"Max Keon" wrote in message
u...
"George Dishman" wrote in message
ps.com...
On 30 May, 00:38, "Max Keon" wrote:
George Dishman wrote:
"Max Keon" wrote in message
...
I'll put back part you snipped as it is the
core of the disagreement:
For example, if an acceleration of 2m/s^2 acts for
50s then the velocity will change by 100m/s.
....
What I said above is true for all objects changing
speed for any reason whatsoever under any circumstances.
Certainly _not_ under any circumstances George.
Yes, under _any_ circumstances Max. The word
"acceleration" is _defined_ as rate of change
of velocity so in a short time dt the velocity
will change from an initial value v_i to a final
value v_f given by
v_f = v_i + a * dt
where a is the acceleration. v_i, v_f and a are
all vectors and can be handled as x and y
components as I do in the code. (You need z too
in general of course but our orbits are two
dimensional).
....
.. it applies to _all_ motion of _any_ nature
whatsoever. That is fundamental to the whole
of dynamics and the process of integrating
acceleration to find velocity is pure maths.
snip application of above until we get
this agreed.
I don't know how to get it through to George, but you are wrong
and I am right. Go and study it properly.
These may help you:
http://www.staff.amu.edu.pl/~romango...eleration.html
http://hep.physics.indiana.edu/~rick...inematics.html
http://www.blurtit.com/q286865.html
You could also look at the "Force and Motion"
topic he
http://learningcenter.nsta.org/produ...e_objects.aspx
It concentrates mainly on position and velocity
and is a bit light on acceleration but it is
well presented and has lots of questions to let
you check your progress. Once you have gone
through that, look again at what I said at the
top:
For example, if an acceleration of 2m/s^2 acts for
50s then the velocity will change by 100m/s.
Do you still deny that?
It seems to me that you may have a problem understanding what is
actually going on here. A gravity anisotropy is something you've
never encountered before and you are trying to explain it using
reasoning that you are accustomed to. That doesn't work.
I have encountered anisotropy in many areas but
actually that doesn't matter, you have done the
work on that side and all I need is your result.
You have told me that the effect is to change
the acceleration due to gravity from Newton's
equation:
a = -GM / r^2
to
a = -GM / r^2 * (1 + v/c)
where v is the radial speed, the rate of change
of the radius, and the direction is the same
as Newton's.
You have specified the acceleration Max, all
that is left is to integrate it once to get
velocity and a second time to get position and
the results fall out of that calculation. I
don't need to know anything more about your
ideas at all. Now if that equation doesn't
represent your idea then obviously my comments
will similarly be misplaced but that is what
you have provided so I am taking it at face
value.
snip diatribe
Do you now understand the importance of knowing the truth?
Of course, so go and look up the definition of
"acceleration" and you will find I have been
telling you the truth throughout. Your analysis
is wrong for numerous reasons that I listed
before and which you have ignored.
I have only made one significant error during
the course of the discussion which was when I
did a very quick simulation of your equation
and accidentally got the sign wrong, I used
(1 - v/c) instead of (1 + v/c) so the
eccentricity increased instead of decreasing
but that was some time ago and the Basic code
has the correct sign for your theory. It is
available for anyone to peer review [1] and
I'll repeat the part relevant to the physics
here since we have changed thread since it was
posted:
r2 = x * x + y * y
radius = SQR(r2)
vr = (radius - lastradius) / dt
lastradius = radius
Newton = GM / r2 ' The constant GM is negative
anisotropy = Newton * (vr / c)
acceleration = Newton + anisotropy
ax = acceleration * (x / radius)
ay = acceleration * (y / radius)
vx = vx + dt * ax
vy = vy + dt * ay
x = x + dt * (vx + 0.5 * dt * ax)
y = y + dt * (vy + 0.5 * dt * ay)
time = time + dt
George
[1] If anyone actually peer reviews that code,
be aware that in the actual version I run,
the acceleration for the next time step
uses a linear prediction from the current
and previous acceleration values to get:
forward_mean = (3 * current - previous) / 2