using fuel converter efficiency map in fuel economy calcs

[bharatwrrr]

Is the fc_eff_map (fuel converter efficiency map) property not used in simdrive.py while calculating fuel economy? The code seems to use the constant kwh_per_gge parameter for conversion. Where else is the fc_eff_map property used?

I am trying to tune a few vehicle models. The glider mass and frontal area are known parameters. I wanted to make a list of other vehicle parameters to tune and was considering fc_eff_map as well.

[kylecarow]

@bharatwrrr The fc_eff_map (fuel converter efficiency map) property is used during the simulation loop to calculate the engine's input power (which is equal to the fuel storage output power)

So, this code:

self.mpgge = self.dist_mi.sum() / (self.fs_kwh_out_ach.sum() / self.props.kwh_per_gge)

is affected by the fc_eff_map by way of self.fs_kwh_out_ach.sum(). For example, if the engine is less efficient, the achieved fuel storage output energy will be higher, thus the fuel economy will be worse.

Does that help?

EDIT:
To note about fc_eff_map when tuning, it is essentially the 'y-values' of the efficiency, where the x-values are the optional parameter fc_pwr_out_perc. The values of fc_pwr_out_perc represent fractions of the peak fc power, fc_max_kw. The default values for these for a spark ignition engine (and other fc_eff_types) are defined in parameters.py:

fc_pwr_out_perc	0	0.005	0.015	0.04	0.06	0.10	0.14	0.20	0.40	0.60	0.80	1.00
fc_eff_map_si	0.10	0.12	0.16	0.22	0.28	0.33	0.35	0.36	0.35	0.34	0.32	0.30

[bharatwrrr]

@kylecarow Yep, I got it. Thanks!
I figured had to recompute attributes like fc_eff_array and fc_kw_out_array while tuning because just changing fc_eff_map did not seem to change fc_eff_array which is ultimately used to get the point efficiencies in simdrive.

I have another question. It seems that the mc_eff_map does not work similarly to fc_eff_map for BEVs. Tuning fc_eff_map works well for 'Conv' vehicles. Is there an analogous attribute that I can tune for BEVs?

I was also trying to output the mpgge values for BEVs but hit a wall. Using the following code for BEVs, the mpgge value that I get is 0:

veh = vehicle.Vehicle.from_vehdb(20) # 2016 Nissan Leaf 30 kWh

cyc = cycle.Cycle.from_file('udds')    
sim_drive = simdrive.SimDrive(cyc, veh)
sim_drive.sim_drive()

sim_drive_post = simdrive.SimDrivePost(sim_drive)
output = sim_drive_post.get_diagnostics()

output

Is there a functionality in place to compute the mpgge for BEVs? If not, I was planning on computing it using the attributes set in set_post_scalars().

[kylecarow]

@bharatwrrr Ah, I see. I suppose the thing I neglected to mention is if you are modifying these values at runtime (after loading a vehicle), you should re-run the vehicle's set_derived method, which will recalculate those arrays for you.

mc_eff_map is indeed the analog for motor efficiency (and mc_pwr_out_perc is the analog to fc_pwr_out_perc), what issue are you seeing there? I feel like calling set_derived as mentioned above would help with that also.

For BEVs the MPG value is hardcoded to output 0. You could calculate equivalent MPG for BEVs by doing something like sd.dist_mi.sum() / (sd.ess_dischg_kj / 3.6e3 / sd.props.kwh_per_gge). You may also want to consider whether to include charger losses in your result, depending on your application.

[bharatwrrr]

@kylecarow Gotcha.
Yep, something similar to the formula you suggested is what i had in mind. Thanks for the clarification! I’d like to contribute functionality to compute MPGGE for BEVs directly in FASTSim, as it seems like a useful feature. Would you be open to reviewing a pull request for this? If there was a rationale of not including this functionality, then I'll leave it as is. I can start by building on the formula you mentioned. Let me know your thoughts.

[kylecarow]

@bharatwrrr If you'd like to make a PR for the fastsim-2 branch that changes the mpgge computation to something like the above we would be happy to review it! That would be something I would want to port to the fastsim-1 branch as well (the pure python version of FASTSim) for consistency.

As an aside, we are working on a pure Rust rewrite of FASTSim in the fastsim-3 branch. This new version will be introducing a greater degree of modularity and flexibility, and moving away from legacy decisions that persist from porting from Excel etc. In FASTSim 3 we will output everything in terms of energy (and energy per distance), rather than presuppose an energy density for fuel.