Logo
Builders' Plan Gallery  |  Hip Pocket Web Site  |  Contact Forum Admin  |  Contact Global Moderator
November 19, 2017, 05:25:54 AM *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with email, password and session length
 
Home Help Search Login Register
Pages: [1]   Go Down
Print
Author Topic: Getting Started in XFLR5 - Airfoil and Wing Analysis  (Read 1227 times)
0 Members and 1 Guest are viewing this topic.
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« on: September 08, 2017, 08:01:28 AM »

Here is an attempt to provide a tutorial for the basics in using XFLR5. I'll stress right from the off that I'm still on the learning curve and there are probably better tutorials out there. I am using version v6.09.01beta - there are newer versions available (https://sourceforge.net/projects/xflr5/files/) but this is the one I learnt with.

XFLR5 is a program that runs Mark Drela's XFoil code for airfoil and wing analysis. XFoil models the airflow over an airfoil and simulates effects of Reynolds numbers including separation bubbles. XFLR5 allows you to take that code and use it to assess airfoils in 2D (like a wind tunnel) but also full designs in 3D. As such you can model different planforms, weights and airfoil selections to assess a wing, but also add tail surfaces and fins for a full plane analysis. (XFLR5 has a beta fuselage model but it's not considered realistic enough to be useful.)

Another function of XFLR5 is the ability to modify and create new airfoils. There are a number of ways to do this but the Inverse Design process will not be covered because I haven't got to grips with it myself yet Smiley


Essentially the program has three stages:

> Airfoil selection and design (CTRL+1 takes you to this page in the program)
> Airfoil analysis (CTRL+5)
> Wing and Plane analysis (CTRL+6)

You can choose different out puts for the graphs of the analysis. This means you can show compare two wings for sink rate or two airfoils for maximum CL or whatever parameter you are interested in.
Attached files Thumbnail(s):
Getting Started in XFLR5 - Airfoil and Wing Analysis
Getting Started in XFLR5 - Airfoil and Wing Analysis
Getting Started in XFLR5 - Airfoil and Wing Analysis
Getting Started in XFLR5 - Airfoil and Wing Analysis
Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #1 on: September 08, 2017, 08:16:55 AM »

Begin by opening XFLR5 and starting a starting a new project

File > New Project



Loading Airfoils

The first step is to load the airfoils you want to assess. These are found in the form of .dat airfoil coordinate files. They can be found online in various repositories but the data in the file must be in the correct format for the airfoil to open properly. The coordinates must run from the upper rear round the LE to the lower rear with no commas or punctuation.

Drela Airfoils http://charlesriverrc.org/articles/drela-airfoilshop/markdrela-ag-ht-airfoils.htm are in this format as are the 'Selig format' files on www.airfoiltools.com.

Save the relevant .dat files to a directory on your computer which you can then access through XFLR5:


Start in the Direct Foil Design page (CTRL+1) and open the airfoil file

File > Open > (select .dat file)

The airfoil should then be displayed.

Load as many airfoils as you want to analyse.
Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #2 on: September 08, 2017, 08:41:30 AM »

It is possible to modify and design airfoils at this stage but I will skip on to analysis for the moment.

XFoil Direct Analysis

This page (CTRL+5) is where you 'wind tunnel test' your airfoils. This can be done for its own sake if you need to look at simulated wind tunnel data, but also to build up a series of polars that are then used to assess a 3D wing in the next stage.

To create a single polar for an airfoil go to

Analysis > Define and Analysis

In the dialog box select Type 1 and a specific Reynolds number. (This type gives the equivalent to standard wind tunnel test data) The Ncrit value is an adjustment for turbulence and can be left at the default of 9. At model scales Mach number is assumed to be zero.

The 'trip position' simulates the forced transition of laminar to turbulent flow as provided by a turbulator. A value of 1.0 means no turbulator. If forced transition is required then the turbulator position is supllied as a fraction of the chord ie 0.12 for a turbulator at 12% chord. Having experimented with forced transition I'm not sure that the results it provides are a realistic model at low Re. I tend not to bother with it. (It's worth noting that the polar is automatically named but this does not indicate the Reynolds number. If this is important I add it to the polar name to avoid confusion later.)

