Hatty

I am building a solar car and have a question about an aerodynamic detail we want to add to it. In the rear corners of the car someone suggested placing a round on it to reduce the votecies. I was wondering how much of an aerodynamic improvement it will give us. Would it be better to add tiny winglets on it (which would have to be mounted perfectly vertical due to a size restriction)? Ive never seen the rounded tip on an aircraft before and thought it might be a good idea if this works.

allanflowers

It is hard to know exactly what you mean by rounding the corners of the car's rear. If it is like the round corners on a wing tip (like a Spitfire?) then maybe it would be better. If you are rounding off the forms of a body shape then it may the worst thing to do.
I have seen vortex generators added to the rear side surfaces to give more energy to the air and bring it around a round form farther before it separates but that is hard to design unless you have a windtunnel with an accurate balance to determine tiny increments of drag.
Although I no longer work for General Motors, I would recommend you do a Google Image search on the GM Sunraycer solar car. It doesn't get much slicker that that.
Allan

Hatty

Here is a picture. The rounds are to the right of the picture. They are not tangent to the rear, but are tangent to the sides. The top and bottom will meet at a point all of the way around the rear.

It is like the Spitfire, but like I said above, the curve meets tangent with the sides and not with the rear. I hope I described it enough.

Tall Paul

The winglets are a better deal than merely rounding off the edges.

mulligan

Guys, he did mention it was a SOLAR car. What is the anticipated speed of the car? If 20-30 mph, I would guess that rounded corners will maintain laminar flow longer and be better for drag. If it moves at significantly higher speeds, where you will get greater flow separation induced wave drag, then the winglets may be better to reduce the vortices and hence wave drag.

BMatthews

Without a wind tunnel or some real world model work using tufts and smoke and stuff it's all a crap shoot. It might help and it might hurt. So much depends on the rest of the shape and if it's got ground control or is set up to slide over the ground air, etc, etc.

By the time the air get back there it's most likely turbulent flow anyhow. As long as you don't have a big blunt or radically curved area then it doesn't reallly make much difference.

allanflowers

Hatty, how about another picture or two. A side view and maybe a perspective rear 3/4 would be very informative.
On the boundary flow issue, it probably WILL go turbulent by the time it gets back there but that is good because you will then have attached flow farther around the curved surface - before it finally separates and forms a wake area. A turbulent boundary layer is good on rear corners.
I would bet that this is for some kind of academic competition in which case you want to be better than the other guys, if only a tiny bit.
Allan

Hatty

Hey, here's our website: http://www.engr.uky.edu/solarcar/

Under Gato del Sol II, you will find construction pictures that will give you more info than the CAD pics that I have saved. I know that the air is going to be turbulent by the time it reaches the rear (and you can verify that). Based on what everyone has said, it couldn't hurt to put the rounds on the rear of the car. We don't have the expertise to run CFD on the design, so, we just stick with standard aerofoils and hope for the best. However, since this is our 2nd car (compared to other's 8th), we are still learning and growing. But I will make sure that CFD is done on the next one.

The car will have a top speed of about 65mph, but the average speed for the 2400 mile journey will be about 30 mph.

Hatty

allanflowers

The pictures are a big help in understanding the forms. I would suggest restricting the rear taper in plan view to 15 degrees, rather than the almost 30 that you showed in your earlier post. Also the side is too flat and the radii along the edges are rather sharp. If they were rounder it would be better.
I don't understand the basic section which looks like an undercambered airfoil which is good for lift but you don't want this to lift. Lift is drag. Same thing goes for excessive downforce since cornering is not a big issue like on a Indy car. I assume that the airfoil is probably pitched down to not make lift? Is that correct?
Another issue is yaw stability. If this thing "hunts" all over it will waste energy to keep it on course. It depends on where your cg is relative to the side shape. With a forward cg your shape is okay but if it is in the middle or toward the rear, you might want to increase the flat area in the rear wheel pants to give it a little more crosswind stability. You might even consider some kind of adjustable rudder surfaces on the backs of the rear pants to compensate for side winds situations.
I have spent hundreds of hours in wind tunnels doing automotive testing so that's where I'm coming from.
Allan

rmh

Going wayyyyy back -- Jim Hall did a Chaparrel (?) Chev design which used a small fan at the rear , to reduce any pressure under the car to zero- Of course it was promptly outlawed but at the time he got my attention as a real thinker.
On the Bonneville racers ( just down the road), the exercize is to reduce drag to bare minimums with -if anything - a bit of downforce.
To me - the electric would do best if weight was first /second and third in importance.
Then a rather flat pumpkin seed shape ( (two blended clamshells) which literally floated (sprung) on the rear wheel (s), allowing any vertical drag imbalance to self align with airflow.
too weird?
This machine is not going to go fast so I would think that any additional area -used ONLY to streamline - would simply be negated by the added weight.
And in this kind of contest - weight is the name of the game .

Tall Paul

Allan is correct about lift. You don't want or need any.
Symmetrical is probably the better choice.

