[WB] Drive Workbench from external program
Joe,
I agree that the External Data is the way to go and I set it up originally that way, but I can't get the WB project to update correctly on my Linux cluster. It refuses to share licenses, or just crashes.
And... my CSV is a cloud of point forces, not pressure. I have a very viscous fluid and the tangential drag is significant. The magnitude of the forces depends on the areas in the CFD model, so it is not a continuous field. The *moper looks like it will interpolate over a continuous field, but that's not what I have.
I know this sounds dumb, but Is there a way to apply a point force that is NOT on a node? Take the force application point, find the closest element face, and distribute the force to the attached nodes?
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer
BAE Systems, Inc.
4050 Peppers Ferry Road, Radford VA 24143-0100
www.baesystems.com
-----Original Message-----
From: Joe Metrisin Joe.Metrisin@kratosdefense.com
Sent: Friday, February 11, 2022 6:34 AM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [Xansys] Re: [External] - Re: [WB] Drive Workbench from external program
External Email Alert
This email has been sent from an account outside of the BAE Systems network.
Please treat the email with caution, especially if you are requested to click on a link, decrypt/open an attachment, or enable macros. For further information on how to spot phishing, access OSI IT Policies and report phishing by forwarding mail to phishing@baesystems.com.
Hi Aaron,
The "external data" object in WB is the best way to map CFD results to a structural model. If you insist on doing it in MAPDL, the *MOPER command is the way to go. You need to create two arrays of data. For example:
Struct array is x,y,z coordinates of the structural model nodes you want to map to.
Cfd array is x,y,z,pressure from your CFD results.
Mapped will be the resulting array with node number and mapped pressure.
*moper, mapped(1,1), struct(1,1), map, cfd(1,4), cfd(1,1), 3,, 0 sffun, pres, mapped(1,1) sf, all, pres
Joseph T Metrisin
Structures Lead
Florida Turbine Technologies, Inc
1701 Military Tr. Suite 110 | Jupiter, FL 33458 USA
+1 (561) 427-6346 Office | +1 (772) 834-4156 Mobile
Joe.Metrisin@kratosdefense.com
Visit our website: https://kratosdefense.com
Confidentiality Note:
The information contained in this transmission and any attachments are proprietary and may be privileged, intended only for the use of the individual or entity named above. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication is strictly prohibited. If you received this communication in error, please delete the message and immediately notify the sender via the contact information listed above.
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
Sent: Thursday, February 10, 2022 6:24 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [External] - [Xansys] Re: [WB] Drive Workbench from external program
CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
Here's a follow-up and another question. Apparently you can drive MAPDL and WB from Matlab using the Ansys as a Service (AaaS) toolbox. Check out Ansys knowledge solutions 2057025, 2058573 on the Ansys Customer Portal. It looks pretty cool, if you are using Matlab. You can also drive MAPDL from Python.
See help page
(https://urldefense.proofpoint.com/v2/url?u=https-3A__ansyshelp.ansys.com&d=
DwIFAw&c=zeCCs5WLaN-HWPHrpXwbFoOqeS0G3NH2_2IQ_bzV13g&r=IOkMURQRtftopTCLYWjAB
vdzlGedqVvAMvUyU2Gzs2I&m=g4dZA3aO5eH_v0TLs6Hj9nqVPb3LbRlmRt_wYpwKHwtLGKNUci3
LOUhrjx5oAoNM&s=nlxaggg3sbDN_oyAeqCR7CbblBrkjNZXUQj0oZvVXF4&e= ):
"Mechanical APDL > Mechanical APDL as a Server User's Guide"
https://urldefense.proofpoint.com/v2/url?u=https-3A__ansyshelp.ansys.com_acc
ount_secured-3Freturnurl-3D_Views_Secured_corp_v221_en_ans-5Faas_ans-5Faas.h
tml&d=DwIFAw&c=zeCCs5WLaN-HWPHrpXwbFoOqeS0G3NH2_2IQ_bzV13g&r=IOkMURQRtftopTC
LYWjABvdzlGedqVvAMvUyU2Gzs2I&m=g4dZA3aO5eH_v0TLs6Hj9nqVPb3LbRlmRt_wYpwKHwtLG
KNUci3LOUhrjx5oAoNM&s=e8TgGFifJxH4XX0_QkliL_d5bPd_bypI_kdDbqH7Mis&e=
My new question is how to read in a CSV of point forces, and map them to the structural model. I've given up on using the WB interface at this point because it is giving me fits on my Linux cluster. So I'm going to try using APDL to do the file read/mapping/solve.
