Frisbee-circular plate aerodynamics

It would seem you guys can do some good computer stuff re: frisbees.
But as this is a boat design site it would be great if someone wanted to do the math on the attached.
I think this boat might have a L/D ratio of 20:1 or better? Twice that of a hovercraft without an inherent speed limit.
Normal planing boats have an attack angle of 5 degrees for best results (5:1 L/D ratio) these disks might preform better at an even lower attack angle?
I did build a tow model with 24" rear disks but due to poor construction one of the disk separate from it's axle on the first turn before we could get any data.
But it did lift quickly out of the water and planed on the outer edge of the disk as predicted with a 5 degree attack angle for 300 yards or so before that first turn.
The model at the site below will be tested soon it's another version of "rotating planing surfaces" (powered rather than passive).

Colin
http://www.ww.xbug.ca

 

Sorry... I just automatically assumed VLM. I never experienced problems with VLM with curved edges, though I never tried all circular "sails". Did try a 470 spinnaker on a reach as I recall, which is pretty close to spherical... the asym on the Cherub has a rounded/raked luff in http://www.wb-sails.fi/news/95_12_Cherub/Cherub.html. The numbers make sense but of course I have no idea how "right" they are. Should run it again with modern software. But MacSail had no LE suction.

 

The plate is 2 mm thick, as per the illustration in the paper, and with sharp edges. Yes, I read from your small compendium there was a little difference with the bevelled edged circular plate in the experimental results. Interestingly, I have 2 copies of your small compendium: In my Ipad, there was this more comprehensive compendium, which also had the CL-CD curves. In my mac at work, there's a stripped version (?) with only lift vs angle of attack. Luckily you put in the more comprehensive one, too, since it was the one that gave me the incentive to do the analysis.

 

A couple of stills from the LES-video.

 

Thanks for the clarification.

There was an early version of the LSP compendium. The later one included some additional planforms. I have another update with some more comparisons with experiments and a "simple" analytical formula that should give about 5 figure accuracy for planar elliptical wings for AR from about 0.2 to infinity. I just need time to check the results before releasing it.

I know it must have been a lot of work for you, but at least you now have a reasonable validated case that you can point to when someone (like me) bothers you.

Maybe you should try to model the Aerobie which was, last I heard, the object that held the record for the longest distance thrown by a human.

Leo.

 

Planingwheel, that "doc" opens nicely on my phone and made me think of those 4 bladed parrot choppers you can control from a smartphone, seen them hoover on water too and they seem a good heli config, not really a hoover or frisbee tho, dont see frisbees on boats directly but who knows what spins off, interesting thread

Btw, old frend of me trew frisbees very far, what distance is that record now?
Once in berkely I almost made worldchampion tablefootball but were getting old

 

If you are not taking account rotation of Frisbee, then the calculations are wrong. The rotation alters the flow of air over surface and make Frisbee longer against the motion. It generates more lift when rotating than flying straight. I studied this in school, unfortunately it was a long time ago and don't remember much more.

 

yes, I have a book about the flying pancake, part of the idea was it had counter rotating props at the wing tips to try to counter the wing tip vortex loss. What killed it, like all prop aircraft, was jet engines. It was also effectively a tailless design, and as such suffered from very narrow CG range (unlike conventional tail-in-back aircraft).

But a rotating disk, one with turbulators (Like most Frisbees have), will significantly change the conditions at the "LE", change lift and stall characteristics (because of higher Reynold's number), etc.. The spin also stabilize the flight. It will not be stable if was not spinning, not without appendages. flow over a spinning disk and a non-spinning disk is very different, so I do not see how a model that does not account for the surface effects (boundary layer) of the spin not be accurate.

 

I agree with everything you have written. Planes with jet engines
(including VTOL) were definitely more deserving of funding than the
Pancake. As I said, those types of planforms are now making a come back.

The validation exercise Mikko attempted was for a non-spinning flat plate
with a circular planform. It was just lucky (for him and me) that some
experiments were also conducted with non-spinning flat disks so we could
compare our predictions on something easier to model.

 

It was about 400m in the late 1980's. That record might have been broken
now. There's probably no drug testing for those "contests"

 

Petros & Mydauphin,

I realize I should have named the thread differently... I don't really have interest in aerodynamics of spinning objects (not many spinning sails or keels around, the innovative Xbug of Planingwheel is probably the first I've seen if you don't count propellers). I merely wanted to do the exercise as validation work for the CFD. As such it is good, and I still want to do the real frisbee later, since in the AIAA paper there are good pressure measurements and surface oil patterns to compare against, too. It is rare to have force measurements & pressure tapping at the same time, and even if the object is circular it does show typical flow characteristics for sails, with high lift, separation & re-attachement and edge vortices. One incentive was someone's sceptisism of CFD's capability to deal with vortex flow, as shown on the dhow sail simulation in another thread.

As for the frisbee and spinning lift, if you read the AIAA paper: "...the lift and drag curves overlay each other for all AdvR 0 to 1.04, which confirms that the lift and drag are unaffected by spin. The effect of spin on the side force, pitching moment and rolling moment (Fig. 8) is small but measurable. The side force (Fig. 8a) is zero throughout the AoA range for low AdvR 0 to 0.35 but for higher AdvR 0.69 & 1.04 becomes positive. The side force is approximately uniform across the entire AoA range, CY = 0.04 & 0.08 for AdvR of 0.69 & 1.04, respectively. The aerodynamic moments (Fig. 8b&c) exhibit similar characteristics, for low AdvR 0 to 0.35 the pitching and rolling moments remain unchanged. However for higher AdvR 0.69 & 1.04 both moments become more negative for typical flight angles of attack, 0° to 10°. The higher advance ratios 0.69 & 1.04 provide a greater nose down pitching moment (Fig. 8b) and a higher trim AoA just above 10°. The rolling moment is zero throughout the AoA range for low AdvR 0 to 0.35 (Fig. 8c) but for higher AdvR 0.69 & 1.04 becomes negative, CR =0.006&0.012at0°AoAforAdvRof0.69& 1.04, respectively."

The effect of the spin is mostly gyroscopic, preventing the frisbee from pitching and rolling to enable it to fly properly.

 

Aero is too easy, Mikko!
You should now try modelling the circular plate as a skimmer board
operating in shallow water.
I'd love to see how CFD handles the leading-edge splash falling back
in front of the plate, and the wave patterns. That might take more
than 24 hours per angle of attack

 

You never give up, do you ;-)

Do you have validation material for that? I have actually been exploring a transom stern planing hull, with reasonable results until about Fn 0,65 (referred to the static WL), but not really for the true planing regime. How important is the depth for that? Maybe my water channel is too shallow and narrow (to save on fluid elements & calc time). The boat tends to pitch too much at Fn>0,7 and that rises the drag.

 

Not so much work really, rather lost computer time from other work. But a good validation, yes, I'm now more confident of my CFD results than before, thanks to you.

It would be nice to understand the aerobie flow, but to mimic the Guinness book validation case for that you would need a 406 m long domain, which is way beyond my computing capacity

 

I'm also looking at a validation case for the motion of a boat (preferably a sailboat) in waves, with a well defined wave pattern (a tanktest?) and easily reproduced hull model, with motion results & forces. If anyone can point me to one I would be greatful.