Handling simulation outputs
In this tutorial we will explore the outputs of a simulation for interesting tasks:
Plot voltage and current curves
Plot overpotentials
Plot cell states in space and time
Save outputs
Load outputs.
Lets start with loading some pre-defined cell parameters, cycling protocols, and running a simulation.
julia
using BattMo, GLMakie
cell_parameters = load_cell_parameters(; from_default_set = "chen_2020")
cycling_protocol = load_cycling_protocol(; from_default_set = "cc_discharge")
model = LithiumIonBattery()
sim = Simulation(model, cell_parameters, cycling_protocol);
output = solve(sim);✔️ Validation of ModelSettings passed: No issues found.
──────────────────────────────────────────────────
✔️ Validation of CellParameters passed: No issues found.
──────────────────────────────────────────────────
✔️ Validation of CyclingProtocol passed: No issues found.
──────────────────────────────────────────────────
✔️ Validation of SimulationSettings passed: No issues found.
──────────────────────────────────────────────────
✔️ Validation of SolverSettings passed: No issues found.
──────────────────────────────────────────────────
Jutul: Simulating 2 hours, 12 minutes as 163 report steps
╭────────────────┬───────────┬───────────────┬──────────╮
│ Iteration type │ Avg/step │ Avg/ministep │ Total │
│ │ 146 steps │ 146 ministeps │ (wasted) │
├────────────────┼───────────┼───────────────┼──────────┤
│ Newton │ 2.32192 │ 2.32192 │ 339 (0) │
│ Linearization │ 3.32192 │ 3.32192 │ 485 (0) │
│ Linear solver │ 2.32192 │ 2.32192 │ 339 (0) │
│ Precond apply │ 0.0 │ 0.0 │ 0 (0) │
╰────────────────┴───────────┴───────────────┴──────────╯
╭───────────────┬────────┬────────────┬──────────╮
│ Timing type │ Each │ Relative │ Total │
│ │ ms │ Percentage │ ms │
├───────────────┼────────┼────────────┼──────────┤
│ Properties │ 0.0440 │ 2.40 % │ 14.9223 │
│ Equations │ 0.6422 │ 50.13 % │ 311.4450 │
│ Assembly │ 0.2640 │ 20.61 % │ 128.0189 │
│ Linear solve │ 0.2783 │ 15.18 % │ 94.3281 │
│ Linear setup │ 0.0000 │ 0.00 % │ 0.0000 │
│ Precond apply │ 0.0000 │ 0.00 % │ 0.0000 │
│ Update │ 0.0555 │ 3.03 % │ 18.8062 │
│ Convergence │ 0.0753 │ 5.88 % │ 36.5324 │
│ Input/Output │ 0.0296 │ 0.70 % │ 4.3196 │
│ Other │ 0.0379 │ 2.07 % │ 12.8584 │
├───────────────┼────────┼────────────┼──────────┤
│ Total │ 1.8325 │ 100.00 % │ 621.2308 │
╰───────────────┴────────┴────────────┴──────────╯Now we'll have a look into what the output entail. The ouput is of type SimulationOutput and contains multiple output quantity dicts, the full input dict and some other structures. Lets print the properties of the SimulationOutput.
julia
propertynames(output)(:time_series, :states, :metrics, :input, :jutul_output, :model, :simulation)In terms of simulation results, we can see that the output structure includes time series data, states data and metrics data. Furthermore, it includes the full input dict, some output structure from Jutul, the model instance, and the simulation instance. Let's for now have a look into the simulation results and see how we can access certain output quantities.
In BattMo, we make a distinction between three types of results:
time series: includes all quantities that only depend on time. For example, time itself, cell voltage, current, capacity, etc.
states: includes all the state quantities like for example, concentration, potential, charge, etc. These quantities can depend on time, position, and radius.
metrics: includes all the from output quantities calculated cell metrics like discharge capacity, charge energy, round trip efficiency, etc. These metrics depend on the cycle number.
The have an overview of all the quantities that are available you can run:
julia
print_info(output)
Case: TIME_SERIES
================================================================================================================================================================
Variable Unit Shape
----------------------------------------------------------------------------------------------------------------------------------------------------------------
CumulativeCapacity Ah (nTime,)
Current A (nTime,)
CycleNumber 1 (nTime,)
NetCapacity Ah (nTime,)
Time s (nTime,)
Voltage V (nTime,)
================================================================================================================================================================
Case: METRICS
================================================================================================================================================================
Variable Unit Shape
----------------------------------------------------------------------------------------------------------------------------------------------------------------
ChargeCapacity Ah (nCycleIndex,)
ChargeEnergy J (nCycleIndex,)
CycleIndex 1 (nCycleIndex,)
DischargeCapacity Ah (nCycleIndex,)
DischargeEnergy J (nCycleIndex,)
RoundTripEfficiency % (nCycleIndex,)
================================================================================================================================================================
Case: STATES
================================================================================================================================================================
Variable Unit Shape
----------------------------------------------------------------------------------------------------------------------------------------------------------------
ElectrolyteCharge C (nTime, nPosition)
ElectrolyteConcentration mol·m⁻³ (nTime, nPosition)
ElectrolyteConductivity S·m⁻¹ (nTime, nPosition)
ElectrolyteDiffusivity m²·s (nTime, nPosition)
ElectrolyteMass g (nTime, nPosition)
ElectrolytePotential V (nTime, nPosition)
NegativeElectrodeActiveMaterialCharge C (nTime, nPosition)
NegativeElectrodeActiveMaterialDiffusionCoefficient m²·s⁻¹ (nTime, nPosition)
