Original+TRO+Purchasing+the+blank+and+hoging+out


 * Purchasing the blank**

In early January 1997 myself and my son David went to 'Specialised glass' of St Helens (after a pre negotiated price in cash of £80 was agreed) to pick up the blank. The blank was surprisingly round and smooth on the edge. Other blanks I have seen with a nipped edge have been very rough and have chunks out of them. This blank did not require any further work to its edge except later on to but on a bevel for the second stage described later.





Above images show my son David getting in the car after purchasing the blank. second shows the blank in the boot carefully packed ready to go home.

Above image shows myself with the blank before any work had started.


 * Setting to work to hog out the blank**

The first task was to create the main spherical curve in the blank. This required the removal of some 130 cubic inches of glass. The traditional method of a glass tool of the same size of the mirror to do the hogging out seemed out of the question due to possible flexure of the thin pieces of glass and also a long-winded process (a second piece would also have to be bought). An idea I had seen back in the early 80's in a Sky and Telescope magazine was adopted. This was to use a radial arm grinder and 6" carberundum wheel. This requires a long pole (scaffold pole or similar) to be fixed by an adjustable ball joint at one end and the carberundum wheel to be attached with a drill at the other. The pole is supported at a point along its length that balances the pole horizontally and is perpendicular to the centre of the mirror. The length of the pole required is to be twice the focal length (i.e. the radius of curvature of the sphere to be ground out). The mirror blank is positioned perpendicular (both in the vertical and horizontal positions) to this and a power drill turns the wheel and cuts into the surface of the glass. The wheel is fed via with a constant steam of water from a lemonade bottle with a small hole cut in the cap feeding over the centre of the disc. The power drill was protected from water splashes and an RCCD protective device used in the supply to the drill. The radial arm is slowly swept across the disk left and right and from top to bottom and is to be moved inward gradually on each vertical pass as the glass is removed. This is achieved by adjusting the ball joint at the far end which is mounted on a threaded rod with locknut to prevent unwanted movement. Vertical movement is achieved by a screw type car jack being fixed to the radial arm at the point of support. A heavy-duty nylon rope the ties to this and the support point (eyebolt) above. The whole assembly just fitted inside my garage. The radial arm being around 6.5 meters long for an F4.32 ratio. A larger mirror would have meant less than F4 and would be out the question. We would have our hands full trying to figure this mirror anyway.

The hogging out process was done in 6 sessions over weekends from January 25 to March 22 and took some 15 hours all told. Geoff Regan myself and Dave Owen all participated in the process, each having their own role to play and all identifying and solving the problems encountered. Geoff Regan did the vertical control of the radial arm by controlled steps of lowering the wheel across the disc. I controlled the sweeping action of the wheel across the disc and Dave Owen added the cooling water and took notes of progress made. Geoff and my father (David Thomson) also did video of the hogging out process. A team effort is an advantage in such a project to ensure a collective discussion on the best solution, also more eyes the better so that things don't get overlooked.



The above images show the adjustable ball joint at the other end of the scaffold pole and the alighnment mirror used to centre the blank to the axis of the pole.

Problems encountered:-


 * 1) Setting up of the blank against the tool face**

Initial setting up of the blank against the tool was done by mounting the blank on an adjustable cell similar to a Newtonian telescope. (This had to be very flat to ensure that no flexture in the blank took place. 40mm worktop was used and checked for flatness in all planes with a straight edge. Irregularities where taken out with filler using the straight edge to spread out the filler evenly. The whole assembly was then well sealed against water. This assembly could then essentially be collimated similar to collimating a Newtonian.

Grinding proceeded and everything seemed all right. The session 1 ground out a 12" diameter sphere. Session 2 then began to show up that the grinding centre was offset to the centre of the blank. This was traced to the fact that the support point for the radial arm was not exactly in line with the centre of the blank. The eyebolt position in the roof was repositioned and grinding proceeded again. It then became obvious that the horizontal position was all right with equal amounts of glass being taken off either side but the vertical position now seemed to be out and was too high. The ball joint mounting point was lowered by a small amount to compensate for the error and the cut then became more even in both horizontal and vertical positions.

Small adjustments to the cell still required to be made the further in the grind went to keep a more concentric cut across the blank with respect to the rim of the blank.


 * 2) Pole flexture**

Variations of depth and unevenness during grinding pointed towards the radial arm pole flexing. This was remedied by fixing a long triangle to the pole out of a piece of 2x3 wood fixed to the main pole using scaffolding clamps to reduce flexure to a minimum over its length. A larger diameter pole would probably work better.


 * 3) Vibration**

When cutting too deep into the disk vibration became a problem and we had to keep backing off. Another source of vibration came from going down the disc too quickly. This would cause ridges in the face of the mirror and the cutting disc would be cutting down into the ridge and bouncing off rather than cutting away from the disk.


 * 4) Tool wear and damage**

When the depth of cut increased it became apparent that the grinding wheel was wearing much faster. This showed itself as an uneven cut across the disc. The sides of the cutting edge would move inwards slowly therefore cutting less out of the glass as the tool face moved back from the cutting face as the tool wore down. We had to keep moving the tool in small amounts by adjusting the ball joint until the tool just touched the surface of the mirror at the point the first cut was taken. This meant that the downward cutting across the mirror had to be taken in steps and revisiting the same area a number of time to ensure that the correct amount of glass had been removed across the face of the mirror for each vertical pass.

Near disaster came at one point when the tool was well worn and vibration became too much. Inspection of the tool revealed that the wheel was fractured and was only just held together. If it had come apart at 2600rpm not only would the mirror be damaged but us seriously also! The wheel was returned and another given in replace at no cost.


 * 5) Bearing wear on the mounting head**

The wheel was attached to a 0.5 inch shaft reduced at one end for the drill to clamp onto. This was mounted onto a heavy channel by two pillow block bearings. The bearings used phosphor bronze bearings for cheapness not ball bearings (only £3.00 each at a local merchant). These had to be cleaned and greased regularly every session and replaced twice over the whole process. The shaft also had a problem in that the drill after a while began to slip on the shaft and wear it down. This caused vibration in the drill and through to the cutting surface. The assembly was removed and re-jigged by removing the reduced section and having the drill clamp onto the 0.5 inch section.


 * 6) Adjusting the depth of cut**

Adjusting the depth of cut was a delicate procedure and care had to be taken to ensure that not too much glass was going to be removed on each cut.


 * 7) Running out of cooling water**

Vibration and raised noise levels could be noticed immediately the water feed ran out. We had to keep topping up the water every minute or so. Ideally at a constant level so that pole flexture was not induced by extra weight of water.

Above image is at the start of the cutting process.

Above image shows the blank after the second cut is half way through.

This image shows the mirror during the grinding process. Note the water slurry.



Video cli[ps of the hogging out process
More video clips will be added soon.

[[file:TRO_005.avi|TRO_005.avi]]
This image shows the mirror when the hogging out process is nearly complete.

Above image shows the mirror fully hogged out.

Original TRO Introduction and concept Original TRO Purchasing the blank and hoging out Original TRO Testing the curve for accuracy Original TRO Making the tools for grinding by hand Original TRO Evening out the irregularities left by rough grinding Original TRO Polishing and testing Original TRO Making the mount and tube assembly Original TRO Making the secondary flat Original TRO Aluminising the primary mirror Original TRO First light