ANSYS QA problems

ahorvat · September 24, 2009, 12:16

Recently completed FSI project in ANSYS v.12 and found 10-20 bugs (depends how a bug is qualified). Does anyone think that ANSYS has serious quality issues ?

denizen · September 25, 2009, 03:02

i think no.
some problems is present, but where the bugs no present?
If you will try to work with other FEA or CFD codes you can see more deep and systematic problems.

regards.

ahorvat · September 25, 2009, 10:15

Thank you for the reply. I hope you are speaking from your own experiences. I cannot. Having two commercial CFD codes is a large expense, therefore, my experiences are limited to ANSYS/CFX.

The only comparison that I can make is a meshing case, which I asked ANSYS UK office to help us with. It took their best ICEM guy most of the day to tet mesh it (no inflation layers).

At the the All Energy Conf. in Aberdeen, STAR-CD sales guys were teasing me and I gave them the same geometry to mesh (to embarrass them). Although, they probably did not have their best meshing guy on a sales conference, they came back in 1.5 h with a mesh with inflation layers (... this is not an invitation for STAR-CD sales guy to contact me and trying to sell me something).

I would encourage other people to explain their experiences.

denizen · September 28, 2009, 02:43

I have experiance with CFX, FLUENT, STAR-CD, STAR-CCM, ICEM, GAMBIT, ANSYS, ANSYS MESHING.

So, i can summarize that tetra+prism (inflation) it is fully automatic and no problems with generation on ICEM, ANSYS Meshig, GAMBIT, and STAR-CCM.

The troubles starts with hexahedral mesh, there only ICEM have tools to make one for complex geometries.

regards.

ahorvat · September 29, 2009, 07:11

My impression is that they (at least ANSYS) are just adding 'capability' after 'capability' with little respect to proper functioning of the software.

ICEM is a typical example. The tools that work for Octree meshes do not work for Delauney meshes and especially not for Hexa-dominated meshes. Also no one bothers to put together a complete capability matrix or clean out different options. I guess because it is just too complicated.

You may object that, but even ICEM support people have large problems with such philosophy. Now, if they cannot train their own people about 'endless' possibilities that the software offers, how do they expect that users will master such tool.

I am an engineer, I use ANSYS tools to improve eng. designs, don't have time to play in order to discover possible short cuts.

denizen · September 30, 2009, 02:27

you right. i recomend you to use the ANSYS Meshing product. This is very easy to use, inuitive user interface and i think they best for tetrahedra+prism.

If you have license to tetra for icem you can try ansys meshing for free (it is promo).

regards.

flattie · October 27, 2009, 15:09

Quote:

Originally Posted by ahorvat

Recently completed FSI project in ANSYS v.12 and found 10-20 bugs (depends how a bug is qualified). Does anyone think that ANSYS has serious quality issues ?

Yes. And I would add no clear direction on integrating all the stuff they bought like ICEM, Fluent, etc. They seem to have the strategy to just buy everything up, hope it sticks, and hope their customer base can stand the pain while they decide what to do.

While not CFD-related, i am seriously regretting not evaluating ABAQUS when we switched from NASTRAN. ANSYS Classic especially needs a GUI update. The DOS era look is horrible.

CFX IMO is the bright spot. Good code and easy to use.

flattie · October 27, 2009, 15:16

I would add poor documentation in ICEM. Also, when you go into the ICEM V12 help and click on tutorials, you get a web link to the customer portal that says they are not available or in progress. This has been the case for months. Not good for a piece of software that costs thousands of dollars. It's enough to possibly be my last ANSYS purchase.

I also deal with multiple ICEM crashes every day and horrid graphics performance (takes over 30 seconds in some cases to blank parts). My ASD basically gave up on correcting that.

I think the WB 2 emphasis will drive away many customers.

ahorvat · October 29, 2009, 05:57

Guys, I would like to continue this discussion, but at present I cannot. The FSI project, which triggered my comments was for an ANSYS reseller and his license to sell ANSYS products was just revoked because of my comments (yes, I do use my real name although I intentionally avoided to name the ANSYS people involved). I feel obliged to do everything in my power get him through.

pimpa · October 29, 2009, 12:40

ahorvat,

ANSYS CFX does have QA (Class3) Error-Reporting for customers on maintenance (TECS).
Actual bugs are reported, remedies are given. Software version containing first appearance of a bug, version for which bug will be fixed is given as well. Each version of CFX does see roughly some ten bugs uncovered (and older ones fixed).

Did you check the Error-Report's?
Did your bugs coincide with the bugs reported there or where these additional (relevant) bugs ?

