Fuel Economy, Hypermiling, EcoModding News and Forum - EcoModder.com - View Single Post

JulianEdgar · 01-09-2021, 11:59 PM

One thing I often notice is that people seem to want to make the subject of car aero modification more complex that it needs to be.

I am sure that this puts off lots of people, because they think that they need to know all of that stuff - stuff that actually doesn't matter much, if at all - in achieving good car modification results.

Here are some:

Turbulent versus laminar boundary layers

First, the boundary layer is just the 'skin' of air nearest the car's body that is flowing slower than the speed the car is moving forward. On a car, the boundary layer is basically all in the form of what is called a turbulent boundary layer. So forget laminar boundary layers as being relevant on any normal road car. (Exceptions? Maybe the first 300mm of car hood, and maybe the mirror housings.)

Don't confuse turbulent boundary layers with separated flow - the latter is where the car's body shape is no longer guiding the flow - a completely different idea.

Reynold's numbers

In effect, this is a number that allows us to compare airflow characteristics (drag / lift) of objects of different size. So, for example, if we were to use a 1:5 model in a wind tunnel, we'd have to ensure that the Reynold's Number was in the same ballpark as it would be for a full size car. (To achieve that, we need to use a proportionally higher airflow speed, ie in this case, 5 times the speed.) But if we're using our full-size car for our on-road testing, Reynold's Numbers are irrelevant - we're already dealing with the full-size model!

Fineness ratio

This relates to how long the object is compared with its greatest frontal projected cross-sectional area. It's very important on objects like tear drop shapes, because at some stage, the smaller wake created by a long tear drop starts to be more than offset by the frictional drag of airflow shearing over the object i.e. frictional drag. (The shearing coming from the different airflow speeds in the boundary layer.)

But if you're modifying your car, the fineness ratio is basically irrelevant - any tapering length you can realistically add at the back to give a smaller wake is going to be good.

Drag coefficient

If you are modifying your car, a drag coefficient (Cd) value is of very limited use. Almost, I'd suggest, relevant only in that if you're starting with a car with a high factory Cd figure, it will be easier to make improvements. That's basically it. Calculating Cd values resulting from your modifications is largely worthless - there is no really accurate way of doing so without access to a full-size wind tunnel with expert staff.

But - and here's the rub - concentrating on Cd can lead you down all sorts of rabbit holes. "That car reduced Cd by 6 per cent with [and pick your modification here]." That doesn't mean that particular modification will do the same on your car, and often nor does it even mean that the original modification reduced Cd by that percentage on that car. (So much published data is unreliable - not all tech papers are equal!)

Furthermore, applying that Cd reduction (6 per cent in this case) to your car's original Cd and saying that, for example, "I've done that modification and so I've now reduced my Cd from 0.28 to 0.2632" has got absolute knobs on it. Why? Well, there's no guarantee that the mod works on your car like it did on the other car (if it did do so even on that other car). Then there's the intellectual and engineering paucity of going from two decimal places to four places - resolution and 'accuracy' where there are actually none!

That will do for now, but I am sure I'll think of some others.

01-09-2021, 11:59 PM	#1 (permalink)
JulianEdgar Banned Join Date: Nov 2017 Location: Australia Posts: 2,060 Thanks: 107 Thanked 1,605 Times in 1,136 Posts	Stuff to not worry about One thing I often notice is that people seem to want to make the subject of car aero modification more complex that it needs to be. I am sure that this puts off lots of people, because they think that they need to know all of that stuff - stuff that actually doesn't matter much, if at all - in achieving good car modification results. Here are some: Turbulent versus laminar boundary layers First, the boundary layer is just the 'skin' of air nearest the car's body that is flowing slower than the speed the car is moving forward. On a car, the boundary layer is basically all in the form of what is called a turbulent boundary layer. So forget laminar boundary layers as being relevant on any normal road car. (Exceptions? Maybe the first 300mm of car hood, and maybe the mirror housings.) Don't confuse turbulent boundary layers with separated flow - the latter is where the car's body shape is no longer guiding the flow - a completely different idea. Reynold's numbers In effect, this is a number that allows us to compare airflow characteristics (drag / lift) of objects of different size. So, for example, if we were to use a 1:5 model in a wind tunnel, we'd have to ensure that the Reynold's Number was in the same ballpark as it would be for a full size car. (To achieve that, we need to use a proportionally higher airflow speed, ie in this case, 5 times the speed.) But if we're using our full-size car for our on-road testing, Reynold's Numbers are irrelevant - we're already dealing with the full-size model! Fineness ratio This relates to how long the object is compared with its greatest frontal projected cross-sectional area. It's very important on objects like tear drop shapes, because at some stage, the smaller wake created by a long tear drop starts to be more than offset by the frictional drag of airflow shearing over the object i.e. frictional drag. (The shearing coming from the different airflow speeds in the boundary layer.) But if you're modifying your car, the fineness ratio is basically irrelevant - any tapering length you can realistically add at the back to give a smaller wake is going to be good. Drag coefficient If you are modifying your car, a drag coefficient (Cd) value is of very limited use. Almost, I'd suggest, relevant only in that if you're starting with a car with a high factory Cd figure, it will be easier to make improvements. That's basically it. Calculating Cd values resulting from your modifications is largely worthless - there is no really accurate way of doing so without access to a full-size wind tunnel with expert staff. But - and here's the rub - concentrating on Cd can lead you down all sorts of rabbit holes. "That car reduced Cd by 6 per cent with [and pick your modification here]." That doesn't mean that particular modification will do the same on your car, and often nor does it even mean that the original modification reduced Cd by that percentage on that car. (So much published data is unreliable - not all tech papers are equal!) Furthermore, applying that Cd reduction (6 per cent in this case) to your car's original Cd and saying that, for example, "I've done that modification and so I've now reduced my Cd from 0.28 to 0.2632" has got absolute knobs on it. Why? Well, there's no guarantee that the mod works on your car like it did on the other car (if it did do so even on that other car). Then there's the intellectual and engineering paucity of going from two decimal places to four places - resolution and 'accuracy' where there are actually none! That will do for now, but I am sure I'll think of some others.