[xiziz]

In a world where everything can be googled, I found something I could not google. Why are all aerodynamic things (Submarines, Aircraft, Raindrops etc) shaped with a blunt front and a long tapering rear end but roofboxes are ALWAYS shaped the opposite way, pointy end first and blunt end back.

And when I want to know something the Internet cant tell me, I will go and find it out for myself! (Hopefully making a few fellow motorists look up from their phones and laugh, not knowing my agenda)

I'm figuring its one of two things,
1. Fluid dynamic interaction with the airflow around the car. (Ie its better with pointy end first on cars because of the way the air interacts with both objects)

2. Roofbox designers are vain and design comes before function.

So I set out to test it, results of my 30km test run was inconclusive(no difference in consumption, subjectively less wind noise with it backwards tough) so tomorrow and sunday are 110km test runs(to work and back home again, once with each setup).

I'll be back

[Dazzler]

That is an interesting exercise! I think it would look better with the skinny end at the front...

[Shambles]

It might look better with the pointy bit forwards, but for aerodynamics check this out...

(click to show/hide)

[The Gonz]

The trailing taper exists to minimise turbulence of the laminar flow leaving the box, which is deemed the major drag factor at speed. A fully comprehensive test would include trying it upside down and at a range of angles of attack but your mounting would have to get creative. Looking forward to your outcome.

[Dazzler]

Yeah, I did think of those helmets too when I saw xiziz's picture. 

[mickd]

The trailing taper exists to minimise turbulence of the laminar flow leaving the box, which is deemed the major drag factor at speed. A fully comprehensive test would include trying it upside down and at a range of angles of attack but your mounting would have to get creative. Looking forward to your outcome.

I was going to say check with the wingman , but as shown reduces drag and creates a little lift,  (same shape & theory as  wing) , reversed would create a little downforce .

[The Gonz]

The trailing taper exists to minimise turbulence of the laminar flow leaving the box, which is deemed the major drag factor at speed. A fully comprehensive test would include trying it upside down and at a range of angles of attack but your mounting would have to get creative. Looking forward to your outcome.

I was going to say check with the wingman , but as shown reduces drag and creates a little lift,  (same shape & theory as  wing) , reversed would create a little downforce .

Indeed, Mick. The unwashed would call them spoilers while the technically inclined would accurately identify them as wings.

[mickd]

The trailing taper exists to minimise turbulence of the laminar flow leaving the box, which is deemed the major drag factor at speed. A fully comprehensive test would include trying it upside down and at a range of angles of attack but your mounting would have to get creative. Looking forward to your outcome.

I was going to say check with the wingman , but as shown reduces drag and creates a little lift,  (same shape & theory as  wing) , reversed would create a little downforce .

Indeed, Mick. The unwashed would call them spoilers while the technically inclined would accurately identify them as wings.

" the unwashed " 

[xiziz]

Seems someone at the metrology office wants to mess with me, forecast changed from dry to 30mm of rain tomorrow, so might have to postpone one of the runs.

Looks really wacky with it on backside first and the boot does not open enough, so it's mostly for fun by now.

[mickd]

Seems someone at the metrology office wants to mess with me, forecast changed from dry to 30mm of rain tomorrow, so might have to postpone one of the runs.

Looks really wacky with it on backside first and the boot does not open enough, so it's mostly for fun by now.

30mm of rain, would be a good test   

[xiziz]

Yeah, turns out I missed the rain, driving down to work they consumed exactly the same amount, 25% or ~7kWh as compared to my average commuting consumption of 21%(5.88kWh). Returning home however yesterday was significantly more consuming using a whole 30% (8.4kWh) than the 23% (6.44kWh) of running with the box backwards.

Thus I don't know what to make of it, as they did the same efficiency going to work, but very different coming home again. I need to do a more controlled, 110kph test on the same day over a fairly long distance.

This weekends results seem to indicate:

Average commute in august, control(nothing on the roof): 115 Wh/km
Roofbox (235cm*90cm*35cm) backwards: 127Wh/km (+10% more than control)
Roofbox correctly mounted forwards: 144Wh/km (+25% more than control or +13% more than with it on backwards)

I did do a short stint of a few km 110kph both runs, they both seem to indicate 180-200Wh/km at that speed (reference is 144Wh/km with bare roof) so +25-40%. Same consumption figures for both directions of box. So my expectation would be that there is no major difference in consumption, just that it gets more sensitive to weather(wind, temp, denser/more humid air) with a box on which is skewing the results.

(I still need to get a good reference consumption at 110kph so I can plan longer journeys with the box on. Mounting it backwards severely restricts the amount one can open the boot hatch, but even facing forwards they intersect so need to see if I can move it a few cm further forward to gain some clearance)

[AlanHo]

Another way of crudely comparing drag is to find a very long and fairly steep downhill straight road on a dry day with no wind.

Start the test at a set speed - say dead on 40 kph - and let the car roll down the hill acceleating under gravity with the gear in neutral. See what maximum speed it achieves.

Reverse the roof box and repeat the exercise to see which configuartion had the least wind resistance by comparing maximum speeds.

Daft idea innit - but it might work.

[Surferdude]

Going back to the op.

Why?

Well, I don't have a roof box but I do carry surfboards.  I'm always aware of two wind forces on the boards which stick out two or three feet in front of the top of the widescreen .
I tie them on deck down because the nose area is designed with a certain amount of lift in it.
Wind comes over the bonnet and up the windscreen area to hit the board which tends to want to rip it up and away. The slight downward angle of the nose tends to negate this by pushing downwards due to the airstream hitting it from straight ahead.
I would think the same principle applies to the roof box.
All the other examples you quote, like subs or maybe airplane wings are only so next to one direction of fluid flow.
The roof box or surfboard has two. One from over the bonnet and up the screen and the other from straight in front of it.

[The Gonz]

Another tip on testing aerodynamics is to test at high speed. Anything below 80kph is largely inconclusive.