Vacuum Polishing



   Vacuum polishing is an attempt to overcome the problem of lap pressure  on the mirror  depressing the surface between the webs and causing the webs to imprint through as depressions in the surface once the lap is removed.  It is basically vacuum bagging the polishing process.  It has three requirements: 1- a suitable vacuum source, 2 - a means of getting the vacuum between the lap and glass and yet allowing the lap to turn, and 3 - confining the vacuum to  much of the area between the lap and glass.  A compressor from and old refrigerator seems to be ideal since the air flow needed is not large and although relatively low vacuum (approx.  .5 psi) is all that is required, it can achieve nearly full vacuum (14.7 psi.) if desired!   1/4" vinyl tubing connects the vacuum pump to the poker pin which is a ¼" pipe nipple.
  Holes were drilled in the sides of the nipple and two 1/8" copper tubes soldered on, one leading to a small bleed valve and the other to a vacuum gauge.  The female part of a air hose connector is on the bottom.  I  disassembled  the connector and removed the stopper valve from the interior.  Lots of Vaseline and the connector seals and rotates nicely.  To make sure the lap could tilt freely, a high tech flex plate, a salsa lid, was attached to the male end of the connector.  The lap itself is three discs of MDF, the top two having circles cut out with a circle cutter to the size of the salsa lid.  Using the drill press and improvised routing bits,( a bolt head filed with teeth and the rounded end of a drill bit}, groves were made in the sides of the hole and the bottom to take o-rings (1/8" vinyl tubing cut to length) which seal for vacuum and allow the weight of the lap to be lifted by the poker pin.  A small hole through the center of the lap allows the vacuum to reach the channels in the lap.  An adjustable cord and bungee cord transmit all of the weight of the lap and over arm to a hook in the ceiling





  The lap was channeled  as shown so that the vacuum extends out to about an inch from the edge and there a complete circle of pitch makes a seal with the mirror to retain the vacuum.  (Remember, Mel is telling us that we don't need to channel our laps so I've probably over done it here.).  The lap can over hang the mirror by an inch before the vacuum is broken.  The initial contouring  and micro faceting with bridal vale  was done with hot pressing and weights.  In operation, the cord is adjusted till the lap  is suspended just above the mirror,  the bleed valve is opened fully, the turn table and vacuum pump started, a slight downward push to mate the  lap and mirror, and the vacuum starts to build.  And darned if the contraption doesn't work!!!  Closing the bleed valve  increases the pressure or I should say vacuum between lap and mirror and can quickly stall the turn table.  In fact I have had to install a larger bleed valve to keep the vacuum  in the .5 psi range.  With the turn table stopped, the bleed valve can be fully  closed, and  I've had the vacuum reach 13 psi which I calculate is equivalent to cold pressing with over 500 lb.. on the poker pin!! and no shattering experiences.
   The first thing that is observed to be different than normal polishing is that the mirror emerges from under the lap in an almost dry state, so the natural reaction is to keep adding slurry ahead of the lap.  However when the lap is lifted off the mirror a puddle of slurry is left on the mirror so it may not be necessary to add more all the time after all.  Moreover, the excess slurry under the lap gets entrained up the tube to the vacuum pump;  not good!, so I've had to add a small glass jar in the line before the pump to catch this excess.
   So now you ask, "have you solved the problem?"; and I have to answer, "I still don't know for sure".  I got distracted at this point by summer and Bath interferometry, and only recently got back to the polishing project and some more revisions.  Ideally, one would like to make the circular channel almost out to the edge of the lap to get uniform vacuum over the entire lap surface, or in other words, the vacuum inside the channel doesn't have to provide the polishing pressure for a large area outside the channel.  The trade off is that only a small amount of overlap with the mirror would be permitted before the vacuum seal would be broken, and since the lap turns in the same direction as the mirror, and there is relatively small velocity difference at the edge, so the edge never gets polished.  The answer seemed to be to add a "power bar" to the over arm to drive the lap in the opposite rotation!  So here is the latest revision.  A  210 rpm gear head motor belt drives the lap at about 15 rpm in the opposite rotation to the mirror.  The  center MDF disc of the lap was originally made a 1/2" smaller in diameter than the top and bottom discs so the lap was already a large pulley. The idler pulley on the spring loaded  arm keeps tension on the belt and now the belt takes most of the side loading off the poker pin.  Note, the vacuum gauge is now on the top of  the poker pin, and the vacuum tube is now opposite a larger bleed valve.    And it all works!!!!  ....   at least for 15 minutes so far.

