Simple aerodynamics question.

[hogfarmer]

I have a simple aerodynamics question that maybe someone with more experience or access to certain software may be able to answer. Please provide an answer sighting references and math where needed. I already have plenty of peoples opinions with no scientific backing, I don't need anymore ;).

-I have two instruments that need to be hung off an aircraft into the airflow to collect data.

-Each instrument is approximately 6 inches long, 3 inches wide, 1.5 inches thick, with a display on the largest side. Instruments are represented by red rectangles in the sketch.

-They must be mounted so each instrument can have its widest side exposed on the same side. In other words they can't be stacked or mounted back to back, so both displays may be viewed at the same time.

-An aerodynamic shaped 'pod' shell will be built from molds, out of composite materials, to house the instruments. The pod is represented by the black shapes in the sketch.

-The pod will be attached by its trailing edge, by a small diameter tube, to the aircraft.

-Airflow will be between 18-70 mph, with special needs to reduce drag at the higher speeds.

The question is, what orientation out of the two possible ones pictured below will yield the least drag? One in front of the other, or side by side? Excuse the poor quality sketch, the intention is to show each layout from top and side views.

[hogfarmer]

So far the two basic opinions I have are:

1. Front and back will cause less drag because there is less frontal area. (induced drag is less)

2. Side by side will cause less drag becuase a foil shape can be maintained with less wetted area. (parasitic drag is less)

Neither opinion has been backed with scientific evidence saying what drag is more important at the Reynolds numbers involved.

[kcaldwel]

Look like hang glider instrument pods! 18 to 70 mph...

I can run an XFLR5 analysis if you give me some sizes and maybe the airfoil cross section for each.

I suspect it will depend more on the cross section than anything. The long way has a better fineness ratio. Both should be pretty similar in wetted area.

The biggest problem with instrument pods is they are never at the right angle for low drag. Beside being mounted at the wrong angle so they are easy to view, the yaw and pitch angle changes on a hang glider will likely mean they are stalled a lot.

And compared to the pilot drag, the instruments are insignificant anyway. Cut your arms off!

Kevin

[hogfarmer]

Look like hang glider instrument pods! 18 to 70 mph...

I can run an XFLR5 analysis if you give me some sizes and maybe the airfoil cross section for each.

Good guess.

I would like to run numbers but I don't know the variables. The sizes are easy, just add up the size of the instruments and extrapolate for the airfoil shape.

Airfoil shape is another matter all together. For side by side the minimum thickness needs to be 1.5 inches for at least 6 inches of cord, for front and back it requires that same minimum thickness for 12 inches, in order to fully in-close the instruments.

The orientation that lends itself to the best airfoil may be the most important aspect to consider. That was a hidden part of my original question ;).

Your angle of attack issue raises a good point. What design will be more tolerant of angle of attack? What design produces less drag when stalled?

My thinking is angle of attack should mainly be considered in pitch, as hang gliders normally fly fairly well coordinated. The side by side arrangement, positioned 'knife edge' at 90 degrees to the horizon, would produce the least amount of drag across all angles of attack.

You are right pilot drag is much more important. But I'm designing and making a home made pod from scratch, and want to start with the best design possible. Even the smallest reduction in drag is desirable if all other things are equal.

[kcaldwel]

Good guess.

My thinking is angle of attack should mainly be considered in pitch, as hang gliders normally fly fairly well coordinated.


Unfortunately, most CFD programs are useless past stall angle. And airfoils basically all stall at the same angle of attack from zero lift. These are really low aspect ratio, which will extend the range a bit. And trips (dimples, or linear trips) could keep the flow attached at higher angles.

If you think hang gliders fly mostly coordinated, try flying with a yaw string and an angle gauge for awhile. I've done it, and the yaw angles can be huge, and the glider rarely is flying straight. HG have very little yaw stability, and generally quite a bit of anhedral so they unstable enough to be weight shift turned. The yaw angles are bigger than the pitch angle changes, especially since the wing is modifying the local flow that close in.

Anyway, I'll do some runs with the basic dimensions.

Kevin

[hogfarmer]

HG have very little yaw stability, and generally quite a bit of anhedral so they unstable enough to be weight shift turned.

HG are very yaw stable, thats why they can make coordinated turns without rudder. Yaw stability in HG has much more to do with leading edge sweep than anhedral. Some high performance gliders sacrifice yaw stability for higher aspect ratio and less sweep.

The yaw angles are bigger than the pitch angle changes, especially since the wing is modifying the local flow that close in.
I'd have to see some data before I believe that. I agree, yawing does happen, especially at higher speed, but not by much unless you really suck as a pilot :D . I think angle of attack changes in pitch are much more important, especially because they are constant for minutes at a time. The differences in angle of attack between stall and full speed is quite dramatic on the newer high performance gliders.

Thanks for running numbers, I'm interested to hear the results.

