CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > CFX

alculating incidence and deviation in turbomachine

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   August 12, 2008, 04:49
Default alculating incidence and deviation in turbomachine
  #1
sven
Guest
 
Posts: n/a
Hello,

I am trying to determine incidence and deviation angle in CFX-Post in a simulated turbomachine. Therefore I need to get the flow angles and the blade angles at leading and trailing edge.

To determine the flow angles I actually use a scalar product from the velocity vector and the axial direction vector and the theta direction vector respectively to get velocities in axial and tangential direction. After that I calculate the angle via inverse tangent.

Does anyone know a better method?

A further problem is: What is the best position for calculating the leading edge flow angle? If I get too close to the leading edge I will have flow stagnation influences and to fare away I will miss some upstream effects from the Rotor.

But my actual main problem is: How to determine the blade leading and trailing edge angles? At the moment I do not have any idea how to calculate the angles with CFX-Post.

any help is appreciated Sven
  Reply With Quote

Old   August 12, 2008, 11:58
Default Re: alculating incidence and deviation in turbomac
  #2
Sungho Yoon
Guest
 
Posts: n/a
I can imagine it would be tricky to calculate incidence and deviation in turbomachines.

I normally look at blade pressure distribution along the axial direction (or meridional direction).

By investigation the pressure distribution near the leading edge, you can estimate whether you have a good incidence or not.

If you really want to have a number (flow angle), I would average flows in pitchwise direction.

If you can calculate the blade angle somehow (it can be tricky though), you can calculate incidence or deviation simply by subtracting one angle from the other.

The axial (or meridional ) position, in order to calculate incidence and deviation, is arbitrary. You need to choose a sensible position.

Again, what really matters in turbomachinery design is blade pressure distribution.

Sungho
  Reply With Quote

Old   August 13, 2008, 04:25
Default Re: alculating incidence and deviation in turbomac
  #3
michel
Guest
 
Posts: n/a
I tried to do this a while back, and indeed, the biggest problem is to get the blade metal angles.

The method I came up with searched for the position of the leading and trailing edges, and their "images" on a user surface that was defined as an offset from the actual blade surface. From the line connecting the actual le/te to their respective images, the blade angles can then be obtained.

However, the obtained spanwise angle distribution usually is not smooth: form one span location to the next, the angles can vary by a few degrees. From what I can tell, this is caused by Turbogrid not doing a good job conserving the geometry, at least the leading/trailing edge shapes. You'd think something important like that would be handled better by such expensive software. (I mimicked the same procedure using a mesh from a different mesher (in-house), and it worked fine. Unfortunately this mesher is not compatible with CFX).

If you come up with somethinng that works better, please do post it!

Cheers,

Michel

  Reply With Quote

Old   August 15, 2008, 01:07
Default Re: alculating incidence and deviation in turbomac
  #4
Sven
Guest
 
Posts: n/a
Hi Michel,

your suggestion seems to be a possible way, but I do not know, how you did search for leading and trailing edge on your airfoil.

Some more hints how to do that are appreciated.

Thanks Sven
  Reply With Quote

Old   August 15, 2008, 04:57
Default Re: alculating incidence and deviation in turbomac
  #5
Sven
Guest
 
Posts: n/a
The other topic is:

Do you know a possibility to improve the conservation of the geometry in Turbogrid, to avoid the mentioned problems, except increasing number of elements?

Greetings,

Sven
  Reply With Quote

Old   August 15, 2008, 06:23
Default Re: alculating incidence and deviation in turbomac
  #6
michel
Guest
 
Posts: n/a
Hi Sven,

The procedure I followed is this:

1) create an Isosurface at a certain span.

2) create Polyline1 as an intersection between Blade High and the Isosurface.

3) create Polyline2 as an intersection between Blade Low and the Isosurface.

4) create Point1 as min(X) on Polyline1.

5) create Point2 as min(X) on Polyline2.

6) determine which of the two Points is the actual Leading Edge (this assumes one of them is the LE, which should be the case). The LE is the point which has the larger X-value of the two. Let's assume Point 1 is the leading edge, and it lies on Blade High.

7) create a User Surface as a Normal Offset to the Blade surface. Make sure this User Surface is very close to the blade.

8) search for the "LE image" on Polyline3:

- create the expression "getdistance" as

getdistance=sqrt((X-(xc))^2+(Y-(yc))^2), where

xle and yle are the coordinates of the actual leading

edge point.

- create the User Variable "distance", defined using the expression "getdistance".

- the LE image is the Variable Minimum of "distance" on Polyline3.

9) get the leading edge metal angle from the line connecting the leading edge and its image.

I hope this is helpful.

Michel
  Reply With Quote

Old   August 15, 2008, 06:29
Default Re: alculating incidence and deviation in turbomac
  #7
michel
Guest
 
Posts: n/a
No idea. Increasing the number of points on the specified profiles doesn't really help either, not does increaseing the numer of mesh points.

I also have the feeling Turbogrid11 has some issues with profile interpolation/surface generation that Turbogrid10 didn't have. In one case I ended up with a hook at the leading edge, in another with an "slot" along the span on the suction surface. Ansys support was notified, but no result/fix has come out of that (yet?).

Michel
  Reply With Quote

Old   September 29, 2009, 07:39
Default
  #8
New Member
 
Magnus Andersson
Join Date: Sep 2009
Posts: 1
Rep Power: 0
Hasumangan is on a distinguished road
Quote:
Originally Posted by michel
;89502
Hi Sven,

The procedure I followed is this:

1) create an Isosurface at a certain span.

2) create Polyline1 as an intersection between Blade High and the Isosurface.

3) create Polyline2 as an intersection between Blade Low and the Isosurface.

4) create Point1 as min(X) on Polyline1.

5) create Point2 as min(X) on Polyline2.

6) determine which of the two Points is the actual Leading Edge (this assumes one of them is the LE, which should be the case). The LE is the point which has the larger X-value of the two. Let's assume Point 1 is the leading edge, and it lies on Blade High.

7) create a User Surface as a Normal Offset to the Blade surface. Make sure this User Surface is very close to the blade.

8) search for the "LE image" on Polyline3:

- create the expression "getdistance" as

getdistance=sqrt((X-(xc))^2+(Y-(yc))^2), where

xle and yle are the coordinates of the actual leading

edge point.

- create the User Variable "distance", defined using the expression "getdistance".

- the LE image is the Variable Minimum of "distance" on Polyline3.

9) get the leading edge metal angle from the line connecting the leading edge and its image.

I hope this is helpful.

Michel
I have reach so far as point 7 then I get somewhat lost. Can you explain what the meaning om "LE image". I guess that polyline 3 is the line between the leading edge and trailing edge?
Hasumangan is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
metallic and deviation angles AndreaDB CFX 0 April 1, 2011 16:17
Incidence and deviation angle for axial turbomachines csmistry CFX 4 October 25, 2010 07:11


All times are GMT -4. The time now is 10:17.