Experiment 1 - my attempt to learn design - RC Groups
Thread Tools
Jan 04, 2011, 05:05 PM
ThomasLee's Avatar
Discussion

Experiment 1 - my attempt to learn design


I have been trying learn how to use XFLR5 on and off for a couple months, but have finally committed myself to sit down, read as many tutorials as I can, go through the (lengthy) threads around RCGroups, and take a chance with the program.

I'm bound to have a LOT of questions as I go on, so I thought I could ask my questions in this thread, and share my progress with anyone who is interested. At the end of this, I hope to learn how to mould (yep, read through most of the threads I can find on moulding DLGs a couple times each already) gliders, and maybe even get parts of this plane built by the end of the year.

Too early to give it a name? Maybee..... but.... I'll call it the Shift for now anyway

DESIGN GOALS

-Zone V2 airfoils
-Optimized for launch and run
-Relatively easy handling
-AUW ~260g, although it is much too early to tell

CONSTRUCTION GOALS

-Relatively easy to mould - no washout Have now decided to do washout... might as well put in that little bit of extra effort!
-1-piece fuse, preferably with stab mount integrated
-Low wing


Goals will likely change over time as the project progresses, we shall see


So to start it off, I have been playing around with the wings, using the Zone V2 foils and its interpolations, and have got it to have a fairly efficient local lift distribution, and what I think should be a fairly tip-stall friendly wing...

However, I don't understand the two graphs in top, the "Induced angle" and "Total angle" graphs.

What do they represent, and how should they be used to design a dlg wing?
Last edited by ThomasLee; Jan 08, 2011 at 02:03 PM.
Sign up now
to remove ads between posts
Jan 04, 2011, 05:19 PM
ThomasLee's Avatar
Here are some very informative readings I have read or am in the process of reading (I'm sure I'm missing some of the stuff I've read, will look for them and this list be updated over time, if you have a good link PM me and I'll add it into the lists):

Tutorials:
http://www.rcsoaringdigest.com/pdfs/...SD-2008-02.pdf
http://www.mjw.co.at/downloads/XFLR5-tut-v1.pdf
http://xflr5.sourceforge.net/docs/Re...Prediction.pdf
http://xflr5.sourceforge.net/docs/Po...t_Envelope.pdf
http://xflr5.sourceforge.net/docs/St...is_Rev_0.5.pdf
https://www.rcgroups.com/forums/show....php?t=1110049
https://www.rcgroups.com/forums/show....php?t=1163553

Build/Design logs:
https://www.rcgroups.com/forums/show...zone+cg&page=2
https://www.rcgroups.com/forums/show....php?t=1008654
https://www.rcgroups.com/forums/show....php?t=1097383
https://www.rcgroups.com/forums/show....php?t=1356201

Misc:
https://www.rcgroups.com/forums/show...&postcount=204
https://www.rcgroups.com/forums/show...g#post13851988


Here's a snapshot of the wing design

Note, there is no washout, I want to stick with something slightly easier first before getting into washout.
Last edited by ThomasLee; Jan 18, 2011 at 01:28 PM.
Jan 04, 2011, 05:34 PM
ThomasLee's Avatar
some other graphs showing the 3 slightly different planforms/distributions I've been playing with today... still trying to make sense of the graphs, still lots of reading!
Latest blog entry: Vibe Assembly Instructions
Jan 04, 2011, 05:41 PM
G_T
G_T
Registered User
Any chance you can zip up and post the project file?

Gerald
Jan 04, 2011, 05:51 PM
Guz
Guz
Gutless wonder
Guz's Avatar
Could you post your list of tutorials?

I too am diving into XFLR for a project, and I need all the help I can get.
Jan 04, 2011, 05:56 PM
G_T
G_T
Registered User
Speaking just for myself, I'm pretty tied up in projects, but I figure if the project file gets posted, I might be able to generate a couple of pictures to show some stuff. I'm sure all designers have their own methods and I have mine, so I don't want to guide too much...

Gerald
Jan 04, 2011, 05:56 PM
ThomasLee's Avatar
Gerald, I'll zip it and upload it later when I get home. It might be a bit messy though, I will likely have to start a new file soon now that I am starting to understand how to use it.

Guz, I'll get a list up tonight, the 'tutorials' I'm looking at are on the XFLR5 site, I'll include them, and also the threads I'm looking at, some are DLGs and some aren't, but they're all worth reading IMHO
Latest blog entry: Vibe Assembly Instructions
Jan 04, 2011, 05:58 PM
ThomasLee's Avatar
Thanks Gerald, that will be really helpful, since I don't know enough about it to have developed my own methods yet.

It will be helpful if other designers can share their methods as well, there's just so much to it that I'm sure most people will learn something from anything
Latest blog entry: Vibe Assembly Instructions
Jan 04, 2011, 06:11 PM
Registered User
I'm certain that a bunch of us will be looking and will certainly appreciate this thread.