[Brandano]

At small sizes Reynold numbers have also to be taken into account. I suspect that the first configuration (longer, more slender) will work better. However, what sort of sensors are these? Won't the second sensor be affected by the passage through the air of the first?

[kcaldwel]

So how is this thing likely to be shaped? A symmetrical airfoil maybe 2" or more at the high point, or a faired nose with flat on one or both sides, with a faired tail? You pick...

I have about 2,000 hours in hang gliders, and did quite a bit of aero experimenting with them.The yaw stability comes entirely from the sweep, which means that it varies with Cl. Once the Cl starts to come down they are very marginal on yaw stability. Even at high Cl, the yaw excursions with roll inputs are very large. I don't have any hard data, since it was all just watching a yaw string on a protractor. I think you'd be amazed at how poorly the turns actually are coordinated. And I was a semi-decent pilot, with a few Canadian HG records, two nationals wins, and a decent placing at the Owens worlds.

Do you fly HG?

K.

[hogfarmer]

So how is this thing likely to be shaped? A symmetrical airfoil maybe 2" or more at the high point, or a faired nose with flat on one or both sides, with a faired tail? You pick...

I have about 2,000 hours in hang gliders, and did quite a bit of aero experimenting with them.The yaw stability comes entirely from the sweep, which means that it varies with Cl. Once the Cl starts to come down they are very marginal on yaw stability. Even at high Cl, the yaw excursions with roll inputs are very large. I don't have any hard data, since it was all just watching a yaw string on a protractor. I think you'd be amazed at how poorly the turns actually are coordinated. And I was a semi-decent pilot, with a few Canadian HG records, two nationals wins, and a decent placing at the Owens worlds.

Do you fly HG?

K.

Well the airfoil shape will no doubt effect the drag. I wish I had the software and the time to learn to use it, so I could figure this out myself. I'm guessing a full airfoil that is thicker would be better than flat sides with flared leading and trailing edges, but I just don't know.

The bottom line is I'm simply interested in whatever shape will give me the least amount of drag and still be able to hold the 2 instruments. Airfoil, size, layout, are all factors.

I won't argue about the yaw stability of HG. Its all very relative. All I'm saying is for an aircraft without any vertical stabilizers or rudders, hang gliders do a pretty decent job of staying coordinated. The newer designs are even better, especially when you get to high performance models that have less L.E. sweep, the improvements in yaw dampening over the older high performance models are notable. I think angle of attack in pitch is a much bigger consideration.

Yes I fly HG. I've flown with a yaw string a lot. I think the coordination is good compared to other aircraft I've flown. You may be surprised to fly in a light airplane or sailplane with a yaw string (I have 100's of hours in those too). They actually don't stay coordinated very well either without a lot of piloting. You also learn the turn coordinator is a heck of a lot more accurate than a piece of string. ;)

[hogfarmer]

At small sizes Reynold numbers have also to be taken into account. I suspect that the first configuration (longer, more slender) will work better. However, what sort of sensors are these? Won't the second sensor be affected by the passage through the air of the first?

Yes I hope people reply using Reynolds numbers calculated with the data I provided. They are no doubt important in figuring which design has less drag at the desired speeds.

The instruments do not need to have clean air flow. The airspeed sensor is a separate device that feeds information to these instruments.

[kcaldwel]

I don't think the output form XFLR5 is going to be of much use. The aspect ratio is too low to really give any believable results, and if I put a straight portion on it, you just have separated flow. I did of course consider the Re.

I don't think the difference in drag on the two configurations is going to be enough to make any difference on a hang glider. A faired instrument deck is probably 1% of the total drag, and you are talking a difference between 1.03%, and 0.98% - not the primary factor in a usable pod design.

I have quite a bit of sailplane and power plane design too. I have never seen a yaw string in a sailplane do anything like it does on a hang glider, except maybe when doing intentional side slips. The problem is you can't do anything about it, where as the sailplane and power aircraft have rudder control. The last glider I had was WW Fusion, so they may have changed some since then, but after 25 years in the sport, I doubt it is that much.

Good luck!

Kevin

[hogfarmer]

I don't think the output form XFLR5 is going to be of much use. The aspect ratio is too low to really give any believable results, and if I put a straight portion on it, you just have separated flow. I did of course consider the Re.

I don't think the difference in drag on the two configurations is going to be enough to make any difference on a hang glider. A faired instrument deck is probably 1% of the total drag, and you are talking a difference between 1.03%, and 0.98% - not the primary factor in a usable pod design.

I have quite a bit of sailplane and power plane design too. I have never seen a yaw string in a sailplane do anything like it does on a hang glider, except maybe when doing intentional side slips. The problem is you can't do anything about it, where as the sailplane and power aircraft have rudder control. The last glider I had was WW Fusion, so they may have changed some since then, but after 25 years in the sport, I doubt it is that much.

Good luck!

Kevin

Thanks for the info Kevin.

