cfd simulation of an uav

prshet · March 4, 2011, 11:40

i've been working on an uav model with a span of around 2m and fuselage lenght around 1m. the inlet velocity(as per the experimental results) is 21m/s.
the results for Cl are matching closely for lower angle of attack, but the problem is i'm not able to get the right stalling angle. i've generated an unstructured mesh(with prism layers) with around 1.2 million elements using ICEM. i've used k-omega turbulence model, also i've done an itterative convergence study for one case. also as a further part of my work i would like to carryout fluttering analysis for the wing, which needs moving mesh. as i'm new to this please suggest what work should i carry out further....?

ghorrocks · March 6, 2011, 17:12

Capturing the stall point is tricky. Have a look in the literature to find out how other simulation researchers do it. It is not my field so I do not know.

prshet · March 7, 2011, 00:08

ya i'm already working on it. i'm try now for a structured mesh, this will help me have a better control over the problem. but for analyzing a flutter problem for the airfoil i'll have to generate moving mesh. can you please suggest me on this, it would be really helpful.

Attesz · March 7, 2011, 07:10

Getting closer to the stall point the steady state results won't match with experimentals, because of transient effects (growing vortex shedding etc). Switching to transient solver can handle this problem. I would use adaptive time step to speed up the simulation.
In flutter I have no experience also, but this is good challenge in cfd

prshet · March 8, 2011, 07:56

i havent tried with the transient solver, but will it be effective even for lower velocities (around 22m/s). and more over i've used an unstructured mesh(tetra) will it be advisable to go for structured mesh for higher angle of attack.....?

Attesz · March 8, 2011, 10:15

Quote:

will it be advisable to go for structured mesh for higher angle of attack

i don't think so...
post pictures of your mesh!

prshet · March 8, 2011, 13:10

ya certainly, i'll mail them to u. thanks a lot for ur suggestions.

prshet · March 10, 2011, 09:28

sir could i please get your email id....... so that i can mail u the contours and other details.

Attesz · March 11, 2011, 06:02

if you are able to do, post pictures here because other people can see it.

prshet · March 11, 2011, 07:30

i would have certainly posted them here, but i cannot upload attachments in the threads. i feel its not permitted to upload attachments. so it would be helpful if u could give me your email id. looking forward for your reply.

Attesz · March 11, 2011, 08:03

felfoldi.attila@gmail.com but try upload here if you can. maybe you should delete the cookies if you have problems with the site.

prshet · March 15, 2011, 10:01

i have been successful in getting the stall angle and the Cl is matching quite well with the experimental results. what further improvements can be done on the problem....? i've been thinking of drag reduction using winglets but i dont have the data for the specific winglet. what further improvements can be done.....?

icemaniac178 · May 18, 2011, 03:17

mr prshet,
can you share with me how have you been successfully getting the stall angle? i believe i had the same problem with u.

prshet · May 20, 2011, 02:32

ya certainly..... can u tell me the mesh details.....? and wht r ur inlet conditions......? if u pls let me know i would be able to help u....

icemaniac178 · May 21, 2011, 01:17

Quote:

Originally Posted by prshet

ya certainly..... can u tell me the mesh details.....? and wht r ur inlet conditions......? if u pls let me know i would be able to help u....

i`m using unstructured mesh of tetrahedral volume on my rectangular domain. my inlet boundary condition was set to be a subsonic flow velocity with Cartesian velocity value. the turbulent intensity was 5%. my inlet was assign at left and bottom face. do you need more info pls tell me.

prshet · May 21, 2011, 02:54

ya thats alright....wht abt ur geometry.....? hw have u placed it.....? have u considered the whole geometry or only the half part (symmetry). and for different angle of attacks are u moving ur geometry.....?

icemaniac178 · May 22, 2011, 05:43

yes my geometry was taking the simplification of symmetrical condition (half part) and my angle of attack changes were determine by cartesian velocity components define at the inlet. i was not moving my geometry angle at all. what else i shoud do to get a good stall angle prediction?

prshet · May 25, 2011, 09:03

when u give the Cartesian velocity components it might give u comparable results for lower angle of attack. if u have to capture the stalling angle then it would be better if u manipulate the geometry itself....pls try once.... when u rotate the geometry please keep in mind the change in conditions for ur inlet velocity. this time u dont have to give the u and v components. u simply have to give the u component.

icemaniac178 · May 29, 2011, 19:31

i will look for forward to do it. thanks a lot. but do i need to switch into transient simulation for high AOA (17-20 degree) in order to capture better stall prediction? i have gone through a lot of discussion in this forum. many people suggest transient simulation for stalling prediction.

baserinia · May 29, 2011, 21:51

Quote:

Originally Posted by prshet

but for analyzing a flutter problem for the airfoil i'll have to generate moving mesh.

Wing flutter is an example of fluid-structure interaction (FSI). You should look at the FSI tutorial in CFX help.

