Static & Transient DataFrames in PyNastran

The Jupyter notebook for this demo can be found in: - docs/quick_start/demo/op2_pandas_multi_case.ipynb - https://github.com/SteveDoyle2/pyNastran/tree/master/docs/quick_start/demo/op2_pandas_multi_case.ipynb

import os
import pandas as pd
import pyNastran
from pyNastran.op2.op2 import read_op2

pkg_path = pyNastran.__path__[0]
model_path = os.path.join(pkg_path, '..', 'models')

Solid Bending

Let’s show off combine=True/False. We’ll talk about the keys soon.

solid_bending_op2 = os.path.join(model_path, 'solid_bending', 'solid_bending.op2')
solid_bending = read_op2(solid_bending_op2, combine=False, debug=False)
print(solid_bending.displacements.keys())
INFO: op2_scalar.py:1459 op2_filename = 'c:\\nasa\\m4\\formats\\git\\pynastran\\pyNastran\\..\\models\\solid_bending\\solid_bending.op2' 
dict_keys([(1, 1, 1, 0, 0, '', '')])

c:nasam4formatsgitpynastranpyNastranop2op2.py:740: FutureWarning:
Panel is deprecated and will be removed in a future version.
The recommended way to represent these types of 3-dimensional data are with a MultiIndex on a DataFrame, via the Panel.to_frame() method
Alternatively, you can use the xarray package http://xarray.pydata.org/en/stable/.
Pandas provides a .to_xarray() method to help automate this conversion.

  obj.build_dataframe()

solid_bending_op2 = os.path.join(model_path, 'solid_bending', 'solid_bending.op2')
solid_bending2 = read_op2(solid_bending_op2, combine=True, debug=False)
print(solid_bending2.displacements.keys())
INFO: op2_scalar.py:1459 op2_filename = 'c:\\nasa\\m4\\formats\\git\\pynastran\\pyNastran\\..\\models\\solid_bending\\solid_bending.op2' 
dict_keys([1])

Single Subcase Buckling Example

The keys cannot be “combined” despite us telling the program that it was OK. We’ll get the following values that we need to handle. #### isubcase, analysis_code, sort_method, count, subtitle * isubcase -> the same key that you’re used to accessing * sort_method -> 1 (SORT1), 2 (SORT2) * count -> the optimization count * subtitle -> the analysis subtitle (changes for superlements) * analysis code -> the “type” of solution

### Partial code for calculating analysis code:

 if trans_word == 'LOAD STEP':  # nonlinear statics
    analysis_code = 10
elif trans_word in ['TIME', 'TIME STEP']:  # TODO check name
    analysis_code = 6
elif trans_word == 'EIGENVALUE':  # normal modes
    analysis_code = 2
elif trans_word == 'FREQ':  # TODO check name
    analysis_code = 5
elif trans_word == 'FREQUENCY':
    analysis_code = 5
elif trans_word == 'COMPLEX EIGENVALUE':
    analysis_code = 9
else:
    raise NotImplementedError('transient_word=%r is not supported...' % trans_word)

Let’s look at an odd case:

You can do buckling as one subcase or two subcases (makes parsing it a lot easier!).

However, you have to do this once you start messing around with superelements or multi-step optimization.

For optimization, sometimes Nastran will downselect elements and do an optimization on that and print out a subset of the elements. At the end, it will rerun an analysis to double check the constraints are satisfied. It does not always do multi-step optimization.

op2_filename = os.path.join(model_path, 'sol_101_elements', 'buckling_solid_shell_bar.op2')
model = read_op2(op2_filename, combine=True, debug=False, build_dataframe=True)
INFO: op2_scalar.py:1459 op2_filename = 'c:\\nasa\\m4\\formats\\git\\pynastran\\pyNastran\\..\\models\\sol_101_elements\\buckling_solid_shell_bar.op2' 
stress_keys = model.cquad4_stress.keys()
print (stress_keys)

# subcase, analysis_code, sort_method, count, isuperelmemnt_adaptivity_index, pval_step
key0 = (1, 1, 1, 0, 0, '', '')
key1 = (1, 8, 1, 0, 0, '', '')

dict_keys([(1, 1, 1, 0, 0, '', ''), (1, 8, 1, 0, 0, '', '')])

Keys: * key0 is the “static” key * key1 is the “buckling” key

Similarly: * Transient solutions can have preload * Frequency solutions can have loadsets (???)

Moving onto the data frames

  • The static case is the initial deflection state
  • The buckling case is “transient”, where the modes (called load steps or lsdvmn here) represent the “times”

pyNastran reads these tables differently and handles them differently internally. They look very similar though.

stress_static = model.cquad4_stress[key0].data_frame
stress_transient = model.cquad4_stress[key1].data_frame

# The final calculated factor:
#   Is it a None or not?
# This defines if it's static or transient
print('stress_static.nonlinear_factor = %s' % model.cquad4_stress[key0].nonlinear_factor)
print('stress_transient.nonlinear_factor = %s' % model.cquad4_stress[key1].nonlinear_factor)

print('data_names  = %s' % model.cquad4_stress[key1].data_names)
print('loadsteps   = %s' % model.cquad4_stress[key1].lsdvmns)
print('eigenvalues = %s' % model.cquad4_stress[key1].eigrs)

stress_static.nonlinear_factor = nan
stress_transient.nonlinear_factor = 4
data_names  = ['lsdvmn', 'eigr']
loadsteps   = [1, 2, 3, 4]
eigenvalues = [-49357660160.0, -58001940480.0, -379750744064.0, -428462538752.0]

