search

ansys / pymapdl

Pythonic interface to MAPDL
https://mapdl.docs.pyansys.com
MIT License
423 stars 120 forks source link

Unable to start ANSYS within 120 sec #180

Closed Irfan995 closed 4 years ago

Irfan995 commented 4 years ago

Hello, I am very much happy to find pyansys. But unfortunately, I am unable to run the example code of your pyansys repository. Whenever, I run the code, I find error after a while. The error is given below. Can you give me any suggestion how can I overcome this??? ansys

akaszynski commented 4 years ago

Interesting. What's your configuration (i.e. version of pyansys and ANSYS)?

anlblk commented 4 years ago

Hi Alex, I have the same problem in last 3 days. Even for the old codes, i am not able to run them now. I have pyansys 39.10, Ansys 2019 r3, academic edition and Windows10. Also, I have problems to create dummy file and remove lock file, therefore i had to inactivate these sections. I am sure mapdl is not open while i am trying to run my code. I re-install pyansys, but still it is same. Do you have any suggestions? Thanks.

akaszynski commented 4 years ago

@anlblk, we’re you able to get the older versions of pyansys to run your scripts? If so, can you try installing an older version of pyansys? For example, 0.39.8?

anlblk commented 4 years ago

@akaszynski I reinstalled everything, python, pyansys and Ansys. Now, if the all files are in the same directory with Ansys, it works, but if only i run it from command line, not with an ide. By the way, it is same with Linus OS. It stalls when you run it with Spyder. But still, i needed to inactivate "create a dummy input file" section and os.remove line for the "Check for lock file" section. This is not necessary with Linux. I will try your suggestion asap, but since I need to finish a project for now, i have to use it in this way for few days. Thanks for your concern.

akaszynski commented 4 years ago

Thanks for testing it out. I’ll debug it and have a release out probably by Friday this week.

natter1 commented 4 years ago

Might be completely unrelated, but I remember having a similar issue after deinstalling an ANSYS version to go back using an older one (19.2). Turned out pyansys tried to run the newer deinstalled version of ANSYS, because the Deinstallation left an entry in the registry somewhere (as I found out by searching in registry). Removing that entry solved the issue

Irfan995 commented 4 years ago

Hello, i am able to solve this problem using ansys 19.2 Thanks everyone

On Mon, 3 Feb 2020, 4:35 am natter1, notifications@github.com wrote:

Might be completely unrelated, but I remember having a similar issue after deinstalling an ANSYS version to go back using an older one (19.2). Turned out pyansys tried to run the newer deinstalled version of ANSYS, because the Deinstallation left an entry in the registry somewhere (as I found out by searching in registry). Removing that entry solved the issue

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/akaszynski/pyansys/issues/180?email_source=notifications&email_token=AH26223QAJTNTGQZGU4EW7DRA5DE5A5CNFSM4KND2PUKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKSC5SA#issuecomment-581185224, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH26223UMGZ7D5IOT6GPYLLRA5DE5ANCNFSM4KND2PUA .

akaszynski commented 4 years ago

Lots of people contributing, thanks! What version of ANSYS were you using @Irfan995 when this error occurred? Actually, if everyone could chime in with the version that pyansys started failing, that would be helpful to track this issue down.

Irfan995 commented 4 years ago

I used ansys 17...

On Mon, 3 Feb 2020, 2:19 pm Alex Kaszynski, notifications@github.com wrote:

Lots of people contributing, thanks! What version of ANSYS were you using @Irfan995 https://github.com/Irfan995 when this error occurred? Actually, if everyone could chime in with the version that pyansys started failing, that would be helpful to track this issue down.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/akaszynski/pyansys/issues/180?email_source=notifications&email_token=AH26223VOO42WYZ5G6LVIT3RA7HSXA5CNFSM4KND2PUKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKS4XKI#issuecomment-581290921, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH26226KUWMFFJMDPKV2XI3RA7HSXANCNFSM4KND2PUA .

akaszynski commented 4 years ago

I've been playing around with this on both Windows and Linux, and have failed to generate this error on 19.5 (2019R3) with the latest version of pyasys=0.39.10. I'll need more help debugging this, and if any of you have figured out a solution that uses CORBA for windows.

I have a feeling that it has to do with something that @natter1 pointed out, but I'm still not sure.

I'm going to release 0.39.11 with the fixes for #179, #178, #177, and #175. Once we're testing with that version. I'd appreciate any inputs regarding code changes (either in code snippets or pull requests).

anlblk commented 4 years ago

@akaszynski thanks a lot for your concern. it could be the issue as you two said, because it happened after i updated Ansys from 19.2 to 19.3 and i did not pay attention for registry files. I will try each suggestions when i have time.

natter1 commented 4 years ago

Hab jetzt nochmal geschaut - wenn ich mich richtig erinnere habe ich zunächst find_ansys() angepasst:

def find_ansys():
    """ Searches for ansys path within enviornmental variables """
    versions = []
    paths = []
    for var in os.environ:
        if 'ANSYS' in var:
            if '_DIR' in var:
                try:
                    path = os.environ[var]
                    if os.path.isdir(path):
                        versions.append(int(var[5:8]))
                        paths.append(path)
                except:
                    continue

I think I put versions.append(int(var[5:8])) inside the isdir-check, to prevent a version without valid path.

akaszynski commented 4 years ago

Das macht sinn. Danke, ich werde das probieren.

akaszynski commented 4 years ago

I'll also check that the executable exists as well. Old paths may exist but without the ansys binaries...

akaszynski commented 4 years ago

Should be fixed in pyansys==0.39.12. Please let me know if it works for you @Irfan995 and @anlblk.

