Making a 250mm f6.4 parabolic mirror.

1: Preparations

First things first: this section lists some initial things to do, such as purchase the raw materials and cut out a glass disk.

000617	Bought two mats of ceramic tiles (20x20x3 mm) from the "Mozaiek" shop in Utrecht (Herenstraat). The mats measure 25x50 cm, and contain 288 tiles each. This amount should be enough for two tools at least. Cost per mat: Hfl 10.=
000719	Fetched the grit and polishing powders from "de Koepel" in Utrecht. They did not have enough pitch, so that's for later. I suspect that the variety of gritsizes is a bit too great: there are about 10 sizes plus CeO polishing powder. It is probably better to buy a bit more of a bit less sizes, as also suggested on various websites. Cost of a 250mm grit-kit: Hfl 73.50
000803	The glass has finally arrived! Normally delivery time would be a week or two, but due to the SE-fireworks explosion they were too busy fitting windows. Anyway, today i fetched the three pieces of 280x280x19 mm, from "Beckers" in Enschede (the minimum order is 0.2 sqm). Cost is Hfl 59.= per piece.
001112	Today we've given it a try to drill-out a glass disk. In the last few weeks my father has deviced a 250mm diameter biscuit cutter from wood and copper, which is powered by an old 650W variable rpm drill. This setup is suspended in a drill-press, and it works OK in principle. However, the drill can hardly produce enough torque: after a while burnt PVC fumes emerge from the cooling slits. Presumably the drill makes too little rpm to cool itself properly. I have to lift the drill slightly instead of pushing it down, otherwise it either stalls or if more power is used it becomes impossible to control rpm by friction alone. After an hour and considerable amount of #80 grit the groove is still only 7mm deep. My father will work his way down, doing a bit every day. I think in the meanwhile i will try to make a set of 1:10 pulleys for a V-belt. I've also started constuction of a grinding bench. The top surface is a circular piece of 25mm beech, leftover from a kitchen-table. I believe this is pretty heavy-duty. The tabletop will sit on a sturdy tripod, like suggested in Texereau. The bench can be stabilized by stacking street tiles inside the tripod.
001201	The disk finally came out clear, only little chips on the edge, although the cutter went through under a slight angle. Apart from the drill-problems it has proven to be pretty easy to do. For the 19mm of glass, the copper blades (1x60mm each) have worn away between 10 and 15 mm. This means that they need to be replaced for each disk. Net grinding time should be less than an hour. The thing to do for the next disk is to make a setup with my new pulleys.
001213	Today i bought a load of weather-proof plaster, which is sold under the name "Duracast" by "De Hazelaar bv - Art Supplies" (Bosstraat 2, Soest). This plaster has a hardness of 300N/mm2, whereas normal plaster has a hardness of only 25. The normal plaster can also be used, but should be sealed first and have the tiles glued on. The Duracast can be poured right over the tiles. (5kg / Hfl 42,=) Other weather-proof products were all cement based, and contained some kind of filler in the form of (fine) sand. There is always the risk of a grain getting loose and harming the glass surface. On the way back made a stop at "De Koepel" to buy a kg of Gugolz pitch (Hfl 27,=).
010221	After making a grinding setup and a tile-tool over the last weeks, now the roughing finally started! The tool is made of an 250mm MDF disk, well sealed with 3 layers of polyurethane floor varnish ("Glitsa"). The tiles were epoxied on, after first arranging the pattern on a piece of sticky paper. The epoxied ("Bison two component ...") MDF disk can then be put up side down on the tiles in one go, thus taking care of the short processing time at your disposal. For the result: click the thumbnail. You can also see the MDF base i made that serves as very flat buffer between the stand and the tool/mirror.

2: Grinding

Grinding starts Feb 21, and takes about two months.