The point force CSV is from my CFD solution, but is on a different mesh than my mechanical simulation. So the forces need to be mapped to the surfaces/nodes of the mechanical model. These are forces at discrete points, not a field, so the interpolation would need to conserve the total force applied.
The CSV is a point cloud of information, with a force-vector/location pair.
No mesh information.
"Traction[i] (N)","Traction[j] (N)","Traction[k] (N)","X (mm)","Y (mm)","Z (mm)"
0.0175519044288876,-0.133241618707105,0.36252376963303,8.48947164677622,-1.6
7156843232078,0.72804856562505
0.0174687670479826,-0.0665973630469475,0.379089332635947,8.58245092723447,-1
.72662388432779,0.713273877000037
Does anyone have suggestions on how to read this into MAPDL, then do the mapping? For the file read, I looked up *TREAD, but this data is not ordered, it's just a point cloud. I could massage the CSV data into a better format if it would make it easier to read into MAPDL.
For the mapping, is this a *MOPER,,,MAP thing? Would that map from my point forces to forces on each node? I am thinking a loop that does:
Loop over all force/location pairs
Find element surface that contains location with ???
Distribute force over nodes with ?shape functions?
Apply forces to nodes with f,<node #>,fx,<portion of force on node> End loop
Thanks in advance,
Aaron
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer II
BAE Systems, Inc. | Ordnance Systems, Inc.
E-mail: aaron.caba@baesystems.com | Mail: 4050 Peppers Ferry Road, Radford VA 24143-0100 www.baesystems.com
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
I have a WB project I'm using in a fluid-structure interaction loop with an external CFD code. The external code is orchestrating the solve loops. It calls WB with:
"runwb2 -B -R SolveForDisplacement.wbjn"
to update the project and save some CSV files with displacements. This is working great, except for the overhead of opening WB every time. It takes about 65 seconds for the whole process, where 5 seconds are ANSYS and the other 60 are opening & closing WB. When the solution requires ~60-80 iterations to converge, that's and extra hour of wall clock time.
Is there a way to connect to a running WB session and send it commands?
Another option would be to re-write the file-read/map/solve/save as APDL commands and avoid the overhead, but I'm more interested in the GUI option for now.
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer II
BAE Systems, Inc. | Ordnance Systems, Inc.
Office: +1 540 639 7086 | Mobile: +1 540 230 3906 | E-mail:
aaron.caba@baesystems.com | Mail: 4050 Peppers Ferry Road, Radford VA
24143-0100 www.baesystems.com
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list _______________________________________________
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list
On Feb 11, 2022, at 4:20 PM, Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org wrote:
Without this mapping piece the whole analysis just doesn't work,
Go for it. Automating this kind of thing can be pretty instructive as well as a big rush. I've frequently over-promised myself about how much effort is truly saved especially when it looks like it might be interesting, and even if it doesn't save time, it's good exercise and may come in handy some day.
Christopher Wright P.E. (ret'd) |"They couldn't hit an elephant at
chrisw@skypoint.com | this distance" (last words of Gen.
| John Sedgwick, Spotsylvania (1864)
http://www.skypoint.com/members/chrisw/
"I know this sounds dumb, but Is there a way to apply a point force that is NOT on a node? Take the force application point, find the closest element face, and distribute the force to the attached nodes?"
Some books on finite elements talk about how to apply a point load which is not located at the node. The element load vector (eg the loads at the nodes) due to the point force is found by evaluating the element shape function vector at the location of the point force and multiplying that by the force. Check out whatever FEM books you have lying around, it might be in there.
Peter
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
Sent: Friday, February 11, 2022 4:24 PM
To: XANSYS Mailing List Home
Cc: Caba, Aaron (US)
Subject: [Xansys] Re: [External] - Re: [WB] Drive Workbench from external program
Joe,
I agree that the External Data is the way to go and I set it up originally that way, but I can't get the WB project to update correctly on my Linux cluster. It refuses to share licenses, or just crashes.