I agree either lay out would be very minor in over all drag, but like I said I'm making molds from scratch so I want to make sure its as good as it can be from the start. I guess the differences are not enough to outweigh considerations like instrument layout for ease of use and visibility?

In modern HG competition at the national and worlds level you simply can not plan on winning unless you have the lowest drag equipment. Even the full race harnesses with single suspension and contour shapes from a few years ago are not good enough any more. There are only a few wings available that are competitive at that level right now. Round down tubes or base bars? Forget about even placing! Everything must be shaped to minimize drag. It sounds silly but its true, almost all the tasks are a full out races because multiple people are almost always making goal. At high speeds above 40 mph all these little things really start to add up to a considerable difference. If you look at the current top pilots and their gliders they have every wire connection flared, they tape over all of the gaps, fill ever void with shaped foam, and even make custom composite pieces to flare out tiny things like VG cleats. It all adds up.

I'm not at that level, but I'd like to be some day :cool: .

[kcaldwel]

This isn't really a simple problem. You'd need a pretty fancy CFD program to handle the 3D flows, or a decent wind tunnel.

We had 100 mile races with 60 gliders at goal most days at the Owens Worlds 15 year ago. I still follow the sport pretty closely - I get Davis Straub's Oz report every day. Rohan Holtkamp just won a comp in Oz on a king posted intermediate glider against all the latest and best.

It hasn't changed. You can chase drag reduction, and the latest glider and equipment - and I certainly did - but the pilot makes all the difference. Davis (edit: Davis Straub, Oz report guy) finishes in just about the same place whether he flies an ATOS against flexies, or a king posted glider against the topless ones like he is in Australia right now.

My best day at the Worlds, my fancy mount that set my instruments at the right angle to the local flow broke, and they hung from the safety line the whole flight. I couldn't use my speed to fly stuff, or even see my altimeter, and I got a top 10 finish on a 95 mile, 3 1/2 hour race. We had 186 pilots at the worlds that year, something that will never happen again.

Design the pod to look cool, and be functional. The pilots will convince themselves it is low drag!

Kevin

[hogfarmer]

Design the pod to look cool, and be functional. The pilots will convince themselves it is low drag!



YEP! :D

Who the heck is "davis"? :eek:

[vintage1]

Frontal area is the dominant drag at sane speeds. Over 70mh I would expect wetted area to become significant.

But you cant reduce that due to the size of the instruments anyway.

Watch out for excesseive side area upsetting lateral stability.

So I would mount them one in front of the other and as flat as possible..i.e. make them into a sort of sesquiplane type layout. You might as well get some lift as well, and make a flat bottomed airfoil out of them.

And, why not use them as the undercarriage as well..

[kcaldwel]

I actually have some results that might be of some use!

First, I had to place a trip at 25% back from the LE to get any sort of attached flow on an airfoil this thick (21%) at the low Re. I haven't had much luck with the flat side version so far.

And the verdict is.... the lowest drag version depends!

At zero angle of attack, the long skinny version has about 30% higher drag (assuming equal "wing" areas) than the wider "span" one, at both 8 and 31 m/sec (about 18 and 70 mph). But at 8 degrees AoA, the highest AoA I can get reasonable results for, the long chord, skinny one, has about 50% less drag than the side by side version, at both speeds.

I thought of another reason the pitch angle is less critical. The hang glider pitches up for low speed, and down for high speed. Since the instruments are attached to the airframe, they pitch down for higher speeds, and up for low speeds. This mean the pitch AoA they see is quite small, if you can set them at the right angle. Mark Drela actually helped me calculate the local flow angles under a hang glider many years ago on-line, and I verified them with in-flight testing. I have an old article I wrote for "Hang Gliding" many years ago that may have the formula in it if I can find it.

Not sure if that is of much help. In my opinion, add a nice techy looking trip at 25% on your pod, and go with the long skinny one.

Kevin

Edit: Please take any of these results as very approximate. The very low aspect ratios mean the actual flows are very 3D, and the results are questionable.

[kcaldwel]

One more plot, of strictly 2D flow around an 18" long thing I thought might look something like your pod design. I can't get anything to converge on a shape like the side by side might have, even at zero angle of attack, so I can't compare.

The drag is pretty low, as long as the AoA is low.

Kevin

[BMatthews]

Can you integrate the instruments into a pod that actually rides in the wake of something that is already on the HG? It's far better aerodynamically to just use an extended fairing shape on an existing strut or component that is already in the breeze than to introduce a whole new shape.

[zitron]

We've done some studies for optimal fairing airfoils for solar powered cars. The issue here is the fatter the airfoil, the more pressure drag, the longer the airfoil, the more skin friction. At our Reynolds numbers (1,000,000ish), the optimal airfoil thickness to cord ratio was found to be about 1:4.

[vintage1]

interesting...so long and slender isn't as good as short and fat..

[kcaldwel]

Ate Re of 1x10^6. You can look that up in "Theory of Flight" from the 40's for 2D flow.

Kevin