Gary
Jan 04, 2011, 06:21 PM
ThomasLee's Avatar
I guess I should document all (or as best I could) the things I have learned so far, and as I continue to learn it. Maybe it'll be boring or too basic for some of you, well too bad

My biggest D'oh moment was that I assumed that when I ran a batch polar analysis for the airfoils in the "X-Foil direct analysis" was that it will run the analysis for ALL the airfoils you have imported for the project... I was wondering why I got the analysis done so quickly! Note that this is NOT the case, and you need to run a batch analysis for ALL the airfoils.

To run the batch analysis, it is under the "Polars" -> "Batch Analysis" menu, and a window will pop up. These are the settings recommended in one of the tutorials:

Type 2 (fixed lift)
Plane mass: 270 grams (I think Gerald uses 300grams, I chose 270 because that is closer to what I have been flying)
x cog: 75mm
VLM2 analysis, with viscous correction

Instead of using a Re range for the analysis, Gerald recommended to use a list, using a series such as:

3K
5K
10K
15K
20K
25K
30K
40K
50K
60K
80K
100K
120K
140K
170K
200K
250K
300K
350K
400K
500K
Last edited by ThomasLee; Jan 04, 2011 at 07:03 PM.
Jan 04, 2011, 06:35 PM
G_T
G_T
Registered User
Looking too briefly at what you have posted, a couple of hints:

1) Watch out for curved or segmented hingelines. When doing wing designs, if working with foils where the flaperons are not in the default position, it is more visually obvious. In natural camber, it is hidden.

2) Don't pay all that much attention to the lift distribution except when around optimal cruise and float settings. Total drag matters more than the distribution of lift in speed mode. For cruise and float you need to generate some additional foils and their polars.

3) You can usually ignore Cm vs alpha and Cl vs alpha, for a wing. Look at Vz vs Vx (sink rate), L/D or gamma vs Vx (glide ratio or glide angle), and the same graphs relative to Alpha. Reduced Vz vs Vx is lower sink rate. Increased L/D vs Vx is better air coverage. A wider plateau of either with respect to Alpha is more forgiving.

4) Total angle is induced angle + geometric angle (washout). Since you have no geometric washout in your wing design, those two graphs are the same. Induced angle is an adjustment to the effective angle of attack seen by the airfoil at each station, due to the influence of the rest of the wing. When the total angle as one approaches the tip is increasing, that may be a warning sign. It can lead to tip stalls when the plane is flown slowly. But the graphs you presented were for a rather low lift coefficient, flying quite a bit faster than optimal for speed camber. You need to look at what the wing does closer to speed camber.

5) Turn on upper surface transition point on your second picture in post #2. Then run animation over the range of angle of attack and watch how the transition point behaves. Ideally, you want the transition point curve to flow smoothly from the trailing edge towards the leading edge as angle of attack increases. The curve should be smooth, not kinky. If it is kinky, then I wouldn't really believe the analysis anyway. If the transition point progresses farther forwards towards the tips then it does towards the root of the wing, that indicates the tip foils are being stressed a bit more than prehaps is good. The solution is typically to use one of these options:

5a) Widen the tips
5b) Increase washout
5c) Redesign the tip foil to delay transition point advance
5d) Shuffle the foils used in the wing design closer to the tip, reducing the tip foil's influence

The reason for 5d is typically tip foils are thinner and less curved than root foils. Thinner, less curved foils have a more rapidly advancing transition point, all else being equal.

6) Take a look at total drag across the span of the wing as Alpha changes. Regions where drag shoots up are regions that could perhaps use some work. Treat it as an indicator of places to consider putting optimization effort. Same thing for lift distribution. That should be near optimal when one is flying at near optimal speeds for the camber settings. But elliptic is NOT really the optimal lift distribution, due to Reynolds numbers effects. Not if the goal is maximizing performance. Elliptic will lead to tips that are too narrow, having higher drag coefficients, reducing the performance of the wing as a whole. You want to minimize the drag of the tips as best you can. I did a lot of that for you with the tip airfoil, but you have to finish the job! Consider L/D for instance. The tips may not have much area, but they function at a lower lift coefficient than the rest of the wing. They also function at a substantially higher drag coefficient. The smaller the chord, the less area, but unfortunately the lower the L and the higher the D coefficients. At some point as one makes a tip smaller, one crosses over into a region where the smaller tip actually hurts the plane's performance.

If one were to take a single airfoil, such as the Zone-V2-20 tip foil, and run a series of polars over a set of Re such as {5K, 10K, 15K, 20K, 25K, 30K, ...}, then one would readily see the non-linear drag coefficient increase as Re decreases. One would also see that at some critical Re, the flow stays fully laminar - even if it separates. Lift is a lot lower then, and drag is higher. You don't really want the foil operating in this range when the wing is being flown at near optimal airspeeds.