Having defined the analysis you then need to run it using the XDirect dialog box.
For a wind tunnel type polar check 'Sequence' and the range of angle of attack (alpha) required.
Then click 'Analyse' and the polar is created.

You can change the colour and style of the polar by clicking in the 'Graph Curve Settings' of the XDirect box.


You can also look at the boundary layer of the airfoil at different angles of attack in the 'Op Point View' (F5) which is useful for low Re numbers.
Attached files Thumbnail(s):
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
« Last Edit: September 08, 2017, 08:57:55 AM by Yak 52 » Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #3 on: September 08, 2017, 08:56:01 AM »

Graph Settings and Views

At this point it's worth noting that you can set the graphs to display any output you want from a list. Just double click on the graph or press G.

You can show 5 graphs at once (shortcut A) or two (T) or go to any particular one by pressing 1 to 5.

The default graphs are
1. Lift/drag polar
2. CL/Alpha
3. Transition point
4. Pitching Moment
5. CL/CD vs Alpha

I also like to look at the power factor [Cl]^(3/2)/Cd which is an indication of the airfoils endurance on minimum power for free flight and sink rate in gliders.


Another useful tool is the screenshot function (CRTL+I) which will save a graph to your computer.
Attached files Thumbnail(s):
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
« Last Edit: September 08, 2017, 09:32:11 AM by Yak 52 » Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #4 on: September 08, 2017, 09:15:29 AM »

Batch Analysis for a 3D Wing

If the purpose is to analyse a 3D wing or plane design then you need to compile a bank of polars that cover the operating envelope of the wing. To do this you use

XFoil Direct Analysis > Analysis > Batch Analysis
(Shortcuts CTRL+5, Shift+F6)

In the dialog box first select 'current airfoil' or select more than one airfoil from the list.

Select Type 1 polars.

Select the range of Reynolds number the wing will encounter, from the tips at the lowest speed to the root at the highest speed. If you do not cover the whole range you will run into problems later with 3D analysis. You can select the min and max Re number and the increment. A smaller increment provides more resolution but is heavy on the computer and takes longer to run. The Re list is a more efficient way with limited processing power for bigger projects.

As before you need to select the range of alpha's to run. Minus 2 degrees to plus 12 degrees covers most applications although -5 to +5 may be appropriate for symetrical airfoils and tailplanes.

At this point click 'Analyse' and go and make a cup of tea.... or several Smiley
Attached files Thumbnail(s):
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #5 on: September 08, 2017, 09:18:41 AM »

Saving

After the airfoil data has run it's a good time to save!

In fact XFLR5 is notorious for crashing and it's important to back up anything important. My version occasionally crashes while saving which corrupts the file so I regularly 'save as' a new iteration so I don't lose all my work.
Logged
USch
Gold Member
*****

Kudos: 17
Offline Offline

Italy Italy

Posts: 874




Ignore
« Reply #6 on: September 08, 2017, 09:40:43 AM »

....heavy breathing....
I just turned my back to the computer and swosh..... Jon started a new topic   Smiley Smiley Smiley

Now I have to find a good excuse to tell my wife that for the next day's I will be occupied with a lot of work.

Thank you Jon to have taken my plea so seriously and started this argument. I will report back as soon I get to open XFLR5.

Urs
Logged

Fast up-Slow down
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #7 on: September 08, 2017, 09:46:33 AM »

No problem Urs, hope some of it's useful. Just the basics so far... More to come.
Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #8 on: September 08, 2017, 11:18:09 AM »

Wing Design

Having created a bank of suitable airfoil data we can now move on to create a wing in 3D in the Wing and Plane Design section [CTRL+6] of XFLR5. Generally this means inputting the geometry from an existing design drawing - panels spans, chords, sweep, twist and airfoil sections.

You can choose your units in the 'Options > Units' dialog box before hand.

