I refer to the following Youtube video of Michael Davis which I have seen with interest and attention, also thanks to the “lock-down” impediments, fortunately so far in good health, successive in these times of crown virus, also considering that the English language dialogues of the films, they constitute an obstacle in Italy for all willing and possibly interested scratch-off glass ATM, who are not familiar with that language.
The movie of 14 minutes (to see maybe after reading this writing) refers to TEST OF PRODUCTION OF A MENISCO BLANK FOR TELESCOPE 406F3.5, following two other tests performed positively for diameters 250 e 300mm.
This is a very good American job in which Michael Davis starts with the concave CNC milling of an insulating sheet of extruded polystyrene "Owens Corning Foamular 150" thick 25mm, with a "Home made" milling machine called "Woodpile" (pile of wood), to make the mold in which to cast the spherical convex core in refractory gypsum, on which the collapse will occur in the oven (slumping) glass, by softening a normal calcium-sodium glass disc, diameter 406mm thick 19mm.
The spherical meniscus made will be the blank whose concave surface is refined and parabolized, it will become a possible parabolic mirror of Newtonian telescope 406F3,75.
A considerable amateur difficulty is given by the use of a numerically controlled milling machine, even if it is also self-built. Potentially overcome by following the simplest centrifugal casting method, that Davis himself has tried in another interesting work of his, also told it in another text here in Grattavetro at the following link: https://www.grattavetro.it/colata-centrifuga-spin-casting-di-stampo-epoxy-per-anima-convessa-in-refrattario-di-collasso-slumping-per-blank-a-menisco/.
THE MOVIE
starts with the milling machine digging from the center, a sequence of concentric circles of increasing radius, each of them, however, with decreasing depth from the initial maximum.
The milled surface then becomes a spherical concave cap with a diameter of 3048mm, corresponding to the curvature of the initial spherical cap impressed in the hot collapsed glass meniscus.
Depth that will differ from center to edge of 7.25mm of the so-called desired arrow.
The milling machine completes the excavation, with a groove on the circumference of the concave surface, which will be the seat of the containment "dam" of the convex casting of refractory gypsum.
The following video shows the total thickness of the refractory gypsum in 13mm. Therefore, taking away from them the depth of the arrow 7.25mm, we can deduce that a “resistant thickness” of the plaster remains, equal to only 5.75mm.. very subtle but which is described by Michael as a precautionary compromise between two important needs:
- to avoid breaking the thicker plaster, occurred in previous works on glass diameter 250 me 300mm, being heated in the oven;
- to ensure sturdiness to the plaster such that it does not break under its own weight when handling it.
SOME "AUTOPTIC" NOTES ABOUT THIS CNC MACHINING:
At the minute 1,10 of the movie, the video of the software control lap-top is framed Mach3 CNC of the Artsoft.
On it, in the center of the window at the top left (which shows the scrolling sequence of the machining G code commands), it reads on the central line with a light background, the command in progress G3 milling of one of the anticlockwise circumferences; while in the preceding and following lines, the sequences of cyclic execution of the three commands are indicated G1 in vertical axis Z (depth axis); G1 in horizontal axis X (radius axis with Y = 0); e G3 (counterclockwise circumference interpolation command), previous and subsequent processing.
That depth on the axis Z has negative values, that is, it cyclically decreases at each milled circumference with "steps to rise" to di (5.500-5.625)= 0.06mm (as can be seen from the difference between two consecutive steps G1 on the coordinate Z).
This decrease in axis Z it makes us understand that the concave excavation must have started in the center of the mold, at its maximum depth and minimum diameter (equally corresponding to the complementary maximum convexity of the collapsing core that will be cast in it in plaster).
ADVANCES OF THE TOOL
The radius of the excavation for concentric circles increases by 0.125mm per revolution (inferred by subtracting the displayed X coordinate values of the G1 commands), gradually widening the milling circle of 2,5 tenths of mm at each turn.
The final surface of the concave spherical shell, it will then present itself with a high "stairway of steps" 0,125 mm and wide 0,25 mm which will then be eliminated by sanding 220 grit and then 400.