And... my CSV is a cloud of point forces, not pressure. I have a very viscous fluid and the tangential drag is significant. The magnitude of the forces depends on the areas in the CFD model, so it is not a continuous field. The *moper looks like it will interpolate over a continuous field, but that's not what I have.
I know this sounds dumb, but Is there a way to apply a point force that is NOT on a node? Take the force application point, find the closest element face, and distribute the force to the attached nodes?
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer
BAE Systems, Inc.
4050 Peppers Ferry Road, Radford VA 24143-0100
www.baesystems.com
-----Original Message-----
From: Joe Metrisin Joe.Metrisin@kratosdefense.com
Sent: Friday, February 11, 2022 6:34 AM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [Xansys] Re: [External] - Re: [WB] Drive Workbench from external program
External Email Alert
This email has been sent from an account outside of the BAE Systems network.
Please treat the email with caution, especially if you are requested to click on a link, decrypt/open an attachment, or enable macros. For further information on how to spot phishing, access OSI IT Policies and report phishing by forwarding mail to phishing@baesystems.com.
Hi Aaron,
The "external data" object in WB is the best way to map CFD results to a structural model. If you insist on doing it in MAPDL, the *MOPER command is the way to go. You need to create two arrays of data. For example:
Struct array is x,y,z coordinates of the structural model nodes you want to map to.
Cfd array is x,y,z,pressure from your CFD results.
Mapped will be the resulting array with node number and mapped pressure.
*moper, mapped(1,1), struct(1,1), map, cfd(1,4), cfd(1,1), 3,, 0 sffun, pres, mapped(1,1) sf, all, pres
Joseph T Metrisin
Structures Lead
Florida Turbine Technologies, Inc
1701 Military Tr. Suite 110 | Jupiter, FL 33458 USA
+1 (561) 427-6346 Office | +1 (772) 834-4156 Mobile
Joe.Metrisin@kratosdefense.com
Visit our website: https://kratosdefense.com
Confidentiality Note:
The information contained in this transmission and any attachments are proprietary and may be privileged, intended only for the use of the individual or entity named above. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication is strictly prohibited. If you received this communication in error, please delete the message and immediately notify the sender via the contact information listed above.
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
Sent: Thursday, February 10, 2022 6:24 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [External] - [Xansys] Re: [WB] Drive Workbench from external program
CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
Here's a follow-up and another question. Apparently you can drive MAPDL and WB from Matlab using the Ansys as a Service (AaaS) toolbox. Check out Ansys knowledge solutions 2057025, 2058573 on the Ansys Customer Portal. It looks pretty cool, if you are using Matlab. You can also drive MAPDL from Python.
See help page
(https://urldefense.proofpoint.com/v2/url?u=https-3A__ansyshelp.ansys.com&d=
DwIFAw&c=zeCCs5WLaN-HWPHrpXwbFoOqeS0G3NH2_2IQ_bzV13g&r=IOkMURQRtftopTCLYWjAB
vdzlGedqVvAMvUyU2Gzs2I&m=g4dZA3aO5eH_v0TLs6Hj9nqVPb3LbRlmRt_wYpwKHwtLGKNUci3
LOUhrjx5oAoNM&s=nlxaggg3sbDN_oyAeqCR7CbblBrkjNZXUQj0oZvVXF4&e= ):
"Mechanical APDL > Mechanical APDL as a Server User's Guide"
https://urldefense.proofpoint.com/v2/url?u=https-3A__ansyshelp.ansys.com_acc
ount_secured-3Freturnurl-3D_Views_Secured_corp_v221_en_ans-5Faas_ans-5Faas.h
tml&d=DwIFAw&c=zeCCs5WLaN-HWPHrpXwbFoOqeS0G3NH2_2IQ_bzV13g&r=IOkMURQRtftopTC
LYWjABvdzlGedqVvAMvUyU2Gzs2I&m=g4dZA3aO5eH_v0TLs6Hj9nqVPb3LbRlmRt_wYpwKHwtLG
KNUci3LOUhrjx5oAoNM&s=e8TgGFifJxH4XX0_QkliL_d5bPd_bypI_kdDbqH7Mis&e=
My new question is how to read in a CSV of point forces, and map them to the structural model. I've given up on using the WB interface at this point because it is giving me fits on my Linux cluster. So I'm going to try using APDL to do the file read/mapping/solve.