7) Speaking of optimal airspeeds, the foils were designed for optimal lift coefficients at certain camber settings. The lift coefficients for the majority of the wing are roughly speaking 0.4, 0.6, and 0.8, for cambers of 0, 2, and 4 degrees. If the wing appears optimal when at 0 camber, but flying at a lift coefficient far off of 0.4, then something is wrong.

To determine roughly the lift coefficients that should be used for the foils, one can generate Type-II polars for each of the foils at each camber setting. Look at L/D vs Cl, and power factor vs Cl. The peak of the former will give the optimal lift coefficient for max L/D, and the peak of the latter will give the optimal lift coefficient for min sink rate. These will differ, sometimes by a lot.

Just for giggles, take a wing design and determine its optimal alpha (or speed) for max L/D for some camber setting. Go over to the operating point, and determine the total angle at each wing station, given that optimal alpha. Take those total angles, and go back to the Type-II polars for these stations, and put a mark at each foil's operating point. It can be most instructive, in pointing out how the foils are being misused! Don't expect it to come out right. Just try to keep it from being ugly!

8) Ncrit affects the results. Lower Ncrit is more forgiving, and probably representative of more microturbulent air (if I can coin a phrase). In the best analysis, in all likelyhood one should use a slightly lower Ncrit value for the polars generated for the foils as one approaches the tips. This is due to the increase in the leading edge sweep angle. This is likely part of the reason why pointy tips can be gotten away with, without severe penalty, in practice. But when I design, because I want the tips friendly, I take the pessimistic approach and assume the leading edge sweep isn't going to help me. So I use the same Ncrit across the wing. I may run results for a range of Ncrit however. It is not necessarily the case that a wing which performs really well at some Ncrit will perform really well at another Ncrit. But presumably the foil designer took steps in this regard...

Gerald

PS - I usually use an Ncrit of 9 for DLG wing analysis. I'm not going to even try to defend that number. Look up posts by me and Kevin and probably Drela and others on the subject. It is not easy to determine an appropriate number to use.
Last edited by G_T; Jan 04, 2011 at 06:47 PM.
Jan 04, 2011, 06:39 PM
G_T
G_T
Registered User
I'd actually recommend running those polars as Type-I rather than Type-II. Type-II is wonderful when designing the foils as it is most representative of what happens when speed changes in level flight. I use them for that.

To speed the process, at some small loss in precision, you can do what I do. Take each foil, and do a global adjustment to 161 points. I don't like going fewer than that. The foil will now process two or three times as fast, and generate results close enough to the original that they can be used for optimization purposes, which is all you really need.

Gerald

PS - I typically use a range of plane weight from as low as 250g to as high as 450g. I optimize often for 300g, a lightly ballasted condition.

PS2 - I presume the 12K in the list is 15K?
Jan 04, 2011, 06:45 PM
G_T
G_T
Registered User
I had thought about starting a thread such as this one many times, but I figured it would slow me down too much to go through the whole process while actually trying to get something done! Kudos Thomas for starting it!

I'm not going to get to look at the project file tonight anyway, so don't waste your time posting it just for me.

Gerald
Jan 04, 2011, 07:02 PM
ThomasLee's Avatar
I've just read through the post, and will try to implement those suggestions when I get home. The one biggest thing I've realized so far (and I've only just started) is the sheer amount of time it takes to tweak, and analyze, and tweak, and..... thank goodness some of the time is just computation work.

Correct, 15K, I'll correct that.

I'm using 9 as NCrit as well, I believe I read that in one of your posts so decided to run with it.

I really hope this thread can help others as well as myself, and push some people who have wanted to design/build their own design but was lost as to how to start, especially because many things that seemed 'kindergarten' for many is quite confusing for those not in the know, and is therefore never brought up in tutorials and threads. Guess what, if I start a thread, that means I'll be forced to commit as well! Can't be half-a$$ed and ditch a project halfway when everyone on RCG knows about it!
Last edited by ThomasLee; Jan 04, 2011 at 07:18 PM.
Jan 04, 2011, 08:41 PM
ThomasLee's Avatar
I will hold off on uploading the file, will be starting off on a new file because the current one is a bit messy from a lot of early experimenting sort of thing... will start working on some of the suggestions Gerald has mentioned above into the new file / wing.

I will be updating the file as I progress along, so that anyone who wants to can look through it to give suggestions, comments, or heck, maybe learn something ?!?!?!???
Latest blog entry: Vibe Assembly Instructions


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Build Log my first attempt at 1/72 scale planes (build log with video) Heliman420 Scratchbuilt Indoor and Micro Models 1257 Jul 14, 2011 01:26 AM
Build Log My design and my first build from scratch Foskje Foamy EDFs 20 Jan 07, 2011 10:08 AM
Discussion My first FPV chase attempt(GoProHD) GoldCraft FPV Talk 5 Nov 14, 2010 10:16 PM
Discussion My attempt at a camera holder [sort of adjustable] - Ooops, Mod 1 = rubber band! turnbui Aerial Photography 6 Sep 21, 2008 05:46 AM