XFLR5 vs Nurflugel (Frank Ranis)

weiß jemand, warum so unterschiedliche Ergebnisse?
does anybody know why so different results?



XFLR5 / Nurflugel

Alpha 1º / 2.3º
CA 0.09 / 0.21
Stabilitätsmass 8.75% / 5.43%
Geschwindigkeit 20.7m/s / 13.67m/s

der gleichen Ebene in beiden Fällen
the same plane in both cases

http://www.das-nurfluegelteam.de/co5_2m.html

Danke

Did you specify a CG location?

Looks like both software calculate different neutral points or MACs.

What positions of neutral points do they calculate?

Seems Nurflügel predicts it a bit more forward than XFLR5.

Kurbel

Hi Edu_74,

Nurflügel does not handle dihedral, while XFLR5 does. I guess that should explain the biggest difference.

Greetings,

Andrés

CG is set at 193mm in both cases and no dihedral

To get the neutralpunkt with XFLR5, you have to move back the CG until the graphic Gcm vs Cl (really Gcm vs alpha) has no slope. At this point it´s easy to get the CG knowing the MAC and stabilitätsmass.

Hola Edu_74,

if you don't have dihedral, then the second screenshot you posted is somewhat misleading, because I see at least winglets .

XFLR5 has three different methods implemented (VLM1, VLM2 and Panel). The Panel method should be the most precise. Compare all three to get an idea of how much XFLR5's results depend on the method. Second, XFLR5 includes viscosity unless you tell it to do an inviscid calculation. Of course, if you compare an inviscid (=Nurflügel) to a viscous solution (=XFLR5), you will get different results. If all three methods of XFLR5 deliver more or less consistent results, I would recommend you to trust rather XFLR5.

Muchos saludos,

Andrés

Hi Edu_74,

I don't think that XFLR5 fixes CG like Nurflügel does. At least I cannot see it in the screenshots you posted. What you usually do with XFLR5 is calculate for one or a series of AOA (you get the curve shown in your third picture from doing the latter) or CL instead of AOA. It doesn't care about a fixed CG at all.
What is shown in your XFLR-Wing-Picture (lower right corner) is the CP (Druckpunkt) position but not the CG.
What you can do to see if both programs match is to calculate with a fixed AOA of 2,3° in XFLR and see if the XCP match or fix the AOA in Nurflügel and see if you get similar CL and XCP.
Anyways, I'd recommend to use the newer FLZ Vortex from Frank Ranis instead of Nurflügel since you can also simulate winglets there. The effect of winglets is that they usually raise the cl in the out wing region which results in a higher overall CL and shifts back the XCP by a small amount.

Friedmar

Andrés:

Here are the results with VLM1 and VLM2, both viscid (inviscid has no sense graphics), 3D panels viscid and inviscid get crazy graphics too. VLM1 and VLM2 get very similar results.

Friedmar:

I´ll try what you say too.



Danke

Hola Edu(ardo?),

it would be easier to help, if you post here your Ranis file.

Saludos,

Andrés

Ja, mein Name ist Eduardo

Hola Eduardo,

mira, si le calculo con mi programa (sin winglets) obtengo sin viscosidad los siguientes valores:

Anstellwinkel (auslegungsAlpha) = 0.857047 grad
Auslegungs CA = 0.108993 CD = 0.000248931 GZ = 437.844 SZ = 144.55
Fluggeschwindigkeit 19.3253 m/s

Estos valores se parecen muchísimo a los del XFLR5. Si hago un cálculo con viscosidad obtengo valores aún más pequeños:

Anstellwinkel (auslegungsAlpha) = -0.384836 grad
Auslegungs CA = 0.0174168 CD = 0.00622994 GZ = 2.79565 SZ = 0.368949
Fluggeschwindigkeit 46.9104 m/s

Usé el centro de gravedad del Nurflügel que era 0,194 m. El valor CA=0.2 del Nurflügel es exagerado. Espero que esto te haya ayudado algo.

Muchos saludos,

Andrés

Andrés, when you say with your program, what program you mean?

Flz Vortex is only a demo http://freenet-homepage.de/frankranis/flz_vortex.html

Danke.

Andres has created his own software for that job.
But it does not yet feature a graphical user interface, so it would not quite fit the needs of the average user.

The free version of Vortex made by Frank Ranis is identical to the full version, except for one point.
That is, you can't save more than one file without changing the name of the foregoing file manually.