The point force CSV is from my CFD solution, but is on a different mesh than my mechanical simulation. So the forces need to be mapped to the surfaces/nodes of the mechanical model. These are forces at discrete points, not a field, so the interpolation would need to conserve the total force applied.
The CSV is a point cloud of information, with a force-vector/location pair.
No mesh information.
"Traction[i] (N)","Traction[j] (N)","Traction[k] (N)","X (mm)","Y (mm)","Z (mm)"
0.0175519044288876,-0.133241618707105,0.36252376963303,8.48947164677622,-1.6
7156843232078,0.72804856562505
0.0174687670479826,-0.0665973630469475,0.379089332635947,8.58245092723447,-1
.72662388432779,0.713273877000037
Does anyone have suggestions on how to read this into MAPDL, then do the mapping? For the file read, I looked up *TREAD, but this data is not ordered, it's just a point cloud. I could massage the CSV data into a better format if it would make it easier to read into MAPDL.
For the mapping, is this a *MOPER,,,MAP thing? Would that map from my point forces to forces on each node? I am thinking a loop that does:
Loop over all force/location pairs
Find element surface that contains location with ???
Distribute force over nodes with ?shape functions?
Apply forces to nodes with f,<node #>,fx,<portion of force on node> End loop
Thanks in advance,
Aaron
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer II
BAE Systems, Inc. | Ordnance Systems, Inc.
E-mail: aaron.caba@baesystems.com | Mail: 4050 Peppers Ferry Road, Radford VA 24143-0100 www.baesystems.com
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
I have a WB project I'm using in a fluid-structure interaction loop with an external CFD code. The external code is orchestrating the solve loops. It calls WB with:
"runwb2 -B -R SolveForDisplacement.wbjn"
to update the project and save some CSV files with displacements. This is working great, except for the overhead of opening WB every time. It takes about 65 seconds for the whole process, where 5 seconds are ANSYS and the other 60 are opening & closing WB. When the solution requires ~60-80 iterations to converge, that's and extra hour of wall clock time.
Is there a way to connect to a running WB session and send it commands?
Another option would be to re-write the file-read/map/solve/save as APDL commands and avoid the overhead, but I'm more interested in the GUI option for now.
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer II
BAE Systems, Inc. | Ordnance Systems, Inc.
Office: +1 540 639 7086 | Mobile: +1 540 230 3906 | E-mail:
aaron.caba@baesystems.com | Mail: 4050 Peppers Ferry Road, Radford VA
24143-0100 www.baesystems.com
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list _______________________________________________
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list
Xansys mailing list -- xansys-temp@list.xansys.org
To unsubscribe send an email to xansys-temp-leave@list.xansys.org
If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list
The more time consuming part would be figuring out what element corresponds to a given point force as it involves a search algorithm. I've done this before for an aeroelastic code and you only want to do it once (and then save the information) for any reasonable sized mesh. Section 8.4 in "Computational Grids" by Graham Carey gives some information on that. All of that said, it probably isn't worth your time trying to do this...LOL.
Peter
From: Attar, Peter J. peter.attar@ou.edu
Sent: Friday, February 11, 2022 11:17 PM
To: XANSYS Mailing List Home
Cc: Caba, Aaron (US)
Subject: [Xansys] Re: [External] - Re: [WB] Drive Workbench from external program
"I know this sounds dumb, but Is there a way to apply a point force that is NOT on a node? Take the force application point, find the closest element face, and distribute the force to the attached nodes?"
Some books on finite elements talk about how to apply a point load which is not located at the node. The element load vector (eg the loads at the nodes) due to the point force is found by evaluating the element shape function vector at the location of the point force and multiplying that by the force. Check out whatever FEM books you have lying around, it might be in there.
Peter
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
Sent: Friday, February 11, 2022 4:24 PM
To: XANSYS Mailing List Home
Cc: Caba, Aaron (US)
Subject: [Xansys] Re: [External] - Re: [WB] Drive Workbench from external program
Joe,
I agree that the External Data is the way to go and I set it up originally that way, but I can't get the WB project to update correctly on my Linux cluster. It refuses to share licenses, or just crashes.
