Mutable Solar Car Fairings- laminar flow airfoils

[smed]

Hello all,
I'm working on designing a solar car, and for our next car, we are investigating "mutable" fairings over the wheels, whose airfoil shape can be optimized depending on wind patterns.
Has anything like this been tried with models (or full-scale planes)? Right now we don't know where to start.
We need a good system to have a piece of the front of the airfoil move, (keeping the seam interface as small as possible), and we would have to analyze the tradeoffs using computers thoroughly.
If you have an general input as to airfoil choices that would also be greatly appreciated.

Thanks!

[Sparky Paul]

Look at the deformable multiple mirrors used in large astronomical telescopes..
These have many small mirrors computer controlled to form the focusing surface..Something like the acutators used could be used on the inside of your fairing to alter the camber and thickness depending on the Re at the time.

[BMatthews]

For the wind to provide a useable thrust it needs to flow around BOTH sides of an airfoiled shape, be it a wing or a sail. There was one car a few years back that had a basically raised flat shape with "struts" that housed the wheels. Is this the sort of car you're working on? If it's much boxier than that then I would suggest that the airflow isn't really going to be able to provide much thrust. Even with the wheel "struts" there's not much area and what there is is going to be fouled by the body. In the end it may not be worth the trouble for the little gained vs the reduction in performance brought about by the required controls and mechanisms.

For example, if the wheels point one way so the car goes where it's supposed to and the fairing around them points another way inorder to try to get some value out of the crosswind component then that would imply that the frontal area of the wheel fairing is going to need to be quite a bit fatter than the wheel. That frontal area is going to add drag. So it'll need to provide more value than cost. I suspect that'll be a tight bit of energy budgeting.

[biber]

There is basically no need to move the strut or parts of it to get thrust out of the airflow component perpendicular to the direction of drive. If your car is one of that flat type, that´s commonly used for such contests, the strut´s pushing(actually it´s more a pulling) effect should be big enough to be considered. Though the aspect ratio of the struts is very small, there are quite effective endplates preventing any tipvortices, called street and the vehicel´s (flat) body.
So the strut should end as close to the street´s surface as possible, and a strut section should be chosen that is not likely to stall at the angles of attack one is expecting to occur. Further should the chosen strut section be capable of delivering good "lift" per drag ratios at that AoAs.
There already have been built working windmills based on this effect.
http://en.wikipedia.org/wiki/Darrieus_wind_turbine

biber

[Tom Harper]

I second Bruce's comments on the wheel covers. Especially since the aspect ratio would be low creating induced drag.

However, for what it's worth:

A cambered section at it's zero Cl angle will generate forward thrust. Since the Cl is zero you don't have an induced drag vector.

Thick sections with round leading edges (Gottingen 434) exagerate this characteristic. Their maximum L/D occurs near the zero lift angle. None of them exhibit a negative Cd. However at this operating point the lift vector is forward. The effect was first noticed on JN4s. In a dive the forward rigging wires went slack.

These sections are not desirable for wings but they might have something to offer for bodies.

[biber]

A cambered section at it's zero Cl angle will generate forward thrust. Since the Cl is zero you don't have an induced drag vector. .Not exactly right. If there is no Cl, there can´t be any forward thrust. And there is no camber needed. See the link in my previous post. To get some things straight, one has to take in count, that the inuitive definitions of lift and drag/thrust applying to this example differ from the common aeroengineer´s definitions. To convert from one to the other definition, one has to transform the coordinate system. The aeroengineer´s reference (x-axis) to split the aerodynamic forces into drag and lift is the direction the airflow is coming from (at a far distant point where the flow is not disturbed by the presence of any objects). For this particular application the intuitive matching definition is refering to the direction of the car´s translatory motion relative to the ground (the x-axis of the car´s coordinate system).

biber

[Salto]

The Australian Geographic entry (which I co-designed and built) in the first World Solar Challenge had a high, thin wing to carry the solar array. Coming down to the ground from this there were two large spats at the rear to fair in the rear wheels and suspension components. These were symmetrical airfoils, about 12 inches thick and 6 feet long, oriented to the longtitudinal axis.