Access this panel from:

Wing-Plane > Define a New Wing
(Shortcuts CTRL+6, F3)

Having created the wing remember to reset the panel mesh. I've also read that its worth increasing the spanwise panel density for a better result. This can be done by doubling the Y-panel values.

Be wary of trying to create complex exact planform shapes, especially at the wing tips. This can increase the panel density too much and introduce super low Reynolds numbers (in an area where the flow is not chordwise anyway) which leads to unrealistic results for span-wise efficiency. It's better to be approximate with 3-4 span-wise wing panels.

Save and Close the dialog box and you now have a wing ready for analysis (Pic 2).

(Note in my version I have to exit and then Rename my wings. Whatever the system you use make sure it works for you. I use something like Wing/Iteration/Airfoil ie E36_1 BE50)


Editing and Duplicating Wings is straightforward enough in the 'Wing-Plane > Current Wing' menu which enables you to make wings with different airfoils, twist or tweak planforms etc.


Again save your work at this point and perhaps consider saving the wing as a separate project.
Wing-Plane > Current Wing > Save as Project...
Attached files Thumbnail(s):
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #9 on: September 08, 2017, 11:50:30 AM »

Wing Analysis

You should now be ready to analyse the wing. Select the wing to be analysed and go to the Wing and Plane Design (CTRL+6) polar view (F8)

Go to

Polars > Define an Analysis (F6)

In the dialog box you have a number of polar Type options but 'Type 2 - Fixed Lift' is the one that simulates a real wing flying. You then have to enter a Plane Mass which equates to the expected all up weight.

Of the three 'wing analysis methods' I use VLM (Vortex Lattice Method) and stick to that. LLT (Lifting line theory) is a simpler method and the 3D method is, I believe a beta method in XFLR5 and less reliable.

Click 'OK' to close the dialog box and you will see the analysis ready to run in the 'Miarex' side bar. As before, check the 'Sequence' box and select the range of angle of attack you want to analyze. If necessary adjust the polar colour and style.

Click 'Analyze' and hopefully, if all has gone right, a 3D wing polar will be generated! Smiley

If no polar shows then you will have to do a little trouble shooting. It's probably because you don't have the 2D data needed for the wing's airfoil, or perhaps the data doesn't cover the required Reynolds number range.
Attached files Thumbnail(s):
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #10 on: September 08, 2017, 12:09:21 PM »

Understanding the Graphs

Now that you've generated a 3D wing polar, it's worth explaining what you are looking at. As before you can set up the four graphs to show almost any parameter (double click or press G to change outputs.) And the same commands (A for all, T for two, 1,2,3 and 4) show particular graphs. Here are a few useful ones.


Sink Rate

In XFLR5 sink rate is given as Vz (velocity in the Z axis.) This equates to the wing with the best glide duration. It equates to 'hang time' in a FF glider albeit in a dead air, straight line flight scenario. In a powered model with a better sink rate will require less power to sustain level flight. In RC gliding it equates to thermalling ability (although other factors such as turning performance are involved.)


Lift/Drag Ratio

Comparing CL/CD against speed (Vx) and alpha is the standard polar for gliders, showing whether it will be a floater (a good CL/CD at low speed) or have better penetration (good CL/CD at higher speed.)


Best Lift/Drag Ratio

This equates to a models best range from a given height, ie it's ability to cover ground. In RC gliders this is actually less important that the penetration - ability to cover ground at a higher speed without loss of height. To assess this I tend to look at Vz vs Vx (forward speed vs sink rate.) And gamma vs alpha (Glide angle vs angle of attack.)


The 'Op Point View' F5

This shows spanwise data for things like lift distribution, induced angle, airfoil transition and bending moment. It's useful for smoothing out and improving efficiency with dihedral breaks, airfoil selection and planform optimisation (perhaps more advanced techniques.)



Basically, you need to know what you are looking for and what 'good' is for a particular model class or aerodynamic task. This enables you to compare models and optimise for a particular parameter, (weight, planform, airfoil) or perhaps for more than one in some cases. For example in DLG and CLG you want a good sink rate but not at the expense of drag at zero lift (use graph of Total Drag vs CL.)

