import numpy as np from numpy.testing import assert_allclose import openpnm as op import openpnm.models.physics as pm class GenericSourceTermTest: def setup_class(self): self.net = op.network.Cubic(shape=[5, 5, 5]) self.phase = op.phase.Phase(network=self.net) self.phase['throat.diffusive_conductance'] = 5e-8 self.phase['pore.mole_fraction'] = 0.0 self.BC_pores = np.arange(20, 30) self.source_pores = np.arange(55, 85) def test_default_values_should_give_zero_rate(self): sources = ["linear", "power_law", "exponential", "natural_exponential", "logarithm", "natural_logarithm"] self.alg = op.algorithms.ReactiveTransport(network=self.net, phase=self.phase) self.alg.settings._update({'conductance': 'throat.diffusive_conductance', 'quantity': 'pore.mole_fraction'}) self.alg.set_value_BC(values=0.4, pores=self.BC_pores) self.alg.set_source(propname='pore.source_term', pores=self.source_pores) # To avoid nans in logarithm-based source term models self.phase['pore.mole_fraction'] = 0.1 for source in sources: self.phase.add_model(propname="pore.source_term", model=getattr(pm.source_terms, source), X="pore.mole_fraction") assert self.phase["pore.source_term.rate"].mean() == 0 assert self.phase["pore.source_term.S1"].mean() == 0 assert self.phase["pore.source_term.S2"].mean() == 0 def test_linear(self): self.phase['pore.item1'] = 0.5e-11 self.phase['pore.item2'] = 1.5e-12 self.phase.add_model(propname='pore.source1', model=pm.source_terms.linear, A1='pore.item1', A2='pore.item2', X='pore.mole_fraction', regen_mode='normal') self.alg = op.algorithms.ReactiveTransport(network=self.net, phase=self.phase) self.alg.settings._update({'conductance': 'throat.diffusive_conductance', 'quantity': 'pore.mole_fraction'}) self.alg.set_value_BC(values=0.4, pores=self.BC_pores) self.alg.set_source(propname='pore.source1', pores=self.source_pores) self.alg.run() self.phase.update(self.alg.soln) self.phase.regenerate_models(propnames='pore.source1') X = self.phase['pore.mole_fraction'] r1 = np.round(np.sum(0.5e-11 * X[self.source_pores] + 1.5e-12), 20) r2 = np.round(np.sum(self.phase['pore.source1.rate'][self.source_pores]), 20) assert r1 == r2 def test_power_law(self): self.phase['pore.item1'] = 0.5e-12 self.phase['pore.item2'] = 2.5 self.phase['pore.item3'] = -1.4e-11 self.phase.add_model(propname='pore.source1', model=pm.source_terms.power_law, A1='pore.item1', A2='pore.item2', A3='pore.item3', X='pore.mole_fraction', regen_mode='normal') self.alg = op.algorithms.ReactiveTransport(network=self.net, phase=self.phase) self.alg.set_value_BC(values=0.4, pores=self.BC_pores) self.alg.set_source(propname='pore.source1', pores=self.source_pores) self.alg.settings._update({'conductance': 'throat.diffusive_conductance', 'quantity': 'pore.mole_fraction'}) self.alg.run() self.phase.update(self.alg.soln) self.phase.regenerate_models(propnames='pore.source1') X = self.phase['pore.mole_fraction'] r1 = np.sum(0.5e-12 * X[self.source_pores]**2.5 - 1.4e-11) r2 = np.sum(self.phase['pore.source1.rate'][self.source_pores]) assert r1 == r2 def test_exponential(self): self.phase['pore.item1'] = 0.8e-11 self.phase['pore.item2'] = 3 self.phase['pore.item3'] = 0.5 self.phase['pore.item4'] = 2 self.phase['pore.item5'] = -0.34 self.phase['pore.item6'] = 2e-14 