And... my CSV is a cloud of point forces, not pressure. I have a very viscous fluid and the tangential drag is significant. The magnitude of the forces depends on the areas in the CFD model, so it is not a continuous field. The *moper looks like it will interpolate over a continuous field, but that's not what I have.
I know this sounds dumb, but Is there a way to apply a point force that is NOT on a node? Take the force application point, find the closest element face, and distribute the force to the attached nodes?
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer
BAE Systems, Inc.
4050 Peppers Ferry Road, Radford VA 24143-0100
www.baesystems.com
-----Original Message-----
From: Joe Metrisin Joe.Metrisin@kratosdefense.com
Sent: Friday, February 11, 2022 6:34 AM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [Xansys] Re: [External] - Re: [WB] Drive Workbench from external program
External Email Alert
This email has been sent from an account outside of the BAE Systems network.
Please treat the email with caution, especially if you are requested to click on a link, decrypt/open an attachment, or enable macros. For further information on how to spot phishing, access OSI IT Policies and report phishing by forwarding mail to phishing@baesystems.com.
Hi Aaron,
The "external data" object in WB is the best way to map CFD results to a structural model. If you insist on doing it in MAPDL, the *MOPER command is the way to go. You need to create two arrays of data. For example:
Struct array is x,y,z coordinates of the structural model nodes you want to map to.
Cfd array is x,y,z,pressure from your CFD results.
Mapped will be the resulting array with node number and mapped pressure.
*moper, mapped(1,1), struct(1,1), map, cfd(1,4), cfd(1,1), 3,, 0 sffun, pres, mapped(1,1) sf, all, pres
Joseph T Metrisin
Structures Lead
Florida Turbine Technologies, Inc
1701 Military Tr. Suite 110 | Jupiter, FL 33458 USA
+1 (561) 427-6346 Office | +1 (772) 834-4156 Mobile
Joe.Metrisin@kratosdefense.com
Visit our website: https://kratosdefense.com
Confidentiality Note:
The information contained in this transmission and any attachments are proprietary and may be privileged, intended only for the use of the individual or entity named above. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication is strictly prohibited. If you received this communication in error, please delete the message and immediately notify the sender via the contact information listed above.
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
Sent: Thursday, February 10, 2022 6:24 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [External] - [Xansys] Re: [WB] Drive Workbench from external program
CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
Here's a follow-up and another question. Apparently you can drive MAPDL and WB from Matlab using the Ansys as a Service (AaaS) toolbox. Check out Ansys knowledge solutions 2057025, 2058573 on the Ansys Customer Portal. It looks pretty cool, if you are using Matlab. You can also drive MAPDL from Python.
See help page
(https://urldefense.proofpoint.com/v2/url?u=https-3A__ansyshelp.ansys.com&d=
DwIFAw&c=zeCCs5WLaN-HWPHrpXwbFoOqeS0G3NH2_2IQ_bzV13g&r=IOkMURQRtftopTCLYWjAB
vdzlGedqVvAMvUyU2Gzs2I&m=g4dZA3aO5eH_v0TLs6Hj9nqVPb3LbRlmRt_wYpwKHwtLGKNUci3
LOUhrjx5oAoNM&s=nlxaggg3sbDN_oyAeqCR7CbblBrkjNZXUQj0oZvVXF4&e= ):
"Mechanical APDL > Mechanical APDL as a Server User's Guide"
https://urldefense.proofpoint.com/v2/url?u=https-3A__ansyshelp.ansys.com_acc
ount_secured-3Freturnurl-3D_Views_Secured_corp_v221_en_ans-5Faas_ans-5Faas.h
tml&d=DwIFAw&c=zeCCs5WLaN-HWPHrpXwbFoOqeS0G3NH2_2IQ_bzV13g&r=IOkMURQRtftopTC
LYWjABvdzlGedqVvAMvUyU2Gzs2I&m=g4dZA3aO5eH_v0TLs6Hj9nqVPb3LbRlmRt_wYpwKHwtLG
KNUci3LOUhrjx5oAoNM&s=e8TgGFifJxH4XX0_QkliL_d5bPd_bypI_kdDbqH7Mis&e=
My new question is how to read in a CSV of point forces, and map them to the structural model. I've given up on using the WB interface at this point because it is giving me fits on my Linux cluster. So I'm going to try using APDL to do the file read/mapping/solve.
