wing area vs L/D

[ciurpita]

i was looking at characteristics of full scale schweizer sailplanes and was surprised that the one with the highest wing loading (1-35) has the best L/D. it is also the lightest, has the highest aspect ratio, and the least wing area. it does not have the longest wing span.

it appears that the gain in performance is primarily due to reducing the wing area which reduces parasitic drag, and maintaining the span which increases the aspect area and reduces induced drag. but reducing the wing area while maintaining the span also reduces the chord and thickness, making it a challenge to construct.

how much of the total drag is due to the wing area? am i correct in concluding that reducing wing area is key element in improving performance? could i improve the performance of by model by reducing the wing chord?

http://www.sailplanedirectory.com/PlaneDetails.cfm?planeID=309

L/D span area empty payload gross loading ballast aspect built name
17 36.00 133.50 243.00 180.00 423.00 3.20 0.00 9.71 2 SGU 1-7
23 52.00 214.00 450.00 410.00 860.00 4.02 0.00 12.64 57 SGS 2-8
24 54.00 237.00 860.00 340.00 1200.00 5.05 0.00 12.30 114 SGS 2-12
16 36.00 170.00 320.00 230.00 550.00 3.25 0.00 7.62 50 SGU 1-19
18 43.00 182.00 400.00 230.00 630.00 3.50 0.00 10.16 3 SGU 1-20
27 51.00 165.00 470.00 250.00 720.00 4.26 266.00 15.76 2 SGS 1-21
17 43.00 210.00 470.00 430.00 900.00 4.80 0.00 8.80 258 SGU 2-22
29 49.20 159.40 470.00 280.00 750.00 4.70 0.00 15.19 74 SGS 1-23
30 55.50 180.00 585.00 200.00 785.00 4.32 0.00 17.11 1 SGS 1-24
32 60.00 231.00 1050.00 400.00 1450.00 4.85 0.00 15.58 1 SGS 2-25
23 40.00 160.00 355.00 220.00 575.00 3.59 0.00 10.00 480 SGS 1-26 A,B,C & D
23 40.00 160.00 445.00 255.00 700.00 4.38 0.00 10.00 210 SGS 1-26 E
34 49.20 153.80 465.00 255.00 750.00 4.87 0.00 15.74 1 SGS 1-29
33 57.00 180.00 831.00 509.00 1340.00 7.40 0.00 18.05 87 SGS 2-32
23 51.00 219.50 600.00 440.00 1040.00 4.73 0.00 11.85 579 SGS 2-33
34 49.20 151.00 550.00 290.00 840.00 5.56 0.00 16.03 93 SGS-1-34
38 49.20 103.80 400.00 260.00 660.00 8.96 0.00 23.32 101 SGS 1-35
31 46.20 140.70 475.00 235.00 710.00 5.05 0.00 15.17 43 SGS 1-36 Sprite
28 59.50 195.70 1200.00 470.00 1670.00 8.99 0.00 18.09 12 SGM 2-37

[kcaldwel]

First, The Schweizer sailplanes were not really performance sailplanes, even when they were current! The 1-35 was their attempt at a higher performance sailplane, and it was by far their cleanest design, and the best airfoils and wings, etc.

At full scale, L/D is proportional to aspect ratio. The only way wing area comes into it, is if the span is limited, like with the 15m span limited classes. Well, I guess even in the open class, structural and handling issues limit the span, so to keep the high aspect ratios they keep the wing area to the minimum.

Stall speed is proportional to wing area, so making the wing area too small makes the stall speed too high for thermalling and landing. Flaps can help a bit with small wing areas.

On a model sailplane, it gets a lot more complicated due to Reynolds Number effects. If the chord gets too small, low Re degrades the airfoil performance. This means there are minimum chord lengths that work best for a given airfoil and flying speed. Dr. Drela optimized his airfoils for a particular Re.

Reducing the wing area to increase the aspect ratio needs careful analysis on an RC sailplane to confirm it helps the overall performance. Lots of time on XFLR can show you the trade offs.

Kevin

Edit: There is a very good article by Dave Register in the Dec 2002 issue of RCSD about optimizing a planform and aspect ratio for a DLG, given the airfoils available then. A similar analysis would be required for other type of RC sailplanes, but the constraints would be different given the different launch and performance targets.

