for them.
 
Scott, thanks for your detailed reply, and for sharing the information from your aero guy. Lots of great information there. Much appreciated!

Jack
 

Scott

Lifetime Supporter
Mesa, the other Scott's car is an Apex

Dave, I had read that thread back in 2018. Your point about the location of the outlet was spot on, but the ensuing discussion left me more confused than before I had read it LOL. There are several SL-C owners who point out that they don't have issues, so just go with stock. That's probably the right answer for most. However, there are a couple of caveats;

(1) There are lot more forced-induction builds now which have a higher Brake-specific Fuel Consumption (BSFC) than a naturally asperated engine. As a rule of thumb NA is is 0.5, turbo is 0.6 and supercharged is 0.65 at WOT. More fuel per HP means more heat that needs to rejected per HP. So comparing the cooling systems of a NA engine and a supercharged engine with the same power works up to the point where it doesn't.

(2)The discussion was mostly around cooling. However the design of the radiator outlet also affects drag and downforce. IMO, the biggest challenge, by far, with a stock SL-C is aero balance (i.e., not enough downforce on the nose). My objective was to maximize mass air flow through the radiator and to improve downforce. I was initially going to have Alan do his LaFerrari-inspired outlet, but the foil inside the duct reduces drag at the cost of down force so my outlet will not have an airfoil. My plan is to develop a splitter that's similar to Superlite's race splitter, but more streetable.

(3)If you look at just about any SL-C that's seriously tracked the radiator outlet is completely different and the stock radiator isn't used. You'll see some pictures of a specific SL-C with a stock outlet and then subsequent pictures where it's been modified. For example,

Factory race car with stock outlet and Gurney flap.
1617073481184.png

Outlet moved forward and it's my guess that the curve in the outlet duct is designed to maximize downforce.
1617075664794.png
 

Terry Oxandale

Skinny Man
So Scott, seems you may have multiple answers based on what your powerplant consists of, and if this is a street car, or all out track car. One would be an aesthetics answer (open it up), the other is optimum performance answers (something different).
 
Thanks for feedback!! I got wrapped up with work and wasn't able to respond. Scott those numbers will definitely help, it's mostly about making things work with my space and the body....more details below. Terry good point, in all honestly this car wasn't designed for the track. This will mostly be street car, but I'd like to build it "right" even if it doesn't see much track time. After all its mostly about the build anyway.

Mesa, yep it's the Superlite Apex. I took a couple year break when I had my first two kids and I've recently pulled it out of storage...I'm a glutton for punishment. When I left off I figured out I needed to cut the transmission & motor mounts to move the engine lower and back and cut off all the coilover mounts since none of them would get me to ride height with the supplied parts. Seemed like a good stopping point. I should have definitely waited and bought the SLC or GTR.

My issue compared to the SLC, is I have a ton of forward faces to get rid of that will cause stagnation especially the side inlets...and the airs got to go somewhere. The radiator opening is 85% of the core size...and the rad is very close to the inlet (if I add a fan shroud especially). I basically don't have much of a choice but to go almost straight in, too sharp of an angle and the air will just detach anyway. The remaining air coming in to the sides of the radiator I'll just channel under the car. For the side inlet areas I've either got to cut away the side face of the body and create more of a post up front with fins channeling are around the side (curtain)...to much work, Or feed a brake duct and maybe cabin air. There's going to be a lot of drag there anyway.

PXL_20210330_041820504.jpg


The exit is small. If I'm lucky I could cut this to about 60% of the core. Only option would be some hood louvers also.
PXL_20210330_041419795.jpg
 
The Porsche 911 GTR- take-your-pick of numeral and letter ... Some of these have HUGE radiators on the outside edge of the nose. Does not outwardly seem like they take any overt methods of extracting that air (?)
 
The remaining air coming in to the sides of the radiator I'll just channel under the car.
If you are looking for high speed stability, I wouldn't dump air to the underside of the car. For a street car that never sees triple digits, you will be fine though.
This book is a great read if you want to try to maximize aerodynamic efficiency.
51CZwjYcGQL._SX353_BO1,204,203,200_.jpg
 
The Porsche 911 GTR- take-your-pick of numeral and letter ... Some of these have HUGE radiators on the outside edge of the nose. Does not outwardly seem like they take any overt methods of extracting that air (?)
if I remember correctly, the exits of those ducts is just ahead of the front wheel wells, aft of the "front corners". As such, I would expect them to be in a high velocity/low pressure area, so the exits wouldn't need any extra help/features. Contrast that with stock SLC, which exits right at the base of the windshield, which is a very high pressure area on nearly every car.
Of course, I've seen too many older Porsches to count get knocked out of races after relatively light corner contact with another car, resulting in radiator failure. Don't know if they've since built up the areas to resist damage.
 
which exits right at the base of the windshield, which is a very high pressure area on nearly every car.
That being said, the most high performance cars in the world (ie dedicated race cars that don't require a frunk for groceries) exit that air through the hood just a little forward of the SL-C opening, which is easily replicated with only minor body work that a builder could knock out on a nice afternoon.
 
