confusion about IDAKLUSolver

[xyyrly]

Hi pybamm team,

I am using the IDAKLUSolver in pybamm, for example, to solve a standard DFN model for just one time step. But why the time dimension of the solution is not 2? Should I take the last value as the solution for the new step?

The following is the simulation and the terminal voltage result. Thank you!

delta_t=20
Nodes=10
M=Nodes-1
model1= pybamm.lithium_ion.DFN()
geometry = model1.default_geometry
param = pybamm.ParameterValues("OKane2022")
param.process_model(model1)
param.process_geometry(geometry)
submesh_types = {
"negative electrode": pybamm.Uniform1DSubMesh,
"separator": pybamm.Uniform1DSubMesh,
"positive electrode": pybamm.Uniform1DSubMesh,
"negative particle": pybamm.Uniform1DSubMesh,
"positive particle": pybamm.Uniform1DSubMesh,
"current collector": pybamm.SubMesh0D,
}
var = pybamm.standard_spatial_vars
var_pts = {var.x_n: Nodes, var.x_s:Nodes, var.x_p: Nodes, var.r_n:M+1 , var.r_p: M+1}
mesh = pybamm.Mesh(geometry, submesh_types, var_pts)
disc = pybamm.Discretisation(mesh, model1.default_spatial_methods)
disc.process_model(model1);
#solver = model1.default_solver
t_eval = np.linspace(0, delta_t, 2)
solver = pybamm.IDAKLUSolver()
solution = solver.solve(model1, t_eval)

[github-actions[bot]]

Hi! We have now moved our discussions to Discourse. Please post your question there.

[MarcBerliner]

Hi @xyyrly, the new IDAKLUSolver stores all the adaptive time steps in addition to those specified in t_eval. You can see the full set of time steps in solution.t.

To only store the first and final points, you can run

 solution = solver.solve(model1, t_eval, t_interp=t_eval)

[xyyrly]

Thank you so much! It helps a lot!@MarcBerliner