A tip for optimising RC gliders for both sink rate (Vz vs CL) and penetration (CL/CD vs Vx) is to adjust the wing loadings of each wing until they all have the exact same sink rate. When this is achieved the one with the best penetration is the better wing.


Ideally for optimisation you need a methodical method for tweaking a particular parameter (airfoil camber, planform in CAD etc) and then create a number of wing iterations to zero in on the best. The more work you are prepared to put in the better the wing design will be.
« Last Edit: September 08, 2017, 12:21:25 PM by Yak 52 » Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #11 on: September 08, 2017, 12:57:48 PM »

Airfoil Modifications

One obvious use of XFLR5 is to 'try' different airfoils on a wing to see which will be best for your model.

Within the program there is a method - Inverse Airfoil Design - for coming up with entirely new airfoils to suit a particular need. This is the best way to design but I'll freely admit I haven't got around to learning this yet so we are stuck with modifying other peoples airfoils to try and improve them for a particular use.

This can be done in the Direct Foil Design section (CTRL+1)


Scaling Camber and Thickness

This allows you to change an airfoil thickness and camber to suit a particular need. For example AG19 is a nice airfoil for RC gliders with 5.4% thickness and 2.27% camber. This is great for a glider with the need to fly fast but more camber could be useful for a FF model.

Select the airfoil then
Foil > Scale Camber and Thickness (F9)
Enter the new values in the dialog box and save under a new airfoil name eg 'AG19_3.5%camber'

Run a batch 2D analysis of the new foil in X-Foil Direct Analysis (CTRL+5) and then make a duplicate wing in Wing and Plane design (CTRL+6) with the new airfoil. Run an analysis of the new 3D wing and see if sink rate is improved.

You can create a number of iterations of an airfoil like this to establish the best camber value for the wing.


Flap Settings

These are fairly self explanatory but this enables you to create a flapped version of any airfoil, run 2D data for it and then add it to a flapped iteration of your wing.


Airfoil Splines

It is possible to visually modify or create an airfoil using the spline airfoil function. This works much like a spline in CAD. You can also load a background image of an airfoil from a drawing or a scanned dihedral break and trace it using the spline. This is a slightly crude method but I have had some surprisingly good results with it. (View > Load Background Image) The spline is then 'normalised' and 'derotated' to bring it into the standard format.


Modified airfoils can be exported to .dat files and saved for further projects. You can also use XFLR5 for blending between two airfoils to give individual rib coordinates. With a little manipulation these can then be exported to CAD for plans.
Attached files Thumbnail(s):
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Re: Getting Started in XFLR5 - Airfoil and Wing Analysis
Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #12 on: September 08, 2017, 01:01:31 PM »

That's about all I have for the moment Smiley A fairly 'whistle stop tour' but hopefully useful for someone wanting to get started in analysing their own wing designs.

Jon
Logged
piecost
Gold Member
*****

Kudos: 6
Offline Offline

United Kingdom United Kingdom

Posts: 262



Ignore
« Reply #13 on: September 08, 2017, 03:14:21 PM »

Jon, Thanks for the posts. I really must try it sometime.
Logged
flydean1
Gold Member
*****

Kudos: 5
Offline Offline

United States United States

Posts: 413



Ignore
« Reply #14 on: September 08, 2017, 08:11:12 PM »

Is there some easy to use software for plotting ribs, spar-slots, LE & TE sections, in addition to plotting angled "geodetic" or "Union Jack" construction?