March 4, 2011, 11:40	cfd simulation of an uav	#1
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	i've been working on an uav model with a span of around 2m and fuselage lenght around 1m. the inlet velocity(as per the experimental results) is 21m/s. the results for Cl are matching closely for lower angle of attack, but the problem is i'm not able to get the right stalling angle. i've generated an unstructured mesh(with prism layers) with around 1.2 million elements using ICEM. i've used k-omega turbulence model, also i've done an itterative convergence study for one case. also as a further part of my work i would like to carryout fluttering analysis for the wing, which needs moving mesh. as i'm new to this please suggest what work should i carry out further....?

March 6, 2011, 17:12		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Capturing the stall point is tricky. Have a look in the literature to find out how other simulation researchers do it. It is not my field so I do not know. zkdkeen likes this.

March 8, 2011, 13:10	pictures of my mesh	#7
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	ya certainly, i'll mail them to u. thanks a lot for ur suggestions.

March 15, 2011, 10:01	further improvements	#12
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	i have been successful in getting the stall angle and the Cl is matching quite well with the experimental results. what further improvements can be done on the problem....? i've been thinking of drag reduction using winglets but i dont have the data for the specific winglet. what further improvements can be done.....?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
CFD simulation techniques	samar	Main CFD Forum	0	November 7, 2007 08:07
Workstation for CFD Simulation	S.Venkat	Main CFD Forum	4	December 5, 2002 11:29
cfd simulation between soild & porous environment	Latifimran Jalil	Main CFD Forum	0	September 12, 2002 18:42
What is the Better Way to Do CFD?	John C. Chien	Main CFD Forum	54	April 23, 2001 08:10
Solidification CFD simulation	Min-Hua Wang	FLUENT	0	February 28, 2001 17:40

March 7, 2011, 00:08		#3
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	ya i'm already working on it. i'm try now for a structured mesh, this will help me have a better control over the problem. but for analyzing a flutter problem for the airfoil i'll have to generate moving mesh. can you please suggest me on this, it would be really helpful.

March 7, 2011, 07:10		#4
Attesz Senior Member Attesz Join Date: Mar 2009 Location: Munich Posts: 368 Rep Power: 17	Getting closer to the stall point the steady state results won't match with experimentals, because of transient effects (growing vortex shedding etc). Switching to transient solver can handle this problem. I would use adaptive time step to speed up the simulation. In flutter I have no experience also, but this is good challenge in cfd

March 8, 2011, 07:56		#5
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	i havent tried with the transient solver, but will it be effective even for lower velocities (around 22m/s). and more over i've used an unstructured mesh(tetra) will it be advisable to go for structured mesh for higher angle of attack.....?

March 10, 2011, 09:28		#8
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	sir could i please get your email id....... so that i can mail u the contours and other details.

March 11, 2011, 06:02		#9
Attesz Senior Member Attesz Join Date: Mar 2009 Location: Munich Posts: 368 Rep Power: 17	if you are able to do, post pictures here because other people can see it.

March 11, 2011, 07:30		#10
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	i would have certainly posted them here, but i cannot upload attachments in the threads. i feel its not permitted to upload attachments. so it would be helpful if u could give me your email id. looking forward for your reply.

March 11, 2011, 08:03		#11
Attesz Senior Member Attesz Join Date: Mar 2009 Location: Munich Posts: 368 Rep Power: 17	felfoldi.attila@gmail.com but try upload here if you can. maybe you should delete the cookies if you have problems with the site.

May 18, 2011, 03:17		#13
icemaniac178 Member iswadi Join Date: Feb 2010 Posts: 44 Rep Power: 16	mr prshet, can you share with me how have you been successfully getting the stall angle? i believe i had the same problem with u.

May 20, 2011, 02:32		#14
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	ya certainly..... can u tell me the mesh details.....? and wht r ur inlet conditions......? if u pls let me know i would be able to help u....

May 21, 2011, 02:54		#16
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	ya thats alright....wht abt ur geometry.....? hw have u placed it.....? have u considered the whole geometry or only the half part (symmetry). and for different angle of attacks are u moving ur geometry.....?

May 22, 2011, 05:43		#17
icemaniac178 Member iswadi Join Date: Feb 2010 Posts: 44 Rep Power: 16	yes my geometry was taking the simplification of symmetrical condition (half part) and my angle of attack changes were determine by cartesian velocity components define at the inlet. i was not moving my geometry angle at all. what else i shoud do to get a good stall angle prediction?

May 25, 2011, 09:03		#18
prshet New Member anonumous Join Date: Mar 2010 Posts: 12 Rep Power: 16	when u give the Cartesian velocity components it might give u comparable results for lower angle of attack. if u have to capture the stalling angle then it would be better if u manipulate the geometry itself....pls try once.... when u rotate the geometry please keep in mind the change in conditions for ur inlet velocity. this time u dont have to give the u and v components. u simply have to give the u component.

May 29, 2011, 19:31		#19
icemaniac178 Member iswadi Join Date: Feb 2010 Posts: 44 Rep Power: 16	i will look for forward to do it. thanks a lot. but do i need to switch into transient simulation for high AOA (17-20 degree) in order to capture better stall prediction? i have gone through a lot of discussion in this forum. many people suggest transient simulation for stalling prediction.