Hatty

Good point. The car does look like it has a lot of camber, and it does. about 4 or 5 (I forgot the exact number) percent with a symmetrical NACA airfoil. And you guesses correctly when you assumed that we were going to give it a negative AOA. By doing this, we get more room for the driver and suspension without having to thicken the airfoil.

As for decreasing the degree of the taper, that is what I decided on doing tonight. I mainly did that in order to make things fit better, so being that it also might decrease the drag, that is a double plus.

The center of gravity in the car is far forward. It will end up putting about 1/3 of the weight of the car on each wheel.

As for the rounds on the corners of the car, we cannot increase the radius on the top part of the car because of the solar cells. We could, however, increase the radius on the side part of the car. Would that make a big difference?

Aerodynamics, what a complex subject. [sm=confused.gif]

And for what you said Dick, weight is very importaint for solar racing. During the shell construction, I would have to yell at people for using too much glue. There is no point in making your car weigh a pound more when it doesn't have to. But I try to get all of the aerodynamic benefits in whenever possible. This is one of those simple changes.

Hatty

Ben Lanterman

What an interesting project - they never had these kinds of things when I had to walk backwards through the blizzards at Purdue. Gripe, Gripe.

All are good comments. I do have to wonder what is the tradeoff of rolling friction vs aero drag of the body shape. Would it be better to put tip plates along the side and improve the body lift and reduce rolling friction?? I don't know but I assume you have done the calculations on that.

If the airfoil could be perfectly contoured a slightly cambered rear section (top and bottom) might be lower drag at zero lift - but with the requirements of solar cells on the top and a place to look out of with wheel fairings and the like, it pretty well eliminates that possibility.

There the more perfectly streamlined shape you can geat the better. Would a added fairing at the nose to get a little more pointy be appropriate at these "airspeeds"?

Looking at the first version the most of the drag is in the cockpit (really barfy) and the wheel/undercarriage fairings (awful). The body looks OK although a slightly more teardrop as you are going to might be better. I would have to look at data which you have probably already done. It might be worth a pound or two of structural weight to minimize the bulk of the wheel fairings. They look bad.

allanflowers

I didn't realize it was a three-wheeler. I pity the drivers because, if the roads are rough, they will hit a lot more bumps. You just can't steer to avoid bumps with a three wheeler. You need a good seat and head room too.
With the requirement on solar cells determining the side radii, you can't do much. Rounding the bottom edge wouldn't hurt.
Allan

BMatthews

Just a thought but I would think the amount of camber and direction of the camber would be mostly determined by the interaction with the airflow between the belly and the road.

Beyond that some wool tuft testing along the sides would show if there is any serious air movement from under to over. Any such movement would suggest that the pressures are not even above and below and that extra drag would be the result. Modifying the shape to achieve as neutral a pressure difference along the whole length of the vehicle would be best I would think. That way it won't matter if the corners are sharp or rounded.

This is just a guess but I suspect the final ideal shape of the upper side would be a nice low camber curve much like you already have but that the lower side may well end up as an oddball hill and valley deal to conform with the pressure wave buildup and release caused by the interaction with the road. Clearly full side skirts would help to isolate the upper and lower flows and the resulting flow through the lower "duct" could be dealt with as a separate issue. Or if such a duct shape could be made to compress the air and release it such that there was little or no drag involved with the lower flow?

With the open nature of your design side winds will be a significant factor but an open side won't try to steer the car like a closed side would. Whether or not one is better than the other from a drag standpoint is up to some measurements to determine. Obviously the open side option would be less work for the driver though.

So many questions and me without my windtunnel..... Either way you have a very interesting project going on there. Best of luck to the team in the competition.

Jimmbbo

Seems that there are three paths to success - maximum efficiency of the power and drive system, minimum weight and minimum aerodynamic drag (DOH!).

Minimum drag would best be determined in a wind tunnel, but excellent information can be obtained (cheaply! ) by tufting one half of the car (assuming bilateral symmetry) and seeing what the tufts tell you. Anything other than straight, smooth flow means drag is being produced, and must be eliminated...

Cheers!

Jim

Hatty

Thank you for your suggestion. I was wanting to do some testing along those lines, but I didn't really know where to begin. That would be an importaint thing to measure and we will do that, even if we cannot examine the whole car.

As for testing with tufts, are there any good places to go to look for information, or is it as straight-forward as it seems? In other words, do I just tape them onto the car and go with the flow (yes, yes that pun was intended ), or is there more thought involved?

I would like to thank everyone for their suggestions. I love aerodynamics, but have not had any proper education in the field, so your suggestions have helped me understand a lot.