NegativeElectrodeActiveMaterialOpenCircuitPotential mol·m⁻²·s⁻¹ (nTime, nPosition)
NegativeElectrodeActiveMaterialParticleConcentration mol·L⁻¹ (nTime, nPosition, nNegativeElectrodeActiveMaterialRadius)
NegativeElectrodeActiveMaterialPotential V (nTime, nPosition)
NegativeElectrodeActiveMaterialRadius m (nNegativeElectrodeActiveMaterialRadius,)
NegativeElectrodeActiveMaterialReactionRateConstant V (nTime, nPosition)
NegativeElectrodeActiveMaterialSurfaceConcentration mol·L⁻¹ (nTime, nPosition)
NegativeElectrodeActiveMaterialTemperature T (nTime, nPosition)
Position m (nPosition,)
PositiveElectrodeActiveMaterialCharge C (nTime, nPosition)
PositiveElectrodeActiveMaterialDiffusionCoefficient m²·s⁻¹ (nTime, nPosition)
PositiveElectrodeActiveMaterialOpenCircuitPotential V (nTime, nPosition)
PositiveElectrodeActiveMaterialParticleConcentration mol·L⁻¹ (nTime, nPosition, nPositiveElectrodeActiveMaterialRadius)
PositiveElectrodeActiveMaterialPotential V (nTime, nPosition)
PositiveElectrodeActiveMaterialRadius m (nPositiveElectrodeActiveMaterialRadius,)
PositiveElectrodeActiveMaterialReactionRateConstant mol·m⁻²·s⁻¹ (nTime, nPosition)
PositiveElectrodeActiveMaterialSurfaceConcentration mol·L⁻¹ (nTime, nPosition)
PositiveElectrodeActiveMaterialTemperature T (nTime, nPosition)
================================================================================================================================================================To get more information on particular output variables, for example all that have concentration in their name:
julia
print_info("concentration", view = "OutputVariable")
----------------------------------------------------------------------------------------------------
📈 Output Variable: NegativeElectrodeActiveMaterialSurfaceConcentration
----------------------------------------------------------------------------------------------------
🔹 Name NegativeElectrodeActiveMaterialSurfaceConcentration
🔹 Category OutputVariable
🔹 Description Concentration of lithium ions at the surface of negative electrode particles.
🔹 Type Vector{Real}
🔹 Shape (nTime, nPosition)
🔹 Unit mol·L⁻¹
----------------------------------------------------------------------------------------------------
📈 Output Variable: PositiveElectrodeActiveMaterialParticleConcentration
----------------------------------------------------------------------------------------------------
🔹 Name PositiveElectrodeActiveMaterialParticleConcentration
🔹 Category OutputVariable
🔹 Description Radial distribution of lithium concentration in positive electrode particles.
🔹 Type Vector{Real}
🔹 Shape (nTime, nPosition, nPositiveElectrodeActiveMaterialRadius)
🔹 Unit mol·L⁻¹
----------------------------------------------------------------------------------------------------
📈 Output Variable: ElectrolyteConcentration
----------------------------------------------------------------------------------------------------
🔹 Name ElectrolyteConcentration
🔹 Category OutputVariable
🔹 Description Concentration of lithium ions in the electrolyte.
🔹 Type Vector{Real}
🔹 Shape (nTime, nPosition)
🔹 Unit mol·m⁻³
----------------------------------------------------------------------------------------------------
📈 Output Variable: PositiveElectrodeActiveMaterialSurfaceConcentration
----------------------------------------------------------------------------------------------------
🔹 Name PositiveElectrodeActiveMaterialSurfaceConcentration
🔹 Category OutputVariable
🔹 Description Concentration of lithium ions at the surface of positive electrode particles.
🔹 Type Vector{Real}
🔹 Shape (nTime, nPosition)
🔹 Unit mol·L⁻¹
----------------------------------------------------------------------------------------------------
📈 Output Variable: NegativeElectrodeActiveMaterialParticleConcentration
----------------------------------------------------------------------------------------------------
🔹 Name NegativeElectrodeActiveMaterialParticleConcentration
🔹 Category OutputVariable
🔹 Description Radial distribution of lithium concentration in negative electrode particles.
🔹 Type Vector{Real}
🔹 Shape (nTime, nPosition, nNegativeElectrodeActiveMaterialRadius)
🔹 Unit mol·L⁻¹
========================================================================================================================As the time series, states, and metrics structures are dicts we can retrieve quantities by accessing their key. Let's for example create a simple voltage vs capacity plot.
julia
voltage = output.time_series["Voltage"]
capacity = output.time_series["CumulativeCapacity"]
fig = Figure()
ax = Axis(fig[1, 1], ylabel = "Voltage / V", xlabel = "Capacity / Ah", title = "Discharge curve")
lines!(ax, capacity, voltage)
display(fig)GLMakie.Screen(...)Or lets plot the lithium concentration versus the active material particle radius of the positive electrode close to the separator at the and of the discharge:
julia
radius = output.states["PositiveElectrodeActiveMaterialRadius"]
positive_electrode_concentration = output.states["PositiveElectrodeActiveMaterialParticleConcentration"]
simulation_settings = output.input["SimulationSettings"] # Retrieve the default simulation settings to get the grid point number that we need.
grid_point = simulation_settings["NegativeElectrodeCoatingGridPoints"] + simulation_settings["SeparatorGridPoints"] + 1 # First grid point of the positive electrode
concentration_at_grid_point = positive_electrode_concentration[end, grid_point, :]
fig = Figure()
ax = Axis(fig[1, 1], ylabel = "Lithium concentration / mol·L⁻¹", xlabel = "Particle radius / m", title = "Positive electrode concentration")
lines!(ax, radius, concentration_at_grid_point)
display(fig)GLMakie.Screen(...)Example on GitHub
If you would like to run this example yourself, it can be downloaded from the BattMo.jl GitHub repository as a script.
This page was generated using Literate.jl.