During pre-race scrutineering Paul MacCready (who had input on the General Motors entry) commented that he considered we should get useful forward thrust from these fairings in favourable cross-wind conditions. During some very windy days in the race we thought we could observe forward thrust, but this was at the expense of huge lateral loads that had to be taken on the skinny tyres. The rear of the car would slide about 1 to 2 feet out to one side in the gusts - not fun to watch!

Our design analysis had indicated that over the expected race wind conditions, adding surface area to make sails was not worth the extra skin drag. The car would mostly operate at very low yaw angles, so keeping the total drag down in this condition was the primary consideration. We needed the rear spats for other reasons, they were not there solely to produce thrust.

Adam, if you intend to use any sort of flexible material, or sliding seams to accomodate morphing the airfoil, I don't think you can count on any significant amount of laminar flow.

Have you looked at the expected wind conditions for the race? If you are travelling at any sort of competitive speed you will need a significant cross-wind to result in a useable yaw angle for generating thrust. As any sailer of fast skiffs will tell you, the relative wind (yaw angle) moves around massively as the boat accellerates and decellerates.

As a guide, sail boats will point up to roughly 30 Deg. off the wind. With a "proper" airfoil you can probably get thrust at a lower yaw angle, but how often will you encounter these sort of yaw angles in the race?

Graham.

[Sparky Paul]

The "mutable panels" could be used to adjust for the lowest drag....

[Salto]

The "mutable panels" could be used to adjust for the lowest drag....
Sparky, if this was in response to my comments, I'd add that there will always be a penalty for any added skin area. You may be able to morph the panels, and orient them to the flow to minimise the drag, but they will still add drag.

Well designed Solar Cars are composed of the smallest possible number of nicely streamlined shapes, and if the vehicle is set up for zero lift or downforce (minimum drag) it should have very little induced drag. So the total drag of the vehicle is highly dependent on the wetted skin area. For this reason it is vital to minimise skin area.

20 years of development has shown that the most successful designs enclose as much of the mechanicals/driver/electronics in one simple shape with the minimum of wheels spats and other stuff added to that.

Adam's idea of obtaining forward thrust from the wheel fairings deserves investigation, but I'd start by looking very carefully at the trade-offs. These include added tyre drag from the considerable sideforce, added wetted area, added weight, added complexity/reliability, added design/development time, etc.

Adam, please let us know how the project develops. Have you studied the winning Dutch entry in the just completed World Solar Challenge here down under? It's a beautifully simple design, with lots of attention to aerodynamic detail (and no sails). I think I'm correct in stating that it was the first to average over 100kph for the race.

Graham.

[Sparky Paul]

Graham, my buddy in Florida, Mujahid Abdulrahim,is working on morphing wings..
Seems to me if vertical surfaces are needed for normal streamlining, it might be possible to trim the surfaces when minimizing the effects of off-axis wind..
Cambering the downwind side as needed, perhaps..
http://news.ufl.edu/2005/08/23/morphing-planes/
Sparky

[Salto]

Sparky,
Yes, I agree. If you have a wheel spat it could be moved to align with the relative wind, and therefore minimise the drag.

If you're only concerned with minimising drag there is no advantage in changing it's camber though. It will always have lowest drag with zero, or close to zero camber.

Adding camber will only potentially give a benefit if you're looking to produce forward thrust, and as I've said above, I'm sceptical about whether chasing thrust from side-wind will genuinly give a net benefit.

Graham.

[smed]

Wow! I never expected such a large response. Thank you all very much for your input.

First of all, yes, our car is the kind that is pretty much an airfoil on wheels with a canopy.

Right now we are just trying to get an idea as to if this is worth spending a lot of energy looking into. We have a lot of time to design and build our next car, so we are giving it a try.
I agree, Salto, that this may end up proving to have too many complications to be worthwhile overall.