Static Table

# Sets default precision of real numbers for pandas output\n"
pd.set_option('precision', 2)

stress_static.head(20)
index fiber_distance oxx oyy txy angle omax omin von_mises
ElementID NodeID Location
6 CEN Top 0 -0.12 5.85e-07 9.73e-06 -1.36e-07 -89.15 9.73e-06 5.83e-07 9.46e-06
Bottom 1 0.12 4.71e-07 9.44e-06 -1.61e-07 -88.97 9.44e-06 4.69e-07 9.21e-06
4 Top 2 -0.12 -6.50e-07 9.48e-06 -1.36e-07 -89.23 9.48e-06 -6.52e-07 9.82e-06
Bottom 3 0.12 -8.37e-07 9.11e-06 -1.61e-07 -89.08 9.12e-06 -8.39e-07 9.56e-06
1 Top 4 -0.12 -6.50e-07 9.98e-06 -1.36e-07 -89.27 9.99e-06 -6.51e-07 1.03e-05
Bottom 5 0.12 -8.37e-07 9.76e-06 -1.61e-07 -89.13 9.76e-06 -8.39e-07 1.02e-05
14 Top 6 -0.12 1.82e-06 9.98e-06 -1.36e-07 -89.05 9.99e-06 1.82e-06 9.21e-06
Bottom 7 0.12 1.78e-06 9.76e-06 -1.61e-07 -88.85 9.76e-06 1.78e-06 9.01e-06
15 Top 8 -0.12 1.82e-06 9.48e-06 -1.36e-07 -88.98 9.48e-06 1.82e-06 8.72e-06
Bottom 9 0.12 1.78e-06 9.11e-06 -1.61e-07 -88.75 9.12e-06 1.78e-06 8.37e-06
7 CEN Top 10 -0.12 7.16e-07 1.02e-05 1.22e-07 89.26 1.02e-05 7.14e-07 9.82e-06
Bottom 11 0.12 7.31e-07 1.04e-05 1.53e-07 89.10 1.04e-05 7.29e-07 1.01e-05
3 Top 12 -0.12 -7.30e-07 1.04e-05 1.22e-07 89.37 1.04e-05 -7.31e-07 1.08e-05
Bottom 13 0.12 -8.05e-07 1.07e-05 1.53e-07 89.24 1.07e-05 -8.07e-07 1.12e-05
2 Top 14 -0.12 -7.30e-07 9.90e-06 1.22e-07 89.34 9.90e-06 -7.31e-07 1.03e-05
Bottom 15 0.12 -8.05e-07 1.01e-05 1.53e-07 89.20 1.01e-05 -8.07e-07 1.05e-05
17 Top 16 -0.12 2.16e-06 9.90e-06 1.22e-07 89.10 9.90e-06 2.16e-06 9.02e-06
Bottom 17 0.12 2.27e-06 1.01e-05 1.53e-07 88.88 1.01e-05 2.26e-06 9.18e-06
16 Top 18 -0.12 2.16e-06 1.04e-05 1.22e-07 89.15 1.04e-05 2.16e-06 9.52e-06
Bottom 19 0.12 2.27e-06 1.07e-05 1.53e-07 88.96 1.07e-05 2.26e-06 9.79e-06

Transient Table

# Sets default precision of real numbers for pandas output\n"
pd.set_option('precision', 3)
#import numpy as np
#np.set_printoptions(formatter={'all':lambda x: '%g'})

stress_transient.head(20)
LoadStep Item 1 2 3 4
EigenvalueReal -49357660160.0 -58001940480.0 -379750744064.0 -428462538752.0
ElementID NodeID Location
6 CEN Top fiber_distance -1.250e-01 -1.250e-01 -1.250e-01 -1.250e-01
Top oxx -3.657e+04 -1.587e+05 -1.497e+05 1.069e+06
Top oyy 2.064e+05 1.084e+06 4.032e+05 6.158e+06
Top txy 2.296e+02 -1.267e+04 4.394e+06 -3.572e+05
Top angle 8.995e+01 -8.942e+01 4.680e+01 -8.601e+01
Top omax 2.064e+05 1.084e+06 4.530e+06 6.183e+06
Top omin -3.657e+04 -1.588e+05 -4.276e+06 1.044e+06
Top von_mises 2.269e+05 1.171e+06 7.627e+06 5.733e+06
Bottom fiber_distance 1.250e-01 1.250e-01 1.250e-01 1.250e-01
Bottom oxx -2.816e+04 -9.555e+04 -1.942e+05 -4.882e+05
Bottom oyy 1.402e+05 7.325e+05 7.017e+03 -2.785e+05
Bottom txy 7.409e+04 -3.522e+04 4.535e+06 -3.533e+05
Bottom angle 6.933e+01 -8.757e+01 4.564e+01 -5.326e+01
Bottom omax 1.682e+05 7.340e+05 4.442e+06 -1.480e+04
Bottom omin -5.611e+04 -9.705e+04 -4.630e+06 -7.519e+05
Bottom von_mises 2.022e+05 7.870e+05 7.857e+06 7.446e+05
4 Top fiber_distance -1.250e-01 -1.250e-01 -1.250e-01 -1.250e-01
Top oxx -9.976e+04 -5.802e+05 -2.925e+05 7.936e+05
Top oyy -1.102e+06 1.461e+06 -3.138e+06 6.441e+06
Top txy 2.296e+02 -1.267e+04 4.394e+06 -3.572e+05