Thread Tools
Sign up now
to remove ads between posts
Oct 01, 2014, 12:56 AM
G_T
G_T
Registered User
Thread OP
I did a quick check on the effect of the thinner tip variant vs the thicker one listed above. This test was with a smaller 2m wing which does not use the inner set of foils; therefore the effect is magnified. Even so, the performance effects were fairly small.

Gerald
Oct 01, 2014, 02:53 AM
Registered User
Oct 01, 2014, 03:50 AM
Registered User
Gerald, this looks like a lot of work based on your similar approach with the wing sections series for DLGs. Thank you for letting the model glider community access to these.

I have one question though: what is the accuracy needed on the wing shape in order to extract every bit of the performances these profils seems to bring? If you could give some indications in % of thickness that would be great.
Oct 01, 2014, 05:03 AM
Registered User
To me an airfoil is kind of like a God, it's basically unreachable, but if you do your best you'll be rewarded.
Oct 01, 2014, 09:49 AM
G_T
G_T
Registered User
Thread OP
Quote:
Originally Posted by fnev
Gerald, this looks like a lot of work based on your similar approach with the wing sections series for DLGs. Thank you for letting the model glider community access to these.

I have one question though: what is the accuracy needed on the wing shape in order to extract every bit of the performances these profils seems to bring? If you could give some indications in % of thickness that would be great.
Essentially if the accuracy is decently within the boundary layer thickness then one should be fine. That thickness is rather small near the leading edge and moderately thick by the trailing edge. Speaking of trailing edges, if it has noticable thickness don't round it off but square it off! The drag should be lower that way.

If one is inaccurate with the foils, it will likely show up as a slight increase in sink rate and slight loss of L/D. Flight behavior at high Cl may well be different, and very high speed performance will likely suffer. Now, there is some possibility that an error in rendering the foils or wing actually makes an improvement for some and even all conditions of flight. The modeling is not perfect and neither is my work!

Probably the worst thing one can do on a wing is have an inconsistent pattern of errors. That will make a plane which has weird flight characteristics; perhaps subtle, but still weird. That will make reading the air and general flying slightly more difficult, placing the pilot at a slight dis-advantage.

For normal and low speed flight, and going up the line, the most critical part of the wing is the leading edge and front third of the top surface, roughly speaking.

For very high speed flight, another critical part might well be the leading edge to the front quarter or so of the bottom surface. But rarely are we flying ballistically!

Errors in leading edge radius will result in differences in how efficiently the wing pulls at high lift coefficients, and changes in drag at high speeds. Small errors would not be very obvious in normal cruising as it would essentially only cause a small change in L/D.

Losing the undercamber in the flaperon area would be bad. The wing would become less floaty and less efficient.

Gerald
Oct 01, 2014, 10:25 AM
Registered User
.wpa ?


Richard
Oct 01, 2014, 12:04 PM
G_T
G_T
Registered User
Thread OP
XFLR5 file format, previous to February this year. Since I've been in development of this series and Synergy-II for F3K, I have frozen my XFLR5 version to one that is not current. I've figured out the bugs that affect results and can work around them. With a new version I'd have to learn a new set of bugs. The bugs are significant in some cases and can lead to completely erroneous but vaguely plausible looking results. For instance, it is easy enough to confuse it between Type-I and Type-II polars.

Gerald

PS - If you are instead asking for when it will get posted, that'll be when I get the time to go in and change the foil names and regenerate polars as needed. This is a very busy week for me.
Oct 01, 2014, 01:08 PM
Registered User
Thank you Gerald for your time and reply.

But, I wish you could give a more "physical" answer as the thickness of the boundary layer you refer to is probably the one the software you are using for your studies/simulations would indicate. Just give us a value in % of chord or, better in % of thickness in the leading edge area and the trailing edge area.

This is of importance as it will show what manufacturing process you need to achieve such accuracy. I do understand your point regarding consistent vs inconsistant pattern errors. This is more related to the manufacturing process. What I am after is the absolute error values at the wing section level where the profile performances you have given start to deteriorate enough to have any influence on the overall performances. Thank you again.

Your statement regarding trailing edge accuracy and shape is certainly very valid and too often overlooked even by "pro" manufacturers. This is definitely an area where you can increase, or hurt, the overall performances of a wing very easily.
Oct 01, 2014, 01:48 PM
G_T
G_T
Registered User
Thread OP
The boundary layer thickness is not a constant but depends on conditions and flight speed.

Gerald

PS -

Building bagged models, getting enough accuracy to really get the wings to perform at their best with this size and AR wing I figure is highly unlikely. However one should be able to do well enough to see differences in flight characteristics vs other models and to produce a fine flying model. If bagging is what you do, go for it! For those doing molded, the people producing the molds are already used to the required accuracy.

I do the majority of my development in inverse modeling and don't have a solid answer to required accuracy. Smoothness and consistency of shape is of greater importance to my mind than absolute accuracy. I could say that I'd like to see accuracy on the order of a thousandth of an inch in the front inch of the wing and that would be true. Can you achieve it? Probably not. Does it affect results? Somewhat. Getting the very leading edge shape wrong in profile rather than just radius has greater effects. The leading edge is NOT just a circular profile merging into a pair of curves.

