Cannot achieve negative wall shear stress values in CFD Post (ANSYS)

SAtkins · August 2, 2016, 11:52

Hello all,

For reference, I'm using ANSYS 17.1 CFX

I am working with a simulation that models 1/6 the geometry of a circular disk (so it basically looks like a slice of pizza). I input a rotational sine wave on the fluid above the disk, which should cause a shear stress equivalent sine wave with an amplitude of +/- 2 Pa. (Labeled Target on graph)

In CFD Post, I am extracting the radial component of the shear stress with the transform Radial Wall Shear Stress = -(0.5 Wall Shear Stress X + 0.866 Wall Shear Stress Y) (from a trig transform).

When the sine wave should be changing direction and going negative down to -2 Pa, instead, the waveform achieves a minimum value of ~ -0.4 Pa (Waveform Labeled "0" on graph). I have tried to over-exaggerate the input angular velocity of the negative portion of the sine wave, and when I do this, I get a curve full of positive "humps" (think the first half period of a sine wave, repeating over and over at the applied amplitude). (Waveform Labeled "1" on graph)

Do you think this could be a problem because I'm modeling 1/6 of the geometry as a symmetry plane? Or is there some problem with my post processing?

TL;DR = Cannot achieve negative wall shear stress values in simulation with a sine wave input (and I'm not exporting the absolute value or magnitudes of shear stress, either).

Any help is much appreciated!

ghorrocks · August 2, 2016, 20:39

The variable "Wall Shear Stress" is the magnitude of the vector so is always positive. But "Wall Shear Stress X/Y/Z" are the components and should be able to go negative when they need to. So I do not know why you are seeing what you are seeing.

Your result also looks wobbly so I am suspicious there is general accuracy problems as well. Have a look at the FAQ: http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

SAtkins · August 3, 2016, 08:14

Right, I am extracting just the Wall Shear X and Wall Shear Y components now - my Wall Shear X goes negative in the appropriate places while my Wall Shear Y is very positive (2 times the magnitude of X). My equation for radial shear stress should take care of the positive values, though, as it adds -0.866 Wall Shear Y.

I've looked at that link - my solutions are definitely converging. I've ran a mesh analysis and don't see any differences in output when changing the grid size. At this point I'm pretty stuck

SAtkins · August 3, 2016, 14:27

Update:

I have figured out the problem - and it's my bad trig. Instead of using the formulas I did to pull out what I called radial, which is really circumferential, shear stress - I just needed to use simple trigonometry.

Circ = cos(30)WSSx - sin(30)WSSy
Circ 1 = cos(3)WSSx + sin(30)WSSy

Circ produced a waveform that was inverted from my target. Circ 1 seemed to produce the output I was looking for (minus, minor inertial effects not taken into the WSS to AV transform I used - which is expected).

Anyways, I wanted to update with the hope that anyone having this problem in the future might get some help on transforming x and y WSS components into a circumferential shear stress component tangential to the direction of a rotational flow.

August 2, 2016, 11:52	Cannot achieve negative wall shear stress values in CFD Post (ANSYS)	#1
SAtkins New Member Sam Join Date: Aug 2016 Posts: 3 Rep Power: 9	Hello all, For reference, I'm using ANSYS 17.1 CFX I am working with a simulation that models 1/6 the geometry of a circular disk (so it basically looks like a slice of pizza). I input a rotational sine wave on the fluid above the disk, which should cause a shear stress equivalent sine wave with an amplitude of +/- 2 Pa. (Labeled Target on graph) In CFD Post, I am extracting the radial component of the shear stress with the transform Radial Wall Shear Stress = -(0.5 Wall Shear Stress X + 0.866 Wall Shear Stress Y) (from a trig transform). When the sine wave should be changing direction and going negative down to -2 Pa, instead, the waveform achieves a minimum value of ~ -0.4 Pa (Waveform Labeled "0" on graph). I have tried to over-exaggerate the input angular velocity of the negative portion of the sine wave, and when I do this, I get a curve full of positive "humps" (think the first half period of a sine wave, repeating over and over at the applied amplitude). (Waveform Labeled "1" on graph) Do you think this could be a problem because I'm modeling 1/6 of the geometry as a symmetry plane? Or is there some problem with my post processing? TL;DR = Cannot achieve negative wall shear stress values in simulation with a sine wave input (and I'm not exporting the absolute value or magnitudes of shear stress, either). Any help is much appreciated! Last edited by SAtkins; August 2, 2016 at 12:06. Reason: Image Attachment

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Natural convection in a closed domain STILL NEEDING help!	Yr0gErG	FLUENT	4	December 2, 2019 00:04
Calculation of Wall Shear Stress	Dave442	CFX	36	July 12, 2016 04:12
Wall shear stress in channel flow	sreeyuth	OpenFOAM Running, Solving & CFD	2	September 5, 2014 10:02
How to get a value of wall shear stress for UDF	shileijerry	Fluent UDF and Scheme Programming	0	June 18, 2013 10:50
WALL SHEAR STRESS	Min-Hua Wang	CFX	0	April 3, 2002 17:40

August 2, 2016, 20:39		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	The variable "Wall Shear Stress" is the magnitude of the vector so is always positive. But "Wall Shear Stress X/Y/Z" are the components and should be able to go negative when they need to. So I do not know why you are seeing what you are seeing. Your result also looks wobbly so I am suspicious there is general accuracy problems as well. Have a look at the FAQ: http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

August 3, 2016, 08:14		#3
SAtkins New Member Sam Join Date: Aug 2016 Posts: 3 Rep Power: 9	Right, I am extracting just the Wall Shear X and Wall Shear Y components now - my Wall Shear X goes negative in the appropriate places while my Wall Shear Y is very positive (2 times the magnitude of X). My equation for radial shear stress should take care of the positive values, though, as it adds -0.866 Wall Shear Y. I've looked at that link - my solutions are definitely converging. I've ran a mesh analysis and don't see any differences in output when changing the grid size. At this point I'm pretty stuck

August 3, 2016, 14:27		#4
SAtkins New Member Sam Join Date: Aug 2016 Posts: 3 Rep Power: 9	Update: I have figured out the problem - and it's my bad trig. Instead of using the formulas I did to pull out what I called radial, which is really circumferential, shear stress - I just needed to use simple trigonometry. Circ = cos(30)WSSx - sin(30)WSSy Circ 1 = cos(3)WSSx + sin(30)WSSy Circ produced a waveform that was inverted from my target. Circ 1 seemed to produce the output I was looking for (minus, minor inertial effects not taken into the WSS to AV transform I used - which is expected). Anyways, I wanted to update with the hope that anyone having this problem in the future might get some help on transforming x and y WSS components into a circumferential shear stress component tangential to the direction of a rotational flow.