[JetPlaneFlyer]

am i correct in concluding that reducing wing area is key element in improving performance? could i improve the performance of by model by reducing the wing chord?

You need to define what you mean by 'performance' and you need to tell us what the model looks like right now.

Increasing aspect ratio and increasing wing loading both improve L/D but there is more to 'performance' then L/D. Increasing wing loading makes the model fly faster which raised Renolds number, which is good for LD but the fact that it flies faster means it descends quicker so despite the better L/D it's duration and thermaling performance will be poorer. Aso consider that narrowing the chord will in itself decrease Reynolds Number, which has a negative effect on performance especially at model scales where Reynolds Number considerations dominate.
Generally small scale models perform better with relatively low aspect ratio as keeping a reasonably large chord is more important than aspect ratio. As the model gets larger aspect ratio can increase. You will see this trend in most high performance gliders.

[kcaldwel]

The 1-35 is the only one of those models with flaps I believe. It was designed to the 15m class span limit. The 1-34 was designed to the standard class rules, that don't allow flaps but still has a 15m span limit.

The flaps on the 1-35 allow a higher aspect ratio (lower wing area because of the 15m span limit). They get the slow speed flight performance for thermalling and landing with the lower wing area by using flap deflections. Negative flap deflections also help the higher speed performance.

Comparing the flapped 1-35 performance to the others isn't really a fair comparison.

Kevin

[ciurpita]

i realize that there are many factors that affect performance, but i'm only interested in considering, L/D. i'm not considering thermaling or handling qualities. since lift is constant, i'm curious if the significant contribution to reducing drag was reducing wing area. it seems that the supra has also minimized it's wing area. how much of the total drag is due to the wing area?

[fnev]

You can’t take one variable and dissociate it from the rest as they ALL interact. To increase the L/D you need to increase the aspect ratio. BUT, are you keeping the same wing area? Are you keeping the SAME wing section? A high aspect ratio wing is VERY difficult to design structurally as flexion and torsion start to be very critical.

You got your answer BUT it is just the basic laws of aerodynamics… To implement this on a flying object, as it is stated above, needs some other considerations. You CAN NOT disregard these…

[kcaldwel]

L/D does not depend on wing area. It is proportional to the aspect ratio, Re effects neglected.

If you take a given full size glider, and increase the aspect ratio by either increasing span with a constant area, or reducing area with a fixed span, the L/D will increase, again neglecting Re. Reducing the zero lift drag coefficient will also increase L/D.

With a model glider, Re is so so important, increasing the aspect ratio with a fixed span (reducing wing area), will usually make the maximum L/D worse, assuming the apect ratio was near optimum for the glider. Increasing the aspect ratio by increasing the span (more wing area), will usually result in an increased L/D. The bigger span may well have handling or structural issues.

The basic methodology of determining the optimum aspect ratio for a given airfoil and zero lift drag is covered pretty well in the article I referenced above.

Kevin

(L/D)max = 1/2* {pi*A / CD0}^1/2

A = aspect ratio
CD0 = zero lift drag coefficient

assumes elliptical lift distribution, neglects Re, planform, aifoil effects.

[lincoln]

Here's one way to make any given design have a higher L/D. First, make it out of unobtanium. Then, load it with lead until it reaches a substantial fraction of Mach 1, or, if it had a laminar foil, until just before laminar flow breaks down. It will have the rate of descent of a brick, but it will cover a LOT of ground doing it, considerably more than the original due to increased Reynold's number. As mentioned above, these things DO interact. A jumbo jet has a max L/D somewhere near 18:1. Want to thermal one?

Even if you can build a practical model with a really high L/D, the Reynolds number will adversely affect the performance of the airfoil. So you can't go to too narrow a chord if you want good performance.

[fnev]

Please, in order to achieve a GIVEN L/D you will need a GIVEN wing loading for a GIVEN plan form (all the rest being the same). The two are working hand in hand. So if you increase the aspect ratio by decreasing the chord, you are increasing your wing loading (as long as the mass stays the same). Is it the plan form modification or the wing loading that influenced MOST the L/D? The formula given for the optimal aspect ratio for a given span is a LARGE approximation. But it gives a pretty good idea.

