Defining new functions in ParameterValues

[ahsadeghi]

Hello,
I am trying to define a new function for calculating (positive) current collector conductivity as

def positive_current_collector_conductivity(T):   
    #Strum(2019)
    sigma_cc = 3.78e7 / (1 + 4.20e-3 * (T - 293.15))
    return sigma_cc

in ORegan2022.py parameters, and I call the function in the end of the file, return { , as

"Positive current collector conductivity [S.m-1]": positive_current_collector_conductivity,

but I get the following error

TypeError: Cannot process parameter '<function positive_current_collector_conductivity at 0x000001E376B9F310>'

I wonder if you would please help to find the reason. Thank you.

Regards,
Hosein

[ahsadeghi]

Dear @brosaplanella
Do you have any comments regarding this issue?

[valentinsulzer]

This is because the current collector conductivity is a Parameter, not a FunctionParameter
It needs to be change from this to this

This should be a super easy fix if you want to try it

[ganmad]

Hi @tinosulzer,

As per your suggestion, I tried to implement temperature dependent EC diffusivity, by replacing self.D_ec = pybamm.Parameter(f"{pref}EC diffusivity [m2.s-1]")
with

def D_eci(self,T):
                pref = self.phase_prefactor
                return pybamm.FunctionParameter(f"{pref}EC diffusivity [m2.s-1]", {"Temperature [K]": T})

self.D_ec = D_eci

I tried different ways, defining the function outside/inside of the _set_parameters, in lithium_ion_parameters.py, but doesn't work. I defined a temperature dependence function locally as well to update parameters. Am I missing something here? You inputs will be much appreciated. Thanks.

[brosaplanella]

You need to define it as a function of the class ParticleLithiumIonParameters so the self.D_ec = D_eci is not needed anymore (see, for example, how it is done with particle diffusivity:

PyBaMM/pybamm/parameters/lithium_ion_parameters.py

Lines 537 to 560 in f81de94

	def D(self, c_s, T, lithiation=None):
	"""
	Dimensional diffusivity in particle. In the parameter sets this is defined as
	a function of stoichiometry (dimensionless), but in the models we use it as a
	function of concentration (mol/m3). We convert from concentration to
	stoichiometry by dividing by the maximum concentration.
	"""
	Domain = self.domain.capitalize()
	sto = c_s / self.c_max
	tol = pybamm.settings.tolerances["D__c_s"]
	sto = pybamm.maximum(pybamm.minimum(sto, 1 - tol), tol)
	if lithiation is None:
	lithiation = ""
	else:
	lithiation = lithiation + " "
	inputs = {
	f"{self.phase_prefactor}{Domain} particle stoichiometry": sto,
	"Temperature [K]": T,
	}
	return pybamm.FunctionParameter(
	f"{self.phase_prefactor}{Domain} electrode {lithiation}"
	"diffusivity [m2.s-1]",
	inputs,
	)

).

You will also have to pass the temperature as an argument everywhere where this parameter is called, but you can do a search in the code base to help you with this.

Note that #3198 does some changes in this part of the code, so you might want to wait until that one is merged (which should happen this week).

[ganmad]

Hi @brosaplanella,

As suggested, I tried implementing a method for D_ec(T) as below:

def D_ec(self, T):
        """Dimensional EC diffusivity as f(T)""" 
        #main = self.main_param
        #domain, Domain = self.domain_Domain
        #phase_name = self.phase_name
        pref = self.phase_prefactor
        if self.domain == "negative":
            inputs = {"Temperature [K]": T}
            #Domain = self.domain.capitalize()
            return pybamm.FunctionParameter(
                f"{pref}EC diffusivity [m2.s-1]", inputs
           )

Also, I commented the D_ec line at l454

But, when I load the model with the following options,
model = pybamm.lithium_ion.SPM({"cell geometry": "arbitrary","thermal": "lumped","SEI":"ec reaction limited"})

I get this NotImplementedError:

Do I have to do anything after defining the method D_ec(self,T)?

Thank you.

[brosaplanella]

Where did you place this new function? It should be in lithium_ion_parameters.py. But you also will have to replace all occurrences of D_EC with D_EC(T), where T needs to be the right temperature for the domain. If are planning on getting this merged to PyBaMM, you can open a pull request so I can see all the code and give more specific comments.

[ganmad]

Hi @brosaplanella ,

I managed to implement this by replacing D_ec with D_ec(T) at all instances. So it is working fine. Thanks for your suggestion. Regards,

[Brushnie33]

@brosaplanella Hello, I managed to change D_ec to D_ec(T) and get it to work. However, if I follow a similar approach to define inner_sei_proportion to be SOC dependent, I got errors. In lithium_ion_parameters.py, I defined the following equation inside class ParticleLithiumIonParameters(BaseParameters):

def cracking_tuning_factor(self, sto):
    domain, Domain = self.domain_Domain
    inputs = {f"{Domain} particle stoichiometry": sto},
    return pybamm.FunctionParameter(
        f"{self.phase_prefactor}{Domain} electrode cracking reaction tuning factor", inputs
    )

And in the sei_growth.py codes, I added the following codes:
sto = variables[f"{Domain} particle concentration"] # variables[f"{Domain} {self.phase_name}particle concentration"]
inner_sei_proportion= phase_param.cracking_tuning_factor(sto)

I believe the errors are associated with sto, could you please suggest how to modify sto in the above codes?