Cl distribution question

[Mihai]

I posted a question on the lift distribution on the "Flying Wings" forum and that seems to be the wrong forum for it (63 views and no idea):

http://www.rcgroups.com/forums/showthread.php?t=867373

Thanks for the help,
Mihai

[BMatthews]

Are you after the spanwise lift or what?

There used to be a sweet little online calculator where you input the root and tip chord and any washout angles and then it showed the spanwise lift coefficient curve. I can't find it at the moment and the link was long ago lost in a system crash.

[JetPlaneFlyer]

Here are a couple of sites that you may find useful:

http://aero.stanford.edu/WingCalc.html
http://www.amadistrictii.org/cjrcc/wing2/wing.html

Also don’t forget that on swept flying wings the twist is usually there to achieve trim in pitch by giving the wing an overall positive pitching moment, not only to prevent tip stall. A flying wing is always something of a compromise between a design that will be stable and trimmed, and a design that will be efficient.

Steve

[richard hanson]

Aflying wing is essentially a very very short coupled airframe -
that is why th goofy add ons required
also it is the worst airframe design for stall recovery

[mlbco]

I posted a question on the lift distribution on the "Flying Wings" forum and that seems to be the wrong forum for it (63 views and no idea):

http://www.rcgroups.com/forums/showthread.php?t=867373

Thanks for the help,
Mihai

The vortex lattice code is calculating the inviscid span efficiency, so you won't see the effect of varying profile drag of the airfoil sections in the results. The method tends to overpredict performance slightly, hence the span efficiencies greater than 1.0 for a planar wing.

The main issue you're dealing with is that a low AR wing is very insensitive to changes in sweep, twist, and planform. This means that all low AR wings tend to be elliptically loaded in the distribution of cl*c. Try re-running your test cases with AR=20 and then see what happens.

Steve

[Mihai]

Are you after the spanwise lift or what?

There used to be a sweet little online calculator where you input the root and tip chord and any washout angles and then it showed the spanwise lift coefficient curve. I can't find it at the moment and the link was long ago lost in a system crash.

Yes, span-wise lift and coefficient of lift. This is what I used:
http://aero.stanford.edu/WingCalc.html

and for a constant chord wing it is always ideal and 100% efficient, which doesn't make much sense.

M.

[Mihai]

Here are a couple of sites that you may find useful:

http://aero.stanford.edu/WingCalc.html
http://www.amadistrictii.org/cjrcc/wing2/wing.html

Also don’t forget that on swept flying wings the twist is usually there to achieve trim in pitch by giving the wing an overall positive pitching moment, not only to prevent tip stall. A flying wing is always something of a compromise between a design that will be stable and trimmed, and a design that will be efficient.

Steve

Thanks! The first one is what I used. The second one doesn't have "my" airfoils (either a MH43 which is on the list but does not work, or a PW51 which is not even on the list), so I can't really use it.

Both my "target" airfoils have positive pitching moment, so I'm not so worried about an overall positive pitching moment. However, I do want to fly this plane low and slow in front of my house, so I am worried about stalling it, so my main reason for using twist is to prevent tip stalls at low speeds.

Thanks,
Mihai

[Mihai]

The vortex lattice code is calculating the inviscid span efficiency, so you won't see the effect of varying profile drag of the airfoil sections in the results. The method tends to overpredict performance slightly, hence the span efficiencies greater than 1.0 for a planar wing.

The main issue you're dealing with is that a low AR wing is very insensitive to changes in sweep, twist, and planform. This means that all low AR wings tend to be elliptically loaded in the distribution of cl*c. Try re-running your test cases with AR=20 and then see what happens.

Steve

Cool! This makes sense. I tried AR=20 and it ain't pretty :).

This means that for low AR the constant chord really works well for tip stalls (i.e., nothing to worry about), and the efficiency of 100% should be interpreted as "efficiency for that given, low, AR is 100%", meaning there is no loss due to the lift distribution beyond the inefficiency of that low AR. Is this right?

If so, I'll just start cutting constant chord templates - it may actually be easier than cutting them of different sizes :).

Thanks a bunch!
M.

