This software is part of the Bolt and Beautiful project funded with subsidy from the Top Sector Energy of the Dutch Ministry of Economic Affairs.

PyFlange Python Package

PyFlange is an open-source Python library for the analysis of large bolted ring-flange connections, with a focus on offshore wind turbine support structures.

The package was developed within the Bolt & Beautiful GROW project by KCI, Siemens Gamesa, JBO and TNO. It provides a computationally efficient implementation of analytical ring-flange models that can be used to predict bolt axial forces and bending moments resulting from shell loading. The core analytical formulation is based on Marc Seidel's polynomial model version which is expected to be published in IEC 61400-6-ED2.

Both L-flange and T-flange ring connections are supported, through the PolynomialLFlangeSegment and PolynomialTFlangeSegment classes respectively.

In addition to flange-segment modelling, PyFlange contains:

• Objects representing standard metric bolts, nuts and washers.
• Gap modelling utilities for flange imperfections and manufacturing tolerances.
• Fatigue assessment utilities for bolted connections.
• Statistical tools and probability distributions.
• Random samplers (pyflange.stats) and a worked-out Monte Carlo simulation example for probabilistic assessments.
• Validation cases and documentation for ring-flange applications.

PyFlange is intended for engineering analyses in which large numbers of flange evaluations are required and where computational efficiency is important.

The rest of this documentation explains how to install the package, create flange-segment models and use the available analysis tools.

Getting Started

PyFlange requires Python 3.9 or later. Its dependencies (numpy, scipy, pandas and metrum) are installed automatically via pip.

You can install PyFlange via pip as follows:

pip install pyflange

After installing the package, you can import it in your python code and start using it. First of all, you need to create a FlangeSegment object as shown below.

# Create the bolt object
from pyflange.bolts import StandardMetricBolt, ISOFlatWasher, ISOHexNut
M80_bolt   = StandardMetricBolt("M80", "10.9", shank_length=0.270, stud=True)
M80_washer = ISOFlatWasher("M80")
M80_nut    = ISOHexNut("M80")

# Define the gap parameters
# (pyflange.gap is deprecated; gap_height_distribution now lives in pyflange.stats)
from pyflange.stats import gap_height_distribution
from math import pi
D = 7.50                        # meters, flange outer diameter
gap_angle = pi/6                # 30 deg gap angle
gap_length = gap_angle * D/2    # outer length of the gap
u_tol = 0.0014                  # flatness tolerance in mm/mm
gap_dist = gap_height_distribution(D, u_tol, gap_length)    # lognormal distribution

# Create the FlangeSegment model
from pyflange.flangesegments import PolynomialLFlangeSegment, Gap
Nb = 120    # number of bolts
fseg = PolynomialLFlangeSegment(
    a = 0.2325,              # distance between inner face of the flange and center of the bolt hole
    b = 0.1665,              # distance between center of the bolt hole and center-line of the shell
    s = 0.0720,              # shell thickness
    t = 0.2000,              # flange thickness
    R = D/2,                 # shell outer curvature radius
    central_angle = 2*pi/Nb, # angle subtented by the flange segment arc

    Zg = -14795000 / Nb,     # load applied to the flange segment shell at rest
                             # (normally dead weight of tower + RNA, divided by the number of bolts)

    bolt = M80_bolt,         # bolt object created above
    Fv = 2876000,            # design bolt preload, after preload losses

    Do = 0.086,              # bolt hole diameter
    washer = M80_washer,     # washer object created above
    nut = M80_nut,           # nut object created above

    gap = Gap(height = gap_dist.ppf(0.95),    # maximum longitudinal gap height, 95% quantile
              angle = gap_angle)              # longitudinal gap length
    )

# Verify that failure mode B is governing for this flange segment, which is
# a requirement for the polynomial model to be applicable. If another
# failure mode governs, a ValueError is raised.
fseg.validate(fy_sh=325e6, fy_fl=295e6)

Notice that a consistent set of units of measurements has been used for inputs, namely: meter for distances, radians for angles and newton for forces. It is not required to always use these units (meter, newton), but you should choose your units and always apply them consistently.

Once you have your fseg object, you can obtain the bolt forces and moments as follows:

Fs = fseg.bolt_axial_force(3500)    # bolt force corresponding to the tower shell force Z = 3500 N
Ms = fseg.bolt_bending_moment(2000) # bolt bending moment corresponding to the tower shell force Z = 2000 N

The argument Z, passed to bolt_axial_force and bolt_bending_moment can also be a numpy array. In that case an array of Fs and Ms values will be returned.

import numpy as np
Z = np.array([2000, 2500, 3000])
Fs = fseg.bolt_axial_force(Z)       # return the numpy array (Fs(2000), Fs(2500), Fs(3000))
Ms = fseg.bolt_bending_moment(Z)    # return the numpy array (Ms(2000), Ms(2500), Ms(3000))

Fatigue analysis

Once a FlangeSegment is available, it can be combined with a load history (expressed as a MarkovMatrix) to perform a bolt fatigue assessment:

from pyflange.fatigue import MarkovMatrix, BoltFatigueAnalysis
import numpy as np

# Markov matrix representing the bending-moment load history acting on the flange
flange_mkvm = MarkovMatrix(
    range    = np.array([50e3, 80e3, 120e3]),   # Nm, load range of each bin
    mean     = np.array([10e3, 15e3, 20e3]),    # Nm, mean load of each bin
    cycles   = np.array([1e6, 5e5, 1e4]),       # number of cycles of each bin
    duration = 25                               # years represented by this matrix
)

# Run the fatigue analysis for the flange segment created above.
# If no custom fatigue curve is given, a BoltFatigueCurve is derived
# automatically from the bolt's nominal diameter, according to IEC 61400-6 AMD1.
fatigue = BoltFatigueAnalysis(fseg, flange_mkvm)

print(fatigue.damage)        # cumulated fatigue damage
print(fatigue.fatigue_life)  # fatigue life, in the same unit as `duration`

Learn More

For more details, read the PyFlange API documentation, which covers the bolts, flangesegments, fatigue and stats modules, as well as a worked-out Monte Carlo simulation example.

Testing

Once you clone this package, you can run the unit tests (assuming you have already the pytest module installed) as follows:

cd <path-to-package>
py -m pytest

Contributing

You can contribute to this project by reporting a bug, highlighting a necessary improvement, submitting a code improvement or by asking a question. For instructions about how to do all these things, please see our contribution guidelines.

Since the typical PyFlange user is expected not to be a professional programmer and probably not familiar with Git, we have created a Git tutorial for non-programmers to make contributing a less intimidating process.

License

pyFlange - python library for large flanges design Copyright (C) 2024 KCI The Engineers B.V., Siemens Gamesa Renewable Energy B.V., Nederlandse Organisatie voor toegepast-natuurwetenschappelijk onderzoek TNO.

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License, as published by the Free Software Foundation, either version 3 of the License, or any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License version 3 for more details.

You should have received a copy of the GNU General Public License version 3 along with this program. If not, see https://www.gnu.org/licenses/.