Command-line interface¶

PyFlange includes a command-line tool that performs deterministic and Monte Carlo bolt fatigue analyses of flange cases described in the PyFlange JSON format, without writing any code. After installing PyFlange, it can be invoked either as pyflange or as py -m pyflange (python -m pyflange on Linux and macOS):

py -m pyflange --help
py -m pyflange COMMAND --help
py -m pyflange --version

Tip

py -m pyflange always works. The shorter pyflange works only if the folder of the Python scripts is in your PATH. The examples below use the short form.

A complete example case file is available here: fatigue-case.json. For a guided introduction, see First analysis without code and Monte Carlo from the command line.

Command	Purpose
`determ`	Deterministic fatigue analysis of a case.
`sample`	Monte Carlo simulation of a case.
`inspect`	Detailed report of one realization of a Monte Carlo sample.
`stats`	Summary statistics of a Monte Carlo sample.

Commands¶

determ¶

Performs a deterministic fatigue analysis and writes a detailed report of the case and its results to a JSON file (see Result files).

pyflange determ case.json                    # writes case-res.json
pyflange determ case.json --xlsx             # also writes case-res.xlsx
pyflange determ case.json --gap-angle 45 --gap-height 1.5 --tilt-angle 0.2 --smf 1.1

Option	Description
`--gap-angle`	Override of the gap angle, in degrees. The shell stiffness is re-interpolated for the new angle.
`--gap-height`	Override of the gap height, in mm.
`--tilt-angle`	Override of the tilt angle, in degrees.
`--smf`	Stress multiplication factor. Defaults to the case `SMF`, or 1.0.
`--odir`	Output folder. Defaults to the current folder.
`--oname`	Output file name. Defaults to `<case>-res[<overrides>].json`.
`--xlsx`	Also write the report to an Excel file, as a key/value table with a named cell per value.

The main results in the report are damage, FatigueLife and, in the FlangeSegment section, BoltPreload, the polynomial points Point1 to Point4 and the curves TransferFunctions.

sample¶

Runs a Monte Carlo simulation, based on the Stats section of the case, and writes one row per realization to a CSV file.

pyflange sample case.json                     # 100000 realizations, case-sample.csv
pyflange sample case.json --size 5000 --seed 1 --xlsx

Option	Description
`--size`	Number of realizations. Defaults to 100000.
`--seed`	Seed of the random number generator. With the same seed and case, the same sample is generated. If omitted, every run is different.
`--odir`	Output folder. Defaults to the current folder.
`--oname`	Output file name. Defaults to `<case>-sample.csv`.
`--xlsx`	Also write the sample to an Excel file.

The CSV columns are:

Column	Unit	Description
`BoltPreload`	N	Bolt preload after losses.
`GapShapeFactor`	-	Gap shape factor.
`TiltAngle`	rad	Flange tilt angle.
`GapAngle`	rad	Gap angle.
`GapHeight`	m	Gap height.
`FatigueClass`	Pa	Knee-point stress range of the random bolt S-N curve (size effect included).
`MomentRangeCoefficient`	-	Factor applied to the Markov matrix ranges.
`Damage`	-	Fatigue damage.
`AllowableDamage`	-	Allowable damage.
`FatigueLife`	as Markov matrix duration	Fatigue life.
`SMF`	-	Stress multiplication factor.
`Utilization`	-	`Damage / AllowableDamage`.

The first column (without header) is the number of the realization, starting from 0, to be used with inspect.

inspect¶

Rebuilds a single realization of a sample and writes its detailed report, like determ does for the deterministic case. This is useful to investigate outliers.

pyflange inspect case.json 1234               # reads case-sample.csv, writes case-realization-1234.json
pyflange inspect case.json 1234 --spath other-sample.csv --xlsx

Option	Description
`--spath`	Path of the sample CSV file. Defaults to `<case>-sample.csv`, in the folder of the case.
`--odir`, `--oname`, `--xlsx`	As for `determ`.

stats¶

Prints minimum, 5% quantile, average, 95% quantile, maximum, standard deviation and coefficient of variation of each column of a sample.

pyflange stats case-sample.csv

Running batches¶

To analyze all the case files of a folder, use a loop of the Windows command prompt:

for %f in (*.json) do pyflange determ "%f" --odir results

(In a .bat file, write %%f instead of %f.) The results folder must exist; create it with mkdir results. Note that the loop also picks up the result files if they are written in the same folder as the cases.

In PowerShell:

Get-ChildItem *.json | ForEach-Object { pyflange determ $_.FullName --odir results }

For more complex batches, for example to collect the results in one table, use Python: see Writing case files from Python.

Migrating from Forrester¶

The PyFlange command-line interface replaces the Forrester tool of the Bolt & Beautiful project:

forrester determ, sample, inspect and stats become pyflange determ, sample, inspect and stats, with the same options. sample adds the --seed option and writes the same CSV columns.
Forrester case files (format version 0.1.0) can be used as they are.
forrester hist has been removed: plot the sample CSV with your tool of choice.
forrester export-sgre-case is not part of PyFlange: the conversion of company-specific file formats to the PyFlange JSON format is left to dedicated tools.