[JetPlaneFlyer]

Both my "target" airfoils have positive pitching moment, so I'm not so worried about an overall positive pitching moment.

The MH43 has a negative pitching moment over most Alaphas. According to Profili using a Re of 100 the Cm only just touches a positive value briefly at 10-11 Deg Alpha, at all other alphas it's considerably negative. I cant find a PW51 in profili so I'm not sure about that one. Certainly if you use the MH43 then you will need some twist, or reflexed elevons (yuk :( )

[Mihai]

Thanks! I was unsure about MH43, as the author (M-H) says that it has a "low moment" of cm=0.007, but doesn't specify if it's positive or negative. The PW51 is a "plank" airfoil and, if I understood it right, the plank airfoils, by definition, have positive (small, but positive) pitching moments, so you don't have to reflex the elevons.

Can you import airfoils in Profili? If so, can you please check the PW51? Apparently it's among the best liked airfoils here on RCG for flying wings: positive pitching moment, OK lift, good inverted, good speed, etc. (almost too good to be true :) ), so if you go through the pain to import it you may use it yourself later :).

I'm attaching the PW airfoils.

Thanks,
Mihai

[Mihai]

The MH43 has a negative pitching moment over most Alaphas. According to Profili using a Re of 100 [...]


Now I saw this - I don't plan to fly at a Re of 100, but rather at about 70000 to 200000 (10mph - 30mph) and up...

Thanks,
M.

[Mihai]

The MH43 has a negative pitching moment over most Alaphas. )

I misspoke. I meant MH45. MH43 is not a "wing'" airfoil.
Sorry for the confusions. Can you please check MH45 for positive pitching moment? I assume that's small and positive like PW51 (with smaller moment, bigger lift upright, but not as good inverted if I recall exactly).

M.

[JetPlaneFlyer]

The Re100 was a typo... I meant 100k i.e. 100,000... about 20mph

I'll see what the other airfoils look like in Profili....

[JetPlaneFlyer]

Reflexed airfoils just dont work well at low Re numbersin terms of generating positive Cm (seperation bubbles?).
Here are some graphs for your selected airfoils at a range of Re numbers... I've thrown a Eppler 330 into the mix for comparison.

Re numbers plotted are
70,000
200,000
400,000

Based on these graphs, if I were designing it (other than for the Eppler 330 at minimum Re of around 300,000 and above) I'd use some twist, but the choice is yours.
I would bet that those guys using the PW51 on a plank have reflexed elevons but just dont like to admit it ;)

Steve

[JetPlaneFlyer]

I would bet that those guys using the PW51 on a plank have reflexed elevons ................


It appears I would have won this bet... here is a quote from a plank flyer using the PW51

"I seem to need at least 4mm up trim on the elevons to keep her level. This is on top of the built in reflex on the wing."

This is the thread it comes from: http://www.rcgroups.com/forums/showthread.php?t=866855&highlight=pw51

It appears to me that the PW51 and MH45, without quite heavily reflexed elevons (which will kill it's performance), are not suitable for a plank. You need an airfoil that has positive Cm throughout it's operating range. This will inevitably mean making compromises with Cl performance but there is no free lunch. So IMHO if you use either of them on a swept wing tailless then add some twist.

[Mihai]

Thanks a lot for the plots of the foils. Is it just me or does it seem that the Eppler 330 "beats" the other two at least at Cl/Cd and moment? What gives? Minimum drag?

Also, thanks for the advice, I'll add a touch of twist to see what happens :-).

Best,
M.

[JetPlaneFlyer]

The thinner MH and PW airfoils give less drag than the Eppler 330 but no point having low drag if the lift is not there.

I've plot off some graphs with the PW51 with 4mm elevon deflection as talked about in the post i found (see attached) You will see that the reflexed elevon gives a nice constant positive Cm (much better in this aspect than the Eppler at low Re), but the reflex also kills off the Cl and Cl/Cd curves.

For swept wings twist rather than elevon reflex is probably the way to go unless you are concerned about inverted performance. If you want a model that's good inverted then the PW51 with reflexed elevons is maybe a good bet.

[Mihai]

Thanks a lot! I feel like you're doing my homework :-). Indeed a reflexed PW51 seems to have a moment that's plenty positive. Isn't that too much? I mean, negative is not good as it's unstable, but too much positive isn't great either (too stable - I'd guess a small performance envelope).

