I don't understand why the flow through the throttling valve in cooling cycles is assumed isenthalpic.

The usual justification is that since no moving parts nor heat transfer, then total energy must be constant, and since kinetic energy is similar for similar input-output cross-sectional areas, enthalpy is approximately the same.

But this is only valid for liquids (constant density), and this is not the case in cooling cycles. Since the liquid is close to saturation, it flashes when throttled, and then density falls greatly with the formation of vapor. Then, by continuity, average speed increases also greatly for equal areas.

With sample calculations for R134a between 9 and 2 bars absolute, it can be seen that density falls from ca. 1000 kg/m3 for HP saturated liquid to 36 kg/m3 for 26.7 % LP vapor quality. Then, average speed increases from 1 to 33 m/s, and 32 m/s squared divided by two is 512 J/kg, much more than the latent enthalpy of evaporation.

Under these conditions, how is it that the difference in kinetic energies can be assumed to be negligible?

Sorry if the question has been posted already somewhere or my analysis have great mistakes; I've been unable to find an answer. Any help will be highly appreciated.