You have to realize that the shape of your plan form is as critical as the aspect ratio of your plan form. The two are intimately linked to the wing section(s) that is (are) chosen as well. As I said in an other tread, it is great to be aware of these interactions and to understand the effect of each variable. Most of the people nowadays have access to some basic (but very good) software to do some analysis. To try to reduce a design to just a few parameters in aircraft design is NOT possible. If it was that simple it should be a lot of excellent performers (models and even full scale) floating around (punt intended)… This is NOT simply the case…

[kcaldwel]

fnev,

Wing loading doesn't have much effect on L/D. Ballasting a sailplane just means the best L/D speed will be higher, it doesn't change the actual best L/D, as long as the airframe deformation is minimal.

Planform effects are small, as long as you are near an elliptical span loading on the lift distribution. Winglets can change that by increasing the span efficiency, but likely not at RC scale.

Kevin

[JetPlaneFlyer]

fnev,

Wing loading doesn't have much effect on L/D.
What about at the scale of lightweight small/medium model gliders where Re number considerations dominate? I'd have expected faster flying speed and the resultant Re increase could make quite a significant difference when dealing with these models?

The light free flight duration models I play with have a pretty hopeless L/D which I always attribute largely to the low Re the operate at.. Glide ratio definately appears to improve significantly if I ballast them up, though at the expense of duration.


Steve

[kcaldwel]

Yes, I should have added the caveat that Re effects make a difference in L/D for model scales. At full size glider sizes that this thread started at, wing loading has virtually no effect on L/D.

Kevin

Edit: That is also covered in the Dave Register article I referenced above.

[fnev]

Yes, I should have added the caveat that Re effects make a difference in L/D for model scales. At full size glider sizes that this thread started at, wing loading has virtually no effect on L/D.

Kevin

Edit: That is also covered in the Dave Register article I referenced above.


Wow…!!! Why full scale gliders have a mass limit???


Sorry but: mass (wing loading) INFLUENCES L/D. As mass (wing loading) increases for a balanced aircraft (glider) with ALL the other parameters IDENTICAL, the speed AND the L/D WILL increase…

Please, look at any glider polars.

[fnev]

I have to be careful here…

I am speaking of balanced polar (taking in account the new equilibrium therefore the new elevator input or trim). This is normally not the one provided by the glider manufacturers. So, when you increase the mass not only the speed will change (increase) but the AoA, the elevator setting as we are speaking of a stable aircraft (glider). When I speak of other variables not changing it is air density/humidity…

The increase of L/D is of the secondary degree (if I remember correctly, might be less): this means it is not proportional to the increase of mass or speed.

I don’t have the time, but since some of you seem to be experts in aerodynamics and flight mechanics, maybe one poster will have the time to do the math with ALL the wonderful aerodynamic equations. AND post them, please.

[fnev]

Again…

Funny enough, the influence of wing loading on L/D is greater as the glider is smaller due to the greater variations of Cl/Cd curves with the Re variation AND the greater variation of the induced drag.

If you go to the extreme (infinite wing span) you are now in a two dimensional flow, without induced drag, but still with the Re influence. Now your reasoning stands 100%.

[kcaldwel]

Yes, full size sailplanes have weight limits, due to structural and handling reasons. They are all tested at the lightest and heaviest certified weights.

I am not talking about infinite span gliders. If the Re effects are small over the speed change, and the airframe is relatively rigid, ballast does not make a difference to the best L/D, only the speed at which it occurs. The elevator trim setting will remain the same at best L/D if the CG is in the same postion, the glider will just be faster at best L/D with ballast:

"When ballast is carried, the best glide ratio is achieved at higher speeds (the glide ratio is not increased)."

http://www.absoluteastronomy.com/topics/Glide_ratio

Pg. 4 of this polar for the ASW 22:

http://www.aoe.vt.edu/~mason/Mason_f/ASW22BLv2.pdf

The line from the graph origin to tangent with the polar is the best L/D line. Notice the same line is tangent to both the ballasted and unballasted polars.

http://home.comcast.net/~verhulst/GBSC/student/ballast.html

Page 3 of this sailplane test:

http://www.dg200.co.uk/Dick%20Johnson%20flight%20test.pdf

RC gliders will see a change best L/D with ballast, but it is entirely due to the higher Re at the higher speeds. Most airfoils get better performance at higher Re, and fuselage drag will likely go down at well. Re is why full size sailplane get L/D's of 60:1, and RC gliders ar more likely in the 20s.