Kurbel

Dear all,

indeed, I have also wittnessed a rather significant difference between XFLR5 and Ranis-Nurflügel:
In fact I have tested two swept wings (27° at nose) with MH62 and TL54mod, respectively. In both cases the CM0 is significantly lower in XFLR5 than in Ranis. The position of the neutral points are identical.

Bei Xwing hatte ich Konvergenzschwierigkeiten wegen unklarer Definition der Segmente auf der Skelettlinie und "Ranis-Vortex" habe ich leider noch nicht getestet.

Gruß,
Andreas

Could you tell me what program (and its setup or parameters) can I use to get accurate results?

Danke

Eduardo,

I sincerly do not think that there is an ideal accurate program. It depends much on what you want to develop.

I am sure that everybody agrees that for easy usage Nurflügel is probably the best program to use. It has some limitations and precission is suboptimal for thick airfoils. However, it is easy to use and many good flying planes have been design with it.

The FLZ_Vortex is a follow up of Nurflügel. It contains some interesting extensions, for example for dihedrals. Nonetheless, people have less experience with it, because it is used less (restricted "demo" version). I'm not sure, if enough Nurflügels have been design using FLZ_Vortex to have an estimate on accuracy.

Both Vortex and Nurflügel use quiet old methods (some variation of the lifting lines method proposed by Prandtl something around 1900). These methods are not accurate for thick airfoils. In theory these methods work only for cambered plates in ideal fluids, i.e. no thickness at all and no viscosity.

XFLR5 offers a panel method and two lifting line methods (probably rather vortex rings). The panel method should in principle be more accurate and should account for thickness effects. The problem is that XFLR5 is not so easy to use as Nurflügel. Several of the functions contained in Nurflügel have to be done manually in XFLR5.

If you get strange results, countercheck with all programs. And use common sense or ask here in the forum

Greetings,

Andrés

Edu_74 schrieb:

To get the neutralpunkt with XFLR5, you have to move back the CG until the graphic Gcm vs Cl (really Gcm vs alpha) has no slope. At this point it´s easy to get the CG knowing the MAC and stabilitätsmass.

Zum Vergrößern anklicken....

Hallo zusammen

ich bin gerade am XFLR5 anschauen. Da ich hauptsächlich Nurflügler anschaue vermisse ich etwas den Button "Stabilitätsmass". Somit kann ich den SP nicht eingeben, was etwas mühsam ist.

Nach der Formel

s = (xNP - xSWP)/lm * 100

lässt sich der SP ja problemlos berechnen. Die einzige "Unbekannte" in der Formel ist der xNP. Ranis und Vortex berechnen den zwar, aber jedes Programm berechnet diesen etwas anders, daher kann und sollte er aus Ranis nicht auf XFLR5 übernommen werden.

Nun die Frage, wo in den berechneten Werten in XFLR5 finde ich den berechneten xNP?

Die Erklären aus dem obigen Zitat läuchtet mir nicht ganz ein.

Parick

Wenn du den SP solange verschiebst, bis im Diagramm das Gesamtmoment (um den SP) über Anstellwinkel bzw. Auftriebsbeiwert konstant ist, hast du den NP gefunden.
Die Linie muss dann waagerecht im Diagramm verlaufen.
Eine Anstellwinkeländerung hat dann gerade kein Gegenmoment als Reaktion zur Folge, was genau der Definition des Neutralpunktes entspricht.

Kurbel

Hallo Kurbel

danke für deine schnelle Antwort, das wäre wohl GCm über Alpha? Das Problem ist ich habe alles durchprobiert aber kein GCm? Nur ein Xcp? Wie heisst das GCm in Worten?

Patrick

Ich kenne XFLR5 nicht im Detail (benutze es selbst nicht).

GCm sollte wohl der Momentenbeiwert des ganzen Modells sein.
Bezüglich des NPs sollte der konstant sein, also unabhängig von alpha.
Bezüglich eines beliebigen anderen Punktes proportional zu alpha.
Bei negativer Steigung der Kurve hat man ein stabiles Verhalten, bei positiver Steigung instabiles.
Entsprechend einem SP vor oder hinter dem NP.

Kurbel

Hallo,

GCm ist der "Global Pitching Moment Coefficient", also der globale Momentenbeiwert. Müsste in der Auswahl der Y-Achse recht weit oben stehen wenn ich mich recht erinnere. Xcp müsste der X-Wert des eingestellten Schwerpunktes sein, den Du in der entsprechenden Analyse definierst.

Kennst Du die Tutorials von der Seite http://xflr5.sourceforge.net/xflr5.htm? Die sind sehr hilfreich beim Einstig in XFLR5, haben mir auch viel geholfen.

Gruß, Alex