My goals for this wing are as follows:
- I want it to be "fast" (use it for a jet trainer).
- I want good inverted performance (to learn well/better to fly inverted - fin drags would be ideal :) )
- If also possible, I'd like to to be stable at low speeds, not only for landing, but also to fly it in my neighborhood (very small space to play in).

Based on what I saw a PW51 without twist and perhaps a small reflex would do just fine.

Thanks for the help.
Now to draw some templates...

Best,
M.

[miniphase]

the wings I've built (all gliders)with the MH45 have all had good inverted performance off the slope(usually with 1.5 to 2 deg of washout built into the wing)

I've heard there's little difference performance wise between MH45 and PW51

[Mihai]

miniphase, what was our planform? Did you have any taper? How much? If I understand exactly there are three types of "twist":
- geometric washout (where you actually twist the tip with the TE up (:edit: ))
- aerodynamic washout (where the tip has less camber, hence a higher stall angle)
- planform washout (not sure if this is a correct term), where the tip is wider than the ideal ellipse planform.

My understanding that any one of those three (or combinations) would do, each with its advantages and disadvantages). If you do taper to about 0.45 (close to ideal elipse), and you keep the airfoil the same, then you need geometric washout, and that may be close to ideal in most situations except very high speed (tips lifting down) and inverted flight at low speed (tip stall).

I chose to go with planform washout (keep the chord constant all the way to the tip) and keep away from the other two understanding that I'll not get great performance from this wing in terms of efficiency, but it should work well both inverted and at high speed (on either side).

Please correct me if I'm wrong (I sort of wrote as I was thinking it).
M.

[BMatthews]

On the aerodynamic washout you've got it wrong. An airfoil with more camber stalls later than one with less unless there's other shaping issues at work.

Also washout implies that you twist the trailing edge UP. Twisting it down increases the angle of attack at the tip and that is called washIN.

The elliptical planform "ideal" only applies to something the size of the Spitfire. In our world with small to critical reynolds numbers almost any taper is a bad thing that leads to tip stalling far too easily thanks to the degradation of the airflow as the reynolds numbers plumet. And of course the numbers are allways critical at the lower speeds such as we find near the stalling point.

This whole reynolds number issue is why our modern model sized sailplanes do not mimic the super high aspect ratios seen on full sized sailplanes. It's also why as you move to the 2 meter class the chord doesn't get much narrower than that found on the open class gliders. Just the span gets crunched down.

[Mihai]

On the aerodynamic washout you've got it wrong. An airfoil with more camber stalls later than one with less unless there's other shaping issues at work.


Well, I'm really not aero-engineer, but I did read Martin Simons book and he says otherwise in Section 7.6 as well as Fig. 7.4 (page 77 on the 4th edition). So, while I can't argue intelligently about this, you'll have to take it up with Simons :).


Also washout implies that you twist the trailing edge UP. Twisting it down increases the angle of attack at the tip and that is called washIN.


True, sorry for the mistake. I'll go and correct the post as I may be able to confuse other noobs like me.


The elliptical planform "ideal" only applies to something the size of the Spitfire. In our world with small to critical reynolds numbers almost any taper is a bad thing that leads to tip stalling far too easily thanks to the degradation of the airflow as the reynolds numbers plumet. And of course the numbers are allways critical at the lower speeds such as we find near the stalling point.

This whole reynolds number issue is why our modern model sized sailplanes do not mimic the super high aspect ratios seen on full sized sailplanes. It's also why as you move to the 2 meter class the chord doesn't get much narrower than that found on the open class gliders. Just the span gets crunched down.

Again I can't really argue with this, however, it seems that where efficiency really matters (e.g., sailplanes - DLG - see the SGII) the wings are tapered (in several sections actually). I can't believe that there isn't a real performance gain for all that pain of making multi-taper wings. Also, in the same (apparently well-respected) book, Martin Simons suggests that the planform should be close to elliptical if possible. I'd imagine that if it doesn't matter he would say so. So, while I can't really argue about it, I can point to people that know what they are talking about (e.g., Simons and Drela) that do taper the wings.