All this aerodynamics, pressure, inlets, outlets and all is really great!

I wonder, for a street-driven car (and, as an American, I must say that I could never drive my SLC in triple-digit velocities on the street, no, sir!) how does one balance high-speed cooling needs with sitting for a long time, in traffic, on a 100-degree F, 90+% humidity day? Or, driving in very cold conditions? It can and does happen, driving on the street. Thanks!
 

Scott

Lifetime Supporter
seems you may have multiple answers based on what your powerplant consists of, and if this is a street car, or all out track car. One would be an aesthetics answer (open it up), the other is optimum performance answers (something different).

Terry, I think that there are multiple answers even when the powerplant and use case are the same. Whatever the approach, once you have a radiator you've locked into the area of the radiator core (A) and the radiator heat conduction coefficient (K) and you don't have any control over ambient temperature and the coolant temperature is the variable you're trying to optimize. The only remaining variable is mass airflow (M) which is driven by pressure differential across the radiator core. A diverging-converging duct is just one proven approach. For a SL-C air is going to diverge in front of radiator unless you significantly shrink the radiator or enlarge the inlet, so the only question is what to do with the air after it exits radiator.

IMO the biggest issue isn't cooling, it's aero balance, especially at high speeds. Like most cars this is only solved with a massive splitter (like Superlite's race splitter). This is especially true if you have a wing. Ducting the outlet helps. If you draw a straight line through the fins in the radiator core (mine are sloped forward 55 degrees which is close to the stock orientation) you'll note that air is directed to the underside of the nose which is going to cause lift. Some of the air is going to hit the vertical face of the foot box, battery, etc. before flowing out. This is why many SL-Cs have heat shielding on the exterior of the footbox.

Allan has built 20+ SL-Cs and he's added a La Ferrari-style radiator outlet to a handful of them.

1617298399357.png


As you can see the opening is moved significantly forward and the stock outlet has been closed. The slats hide most of the interior of the nose. In theory they reduce drag at the cost of downforce. There is no ducting underneath. He's made a few observations.
  • When standing at the nose of the car, he can feel airflow when the fans are on which doesn't happen with the stock nose. This is because the air isn't slamming into the underside of the nose before flowing out.
  • The foot box is much cooler. I don't think he adds heat shielding to the exterior of the foot box on cars with the modified outlet.
  • Engine temps seem to be a bit lower, but this is anecdotal compared to the previous two observations.
 

Scott

Lifetime Supporter
how does one balance high-speed cooling needs with sitting for a long time, in traffic, on a 100-degree F, 90+% humidity day?
You need a shroud with good fans while sitting in traffic. The shroud restricts airflow when the car is at speed so flaps should be added where appropriate.

Or, driving in very cold conditions?
The thermostat will manage engine temp. You could restrict airflow through the radiator to better optimize engine temp. I've seen people put cardboard over the core which will reduce cooling, but add a lot of drag. For a street car, I'd be focused on ensuring it's properly warmed up and once at temp I wouldn't worry about it.
 
Last edited:
All this aerodynamics, pressure, inlets, outlets and all is really great!

I wonder, for a street-driven car (and, as an American, I must say that I could never drive my SLC in triple-digit velocities on the street, no, sir!) how does one balance high-speed cooling needs with sitting for a long time, in traffic, on a 100-degree F, 90+% humidity day? Or, driving in very cold conditions? It can and does happen, driving on the street. Thanks!
it's pretty simple - street cars don't need to be 100% optimized for down force and drag, and can carry unrestricted horsepower to make up for any inefficiencies. So you put in more radiator than physics dictates, taking into account less than 100% efficient cooling. Add a thermostat and electric fans, and you've covered most of the bases, as long as the rest is reasonably close.
 

Howard Jones

Supporter
Over the next couple of months, I intend to properly duct the nose of my SLC track car based on everything we have learned so far. I started a couple of years ago with the radiator and its exit ducting. These two changes have made a significant difference in managing coolant temp so far but there is more progress to come with the inlet ducting.

Everything I have done so far has cost very little money, mostly being fiberglass material, with the exception of the radiator. I will post many pictures and describe in full the basis of my modifications.

Watch my build page.
 
Amazing thread, Howard has a nice exit duct set up, looking forward to see what improvements he does so I can follow suit. Thanks for all the technical info Scott, super helpful even though mine is a street car. Simple outlet modifications can help, especially here in the West Texas heat. I did maximize the radiator fan size . Now I will do the best I can with the radiator outlet, creating downforce and minimizing heat transfer to footbox are nice added benefits.
 
Back
Top