The point force CSV is from my CFD solution, but is on a different mesh than my mechanical simulation. So the forces need to be mapped to the surfaces/nodes of the mechanical model. These are forces at discrete points, not a field, so the interpolation would need to conserve the total force applied.
The CSV is a point cloud of information, with a force-vector/location pair.
No mesh information.
"Traction[i] (N)","Traction[j] (N)","Traction[k] (N)","X (mm)","Y (mm)","Z (mm)"
0.0175519044288876,-0.133241618707105,0.36252376963303,8.48947164677622,-1.6
7156843232078,0.72804856562505
0.0174687670479826,-0.0665973630469475,0.379089332635947,8.58245092723447,-1
.72662388432779,0.713273877000037
Does anyone have suggestions on how to read this into MAPDL, then do the mapping? For the file read, I looked up *TREAD, but this data is not ordered, it's just a point cloud. I could massage the CSV data into a better format if it would make it easier to read into MAPDL.
For the mapping, is this a *MOPER,,,MAP thing? Would that map from my point forces to forces on each node? I am thinking a loop that does:
Loop over all force/location pairs
Find element surface that contains location with ???
Distribute force over nodes with ?shape functions?
Apply forces to nodes with f,<node #>,fx,<portion of force on node> End loop
Thanks in advance,
Aaron
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer II
BAE Systems, Inc. | Ordnance Systems, Inc.
E-mail: aaron.caba@baesystems.com | Mail: 4050 Peppers Ferry Road, Radford VA 24143-0100 www.baesystems.com
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
I have a WB project I'm using in a fluid-structure interaction loop with an external CFD code. The external code is orchestrating the solve loops. It calls WB with:
"runwb2 -B -R SolveForDisplacement.wbjn"
to update the project and save some CSV files with displacements. This is working great, except for the overhead of opening WB every time. It takes about 65 seconds for the whole process, where 5 seconds are ANSYS and the other 60 are opening & closing WB. When the solution requires ~60-80 iterations to converge, that's and extra hour of wall clock time.
Is there a way to connect to a running WB session and send it commands?
Another option would be to re-write the file-read/map/solve/save as APDL commands and avoid the overhead, but I'm more interested in the GUI option for now.
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer II
BAE Systems, Inc. | Ordnance Systems, Inc.
Office: +1 540 639 7086 | Mobile: +1 540 230 3906 | E-mail:
aaron.caba@baesystems.com | Mail: 4050 Peppers Ferry Road, Radford VA
24143-0100 www.baesystems.com
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list _______________________________________________
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an email to xansys-temp-leave@list.xansys.org If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list
Xansys mailing list -- xansys-temp@list.xansys.org
To unsubscribe send an email to xansys-temp-leave@list.xansys.org
If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list
Xansys mailing list -- xansys-temp@list.xansys.org
To unsubscribe send an email to xansys-temp-leave@list.xansys.org
If you are receiving too many emails from XANSYS please consider changing account settings to Digest mode which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to xansys-mod@tynecomp.co.uk and not to the list
The remote load feature in WB does this by using a contact/target pair
(174,170 elements). The target element is on a single node and is where you
apply your point force. The contact element is on the face of the elements
where you want this load distributed. I'm not sure if this will work with
the shear volume of contact target pairs you would end up generating in your
case though.
Joseph T Metrisin
Structures Lead
Florida Turbine Technologies, Inc
1701 Military Tr. Suite 110 | Jupiter, FL 33458 USA
+1 (561) 427-6346 Office | +1 (772) 834-4156 Mobile
Joe.Metrisin@kratosdefense.com
Visit our website: https://kratosdefense.com
Confidentiality Note:
The information contained in this transmission and any attachments are
proprietary and may be privileged, intended only for the use of the
individual or entity named above. If the reader of this message is not the
intended recipient, you are hereby notified that any dissemination,
distribution, or copying of this communication is strictly prohibited. If
you received this communication in error, please delete the message and
immediately notify the sender via the contact information listed above.
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
Sent: Friday, February 11, 2022 5:25 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [Xansys] Re: [External] - Re: [WB] Drive Workbench from external
program
CAUTION: This email originated from outside of the organization. Do not
click links or open attachments unless you recognize the sender and know the
content is safe.