M.

[JetPlaneFlyer]

Aerodynamic washout is, if I'm not mistaken, where the tip airfoil is selected to have a higher zero lift angle than the root.

So lets say the root uses a Clark Y that has (say) -3 Deg zero lift angle, and the tip uses a symmetrical airfoil that has a 0 Deg zero lift angle.... then this is 3 degrees aerodynamic washout.

I agree with BMathews on the elliptical wing issue... They are not good for small/medium models because Reynolds number issues dominate at smaller scale and low airspeed. This is not to say that some taper is not desirable, for maximum efficiency it still is, but just not as much taper as a true ellipse. Scale models that use elliptical wings (eg Spitfire) invariably have quite a bit of washout built in to make their flying characteristics manageable... This of course destroys the theoretical benefit of the elliptical platform but without it they would be all but un-flyable.

I can only assume that the Simons book is overlooking Reynolds number considerations for the sake of simplicity.

BTW... I see where you are going with the description of geometric washout being 'TE Up' at the tip, but this is a little misleading because it could just as easy be 'LE down' at the tip. Geometric Washout is a twist in the wing such that the angle of incidence decreases as you move toward the tip... period.

Steve

[miniphase]

miniphase, what was our planform? Did you have any taper? How much? If I understand exactly there are three types of "twist":
- geometric washout (where you actually twist the tip with the TE up (:edit: ))
- aerodynamic washout (where the tip has less camber, hence a higher stall angle)
- planform washout (not sure if this is a correct term), where the tip is wider than the ideal ellipse planform.

My understanding that any one of those three (or combinations) would do, each with its advantages and disadvantages). If you do taper to about 0.45 (close to ideal elipse), and you keep the airfoil the same, then you need geometric washout, and that may be close to ideal in most situations except very high speed (tips lifting down) and inverted flight at low speed (tip stall).

I chose to go with planform washout (keep the chord constant all the way to the tip) and keep away from the other two understanding that I'll not get great performance from this wing in terms of efficiency, but it should work well both inverted and at high speed (on either side).

Please correct me if I'm wrong (I sort of wrote as I was thinking it).
M.


I have worked through various designs, mostly constant chord 20-26 deg sweep usually between 1.5-2 degs washout (twist)at the tip-ie trailing edge
lifted but several mm at the tip, the same section at root and tip, some with taper but I prefer the look of the constant chord

[Mihai]

That's a sweet looking plane! And yes, it seems that for that large of an AR you do need a bit of twist to bring things back in shape. My toy is on the smallish side with about 30" span, so it's a stubby thing. That program indicates that I don't need washout for this small of a wing. However, I'll certainly add some if I get at that high ARs.

Simons' book is looking at Reynolds numbers all over the place (especially for designing wings :) ). It's true that I didn't actually measure the wing on an SGII to see if it's close to elliptical or not, I just noticed that it's tapered and made of multiple panels (2 I think, and 3 for larger sailplanes like Supra if I recall exactly).

Thanks,
Mihai

[nmasters]

Well, I'm really not aero-engineer, but I did read Martin Simons book and he says otherwise in Section 7.6 as well as Fig. 7.4 (page 77 on the 4th edition). So, while I can't argue intelligently about this, you'll have to take it up with Simons :).
You misunderstood that section, read the last paragraph of 7.6 again. A symmetrical airfoil section has a narrower operating range than a cambered section of the same thickness distribution. The same is true of reflexed sections i. e. decreasing the camber decreases Clmax and causes the stall to occur earlier as measured from the zero lift AoA not the geometric AoA. Compare the airfoil sections in the attachment. I've marked the points of zero and maximum lift with red lines. Notice that although the heavily cambered section stalls at a lower geometric AOA than the lightly cambered one its range of operation is 3 degrees wider and the lift characteristics are vastly different. I've also marked a couple of interesting points on the L/D curve with green lines. Notice that at Clmax the Eiffel 401 has a lift to drag ratio of ~7 at CLmax and the Eiffel 400 has L/D ~9. Since the 401 never reaches the high CL of the 400 let's see what the L/D of the 400 is at the Clmax of the 401. Wow, at CL= 0.9 the Eiffel 400 has L/D=16, nearly twice the 401's. This high L/D at high CL translates into low sinking speed for gliders or high angle of climb for powered planes. If I were going to build a wing with these airfoils I'd want the highly cambered section at the tip with 3 degrees of twist to get the AoA of zero lift aligned (zero aerodynamic washout). That way you never have the tip lifting downward and the root stalls first.