We have carefully studied the Dutch Nuna car and we have come to the conclusion that our car has a lower aerodynamic drag but that the main reason for their record time was their solar cells: they are the only team that had real, satellite-grade cells. In fact, their design makes some aerodynamic tradeoffs in order to maximize solar array area. That is very impressive that you co-designed and built a solar car. What a small world it is! I will try to find some pictures of your design.

Sparky- I’m not sure what you mean by “The "mutable panels" could be used to adjust for the lowest drag....”
Do you mean that we should keep them in the shape of the lowest drag for most of the time, except where we would like to get thrust from the cross-winds?

Salto- why would a flexible, yet stiff material lose its laminar flow? (I understand why a seam would trip the flow). I see the problem of the lateral loads. Instability of a 550 lb car traveling at 70+ mph can be very dangerous.

Thanks for the ideas on making the fairings morph mechanically; I will look into them all thoroughly.

[Sparky Paul]

Sparky- I’m not sure what you mean by “The "mutable panels" could be used to adjust for the lowest drag....”
Do you mean that we should keep them in the shape of the lowest drag for most of the time, except where we would like to get thrust from the cross-winds?
.
Smed,, yes... some sort of pressure sensing to equalize the pressure on each side of the fairing.. probably total pressure probes on each side that would sense a pressure difference between the sides and control the shaping or offset of the panels autonomously.

[Salto]

Salto- why would a flexible, yet stiff material lose its laminar flow?
Because any surface with ripples, waves, bumps, roughness, etc. would not maintain laminar flow for much of the chord. You'd also need some clever flexible or sliding interface to the underside of the wing to stop this intersection upsetting the flow.

Have you measured how much laminar flow you get on top of the current car? And underneath?

Nice car BTW.

Graham.

[Purdue Aero Man]

A highly iteresting thread. I used the IE text search and noticed that nobody has brought up friction. Bearing friction, rolling friction, etc... What would the ability to morth the shapes of the car weigh? Would the reduction in profile drag be worth the added weight and complexity? Would the increase in the wetted area negate it? Forward lift by morphing? Sure, when you're using a large sail, or a combination of sails. For a small light-weight solar car, I can't see any real benefits to what you're looking for. But you're thinking, and that's reassuring. :)

[Salto]

Purdue Aero Man,

As I mentioned in a previous post, I agree that the increase in tyre drag (rolling resistance) of a side loaded tyre needs to be carefully evaluated. This alone, may kill the idea of using sails to produce forward thrust.

With these vehicles, the energy loss at cruising speed is dominated by aerodynamic drag. So the balance of what weight penalty you can suffer for an improvement in aerodynamics swings heavily towards adding weight. The tyres used will have extremely low rolling resistance, so adding weight consumes surprisingly little extra energy.

Graham.

[smed]

Hey guys, just thought I'd update on the status of this project:

I recently conducted a wind tunnel test at my university. At V=30m/s and with a full-scale "mutable" fairing angled at varying angles into the air flow, the mutable fairing had more drag in all cases. This contradicts CFD analysis that I conducted and news of positive results of a similar wind tunnel test that I heard of through the grapevine.

However, there are some questions as to the validity of my test: namely, seams in the fairing may have affected the results, as they were not as minimal as desired and draggy vortexes may have been created because of the implementation of the fairing's deformation. Lastly, the fixture in the wind tunnel contributed to the majority of the overall drag, hindering satisfactory analysis of the results. If I have time this summer, I'd like to machine foam fairings that have what I decide is an "ideal" shape and mount them on a smaller fixture. This should help to determine if the results reflected simply a problem with the fairing model or a problem with the fundamental aerodynamics theory.

[biber]

Are there any pics available of that full-scale "mutable" fairing you observed in the wind tunnel?
I'm curious how it looks like and why you seem to be unhappy with their implementation
(except for the discourageing wind tunnel data it produced, of course).

biber