010221	I use a flat square wooden dish with cleats inside, covered in PUR varnish, that can hold either the too or mirror. This is mounted on a tripod table 100cm high, with a rubber soft and thins anti-slip mat in between. The tool is a disk of MDF, with tiles epoxied to it. The grinding stand is a sturdy wooden tripod, with a low shelf to carry stones or some other weight. The surface has been flattened to a fraction of a mm, the rest of the irregularities will be taken away by a foam anti-slip mat. Started with 3 laps around the stand, each comprising of 8 wets. Grinding time was 0:45 hours with SiC #80, using chordal strokes and heavy pressure: 1mm sagitta.
010507	Did 5 more laps of 8 wets in 1:15 hours, SiC #80: 2mm sag. The tool seems to wear on the outer ring only, much less than i expected. My conclusion is that the time spent on making a tile tool for hogging was wasted.
010508	Did 2 more laps of 8 in 0:30, the last lap slowly moving to wide W strokes. The sagitta is now 2.3 mm, which corresponds with F/6.5. A preliminary measurement of the focal length corroborates this calculation, so I'll stop hogging and start fine grinding.
010314	Tile tool for fining and smoothing: MDF disk with lots of screws partially screwed in, which forms the base of the Duracast plaster and embedded tiles. Made it by putting plastic foil on the mirror, arranging the tiles, pouring on the plaster and float the disk on top. The disk was sealed with PUR varnish first. The amount of plaster i made was slightly too much (2kg in 0.5l water), next time 1-1.5 kg wil be enough. Before completely curing i deepended the inter-tile grooves with a steel brush. (time spent 1:00)
010318	Filled the grooves between tiles with PUR varnish to prevent SiC getting trapped in some small holes. This was a bad idea! The PUR does not attach very well to the plaster, so during grinding it came loose in small bits. i hope that the figure doesn't suffer from it. A better way is to enlarge holes with a knife or the like. (wasted 0:15)
010319	Started fine grinding today, with SiC #150. After one lap of 8 wets, the tool and mirror have full contact. The wet mirror now gives a reasonable image of a light bulb. (0:30) After a thorough cleanup moved to SiC #240. One lap of 8 (another 0:30) and there are no deep pits left over. Marked some pits with a water-proof marker, and they ground out rapidly in one wet. Finish a lap of 4 wets, cleanup the stuff and store (0:15). The wet mirrorgives a pretty good image of the lamp. Now the mirror is ready for the SiC #600 (marked 3 pits to be sure it's OK). The plaster and tile tool after SiC #240. The blank after SiC #240, three pits were marked but can barely be seen on the image.
010322	Did one lap of 8 wets with SiC #600 (0:45). The three marked pits are gone, surface looks very smooth, licht filament is reflected at about 10°.
010323	Did one lap of 4 wets, with SiC #600 (0:30). Before starting marked the ten largest pits i could find, but the marks vanished quickly in the first wet. Reflection angle is still appr. 10°. Next grit size is #1200. A more precise (accurate to a few cm) measurement of the ROC yielded 305 cm. This means that the focal length is 1.55 m, which corresponds to F/6.2, very close to the intended 6!
010326	Made a base for a pitch lap with Duracast, 2kg plus 0.5l gives a thickness of 3cm. I planned to use the roughing tool as basis, a la the fine grinding tool. This failed because of size mismatch (couldn't get it neatly inside the duct tape wall). When the casting was sort of cured after half an hour, i scraped of all irregularities, mainly on the rim, to make a flat back.
010328	Started smoothing with SiC #1200, 1 lap of 8 wets (1:00). The light filament is now reflected at almost 20°, so we're getting near. The glass surface starts getting transparent again, you can make out details through it. I expect to do two more laps, and maybe some work with toothpaste, before polishing.
010329	One more round of 4 wets, and an additional attempt with "Colgate sensational whitening". The toothpaste is a bad idea: it acts like glue, mixes slowly with water and stays in a blob too long, and you risk therefore scratches. SiC on the other hand is hygroscopic, and blends quickly into a creamy slurry which forms a lubricating film of abrasive and water. The experiment resulted in some superficial scratches. Total time spent was 0:45.
010402	Did a round of 8 wets, time spent was 1:00. The scratches are gone, reflection of the filament of a lightbulb was visible under appr. 20°. The idea is to do one more session, and see if this changes significantly. If not, then i have to prepare a lap for polishing.
010409	Final smoothing session took 0:45. This time i used a different method; walk around the stand continuously while rotating the mirror in the opposite direction, using 1/3 COC strokes like otherwise. You can do this by repeatedly taking a big step left and then moving to the right along with the mirror. Apply very light pressure and extend the wet as much as possible. When it dries out, moisten the tool close to the edge of the mirror. After 10 to 15 minutes there is only a very thin film of worn-out abrasive and water left, the grinding makes a singing or whistling sound. Did three wets in the 45 minutes, and then cleaned up the lot. Reflection is still at 20°, or possibly slightly more. I guess it's not going to improve with the #1200. Next step: polishing!

