We call this as "no-refinement" simulation.
This "no-refinement" simulation used 458.0 secs in computer time.
We once again run three simulations, respectively, with adaptive, uniform mesh refinement and no-refinement.
The action potential from the "no-refinement" simulation has least accurate results.
Potential oscillation is even observed at point "7" (see Figure 6) in the "no-refinement" simulation.
It is observed that, in each of the adaptive, uniform refinement and no-refinement simulations, the action potential successfully propagates and goes through the whole domain.
The adaptive simulation roughly uses the same computer time as the "no-refinement" one but yields much more accurate results and gains about 15.5 times speedup over the uniform one.
The action potential successfully propagates across the whole computational domain in each of the adaptive and uniform refinement simulations while it fails to pass across branching points where the fiber radius changes from small to large in the "no-refinement" simulation (see Figure 7).
No refinement of crystallite size by the addition of Si was observed in the present series of Ti Si N films.