As for boundary layer thickness, it is perhaps a bit of a mis-nomer. At the molecular level right next to the skin of the wing we presume that air motion is essentially zero parallel to the surface. As one looks away from the surface one sees an increase in airspeed relative to the surface, until sufficiently far away from the wing we see just the undisturbed freestream. The layer near the wing where the air is moving more with the wing than with the freestream we call the boundary layer. If there is a feature on the wing which resides in an area where air motion is small and the feature itself is small enough it is NEARLY invisible to the air flow about the wing. The boundary layer is thicker on the top and thinner on the bottom, and thinnest at the leading edge. That's why leading edge shaping is important as even a small feature there affects flow downstream. However a similarly sized feature near the hingeline on top would be irrelevant.
Last edited by G_T; Oct 01, 2014 at 02:07 PM.
Oct 01, 2014, 02:48 PM
Registered User
I am new to RC glider. I am not new to glider design and airfoil design. It has been years since I tried to mess with airfoil shaping. I use to design my own Nordic Gliders and free flight planes for years. I found out that using normal wing construction for free flight did not produce the same airfoil performance when several planes were built to the same design. Minor changes in the construction such as leading edge shape, spar height, high point location (sanding of rib and finishing sanding of wing, surface roughness ) produced big difference in performance. I had one glider that was covered with Jap Tissue and it performed much different than those covered with Mono Coat. The differences between the two covering was more than surface roughness of the covering. Jap Tissue covering laid across spars different on the top surface which acted as small trips when compared with the the other covering. I also modified airfoils high point location, radius of leading edge, thickness distribution. Sometime with good results and sometimes with no change and other times bad results. All I learned is each case is different and what worked in one regime did not work in another (size, type, weight). My best one was with an FAI Power model and the airfoil was modified for structure more than anything else and moved the high point farther back (from 28% to near 35%). The change in structure to keep a 6% 13/1 aspect ratio from fluttering. It worked and the performance was a leap in performance over all other FAI power models I had. (like from 3:30 dead air time to 6 minute dead air time both flown in same air and same motor performance).

I am planing to build a new ALES glider and am having trouble deciding on best airfoil to use. Thinking of using about a 6% airfoil using built up structure.

I have tried the XFLR5 airfoil program and found I have trouble believing the results. I seems it has some issues with some airfoils as to providing usable results. Two airfoils that you can not hardly tell the difference between them and one has trouble converging.
Oct 01, 2014, 02:56 PM
G_T
G_T
Registered User
Thread OP
Try refining globally to about 150 points and see if that solves the convergence issue. It can also be surface smoothness at a level which is hard to see. If one uses XFLR5's built-in tools to move the thickness peak or camber peak than it destroys the smoothness of the airfoil in an essentially unrepairable fashion. Take a glance at it in inverse modeling to see what I mean.

Gerald
Oct 01, 2014, 06:01 PM
G_T
G_T
Registered User
Thread OP
Here is the XFLR5 project file (.wpa) for the SynerJ reference wing design. Airfoil polars are computed for three camber settings, and wing models are provided for each. Also, derotated versions of the foils are available in cruise camber, for those who might choose for whatever reason to loft or cut the wing in cruise mode rather than the native speed mode.

I've also attached a few screen captures of various graphs from XFLR5 showing a bit about how the wing performs at 6.5m/s in cruise camber, which is not far from best L/D and min sink speeds.

For those doing structural design, pay some attention to the bending moment graph. By the time one is out at half span, the spar needs only a fraction of the strength and stiffness of the root. There is potential for weight savings there.

Gerald Taylor
Oct 01, 2014, 06:22 PM
Registered User



Richard
Oct 01, 2014, 11:03 PM
Registered User
Gerald, thank you again. I do agree with what you say and your approach/analysis regarding the design of a high performance "small" wing. I was interested, in pushing you a bit, to see what order of magnitude you were talking about. Too many people do not realize the degree of manufacturing accuracy needed to extract the performances of a specific wing section, or suite of wing sections.

One thousandth of an inch is equivalent to around 2.5 hundredth of a millimeter. Around the leading edge this translates in a pretty high percentage of error as the thicknesses are rather small. This accuracy is achievable with the right material (for tooling) and the right machinery, but at a very high cost. One shouldn't forget the finishing of the tooling surface that can introduce greater inaccuracy than the machining itself... Unless you go for fancy processes that are pretty much (still) out of reach for model making...

Then you have to deal with the manufacturing process of the wing itself and at such level of accuracy it is very easy to introduce larger than expected inaccuracies (just during the mold closure as an example, but they are other areas where errors might be introduced as well). The curing process and the materials used booth for the mold and in the wing structure will have an influence on the product final accuracy (i.e. different thermal expansion/retraction). As well as the wing structure design and the materials used in the structure. BTW this is an area where I personally think that a hollow wing might be at a disadvantage vs a foam cored wing (both with the same level of design seriousness/requirement of course).

All of this is to underline the difficulty (and the costs involved) to achieve the "perfect" product that these high performances wing sections and wing designs (planform, twist, dihedral(s),etc...) require.

So the debate may shift to a more "philosophical" one: isn't it better to have a "slightly" lower performance wing but more "forgiving" if inaccuracy due to manufacturing or/and handling/use do not degrade the performances as drastically?


Quick Reply
Message:

Thread Tools