Joe,
I agree that the External Data is the way to go and I set it up originally
that way, but I can't get the WB project to update correctly on my Linux
cluster. It refuses to share licenses, or just crashes.
And... my CSV is a cloud of point forces, not pressure. I have a very
viscous fluid and the tangential drag is significant. The magnitude of the
forces depends on the areas in the CFD model, so it is not a continuous
field. The *moper looks like it will interpolate over a continuous field,
but that's not what I have.
I know this sounds dumb, but Is there a way to apply a point force that is
NOT on a node? Take the force application point, find the closest element
face, and distribute the force to the attached nodes?
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer
BAE Systems, Inc.
4050 Peppers Ferry Road, Radford VA 24143-0100 www.baesystems.com
-----Original Message-----
From: Joe Metrisin Joe.Metrisin@kratosdefense.com
Sent: Friday, February 11, 2022 6:34 AM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [Xansys] Re: [External] - Re: [WB] Drive Workbench from external
program
External Email Alert
This email has been sent from an account outside of the BAE Systems network.
Please treat the email with caution, especially if you are requested to
click on a link, decrypt/open an attachment, or enable macros. For further
information on how to spot phishing, access OSI IT Policies and report
phishing by forwarding mail to phishing@baesystems.com.
Hi Aaron,
The "external data" object in WB is the best way to map CFD results to a
structural model. If you insist on doing it in MAPDL, the *MOPER command is
the way to go. You need to create two arrays of data. For example:
Struct array is x,y,z coordinates of the structural model nodes you want to
map to.
Cfd array is x,y,z,pressure from your CFD results.
Mapped will be the resulting array with node number and mapped pressure.
*moper, mapped(1,1), struct(1,1), map, cfd(1,4), cfd(1,1), 3,, 0 sffun,
pres, mapped(1,1) sf, all, pres
Joseph T Metrisin
Structures Lead
Florida Turbine Technologies, Inc
1701 Military Tr. Suite 110 | Jupiter, FL 33458 USA
+1 (561) 427-6346 Office | +1 (772) 834-4156 Mobile
Joe.Metrisin@kratosdefense.com
Visit our website: https://kratosdefense.com
Confidentiality Note:
The information contained in this transmission and any attachments are
proprietary and may be privileged, intended only for the use of the
individual or entity named above. If the reader of this message is not the
intended recipient, you are hereby notified that any dissemination,
distribution, or copying of this communication is strictly prohibited. If
you received this communication in error, please delete the message and
immediately notify the sender via the contact information listed above.
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
Sent: Thursday, February 10, 2022 6:24 PM
To: XANSYS Mailing List Home xansys-temp@list.xansys.org
Cc: Caba, Aaron (US) Aaron.Caba@baesystems.com
Subject: [External] - [Xansys] Re: [WB] Drive Workbench from external
program
CAUTION: This email originated from outside of the organization. Do not
click links or open attachments unless you recognize the sender and know the
content is safe.
Here's a follow-up and another question. Apparently you can drive MAPDL and
WB from Matlab using the Ansys as a Service (AaaS) toolbox. Check out Ansys
knowledge solutions 2057025, 2058573 on the Ansys Customer Portal. It looks
pretty cool, if you are using Matlab. You can also drive MAPDL from Python.
See help page
(https://urldefense.proofpoint.com/v2/url?u=https-3A__ansyshelp.ansys.com&d=
DwIFAw&c=zeCCs5WLaN-HWPHrpXwbFoOqeS0G3NH2_2IQ_bzV13g&r=IOkMURQRtftopTCLYWjAB
vdzlGedqVvAMvUyU2Gzs2I&m=g4dZA3aO5eH_v0TLs6Hj9nqVPb3LbRlmRt_wYpwKHwtLGKNUci3
LOUhrjx5oAoNM&s=nlxaggg3sbDN_oyAeqCR7CbblBrkjNZXUQj0oZvVXF4&e= ):
"Mechanical APDL > Mechanical APDL as a Server User's Guide"
https://urldefense.proofpoint.com/v2/url?u=https-3A__ansyshelp.ansys.com_acc
ount_secured-3Freturnurl-3D_Views_Secured_corp_v221_en_ans-5Faas_ans-5Faas.h
tml&d=DwIFAw&c=zeCCs5WLaN-HWPHrpXwbFoOqeS0G3NH2_2IQ_bzV13g&r=IOkMURQRtftopTC
LYWjABvdzlGedqVvAMvUyU2Gzs2I&m=g4dZA3aO5eH_v0TLs6Hj9nqVPb3LbRlmRt_wYpwKHwtLG
KNUci3LOUhrjx5oAoNM&s=e8TgGFifJxH4XX0_QkliL_d5bPd_bypI_kdDbqH7Mis&e=
My new question is how to read in a CSV of point forces, and map them to the
structural model. I've given up on using the WB interface at this point
because it is giving me fits on my Linux cluster. So I'm going to try using
APDL to do the file read/mapping/solve.
