search

ansys / pyadditive

Pythonic interface client to Ansys additive simulation service
https://additive.docs.pyansys.com
MIT License
2 stars 1 forks source link

Examples in pyadditive v0.17.0 not working when display language uses `,` for decimal separator #261

Closed maxcarrat90 closed 5 months ago

maxcarrat90 commented 5 months ago

@pkrull-ansys Thank you very much for you hints. Now I am able to correctly run the examples/00_additive_single_bead.py. However, using material = additive.material("316L") , machine.scan_speed = 1, and machine.laser_power = 195 I got zero values in all Melt Pool Dimensions, see plot below: MP_plot

Originally posted by @maxcarrat90 in https://github.com/ansys/pyadditive/discussions/256#discussioncomment-8472900

pkrull-ansys commented 5 months ago

@maxcarrat90 This is what I'm seeing using pyadditive v0.17.0 and Ansys 24.1 when running 00_additive_single_bead.py and changing the following values

material = additive.material("316L")
machine.scan_speed = 1  # m/s
machine.laser_power = 195  # W

single-bead-316l-1-195

Were any other edits made to the python file? Line 120 converts from meters to microns. Did it get changed by chance? What does df.head() show (line 145)? Do you see only zeros or nan's for the values?

maxcarrat90 commented 5 months ago

@pkrull-ansys I have cloned the latest pyadditive version and reinstalled everything to be sure that the python's script are all clean. I attach the melt_pool.csv file and here is what I get on the Python Console at the end of the simulation:

C:\Users\Massimo\anaconda3\envs\pyAnsys\python.exe "C:\Program Files\ANSYS Inc\PyAnsys\PyAdditive\pyadditive-0.17\examples\00_additive_single_bead.py" 
user data path: C:\Users\Massimo\AppData\Local\Ansys Inc\pyadditive
Client 0.18.dev5, API version: 1.5.0
ServerConnectionStatus(connected=True, channel_str='127.0.0.1:64708', metadata={'Version': '24.1.0-2', 'API version': '1.4.1'})
2024-02-16 09:00:53 Connected to 127.0.0.1:64708
AdditiveMachine
laser_power: 195 W
scan_speed: 1.0 m/s
heater_temperature: 80 °C
layer_thickness: 5e-05 m
beam_diameter: 0.0001 m
starting_layer_angle: 57 °
layer_rotation_angle: 67 °
hatch_spacing: 0.0001 m
slicing_stripe_width: 0.01 m

Single input
Single bead simulation: 100%|██████████| 100/100 [01:24<00:00,  1.19it/s]

Process finished with exit code 0

melt_pool.csv

pkrull-ansys commented 5 months ago

@maxcarrat90 I'm a bit baffled. I've duplicated your configuration and run the example.

(.venv) 08:35 AM pyadditive $ python examples/00_additive_single_bead.py 
2024-02-16 08:36:01 Connected to 127.0.0.1:59008
user data path: C:\Users\pkrull\AppData\Local\Ansys Inc\pyadditive
Client 0.18.dev5, API version: 1.5.0
ServerConnectionStatus(connected=True, channel_str='127.0.0.1:59008', metadata={'API version': '1.4.1', 'Version': '24.1.0-2'})
AdditiveMachine
laser_power: 195 W
scan_speed: 1.0 m/s
heater_temperature: 80 °C
layer_thickness: 5e-05 m
beam_diameter: 0.0001 m
starting_layer_angle: 57 °
layer_rotation_angle: 67 °
hatch_spacing: 0.0001 m
slicing_stripe_width: 0.01 m

Single input
Single bead simulation: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 100/100 [00:43<00:00,  2.29it/s]

Here are the melt pool statistics I get. melt_pool.csv

Questions:

  1. Do you get zero values with 316L when using a higher laser power and slower scan speed, say 700 W and 0.5 m/s?
  2. Do you get non-zero values for IN718 using laser power = 195 and scan speed = 1 m/s?
maxcarrat90 commented 5 months ago

@pkrull-ansys Thanks for your help! Regarding your questions:

  1. MP_plot_700_05

  2. File "C:\Program Files\ANSYS Inc\PyAnsys\PyAdditive\pyadditive-0.17\examples\00_additive_single_bead.py", line 108, in <module>
raise Exception(summary.message)
Exception: Exception: INVALID STATE: The melt pool has become too wide and deep to produce valid results (current width = 825.000000 microns, current depth = 1200.000000 microns). We recommend you reduce the energy input by reducing Laser Power and/or increasing Scan Speed.
2024-02-16 18:14:44 Uncaught exception
Traceback (most recent call last):
File "C:\Program Files\ANSYS Inc\PyAnsys\PyAdditive\pyadditive-0.17\examples\00_additive_single_bead.py", line 108, in <module>
raise Exception(summary.message)
Exception: Exception: INVALID STATE: The melt pool has become too wide and deep to produce valid results (current width = 825.000000 microns, current depth = 1200.000000 microns). We recommend you reduce the energy input by reducing Laser Power and/or increasing Scan Speed.
pkrull-ansys commented 5 months ago

@maxcarrat90 The server seems to be misinterpreting the parameters sent to it. On the host where the server is running, what locale is used? More specifically, what's the default character used for the decimal point? The server should account for different locales but this hasn't received much testing yet.

maxcarrat90 commented 5 months ago

@pkrull-ansys here what I get checking locale on the powershell:

(pyAnsys) PS C:\Program Files\ANSYS Inc\PyAnsys\PyAdditive> Get-WinSystemLocale

LCID             Name             DisplayName                                                                                                                                      
----             ----             -----------                                                                                                                                      
1033             en-US            Inglese (Stati Uniti)                                                                                                                            

(pyAnsys) PS C:\Program Files\ANSYS Inc\PyAnsys\PyAdditive>
pkrull-ansys commented 5 months ago

@maxcarrat90 I switched my display language to Italian and think I found what the problem is. Material data is stored in comma separated value files by the server. When the user chooses a material for a simulation, that data is transferred to the client then back to the server as part of the simulation inputs. This is done so that the user can modify material parameters if they choose. Regardless, when the display language does not use . for the decimal separator, the material data is mishandled by the client. Until I can get this fixed, you will need to modify your display language settings to use . for the decimal separator or switch to English to successfully run the examples in pyadditive.

maxcarrat90 commented 5 months ago

@pkrull-ansys Thank you very much for you support! The issue is now fixed. Now, I would like to run a parametric study using a user defined material and laser beam shape (multi-mode). Would this be possible?

pkrull-ansys commented 5 months ago

@maxcarrat90 User defined materials are supported. Shaped lasers are not yet supported.

For user defined materials, the pyadditive documentation is a little light and refers you to documentation for our legacy application. Let me know if you need help and I'll answer your questions or put you in touch with someone who can.

PyAdditive examples

Custom material further information

maxcarrat90 commented 5 months ago

@pkrull-ansys Thank you for the links! Would it be possible to contribute in developing shaped laser functions? I need this feature within a research project framework and I can work on its implementation.

pkrull-ansys commented 5 months ago

@maxcarrat90 Unfortunately, I think it would likely be used in our thermal solver, which is not open source. We do have a shaped laser feature in our backlog. If you can write up a feature request as an issue for pyadditive, I can see if what we intend meets your needs or whether we can modify it to meet your needs.

pkrull-ansys commented 5 months ago

This bug is in the Additive server. It is being fixed and will be available in the Ansys 2024 R2 release.