My new way of measuring on-road drag

[Yippeekyaa]

Awesome information! Just scrolled the all the possible obd pids and found the map psi and the vehicle load % pids. Set up a screen showing those, plus speed, rpm and throttle position command value and will experiment with it. Fingers crossed that there is a usable method to make the throttle stop testing method work. Will be over a week before I have results as am leaving in a couple days for a short trip in my old RV.

[JulianEdgar]

Originally Posted by JulianEdgar

And now I have tested on a day with gusting 25-30 km/h crosswinds. And, in those conditions, the Edgarwits do absolutely nothing to reduce drag. The throttle-stop speeds with and without the aerofoils is identical.

So that behoves the question: do any air curtains work in gusty crosswinds? I have never seen any published research on this, so I have asked an expert I know. It will be interesting to see what he says.

The expert got back to me (he is head of aero at a major car company and had already given me feedback on the throttle-stop drag testing method).

I first asked him if the window up / window down results I was getting in different wind conditions were reasonable. He said:

Your window-opening results look plausible to me (not that there’s much data around).

I then asked him about the Edgarwits air curtains not appearing to work in crosswinds. He said:

I like your “air curtain”. I’d call it a turning vane, as it’s external. Looks similar in concept at least to something I experimented with a while back. I think your results are about right for that kind of system. It’s always going to be a challenge at yaw as you’ll get separation off the leading edge of the vane. I’d think it should be good to 5 degrees or so. The interesting thing here is given environmental conditions/vehicle operation profile is it a benefit or detriment? Only a long-term evaluation could really say.

Our internally ducted (behind the front fender) air curtains tend to do better at yaw than straight-ahead.

He went on:

With these kinds of intervention, the benefit always depends on the place you start from: the better the front corner design and its integration with the wheel arch, the less scope for benefit.

A device through or over which the flow passes always has a local drag of its own (device drag). It’s key to minimise that. With your vane arrangement it might be worth looking at whether the internal channel should expand (diffusing), contract or remain constant. A diffusing flow will give the lowest device drag due to pressure recovery. A contracting channel will accelerate the flow and give more downstream effect. The question is: where does the best balance lie?

On the gusty element: vehicle aerodynamics as practised is overwhelmingly (/disappointingly) about performance at zero yaw in low-turbulence flow. The only OEM I know that use “gusty” onset flow conditions for drag development is [redacted]. They do this through their CFD models. The heavy truck industry tend to do a bit better (because they have to) in this regard and use at least an approximation to the wind-averaged drag (a function of range of static yaw angles).

And he was nice enough to finish off:

Anyway, interesting work and results that make sense.

[freebeard]

Originally Posted by The expert

A device through or over which the flow passes always has a local drag of its own (device drag). It’s key to minimise that. With your vane arrangement it might be worth looking at whether the internal channel should expand (diffusing), contract or remain constant. A diffusing flow will give the lowest device drag due to pressure recovery. A contracting channel will accelerate the flow and give more downstream effect. The question is: where does the best balance lie?

I'll jump right on those coat-tails.

Converging vs diverging? You pointed to the fences on the openings of the Porsche Cayenne in some thread. Wasn't that for yaw angles?

A simple experiment might be to invert (and swap) the airfoil extrusions, and compare the results.

[JulianEdgar]

Originally Posted by freebeard

I'll jump right on those coat-tails.

A simple experiment might be to invert (and swap) the airfoil extrusions, and compare the results.

Good idea; I hadn't thought of that.

Edit: Not so easy. Looking at them, I'd have to make and attach new end plates to get the geometry right.

[freebeard]

Once the geometry is right you can make bead-rolled end plates to fillet the top and bottom corners.

[JulianEdgar]

Edgarwit with my 'eroding clay' visualization technique.

Separation line marked on outside of aerofoil:



Vortex generated by sharp leading edge of endplate:



Completely clean inside surface of aerofoil:



Lack of removal of clay on bumper:



Testing done in windy conditions.

[j-c-c]

The "expert" is correcting the terminology here I see. In that the device in use is effectively a "wing" designed to induce differential pressure, being used vertically when differential pressure has little use, which only adds to drag, size, weight, complexity, etc. An infinitely thin correctly configured "vane" would reduce all these issues, as I see it, what say the "expert"?

I see he included the description "constant" (regarding crossection?), that would be my starting place.

[JulianEdgar]

Originally Posted by j-c-c

The "expert" is correcting the terminology here I see. In that the device in use is effectively a "wing" designed to induce differential pressure, being used vertically when differential pressure has little use, which only adds to drag, size, weight, complexity, etc. An infinitely thin correctly configured "vane" would reduce all these issues, as I see it, what say the "expert"?

I see he included the description "constant" (regarding crossection?), that would be my starting place.

The beauty of this world is that we can all try out different approaches and then share the results. If holding a constant cross-section and not using an aerofoil are your preferred approach, why not do both those things and tell us how you go?

[j-c-c]

I will, and while I do my pre test design, I will keep a close eye on F1 plethora of vertical "turning vanes" they use of mostly constant crossections.

[JulianEdgar]

Great! I look forward to seeing what results you get.