In the "Tool information" box you can read the current working time of 54,12 minutes corresponding to the milling position that appears almost to the eye 70% of the mold diameter, but whose working time grows very quickly with the increase of the milling radius, until the final time of 2 hours and 5 minutes required to complete the remainder 30% of work.
HOW DOES THE MILLING OF A CONVEX SPHERICAL SURFACE WORK
Normally, as in the movie, takes place starting from the center of the desired diameter of that surface, corresponding to the plane of the raw material to be milled, digging on it a series of concentric circles increasing in diameter, step by step of tool advancement, from the center to the outer diameter of the surface, each of which circumferences will grow in turn deep, step by step advancement, from the zero of the surface of the raw material, to the final depth reached at the maximum circumference. Depth that will be equal to the size of the arrow that the glass disc must have at the end of the sagging process (slumping) hot.
It will therefore be necessary to choose the length of the advancement steps, reaching an acceptable compromise between the degree of roughness left by milling, which will be as much better as the advancement steps are short in decimal or centesimal sense. But in that case, the number of circumferences to be completed will increase proportionally, and with it the processing time will increase very heavily. However, the advantage achieved would be that manual correction of the surface roughness will no longer be necessary.
Another advantage for less roughness would be offered by the choice of a cylindrical cutter with a diameter of approximately 8mm (as are the 5/16 ”of the movie cutter) but with a round tip.
At the end of the job, such a tool would produce a surface only to be painted with polyurethane spray paint to obtain a smooth and hard surface, only to be waxed with non-stick to facilitate future demoulding from the mold.
The processing continues with sanding to eliminate the steps left by the milling, using a grit abrasive sponge 220, which he rounded the edges, as they created scratches in the delicate polystyrene surface.
Stripes that suggested filling the areas with white wood filler, and then sand grit 400 and finally restore the smoothness of the surface, making it shiny with two coats of hard, quick-drying, transparent polyurethane spray paint, interspersed with grit sanding 400. Polyurethane was chosen because it lacks the solvents capable of dissolving the polystyrene of the milled foam plate.
The containment dam of the plaster casting is then built, inserting a 25mm high draft-proof rubber band in the special circular groove
This is followed by the installation of the polystyrene mold on a well-leveled platform by means of three adjustable screw feet.
And to the minute 6,51 the refractory gypsum of a specific type for glass casting takes place (Ransom & Randolph glass cast 400 mold mix), available in equivalent of another brand also from us, in fine art supplies stores.
From the minute 10,42 follow operations and waiting for a few days , aimed at making the plaster dry well flat and in a natural way (in order to avoid deformations induced by forced hot drying), placing it on a new polystyrene foam base (pink), with a newspaper as absorbent, checking over time the return of any deformations in the transition from wet to dry, by means of a crossbar equipped with a comparator.
Michael then shows off his largest Clarkson pit furnace, from 60 Amps (that a 127 volts do 7 kilowatts of power).
On the bottom of the oven well leveled, the plaster core was deposited with the glass overlaid by 406 x 19mm.
That "thing" that is seen in black that is around the whole, it is a metal plate, placed vertically to form a circle of shielding the radiation directly incident on the near edge of the glass, placed too close to the heating resistances, because in the past works the fusion of the glass at the edge had occurred, casting disastrously under the refractory mold.
Michael later starts slumping after programming in an unspecified manner, the controller cycle of the oven rise time-temperature ramp and the descent from the softening temperature (temperature which technically is read to be around 750 ° for the calcium-sodium glass and 820 ° for the boron-silicate).
At the minute 13,35 then presents the meniscus produced, which he notes is beautiful in appearance and better than previous experiments, but in fact with a good curvature only up to a diameter of 365mm (14″). Because in the extreme 41mm there is missing 5.2mm of curvature at the edge.
So the blank is a good 14 ″, but being a diameter 406, in fact it remained a little more open at F3.75 instead of the desired F3.5.
At the minute 14,33, he says that after checks have been made, a slight astigmatism is present, found by rotating the bar with comparator, but that could be eradicated with a little rough abrasion work, followed by the fine-grained one.