The program presented here is a simulator for polishing/figuring strokes on a mirror.
I plan to use it to control the figuring process of my next 18" mirror. It might also be useful for others to evaluate different stroke patterns.
The prototype can be downloaded or executed from here (1.3 MByte download):
As a next step I plan to simulate many different strokes changing the parameters:
- Turntable Speed
- Strokes per Minute
- Stroke Length
- Stroke Offset
The diameter of my subdiameter tool is always 6".
Each simulation run calculates how much glass will be removed in each zone (only relative to the other zones).
These datasets will be saved in a spreadsheet and later used as a 'toolbox' of figuring strokes.
My new polishing machine (turntable + linear stroke) has a steppper motor to change the stroke offset during operation.
So figuring sessions are possible combined of strokes with different offsets.
The other parameters
- Turntable Speed
- Strokes per Minute
- Stroke Length
will be constant in a figuring session.
One of my telescope controllers drives the stepper motor getting positioning commands from a spreadsheed on a PC.
This spreadsheet runs in a loop, changing the stroke offset in a predictable way, for example:
Offset [mm]
Working Time [seconds]
0 - 10
2
10 - 20
3
20 - 30
1
30 - 40
1.5
40 - 50
1
50 - 60
0.5
The goal of a figuring session is to eliminate measured surface errors.
This leads to the problem how these strokes with different offsets must be combined optimally.
Fortunately the combination of strokes and their glass removal per zone is a linear problem and can be solved very fast on a PC.
Finally the complete tool chain could be like this:
- Simulated set of figuring strokes with their glass removal per zone
- Foucault measurements giving positive surface errors per zone
- Linear equation system to reduce the surface errors with a combination of strokes
- Automatic machine working each stroke offset for the calculated time