[sumit_raised]

Hello everyone,

I have this basic question kept striking my mind that whether we can predict the coefficient of lift of a cambered airfoil (symmetric/non tapered) without first actually getting it tested in a wind tunnel?

If so, anyone please help me with the most basic doubt of mine and help me find a solver that best suits it. ( I basically look forward to FLUENT AND CFX).

I am a total novice in this field and desperate to know it.

Thanks to everyone.

[ricardo.costa]

Originally Posted by sumit_raised

Hello everyone,

I have this basic question kept striking my mind that whether we can predict the coefficient of lift of a cambered airfoil (symmetric/non tapered) without first actually getting it tested in a wind tunnel?

If so, anyone please help me with the most basic doubt of mine and help me find a solver that best suits it. ( I basically look forward to FLUENT AND CFX).

I am a total novice in this field and desperate to know it.

Thanks to everyone.

Hello,

Try to get Xfoil software. You can download it from

http://web.mit.edu/drela/Public/web/xfoil/

You can simulate modified profiles getting the new Cl and Cd curves.

For this simply case, its not necessary fluent or CFX.

Ricardo

[sumit_raised]

Thanks a lot Ricardo for giving your time to write me.

But does that mean that I should convince myself that there is a possibility to get the lift coefficient through FLUENT or CFX?

Thanking You

[truffaldino]

What is the range of reynolds numbers and type of airfoils you are going to test?

If your airfoil is more or less standard and flow is sub-sonic, it is better to use x-foil, since you will get much less accurate result with RANS modelling of turbulence in FLUENT or similar CFD software.

Flow separation with laminar-turbulent transition at low Re numbers is a very weak point of software that uses RANS models.

Truffaldino

[sumit_raised]

Thank you truffaldino

My reynolds number is definitely not to exceed laminar constraints and flow is subsonic. But you gave me an excellent idea about the two software I was confused with.

When I switched to the above link [as stated by ricardo.costa], I found link to many other softwares like xfoil, Javafoil, Designfoil, XLR5 as well.

So I wonder, what models do they use to get all those readings. And moreover if the data of each and every airfoil is predefined in the program then how can they have such low size in the disk.?

Thanking you

sumit_raised

[sumit_raised]

Thank you truffaldino

My reynolds number is definitely not to exceed laminar constraints and flow is subsonic. But you gave me an excellent idea about the two software I was confused with.

When I switched to the above link [as stated by ricardo.costa], I found link to many other softwares like xfoil, Javafoil, Designfoil, XLR5 as well.

So I wonder, what models do they use to get all those readings. And moreover if the data of each and every airfoil is predefined in the program then how can they have such low size in the disk.?

Thanking you

sumit_raised

[truffaldino]

Hello Sumit_raise,

Not of all software you found is quite good. I do not advice to use Javafoil as it uses a quite primitive boundary layer method and could be very innacurate.

XFLR5 and Xfoil are essentially the same: XFLR5 uses Xfoil for each 2D section to get a polar of the whole wing/aircraft.

Xfoil is indeed of a very small size and very fast: basically you could spent weeks of run-time doing Large Eddy Simulation with CFD-sotware like FLUENT, while X-foil will give you a result with the same accuracy just in a fraction of second.

Since Xfoil is based on different principles and does not solve directly Navier-Stokes equation directly on big meshes, it is faster and much more smaller, but it only works for flows with moderate separation regions, where viscous-inviscid bl interaction approximation is valid. To locate laminar-turbulent transition it uses some tabulated data from orr-sommerfeld theory in cojunction with empirical e^n method.

This software gives a quite accurate results for more or less conventional airfoils, but it is not applicable to a flow past cylinders, squares, stepped airfoils and other exotic stuff.


Truffaldino

[sumit_raised]

Thank You so much Truffaldino.

That is very informative for me as a beginner to differentiate them. Thanks a lot.

[akku]

Originally Posted by truffaldino

Hello Sumit_raise,

Not of all software you found is quite good. I do not advice to use Javafoil as it uses a quite primitive boundary layer method and could be very innacurate.

XFLR5 and Xfoil are essentially the same: XFLR5 uses Xfoil for each 2D section to get a polar of the whole wing/aircraft.

Xfoil is indeed of a very small size and very fast: basically you could spent weeks of run-time doing Large Eddy Simulation with CFD-sotware like FLUENT, while X-foil will give you a result with the same accuracy just in a fraction of second.

Since Xfoil is based on different principles and does not solve directly Navier-Stokes equation directly on big meshes, it is faster and much more smaller, but it only works for flows with moderate separation regions, where viscous-inviscid bl interaction approximation is valid. To locate laminar-turbulent transition it uses some tabulated data from orr-sommerfeld theory in cojunction with empirical e^n method.

This software gives a quite accurate results for more or less conventional airfoils, but it is not applicable to a flow past cylinders, squares, stepped airfoils and other exotic stuff.


Truffaldino

Hi truffaldino,

Is Xfoil a good software for obtaining lift and drag coefficients of aerofoils that are in low Reynolds number regime (typical of vertical axis wind turbines) ??.

Thanks

[truffaldino]

Originally Posted by akku

Hi truffaldino,

Is Xfoil a good software for obtaining lift and drag coefficients of aerofoils that are in low Reynolds number regime (typical of vertical axis wind turbines) ??.

Thanks

To my knowledge it is the best software for getting Cl and Cd of 2D airfoils for Re about 10^5 and above

As for other softwares: I have tried RANS with Fluent and it is much less accurate, LES with fluent takes an enormous time and seeems to be also less accurate than Xfoil

[akku]

Hi truffaldino,

Thanks. This reply might be coming after 9 months but nevertheless, it is on course because ever since your last response, I have downloaded XFLR5 and have started playing around with it.

The following are the three methods to find lift and drag coefficients

by RANS method using OpenFOAM
by LES method using FLUENT
by Vortex lattice method using XFLR5

My questions are the following:
1) Are there any other methods to find the lift and drag co-efficient other than those three mentioned above
2) For LES method, is there any Open source free software which is equivalent to FLUENT

Looking forward to your response.

Thanks in advance

Regards,
Abdul Akbar

[truffaldino]

Open Foam has a large libray of LES models. I am unaware of other free software implementing LES. Maybe other people on this forum knows more about this matter?