       Interestingly, now the leading edge of the lap is moving in the opposite direction as it turns and slurry dropped ahead of the lap is moved towards the center of the mirror rather than being shoved overboard. In fact, almost no slurry is wasted over the edge.  Remember that jar that traps slurry before the vacuum pump?  It is on the shelf with the vacuum pump about three feet above the mirror.  Now  it takes 1 psi vacuum to raise water 2 feet, so if the vacuum is kept less than 1.5 psi, slurry should run down a tube, (once the tube is full) from the jar to the mirror surface.  I've drilled a hole in the bottom of the jar and a small tube and valve from an intravenous bag will hopefully recirculate slurry to the mirror.  The plan is to put say 1/2 a cup of slurry in the jar and let the tube fill down to the mirror before starting the vacuum.  Bleed air passing through  the jar should keep the slurry stirred up.  We'll see if it works.  {Later} And it does work!!!  Initially I didn't have enough slurry in the jar and the air passing through frothed the slurry and  when enough of this froth entered the tube the siphoning action stopped and air started going up the tube.  Adding more slurry  to the jar overcame this problem and polishing went along very nicely on it's own, slurry continuously dripping on the mirror ahead of the lap.
         In the realm of speculation, I've been wondering, since the pressure for polishing is being provided by the atmosphere more or less evenly over the entire lap, could the lap backing be made  semi flexible, say from a 1/2" Plexiglas or 1/8" aluminum disc.   We make our lap backing solid as a rock mostly to spread weight from our hands, or poker pin loading, or weights on the back of the lap itself more or less evenly around, then ask the pitch to flow in and out of the changing curve of the parabola.  And the pros  flex their laps, all be it, with computer control.  On the other hand, looking at Mel's pictures of over size laps polishing large thin fast mirrors, it probably isn't needed.  Just a thought;  maybe another experiment.
       The next interesting development.  In my zeal to try out the powered lap, I gave the lap a couple of hot presses to micro facet and then started polishing, totally forgetting to give things a cold bath.  After 15 minutes of polishing, to my disbelief, I had totally reversed the situation on the mirror face.  Where there had been rather broad valleys over the ribs, there are now sharp ridges!  More theorizing says that  heat was being conducted away by the ribs thus shrinking the glass above and when everything came to equilibrium later, this glass was expanded relative to the glass in between the ribs; a very dramatic demonstration how important temperature is in the figuring process!  I'm thinking that the next step may be to make the turn table into a shallow pan and fill it with water up to the under side of the top plate glass to conduct the heat of polishing away.  Over all though, I'm moderately encouraged that some, if not all of these experiments may be useful not only in figuring thin mirrors, but in polishing out as well.

Almost a year later!  A few changes and a few sessions of polishing  in the mean time.  The bungee cord has been eliminated and a counter weight that is easily adjustable  substituted, (thanks to a picture of one of Carl Zambuto's machines).  A six foot water manometer has been teed into the vacuum line for a more accurate reading up to 3 psi of vacuum.  Smaller and thinner laps have been tried.  One session with the mirror immersed in water has been tried.   This seemed to be working as the print through is barely perceptible.  It also allowed me to warm the mirror up uniformly to about 80º F since my shop is rather on the cool side.  Unfortunately,  I think the water allowed the mirror to slip or turn slightly on the closed cell foam pad and a tapered end of a rib must have wedged up on one of the retaining blocks and the resultant hole that was polished in in this vicinity is amazing!  I may have to go back to fine grinding to remove it.  I also found that there is just enough play in the air hose coupling that a flex plate may not be needed, however the quill must  be very near vertical or the lap would  pop off the mirror surface at one end of the stroke. 



Vacuum polishing  -  Mar 2010

                                                                         New
                                                             Lead counterbalance -  at rear
                                                             Manometer  -  at rear
                                                             Water tray
                                                             Star lap
                                                             Thinner lap with no flex plate or ball joint
                                                             All brass quill


back to home page

Last revised - 17 Mar 2010
.