In this second of two parts, the fundamentals of convective wall heat transfer losses are elucidated in the context of the desired objective toward its reduction in a direct-injected, hydrogen-fueled internal combustion engine. A comparative, transient 2D CFD analysis evaluated at 4500 RPM between a combustion chamber design representing current practice and the here-introduced “vortex-stratified combustion” process finds an approximately 50% reduction in the convective flux with the latter. The simulation results show that reduced heat flux of the vortex approach is driven by the combination of two effects. The first is finite-time diffusive mixing getting outpaced by the replenishment of pure air being introduced preferentially along the circumference of the combustion chamber due to the Coandă effect; this results in a distinct radial charge stratification during mixture preparation in the compression stroke, with a fuel-concentrated center and essentially pure air at the periphery. The second effect is the forced-segregation of different density reactants during the course of the combustion process caused by large body forces that result from the gravitational acceleration of the rapidly rotating charge, thereby constraining the combustible mixture and the flame to some distance from the walls. Evidence for this is observed by hot, low-density hydrogen being forced to remain near the center and cooler, heavier oxygen being inhibited from migrating from the outer periphery to react with the aforementioned hydrogen, and the distinct curvature of the radial gas temperature profile at a substantially greater distance from the wall than the thermal boundary layer thickness.