--Norm

[Brandano]

same as adding stall strips on the inboard section as they do on some aerobatic planes. At low AOA the drag is negligible, they sit in the bubble formed by the separation of the airflow above and below the wing. At high AOA they force flow separation. And on a symmetric profile the effect is the same whichever way up the plane is facing.

[Mihai]

You misunderstood that section, read the last paragraph of 7.6 again. A symmetrical airfoil section has a narrower operating range than a cambered section of the same thickness distribution. The same is true of reflexed sections i. e. decreasing the camber decreases Clmax and causes the stall to occur earlier as measured from the zero lift AoA not the geometric AoA. Compare the airfoil sections in the attachment. I've marked the points of zero and maximum lift with red lines. Notice that although the heavily cambered section stalls at a lower geometric AOA than the lightly cambered one its range of operation is 3 degrees wider and the lift characteristics are vastly different. I've also marked a couple of interesting points on the L/D curve with green lines. Notice that at Clmax the Eiffel 401 has a lift to drag ratio of ~7 at CLmax and the Eiffel 400 has L/D ~9. Since the 401 never reaches the high CL of the 400 let's see what the L/D of the 400 is at the Clmax of the 401. Wow, at CL= 0.9 the Eiffel 400 has L/D=16, nearly twice the 401's. This high L/D at high CL translates into low sinking speed for gliders or high angle of climb for powered planes. If I were going to build a wing with these airfoils I'd want the highly cambered section at the tip with 3 degrees of twist to get the AoA of zero lift aligned (zero aerodynamic washout). That way you never have the tip lifting downward and the root stalls first.

--Norm

Actually your figure reinforces what I said: you can have three types of washout (call them whatever you want) and you can mix and match them to achieve the desired performance. It's pretty clear that any single washout method, by itself, has severe limitations, and that's why Simons (and you, above) combine two of the methods (airfoil and twist) to achieve better performance than any single method by itself. However, if are to correct the distribution by airfoil alone (not in combination with twist), then you'll have to have a less cambered airfoil at the tip - again, by keeping the twist zero, i.e., at the same geometric angle (as shown in the graphs), i.e., at different zero angles.

I think that there is no argument about this (i.e., we all agree).

However, there is one point I'd like to argue about - about aligning the zero lift angles after you change the airfoil at the tip. I think that this optimizes high speed behavior and if that's what you want (e.g., for a jet) it's OK. However, if you want min-sink (like a glider), you should probably align the twist such that the min sink of the two are aligned. If you want stall behavior, you should align the stall angles (plus a little reserve for the tip).

Does it make sense?
M.

[CloudyIFR]

I haven't followed this real close but has anyone mentioned Dr. Panknin's theorm on zero lift angles and pitching moments and how he computes the required twist for whatever airfoil is chosen?

I have a spreadsheet on my website that's called Flying Wing Calc that will do all the math for you.

http://h1.ripway.com/cloudyifr/files.htm

Curtis
Montana

[Mihai]

I think that Panknin is getting the pitching moment of the wing equal to zero at some angle of attack (zero lift?). I think that Martin Simons tries to avoid tip stall. I may be wrong (on both accounts) :).
M.

[JetPlaneFlyer]

However, there is one point I'd like to argue about - about aligning the zero lift angles after you change the airfoil at the tip. I think that this optimizes high speed behavior and if that's what you want (e.g., for a jet) it's OK. However, if you want min-sink (like a glider), you should probably align the twist such that the min sink of the two are aligned. If you want stall behavior, you should align the stall angles (plus a little reserve for the tip).

Does it make sense?
M.

You may have a point but for my part all I was quoting is the 'official' definition of Aerodynamic washout, which is related to the airfoils zero lift AoA... If this produces a more tip stall resistant wing or not would depend on the chosen airfoils.