def compute_Ts(self, PD_deck, PD_problem):
Ts = np.zeros((int(PD_deck.Num_Nodes)))
for x_i in range(0, self.len_x):
index_x_family = PD_problem.get_index_x_family(x_i)
for x_p in index_x_family:
Ts[x_i] = Ts[x_i] + self.T[x_i, x_p] - self.T[x_p, x_i]
Ts[x_i] = Ts[x_i] * PD_deck.Volume
self.Ts = Ts
def compute_ext_state(self, PD_deck, PD_problem, y):
# Initialization for e
e = np.zeros((int(PD_deck.Num_Nodes), int(PD_deck.Num_Nodes)))
for x_i in range(0, len(PD_problem.x)):
index_x_family = PD_problem.get_index_x_family(x_i)
for x_p in index_x_family:
e[x_i, x_p] = np.absolute(
y[x_p] - y[x_i]) - np.absolute(PD_problem.x[x_p] - PD_problem.x[x_i])
self.e = e
def compute_T(self, PD_deck, PD_problem, y):
w = PD_deck.Influence_Function
M = PD_problem.compute_m(PD_deck.Num_Nodes, y)
tscal = np.zeros((int(PD_deck.Num_Nodes), int(PD_deck.Num_Nodes)))
for x_i in range(0, self.len_x):
index_x_family = PD_problem.get_index_x_family(x_i)
for x_p in index_x_family:
tscal[x_i,
x_p] = (w / PD_problem.weighted_function(PD_deck,
PD_problem.x,
x_i)) * self.Modulus * self.e[x_i,
x_p]
self.tscal = tscal
T = np.zeros((int(PD_deck.Num_Nodes), int(PD_deck.Num_Nodes)))
for x_i in range(0, self.len_x):
index_x_family = PD_problem.get_index_x_family(x_i)
for x_p in index_x_family:
T[x_i, x_p] = tscal[x_i, x_p] * M[x_i, x_p]
self.T = T