Thanks again,
Hatty

allanflowers

I think the best thing you could do is to allow for some adjustment of the angle and height at which your car runs. Even a little bit of adjustment would be worth while. Then if you can figure out some kind of valid "coast down" test where you measure the speed of the vehicle at various points in the coast down, you may be able to determine if one body angle is better than another with respect to drag. The problem with coast down testing is there are so many variables to eliminate like wind, etc. You just have to run a bunch of tests before you know anything. I have never been involved in coast down testing because the tunnels my company used had sophisticated balances to measure any forces.
Tuffs can certainly be taped onto the form (light weight yarn and magic mending tape) to look for problem areas in the boundary layer. Interpretation of the tuffs is a little difficult unless there is a definite flow reversal or whipping around of the tuffs.
Also, as has been mentioned, WEIGHT is extremely important. Also any loss of energy due to rolling resistance or trying to maintain a straight path is important.
Knowing now that your cg is well forward - due to the 3-wheel config - should eliminate any major concerns about sidewind stability.
Using moped tires should give you some reliability but as you increase the tire pressure you will experience lower rolling resistance and.. the likelyhood of a tire blowing out. This may be worth doing some testing beforehand on this issue; you could do it on another vehicle well before your main-machine is ready to go.
Allan

rmh

So - no value in using a shape which could self center the drag?
Seems to me to be a simple way to equalize/reduce drag.
as for fairings - I would think - chuck em - go with plate wheels like the pedal bike guys use. Just leave room for the wheels to rotate.
Reduce frontal area as much as possible

BTW--- three wheelers -two up front -one in the rear is far more stable - done right than almost any other setup .
You can decelerate into a turn without unloading rear inside wheel
Among my past toys (1962) was an A100 AH roadster with a highly modified 265 " Chev V8.
You learn about handling --fast -- with these setups -It was better than my TR3B-but what wasn't -- Mine was a bit of a squirrel tho- simply because I didn't have enough rubber under it .
My last toy - a Volvo with a Ford HO Mustang - handled better ---tho not as fast --
But the Po leece also never gave it a second glance.

allanflowers

Dick,
I don't know if the self leveling body would reduce drag or not. I guess it makes sense that it would "find" the best angle. How would one structure such a body without adding weight though? The body couldn't carry the chassis load so there would be some duplication of structure, unless I am overlooking something.
I once was told by a noted aerodynamicist, Dr. Bettis, that the disc type wheels do create more drag than fairings. If you look at airplanes (everybody remember them?) the WW I planes used disc type wheels but, as the speeds went up in the 30s, it was all fairings/wheelpants.
Another student and I built a three wheeler in college for a magazine article in Pop. Mechanics. It had the two wheels in front, was tremendous fun to drive but too high off the ground. It would lift the inside wheel if you pushed it and, looking back. was very dangerous. They had a body design contest and the "winner" got the thing.
Allan

rmh

The sprung setup -as I see it would have a pneumatic shock at the rear - which could be regulated .
I would circle track test the car to prove which setting drew least amps . The rear wheel (driver) would be on a long swing arm-ala dirt bike setup .
The car would be totally sprung except for rear arm and a solid tube front axle suspension.
these setups as simple as they are , are good flat track setups
I did some carting in the 1960's also - and the flat chassis Moss kart and subsequent electric model cars convinced me that simple light minimum setups are the quick way -

I don't see how wheel fairings -at speeds -- under say 60 mph could beat the losses incurred from increased cross section and wetted area.
I would put my money in this dogfight on minimum cross sections and lowest weights.
The driver would get to recline or lay on his belly.
I would love to have time and a few bucks to do one of these.

Hatty

Now, remember that the fairings are designed to snugly right around the brake and wheel assembly. The Gato del Sol 1 has fairings that are large due to the chassis coming down to the wheel instead of the suspension coming down to the wheel. Check out some of the pictures at www.americansolarchallenge.org. The fairings on the cars are as thin as they can get. They are designed to open when the car turns, so you can get away with having the fairing run right along the wheel. This is how the fairings on the Gato del Sol 2 will work.

As for the small cross section, that is the way many cars are designed. They look like they could slice you in two if you jumped in front of one of them. It is really hard to explain until you go and look at some of the pictures. Also, while you are there, check out the regulations. It will give you an idea on the restrictions these cars face.

allanflowers

As far as fairings are concerned, they are much better than a round cross section, like a tire or cable.
A 1" pipe section has the SAME drag as the 10" airfoil section. This is why a bare wheel, even with skirting on it, is not as good. This is why the wires on old planes are so draggy.
Please pardon my ratty sketch.
Allan

rmh

I understand - but the wheel/tire combo is a disc not a round section .
The fineness ratio is not per the 10-1 .-but should be pretty clean -depending on how it intersects with the shell.
I will have to take a look at the rules .
My idea of an ideal fairing -would be sorta like the fuselage fairing on a vertical fin .
Not like the wheel pants on an old Waco -
I have done lots of plugs/glass molds -so I don't think the mechanics of making the slim transitions would be too bad.

BMatthews

Hatty, the wool or string tuft thing is as simple as it sounds. You tape on a whole pattern of light constrasting color fibers and then have someone drive alongside and video or film the action of the tufts. The patterns show very clearly local flow direction and turbulence levels. Smoothly laying fibers are most likely in a laminar area and dancing fibers are being subjected to turbulence. If they all angle up or down you have flow transitioning from one side to the other. Changing the trim angle can then be attempted to help minimize or nuetralize the upper to lower flow.

Incidentally, what about some onboard water manometers to help find high and low pressure zones? Here again the trim angle of the car could be adjusted to try to minimize any adverse pressures.