It is a valid model to explain why an object inherits two times the momentum of an incoming photon that leaves the object in the opposite direction.
I was not striving to modell every aspect of reflections. In fact, I was only going for the aspect relevant to our discussion. I choose a modell suited to adequately explain the momentum transfer and the amount of transfered momentum.
But that's my point - it is not a valid (alternative) model to discribe that.
Absorbing/re-emitting of a photon is entirely different to reflection, even if it sounds like a valid description, it is not the way it works, nor a valid way to explain it.
You can view momentum as a vector or as an amount. The amount would be the length of the vector. You can also destinguish between absolute and relative amounts.
Yes, that's right - there is just no such thing as a "negative amount".
Again, you are mixing up a couple of things: Vectors, abolute amounts and relative amounts. I do not get any joy from your hair splitting that even contains false hairs. It's more like setting a toupée on fire.
Well, you started with the hairsplitting (no NOTICEABLE recoil), so i guess i'm entitled to split some hairs, too.
This isn't about relative or absolute amounts (movement is always relative anyway, there is no such thing as absolute movement anyway) - if you inverse the direction, that's fine (inverse vector), but you cannot inverse the entire momentum (which is a derived unit from mass and speed)
Read what I wrote. I am talking about the normal vector of the surface at the point of impact. "My dear". It's cosine. For normal vectors see here:
http://mathworld.wolfram.com/NormalVector.html
Vector "normal to" - didn't sound like "normal vector"... yes, if you take that angle it's cosine.
(But i made it clear that i am talking about angle between incoming vector and surface, not "normal to surface" a.k.a. "normal vector" - english is your first language, i take it? Bit unfair, isn't it?
)
If you modell the incoming photon as a vector and calculate the angle between this vector and the normal vector of the surface you might have to take care of the sign. In a head-on collision the vectors would point in opposite directions. I don't care enough to check the math for the sign.
Take care of the sign?
Cosine becomes negative for angles < or > 90 degree.
If the incoming ray has an angle of > or < 90 degree to the normal vector, you are on the other side of the plane.
So you either don't have to care for the sign at all (normal case), or take it as indicator that you are already on the other side of the surface.
It doesn't inverse the incoming/reflected vector though, but the normal vector (inverse normal vector = normal vector on the opposite site of the plane)
However, we digress.
Th reflected photons lose momentum (lose energy), which means that the image we see in (through) a mirror is a tiny bit "miscoloured" - but can it really work that way?
If we trap a single photon in between two mirrors, it would eventually cease to exist, because it loses energy on every "bounce"... it's just, scientist have trapped single photons (in between reflective surfaces), and kept them in the trap for quite a while, so the loss of energy is either very minimal (or percentual), or there is something we didn't pay attention to.