Irfan995 commented 4 years ago

I am facing another problem... I converted an apdl script to python script but while running the python script , I am getting an error... the error is "unable to remove lock file"... how can I solve this error???

anlblk commented 4 years ago

Should be fixed in pyansys==0.39.12. Please let me know if it works for you @Irfan995 and @anlblk.

@akaszynski for me, it works with Ide, too now, thanks. I still have Ansys 19.3.

for this lock file issue still remains. The thing about it, it occurs if only the code cannot complete the task and somehow if Ansys cannot exited. If there is an error on the code, it always happens. After that, when you try to run something, it keeps giving error says lock file is using, so it's not reachable.

When it does, i should run Ansys APDL itself, so he asks if i want to override the lock file, i go with yes and then close APDL, and run pyansys again. If there is no error, I can use pyansys without any problem. I don't know if this rings any bell, but I assume @Irfan995 has the same issue.

Irfan995 commented 4 years ago

error *TREAD is not recognized PREP7 command.. this error was given while running a converted python script... but tread is a prep7 command

akaszynski commented 4 years ago

I'll have to investigate the lockfile issue. The real issue is that the windows process isn't closing, which must be annoying if something fails. I'll investigate that.

akaszynski commented 4 years ago

As for *TREAD, which version of ANSYS are you using @Irfan995? Any command in MAPDL should also be in pyansys, and in fact that error is coming right from MAPDL.

Irfan995 commented 4 years ago

I am using ansys version 19.2

On Thu, 13 Feb 2020, 5:10 am Alex Kaszynski, notifications@github.com wrote:

As for *TREAD, which version of ANSYS are you using @Irfan995 https://github.com/Irfan995? Any command in MAPDL should also be in pyansys, and in fact that error is coming right from MAPDL.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/akaszynski/pyansys/issues/180?email_source=notifications&email_token=AH26226YI7NKBD4AXXOJ5C3RCR6XLA5CNFSM4KND2PUKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOELSXXXA#issuecomment-585464796, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH2622ZADTIDTMQ33N2PMXDRCR6XLANCNFSM4KND2PUA .

akaszynski commented 4 years ago

Looks like the command is depreciated. I'm getting this 15.0

 *** NOTE ***                            CP =       0.348   TIME= 10:06:18
 The TREAD command will be removed in a future release.  Please use the  
 LDREAD command instead.

Also, in the future, could you please open a new issue when there's a different issue? That way if someone's searching for something on google, they see a clear issue/resolution thread.

Irfan995 commented 4 years ago

But the APDL scripts ran well in MAPDL GUI... after converting the APDL scripts to python scripts, when I ran the python scripts, there the errors come... Now VSBA, VOVLAP, VDELE are also giving the same error...

On Thu, 13 Feb 2020 at 23:08, Alex Kaszynski notifications@github.com wrote:

Looks like the command is depreciated. I'm getting this 15.0

NOTE CP = 0.348 TIME= 10:06:18 The TREAD command will be removed in a future release. Please use the LDREAD command instead.

Also, in the future, could you please open a new issue when there's a different issue? That way if someone's searching for something on google, they see a clear issue/resolution thread.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/akaszynski/pyansys/issues/180?email_source=notifications&email_token=AH26223Q2TO4EC4WPYGXRM3RCV5ALA5CNFSM4KND2PUKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOELVZQII#issuecomment-585865249, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH262262ALCJWYV4HPG67XTRCV5ALANCNFSM4KND2PUA .

akaszynski commented 4 years ago

This is quite strange. Would you mind uploading your script here? Both the input file and MAPDL file? You can also email me at my ANSYS email address alexander.kaszynski@ansys.com

Irfan995 commented 4 years ago

okay... I have sent the files to your email....

On Mon, 17 Feb 2020 at 03:10, Alex Kaszynski notifications@github.com wrote:

This is quite strange. Would you mind uploading your script here? Both the input file and MAPDL file? You can also email me at my ANSYS email address alexander.kaszynski@ansys.com

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/akaszynski/pyansys/issues/180?email_source=notifications&email_token=AH262255ZLY7FCL2X36GNJ3RDGTUXA5CNFSM4KND2PUKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEL4SMEA#issuecomment-586753552, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH26225NFTTDSBQL4XNN6STRDGTUXANCNFSM4KND2PUA .

Irfan995 commented 4 years ago

okay somehow I managed to solve other commands but *tread is still giving error as it is not recognised command

On Mon, 17 Feb 2020 at 11:31, Fahim Ahmed Irfan fahimirfan70@gmail.com wrote:

okay... I have sent the files to your email....

On Mon, 17 Feb 2020 at 03:10, Alex Kaszynski notifications@github.com wrote:

This is quite strange. Would you mind uploading your script here? Both the input file and MAPDL file? You can also email me at my ANSYS email address alexander.kaszynski@ansys.com

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/akaszynski/pyansys/issues/180?email_source=notifications&email_token=AH262255ZLY7FCL2X36GNJ3RDGTUXA5CNFSM4KND2PUKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEL4SMEA#issuecomment-586753552, or unsubscribe https://github.com/notifications/unsubscribe-auth/AH26225NFTTDSBQL4XNN6STRDGTUXANCNFSM4KND2PUA .

akaszynski commented 4 years ago

Please read the above comment regarding *tread. It's no longer a valid command:

 *** NOTE ***                            CP =       0.348   TIME= 10:06:18
 The TREAD command will be removed in a future release.  Please use the  
 LDREAD command instead.
mikerife commented 1 year ago

@Irfan995 Sorry about coming so late to this - I came across this discussion researching another issue. TREAD was deprecated way back in the version 5.x days. But TREAD is still an valid APDL command (read data into a table array). I think the APDL to PyMPDL translator might have stripped out the . Or something. Going forward see the PyMAPDL load_table command. Mike