deeplearning/modulus/modulus-sym-v130/_modules/modulus/sym/geometry/adf.html
Source code for modulus.sym.geometry.adf
# Copyright (c) 2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import torch
import numpy as np
from typing import List, Tuple, Dict
[docs]class ADF(torch.nn.Module):
"""
Used for hard imposition of boundary conditions.
Currently supports 2d geometries and Dirichlet boundary conditions.
Contributors: M. A. Nabian, R. Gladstone, H. Meidani, N. Sukumar, A. Srivastava
Reference: "Sukumar, N. and Srivastava, A., 2021.
Exact imposition of boundary conditions with distance functions in physics-informed deep neural networks.
Computer Methods in Applied Mechanics and Engineering, p.114333."
"""
def __init__(self):
super().__init__()
self.mu: float = 2.0
self.m: float = 2.0
self.eps: float = 1e-8
[docs] def forward(self, invar):
raise RuntimeError("No forward method was defined for ADF or its child class")
[docs] @staticmethod
def r_equivalence(omegas: List[torch.Tensor], m: float = 2.0) -> torch.Tensor:
"""
Computes the R-equivalence of a collection of approximate distance functions
Parameters
----------
omegas : List[torch.Tensor]
List of ADFs used to compute the R-equivalence.
m: float
Normalization order
Returns
-------
omega_E : torch.Tensor
R-equivalence distance
"""
omega_E = torch.zeros_like(omegas[0])
for omega in omegas:
omega_E += 1.0 / omega**m
omega_E = 1.0 / omega_E ** (1.0 / m)
return omega_E
[docs] @staticmethod
def transfinite_interpolation(
bases: List[torch.Tensor], indx: int, eps: float = 1e-8
) -> torch.Tensor:
"""
Performs transfinite interpolation of the boundary conditions
Parameters
----------
bases: List[torch.Tensor]
List of ADFs used for the transfinite interpolation.
indx: int
index of the interpolation basis
eps: float
Small value to avoid division by zero
Returns
-------
w : torch.Tensor
Interpolation basis corresponding to the input index
"""
bases_reduced = [bases[i] for i in range(len(bases)) if i != indx]
numerator = torch.prod(torch.stack(bases_reduced), dim=0)
denominator = 0.0
for j in range(len(bases)):
denom_term = [bases[i] for i in range(len(bases)) if i != j]
denominator += torch.prod(torch.stack(denom_term), dim=0)
w = torch.div(numerator, denominator + eps)
return w
[docs] @staticmethod
def infinite_line_adf(
points: Tuple[torch.Tensor], point_1: Tuple[float], point_2: Tuple[float]
) -> torch.Tensor:
"""
Computes the pointwise approximate distance for an infinite line
Parameters
----------
points: Tuple[torch.Tensor]
ADF will be computed on these points
point_1: Tuple[float]
One of the two points that form the infinite line
point_2: Tuple[float]
One of the two points that form the infinite line
Returns
-------
omega : torch.Tensor
pointwise approximate distance
"""
L = ADF._distance(point_1, point_2)
omega = (
(points[0] - point_1[0]) * (point_2[1] - point_1[1])
- (points[1] - point_1[1]) * (point_2[0] - point_1[0])
) / L
return omega
[docs] @staticmethod
def line_segment_adf(
points: Tuple[torch.Tensor], point_1: Tuple[float], point_2: Tuple[float]
) -> torch.Tensor:
"""
Computes the pointwise approximate distance for a line segment
Parameters
----------
points: Tuple[torch.Tensor]
ADF will be computed on these points
point_1: Tuple[float]
Point on one end of the line segment
point_2: Tuple[float]
Point on the other ned of the line segment
Returns
-------
omega : torch.Tensor
pointwise approximate distance
"""
L = ADF._distance(point_1, point_2)
center = ADF._center(point_1, point_2)
f = ADF.infinite_line_adf(points, point_1, point_2)
t = ADF.circle_adf(points, L / 2, center)
phi = torch.sqrt(t**2 + f**4)
omega = torch.sqrt(f**2 + ((phi - t) / 2) ** 2)
return omega
[docs] @staticmethod
def circle_adf(
points: Tuple[torch.Tensor], radius: float, center: Tuple[float]
) -> torch.Tensor:
"""
Computes the pointwise approximate distance for a circle
Parameters
----------
points: Tuple[torch.Tensor]
ADF will be computed on these points
radius: float
Radius of the circle
center: Tuple[float]
Center of the circle
Returns
-------
omega : torch.Tensor
pointwise approximate distance
"""
omega = (
radius**2 - ((points[0] - center[0]) ** 2 + (points[1] - center[1]) ** 2)
) / (2 * radius)
return omega
[docs] @staticmethod
def trimmed_circle_adf(
points: Tuple[torch.Tensor],
point_1: Tuple[float],
point_2: Tuple[float],
sign: int,
radius: float,
center: float,
) -> torch.Tensor:
"""
Computes the pointwise approximate distance of a trimmed circle
Parameters
----------
points: Tuple[torch.Tensor]
ADF will be computed on these points
point_1: Tuple[float]
One of the two points that form the trimming infinite line
point_2: Tuple[float]
One of the two points that form the trimming infinite line
sign: int
Specifies the trimming side
radius: float
Radius of the circle
center: Tuple[float]
Center of the circle
Returns
-------
omega : torch.Tensor
pointwise approximate distance
"""
assert sign != 0, "sign should be non-negative"
f = ADF.circle_adf(points, radius, center)
t = np.sign(sign) * ADF.infinite_line_adf(points, point_1, point_2)
phi = torch.sqrt(t**2 + f**4)
omega = torch.sqrt(f**2 + ((phi - t) / 2) ** 2)
return omega@staticmethod
def _distance(point_1: Tuple[float], point_2: Tuple[float]) -> torch.Tensor:
"""
Computes the distance between two points
point_1: Tuple[float]
The first point
point_2: Tuple[float]
The second point
Returns
-------
distance : torch.Tensor
distance between the two points
"""
distance = np.sqrt(
(point_2[0] - point_1[0]) ** 2 + (point_2[1] - point_1[1]) ** 2
)
return distance
@staticmethod
def _center(point_1: Tuple[float], point_2: Tuple[float]) -> Tuple[float]:
"""
Computes the center of the two points
point_1: Tuple[float]
The first point
point_2: Tuple[float]
The second point
Returns
-------
center : torch.Tensor
Center the two points
"""
center = ((point_1[0] + point_2[0]) / 2, (point_1[1] + point_2[1]) / 2)
return center