I'm not interested in analyzing anything.  I just want to enter coordinates and plot ribs for a model.
Logged
Hepcat
Platinum Member
******

Kudos: 207
Offline Offline

United Kingdom United Kingdom

Posts: 1,678



Ignore
« Reply #15 on: September 08, 2017, 08:38:57 PM »

Flydean (#14),
I would expect any CAD program on sale nowadays would do that sort of thing with no problem.  I use 'DesignCAD' and I think many modellers use'Draftsight' because it is available as a free download (although it has never worked for me!).
John
Logged
fred
Gold Member
*****

Kudos: 2
Offline Offline

Canada Canada

Posts: 265



Ignore
« Reply #16 on: September 12, 2017, 03:33:16 AM »

Clearly Someone has taken a gulp of the Kool Aid :-)
 I did too once.. I fooled with R5 years ago.
  Quite the Bunny hole it was too. 
Sucked up More of my life hours... than I'm willing to admit to.
Frustratingly it proved neither overly accurate nor real world useful.
 The resultant wing simply fell far short of Screen expectations. 
As in the darned Model didn't do half as well as predicted.
My interest was a Horten.. so not any easy challenge for either me or the software.
Plus I could easily have gotten it All wrong.. as could have, the software sim.
  Hopefully others  experiences may be more fruitful.
 Our friend Mr Reynolds continues to have unchallenged Dominion.
Prediction software (at least in the 'Free' varieties that can run on a typ home processor ) are still more show than go  IMO.
 Just MY 2 cents is all.
Logged
Yak 52
Titanium Member
*******

Kudos: 53
Online Online

United Kingdom United Kingdom

Posts: 2,161


Topic starter
Free Flight Vagrant



Ignore
« Reply #17 on: September 12, 2017, 04:32:56 AM »

Hopefully others  experiences may be more fruitful.

Yes, mine have been. I've used this tool for a number of years alongside other design work (CLG, Bostonian, 36" Glider, F3-RES RC gliders) and found it useful and interesting.

The real power of XFLR5 is in comparison of wing designs rather than simulation of real world predictions, and also in designing airfoils that control laminar separation bubbles for reducing profile drag at low Reynolds numbers. The skill is in understanding what it is telling you about the real world result and yes, putting in the time to improve a design.

If there is any doubt as to the real world impact I would suggest a bit of digging into the work of Gerald Taylor and others who have taken F3K (1.5m discus launched glider)and latterly F5J (electric launched glider) design to the next level by means of wing design optimisation in XFLR5.
Logged
USch
Gold Member
*****

Kudos: 17
Offline Offline

Italy Italy

Posts: 874




Ignore
« Reply #18 on: September 12, 2017, 01:55:46 PM »

I think a common error of us all, me included, is to expect a software which tells you:
DO IT LIKE THIS AND THAT AND YOU GET THE WORLD-BEATER YOU WANTED.

Hard to understand that no scientific software will do that, but it will only tell you what your design might be able to do in terms of efficiency-speed-glide ratio and so on. The software will tell you what it does, not how to design it to get better. You must do a second, third model and compare the numbers, iteration they call it.

I understood this sort of thinking time ago working with a yacht designer. For a new project he designed, with the help of his computer of corse, 800 hulls and had them competing one against the others in different, virtual conditions of wind and waves. At the end of this long process he build a good yacht, but not an exceptional one. Now for us modellers maybe 3-4 virtual models are enough to judge and choose what may work pretty well.

Fortunately men's brain and intuition has still a lot to offer  Wink

Urs
Logged

Fast up-Slow down
mikeyc38
Nickel Member
*

Kudos: 0
Offline Offline

United Kingdom United Kingdom

Posts: 1



Ignore
« Reply #19 on: October 24, 2017, 10:29:27 AM »

Is there some easy to use software for plotting ribs, spar-slots, LE & TE sections, in addition to plotting angled "geodetic" or "Union Jack" construction?

I'm not interested in analyzing anything.  I just want to enter coordinates and plot ribs for a model.

Hi Flydean1 you should have a look at DevWing http://www.devcad.com/eng/devwing.asp. You don't even need to enter co-ordinates, just select the aerofoils you require and design away. The CAM version will generate GCode to go straight to your CNC router or laser Cutter. The basic version exports to .dxf and you load this into your CAM program to generate Gcode. Failing that, just print the wing sections to paper and stick them on the wood. I use Profili Pro for years which is not as sophisticated in terms of the wing design but good enough for me at the moment. I will upgrade to DevWing soon however as I will have the need for more sophisticated wing plan forms.

Regards
Mike
Logged
Pages: [1]   Go Up
Print
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!