analyst5 said:
So can you please explain the time dilation regarding the moving twin from the stationary twin's reference frame.
IMO, the best way to conceptualize what's going on is to discard the concept of "time dilation" altogether. But before going into that, I'll answer your question as you asked it; see below.
analyst5 said:
I know that during the intertial motion the moving twin's clock is slowed down, but what happens when the moving twin accelerates? Does the stationary twin 'perceive' (and by that I don't mean see, but has a specific segment of the other twin in his reference frame) the increase of time dilation when the moving twin is accelerating to get from the position of rest to the position of motion relative to the stationary twin?
If you are using the concept of "time dilation", then yes, the moving twin's time dilation, as "perceived" by the stationary twin, depends on the moving twin's velocity in the stationary twin's rest frame; the higher the velocity, the more time dilation.
However, there are a number of issues with looking at things this way, which I won't go into in detail, but just summarize as follows: "time dilation", as a concept, does not generalize well, because it's not fundamental; it's a derived concept that works OK for certain scenarios, but that's all. So trying to analyze things using "time dilation" as your fundamental concept doesn't work well.
The fundamental concept is that spacetime is a 4-dimensional geometric object, and different curves in this geometry will have different lengths. The stationary twin follows one curve between the two points where the two twins meet; the moving twin follows another, different curve. Since the curves are different, their lengths are different;, and the length of a timelike curve (i.e., of the worldline of an object with nonzero rest mass, like either twin) is just the proper time experienced by an observer who follows the curve. So different lengths of curves means different proper times experienced.
This concept is completely general; it covers all the different variations on "twin paradox" type scenarios in flat spacetime, and it also generalizes to curved spacetime, when gravity is present (i.e., to general relativity as well as special relativity). Also, once you have a scenario analyzed in terms of spacetime geometry and lengths of curves, you can easily "read off" the usual stuff people talk about with relativity from the analysis: time dilation, length contraction, relativity of simultaneity, etc. You can also easily see the limitations of all those other concepts.
A good presentation of all this with regard to the twin paradox is given in this Usenet Physics FAQ article:
http://math.ucr.edu/home/baez/physics/Relativity/SR/TwinParadox/twin_paradox.html
The "Spacetime Diagram Analysis" is the geometric viewpoint I have described above; the FAQ article also shows how this analysis serves as a common framework for deriving all the other concepts.