self.phase.add_model(propname='pore.source1', model=pm.source_terms.exponential, A1='pore.item1', A2='pore.item2', A3='pore.item3', A4='pore.item4', A5='pore.item5', A6='pore.item6', X='pore.mole_fraction', regen_mode='normal') self.alg = op.algorithms.ReactiveTransport(network=self.net, phase=self.phase) self.alg.set_value_BC(values=0.4, pores=self.BC_pores) self.alg.set_source(propname='pore.source1', pores=self.source_pores) self.alg.settings._update({'conductance': 'throat.diffusive_conductance', 'quantity': 'pore.mole_fraction'}) self.alg.run() self.phase.update(self.alg.soln) self.phase.regenerate_models(propnames='pore.source1') X = self.phase['pore.mole_fraction'] r1 = np.sum(0.8e-11 * 3 ** (0.5 * X[self.source_pores]**2 - 0.34) + 2e-14) r2 = np.sum(self.phase['pore.source1.rate'][self.source_pores]) assert_allclose(actual=r2, desired=r1, rtol=1e-7) def test_natural_exponential(self): self.phase['pore.item1'] = 0.8e-11 self.phase['pore.item2'] = 0.5 self.phase['pore.item3'] = 2 self.phase['pore.item4'] = -0.34 self.phase['pore.item5'] = 2e-14 self.phase.add_model(propname='pore.source1', model=pm.source_terms.natural_exponential, A1='pore.item1', A2='pore.item2', A3='pore.item3', A4='pore.item4', A5='pore.item5', X='pore.mole_fraction', regen_mode='normal') self.alg = op.algorithms.ReactiveTransport(network=self.net, phase=self.phase) self.alg.set_value_BC(values=0.4, pores=self.BC_pores) self.alg.set_source(propname='pore.source1', pores=self.source_pores) self.alg.settings._update({'conductance': 'throat.diffusive_conductance', 'quantity': 'pore.mole_fraction'}) self.alg.run() self.phase.update(self.alg.soln) self.phase.regenerate_models(propnames='pore.source1') X = self.phase['pore.mole_fraction'] r1 = np.sum(0.8e-11 * np.exp(0.5 * X[self.source_pores]**2 - 0.34) + 2e-14) r2 = np.sum(self.phase['pore.source1.rate'][self.source_pores]) # assert_allclose(actual=r2, desired=r1) def test_logarithm(self): self.phase['pore.item1'] = 0.16e-13 self.phase['pore.item2'] = 10 self.phase['pore.item3'] = 4 self.phase['pore.item4'] = 1.4 self.phase['pore.item5'] = 0.133 self.phase['pore.item6'] = -5.1e-13 self.phase.add_model(propname='pore.source1', model=pm.source_terms.logarithm, A1='pore.item1', A2='pore.item2', A3='pore.item3', A4='pore.item4', A5='pore.item5', A6='pore.item6', X='pore.mole_fraction', regen_mode='normal') self.alg = op.algorithms.ReactiveTransport(network=self.net, phase=self.phase) self.alg.set_value_BC(values=0.4, pores=self.BC_pores) self.alg.set_source(propname='pore.source1', pores=self.source_pores) self.alg.settings._update({'conductance': 'throat.diffusive_conductance', 'quantity': 'pore.mole_fraction'}) self.alg.run() self.phase.update(self.alg.soln) self.phase.regenerate_models(propnames='pore.source1') X = self.phase['pore.mole_fraction'] r1 = np.sum(0.16e-13 * np.log(4*X[self.source_pores]**(1.4) + 0.133) / np.log(10) - 5.1e-13) r2 = np.sum(self.phase['pore.source1.rate'][self.source_pores]) assert_allclose(actual=r2, desired=r1) def test_natural_logarithm(self): self.phase['pore.item1'] = 0.16e-14 self.phase['pore.item2'] = 4 self.phase['pore.item3'] = 1.4 self.phase['pore.item4'] = 0.133 self.phase['pore.item5'] = -5.1e-14 self.phase.add_model(propname='pore.source1', model=pm.source_terms.natural_logarithm, A1='pore.item1', A2='pore.item2', A3='pore.item3', A4='pore.item4', A5='pore.item5', X='pore.mole_fraction', regen_mode='on_demand') self.alg = op.algorithms.ReactiveTransport(network=self.net, phase=self.phase) self.alg.set_value_BC(values=0.4, pores=self.BC_pores) self.alg.set_source(propname='pore.source1', pores=self.source_pores) self.alg.settings._update({'conductance': 'throat.diffusive_conductance', 'quantity': 'pore.mole_fraction'}) self.alg.run() self.phase.update(self.alg.soln) self.phase.regenerate_models(propnames='pore.source1') X = self.phase['pore.mole_fraction'] r1 = np.sum(0.16e-14*np.log(4*X[self.source_pores]**1.4 + 0.133) - 5.1e-14) r2 = np.sum(self.phase['pore.source1.rate'][self.source_pores]) # assert_allclose(actual=r2, desired=r1) def test_general_symbolic(self): # natural log function y = "a*ln(b*x**c + d)+e" self.phase['pore.item1'] = 0.16e-14 self.phase['pore.item2'] = 4 self.phase['pore.item3'] = 1.4 self.phase['pore.item4'] = 0.133 self.phase['pore.item5'] = -5.1e-14 self.phase.add_model(propname='pore.source1', model=pm.source_terms.natural_logarithm, A1='pore.item1', A2='pore.item2', A3='pore.item3', A4='pore.item4', A5='pore.item5', X='pore.mole_fraction', regen_mode='normal') arg_map = {'a': 'pore.item1', 'b': 'pore.item2', 'c': 'pore.item3', 'd': 'pore.item4', 'e': 'pore.item5'} self.phase.add_model(propname='pore.general', model=op.models.physics.source_terms.general_symbolic, eqn=y, x='pore.mole_fraction', **arg_map) assert np.allclose(self.phase['pore.source1.rate'], self.phase['pore.general.rate']) assert np.allclose(self.phase['pore.source1.S1'], self.phase['pore.general.S1']) assert np.allclose(self.phase['pore.source1.S2'], self.phase['pore.general.S2']) def test_butler_volmer_kinetics(self): np.random.seed(10) self.net["pore.reaction_area"] = np.random.rand(self.net.Np) self.phase['pore.electrolyte_voltage'] = np.random.rand(self.net.Np) * 0.2 self.phase['pore.solid_voltage'] = 1.1 self.phase['pore.open_circuit_voltage'] = 1.2 self.phase['pore.electrolyte_concentration'] = np.random.rand(self.net.Np) BV_params = { "n": 4, "i0_ref": 1e-3, "c_ref": 1000, "beta": 0.5, "gamma": 1 } self.phase.add_model(propname='pore.rxn_BV_c', model=pm.source_terms.butler_volmer_conc, X="pore.electrolyte_concentration", **BV_params) self.phase.add_model(propname='pore.rxn_BV_v', model=pm.source_terms.butler_volmer_voltage, X="pore.electrolyte_voltage", **BV_params) # Check Butler-Volmer model (concentration) S1_BV_c = self.phase["pore.rxn_BV_c.S1"] S2_BV_c = self.phase["pore.rxn_BV_c.S2"] rate_BV_c = self.phase["pore.rxn_BV_c.rate"] eps = np.finfo(float).eps * np.abs(rate_BV_c).max() assert_allclose(S2_BV_c.mean(), 0, atol=eps) assert_allclose(S1_BV_c.mean(), -0.0008190647392381356) assert_allclose(rate_BV_c.mean(), -0.00040556020470030826) # Check Butler-Volmer model (voltage) S1_BV_v = self.phase["pore.rxn_BV_v.S1"] S2_BV_v = self.phase["pore.rxn_BV_v.S2"] rate_BV_v = self.phase["pore.rxn_BV_v.rate"] assert_allclose(S2_BV_v.mean(), 2104.1093852527683) assert_allclose(S1_BV_v.mean(), -12190.386173792602) assert_allclose(rate_BV_v.mean(), -156.52244418065774) # The two Butler-Volmer models must only differ by n*F (unit conversion) assert_allclose(rate_BV_v, rate_BV_c * BV_params["n"] * 96485.33212) if __name__ == '__main__': t = GenericSourceTermTest() self = t t.setup_class() for item in t.__dir__(): if item.startswith('test'): print(f"Running test: {item}") t.__getattribute__(item)()