September 24, 2009, 12:16	ANSYS QA problems	#1
ahorvat New Member Andrej Horvat Join Date: Mar 2009 Posts: 7 Rep Power: 17	Recently completed FSI project in ANSYS v.12 and found 10-20 bugs (depends how a bug is qualified). Does anyone think that ANSYS has serious quality issues ?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Command Line Run for ANSYS	Liaquat	ANSYS	11	April 19, 2018 10:49
Ansys installation on VISTA 32/64 bit sol'n!	Andrzej F	CFX	0	September 22, 2008 17:51
Ansys CFX 11 SP1 runs on Vista with problems	kayp	CFX	3	February 5, 2008 17:49
Exporting results from CFX to ANSYS ??	sohail ahmed	CFX	1	December 20, 2007 02:10
problems of ansys icem10	zuozicheng	CFX	2	September 26, 2005 10:22

September 25, 2009, 03:02		#2
denizen New Member Join Date: Mar 2009 Posts: 25 Rep Power: 0	i think no. some problems is present, but where the bugs no present? If you will try to work with other FEA or CFD codes you can see more deep and systematic problems. regards.

September 25, 2009, 10:15		#3
ahorvat New Member Andrej Horvat Join Date: Mar 2009 Posts: 7 Rep Power: 17	Thank you for the reply. I hope you are speaking from your own experiences. I cannot. Having two commercial CFD codes is a large expense, therefore, my experiences are limited to ANSYS/CFX. The only comparison that I can make is a meshing case, which I asked ANSYS UK office to help us with. It took their best ICEM guy most of the day to tet mesh it (no inflation layers). At the the All Energy Conf. in Aberdeen, STAR-CD sales guys were teasing me and I gave them the same geometry to mesh (to embarrass them). Although, they probably did not have their best meshing guy on a sales conference, they came back in 1.5 h with a mesh with inflation layers (... this is not an invitation for STAR-CD sales guy to contact me and trying to sell me something). I would encourage other people to explain their experiences.

September 28, 2009, 02:43		#4
denizen New Member Join Date: Mar 2009 Posts: 25 Rep Power: 0	I have experiance with CFX, FLUENT, STAR-CD, STAR-CCM, ICEM, GAMBIT, ANSYS, ANSYS MESHING. So, i can summarize that tetra+prism (inflation) it is fully automatic and no problems with generation on ICEM, ANSYS Meshig, GAMBIT, and STAR-CCM. The troubles starts with hexahedral mesh, there only ICEM have tools to make one for complex geometries. regards.

September 29, 2009, 07:11		#5
ahorvat New Member Andrej Horvat Join Date: Mar 2009 Posts: 7 Rep Power: 17	My impression is that they (at least ANSYS) are just adding 'capability' after 'capability' with little respect to proper functioning of the software. ICEM is a typical example. The tools that work for Octree meshes do not work for Delauney meshes and especially not for Hexa-dominated meshes. Also no one bothers to put together a complete capability matrix or clean out different options. I guess because it is just too complicated. You may object that, but even ICEM support people have large problems with such philosophy. Now, if they cannot train their own people about 'endless' possibilities that the software offers, how do they expect that users will master such tool. I am an engineer, I use ANSYS tools to improve eng. designs, don't have time to play in order to discover possible short cuts.

September 30, 2009, 02:27		#6
denizen New Member Join Date: Mar 2009 Posts: 25 Rep Power: 0	you right. i recomend you to use the ANSYS Meshing product. This is very easy to use, inuitive user interface and i think they best for tetrahedra+prism. If you have license to tetra for icem you can try ansys meshing for free (it is promo). regards.

October 27, 2009, 15:16		#8
flattie New Member Craig Hildreth Join Date: Mar 2009 Posts: 22 Rep Power: 17	I would add poor documentation in ICEM. Also, when you go into the ICEM V12 help and click on tutorials, you get a web link to the customer portal that says they are not available or in progress. This has been the case for months. Not good for a piece of software that costs thousands of dollars. It's enough to possibly be my last ANSYS purchase. I also deal with multiple ICEM crashes every day and horrid graphics performance (takes over 30 seconds in some cases to blank parts). My ASD basically gave up on correcting that. I think the WB 2 emphasis will drive away many customers.

October 29, 2009, 05:57		#9
ahorvat New Member Andrej Horvat Join Date: Mar 2009 Posts: 7 Rep Power: 17	Guys, I would like to continue this discussion, but at present I cannot. The FSI project, which triggered my comments was for an ANSYS reseller and his license to sell ANSYS products was just revoked because of my comments (yes, I do use my real name although I intentionally avoided to name the ANSYS people involved). I feel obliged to do everything in my power get him through.

October 29, 2009, 12:40		#10
pimpa New Member Join Date: Mar 2009 Posts: 9 Rep Power: 17	ahorvat, ANSYS CFX does have QA (Class3) Error-Reporting for customers on maintenance (TECS). Actual bugs are reported, remedies are given. Software version containing first appearance of a bug, version for which bug will be fixed is given as well. Each version of CFX does see roughly some ten bugs uncovered (and older ones fixed). Did you check the Error-Report's? Did your bugs coincide with the bugs reported there or where these additional (relevant) bugs ?