The point force CSV is from my CFD solution, but is on a different mesh than
my mechanical simulation. So the forces need to be mapped to the
surfaces/nodes of the mechanical model. These are forces at discrete
points, not a field, so the interpolation would need to conserve the total
force applied.
The CSV is a point cloud of information, with a force-vector/location pair.
No mesh information.
"Traction[i] (N)","Traction[j] (N)","Traction[k] (N)","X (mm)","Y (mm)","Z
(mm)"
0.0175519044288876,-0.133241618707105,0.36252376963303,8.48947164677622,-1.6
7156843232078,0.72804856562505
0.0174687670479826,-0.0665973630469475,0.379089332635947,8.58245092723447,-1
.72662388432779,0.713273877000037
Does anyone have suggestions on how to read this into MAPDL, then do the
mapping? For the file read, I looked up *TREAD, but this data is not
ordered, it's just a point cloud. I could massage the CSV data into a
better format if it would make it easier to read into MAPDL.
For the mapping, is this a *MOPER,,,MAP thing? Would that map from my point
forces to forces on each node? I am thinking a loop that does:
Loop over all force/location pairs
Find element surface that contains location with ???
Distribute force over nodes with ?shape functions?
Apply forces to nodes with f,<node #>,fx,<portion of force on node> End
loop
Thanks in advance,
Aaron
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer II
BAE Systems, Inc. | Ordnance Systems, Inc.
E-mail: aaron.caba@baesystems.com | Mail: 4050 Peppers Ferry Road, Radford
VA 24143-0100 www.baesystems.com
-----Original Message-----
From: Caba, Aaron (US) via Xansys xansys-temp@list.xansys.org
I have a WB project I'm using in a fluid-structure interaction loop with an
external CFD code. The external code is orchestrating the solve loops. It
calls WB with:
"runwb2 -B -R SolveForDisplacement.wbjn"
to update the project and save some CSV files with displacements. This is
working great, except for the overhead of opening WB every time. It takes
about 65 seconds for the whole process, where 5 seconds are ANSYS and the
other 60 are opening & closing WB. When the solution requires ~60-80
iterations to converge, that's and extra hour of wall clock time.
Is there a way to connect to a running WB session and send it commands?
Another option would be to re-write the file-read/map/solve/save as APDL
commands and avoid the overhead, but I'm more interested in the GUI option
for now.
Aaron C. Caba, Ph.D.
Sr. Principal R&D Engineer II
BAE Systems, Inc. | Ordnance Systems, Inc.
Office: +1 540 639 7086 | Mobile: +1 540 230 3906 | E-mail:
aaron.caba@baesystems.com | Mail: 4050 Peppers Ferry Road, Radford VA
24143-0100 www.baesystems.com
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an
email to xansys-temp-leave@list.xansys.org If you are receiving too many
emails from XANSYS please consider changing account settings to Digest mode
which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to
xansys-mod@tynecomp.co.uk and not to the list
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an
email to xansys-temp-leave@list.xansys.org If you are receiving too many
emails from XANSYS please consider changing account settings to Digest mode
which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to
xansys-mod@tynecomp.co.uk and not to the list
Xansys mailing list -- xansys-temp@list.xansys.org To unsubscribe send an
email to xansys-temp-leave@list.xansys.org If you are receiving too many
emails from XANSYS please consider changing account settings to Digest mode
which will send a single email per day.
Please send administrative requests such as deletion from XANSYS to
xansys-mod@tynecomp.co.uk and not to the list