If you go back to my L/Dmax equation above, which is not for infinite wings, you will see it only carries terms for the aspect ratio and zero lift drag. There is no mass term, or wing loading term. While this neglects Re effects, it is the governing equation for L/D.

Kevin

[fnev]

We are going to agree to disagree… It is becoming too pedantic here.

The art of aerodynamic is to know what to “ignore” and when you want (or need) to ignore…!!!

As our calculation methods improve (over time) we are now in the position to compute with variables that we were “ignoring” a few years back. Our understanding of aerodynamic because of this is more acute (look at the progresses in wing sections and wing plan forms). I am lucky to work with aerodynamic packages that are based on the vortices lattices and I must say it opened my eyes on a lot of issues as it is so much more accurate by taking in account so many more variables.

I will agree with you that when speaking of a full scale glider the L/D is not affected by a large amount (or easily measurable). On model gliders it is much more sensitive and you can actually measure the improvement.

Two more points:
- You didn’t give me any real calculations… (I was waiting for the “approximations”!!!)
- You have some well known books or publications that carry some major flaws (mistakes or misleading information, some plain wrong)

[kcaldwel]

We are going to agree to disagree… It is becoming too pedantic here.

- You have some well known books or publications that carry some major flaws (mistakes or misleading information, some plain wrong)

Please let me know what these "well known books or publications" are, and the "major flaws", so that I can make sure they are corrected...

Kevin

[markschaffin]

In simple terms:
CD=CDo + CDi
where CDo=CDmin (not necessarily (and almost never) at zero lift)
CDi = induced drag

Think of CDo as being tied to wetted area and Re.
Think of CDi as being tied to AR.

L/Dmax happens when CDo=CDi.
If Re is low enough that CDo is high, decreasing CDi no longer improves L/D => lower AR. This is evident by looking at the AR of model sailplanes as they decrease in size.

If Re is high enough that CDo is low, increasing CDi improves L/D => higher AR.

Again, this is simplified. In reality, there are other so many other things that must be considered that it is silly. Root bending moment, structures, weight, marketing...

Mark

[fnev]

Sorry, I couldn't resist:

(Extract from the DG-Flugzeugbau.de website)

L/D ratios in gliders
It is almost impossible to describe the performance of a glider using just the maximum glide or L/D ratio. For this reason one of the glider manufacturers stopped quoting L/D ratios several years ago.

We agree that a single number says very little about the actual overall performance of a glider and prefer analyzing the differences in performance of different gliders by comparing them in flight. After all, with this method you can test the glider through the range of the complete polar curve. You might not get any absolute values but a good comparison to the other glider used for reference.

However, our customers would like to see a maximum L/D ratio figure, so the following might explain that.

In general we publish a calculated polar curve and a maximum L/D ratio. This calculation is based on the profile polars which have been measured in the wind tunnel. The induced drag, the form drag of the fuselage, and the interference drag are added to the profile drag values of wings and elevator. Unfortunately not all these drag values are known, and the wind tunnel measurements might not fully reflect reality either.

Therefore the engineer will compare calculations and measurements on previous gliders and estimate a correction drag value.

It is also important to interpret the L/D ratio figures correctly. Based on the assumption that the correction factor wasn't "reduced for marketing purposes" the glider will reach the quoted value in optimal conditions - i.e.:




taped correctly
max. wing loading
optimal c of g position
polished and totally bug-free
in absolutely calm air, i.e. no turbulences (which means nowhere near clouds)
with closed and locked undercarriage and engine doors and airbrakes (all flush)
without corrective control movements
in ISA standard atmosphere conditions
Calculating polar curves in a test flight is very difficult and prone to errors. There are for instance polar curves for the same glider but from different years which are not identical. Particularly the best L/D ratio changes significantly due to minute measuring errors. For example a difference of only 1 cm per second in the sinking speed causes an L/D ratio difference of 1 point. Measuring errors in a comparative flight can however be much bigger.

So you see that the quoted L/D ratio is normally merely a theoretical value that you shouldn't use for your final glide calculations. In real life you should use a value of roughly 4 - 8 % below the theoretical best L/D ratio.

- friedel weber & w-dirks