On Tue, 5 Feb 2019 00:03:14 -0500, bitrex wrote:

https://www.jstor.org/stable/44611429?seq=1#page_scan_tab_contents

It's behind a pay-wall but I can probably get my hands on a copy

Having authored peer-reviewed papers myself (in a different field), I took
a quick peek at the abstract of that paper, titled:
o Computer Simulation of an I.C. Engine During Cranking by a Starter Motor
https://www.jstor.org/stable/44611429?seq=1#page_scan_tab_contents

"A mathematical model is developed to study the transient behavior of a two
stroke or four stroke, single cylinder I.C. engine during cranking and
starting by a starter motor. The engine model includes forces due to
inertia of reciprocating and rotating parts of engine, gas pressure,
frictional loss while starter motor dynamics is determined by the motor's
torque versus speed behavior. The numerical results of the analysis when
compared with the experimental results showed close correlation.
Engine starting by three models of starter motor is presented for a given
battery. Effect of different parameters like engine inertia and reduction
ratio between engine and a starter motor is described. It is shown here how
this analysis can be effectively used as a first step by an engine designer
for determining a suitable starter motor characteristic and its related
transmission parameters."

Hmmm... they _might_ cover the charge payback component, but I suspect
likely it will only be an ancillary input to the mathematical model, and
certainly it won't apply _directly_ to an 87 Buick Regal.

We should note that the given "battery" is seemingly incidental in this
paper, which seems to be aimed more toward designing starter motors, and,
specifically between choosing among three different types of fundamental
starter motor designs.

Still, it may be an interesting read, where, I'd be curious how the three
types of starter motors affected the model - but - I hazard a guess that we
won't find a direct answer for our charge component in that paper.