3: Polishing

Polishing starts April 18 2001, and takes about eight months. First attempt has led to scratches and bad figure: had to go back to SiC#320. Second attempt went better, total polishing time is 13:45 (26-Nov). This could have been shorter, when more attention was given to fine grinding a good sphere, micro facets the pitch lap, use of heavy pressure and dryish lap.

010412	Finished the pitch lap today; 2:00 hours spent. Moisten the plaster base with turpentine, which it quickly absorbs (very nice smell). Made the CeO slurry and added a few drops of some dish wash stuff. Melted the pitch (Gugolz 64), until a quite thick syrup. Poured it on, it stopped flowing right at the edges! I thought that this was great, but putting the coated mirror on it to pre-press made the pitch already too hard to put all the channels in (using a aluminium rules of 3mm thick). A remedy for this is the paint stripper, just play it over the surface lightly, and push in the ruler until the squares are OK. Final pressing while the lot cools down, and then the lot was stored for later use. My squares are approximately 25mm wide, but not too precise. Channels are 3mm, squares in no way centered on the tool.
010418	OK, today i'm on: it's polishing time. Warmed the lap with the stripper again, and let the tool and lap come in contact with a cold press of 0:15. Initial polishing shows that the outer rim quickly becomes transparent, that's fast! Cut off the excess pitch from the edges, heat and press again, and then the fun can start. Polished for 0:30, tool on top and center over center 1/3 total amplitude. Press another 15 minutes, and then polished another 0:30, same technique. after the hour, the outer 4 cm seem to be pretty transparent now, and also the central plane is polishing but at a much lower rate. Do another session before initial testing.
010419	Some of the channels had closed, so i started with rechanneling the lap. The paint stripper works great for this, only you have to take care not to overheat some spopts (keep moving it around). Start after a short press, ahen the temperature has more or less stabilized (5 min). Polished only 0:15.
010423	Pressing the lap with an orange-net for micro facets. The plastic bag prevents drying out when the mirror and lap are stored. Cleaning and storing separately appeared to be better for preventing scratches. The center still shows considerable haziness, which wouldn't go away easily (surface is not sufficiently spherical). Did 1:00 of polishing, TOT and COC. When i let the mirror rest on the lap for a while (to take a sip of coffee) the mirror almost got stuck: so much for good contact! After the session the haze in the center seemed to be decreased significantly, so a preliminary test should be possible. ROC is approximately 318 cm. Ronchi gives sharp outward hooks when inside ROC. Foucault shows a grey center with a bright rim on the right side. You can also see some air turbulence, maybe some warm object under the light path. This is all looking very interesting, but the conclusion must be that the mirror has a turned-up edge. I guess (and so does most of the list) i just need to go on with polishing. Since the center is also getting more transparant, i have good hope that i don't need to go back to grinding.
010501	Today i discovered two scratches on the mirror, don't know how they got there. I suppose it is caused by uncautiously putting together the potentially polluted mirror and tool to carry them upstairs for a rinse. Or maybe it was when i had to force the disks from eachother last week, when they got stuck during cold pressing. I hope the scratches will polish out, but at least one of them looks too deep for that. See what more polishing brings, maybe i have to go back to grinding anyway... Put in one more hour of polishing, and i have the feeling that the drag is more constant during strokes. Another Fourier test showed that the turned edge has decreased, but not so much. ROC is about 6mm shorter on the edge than it is in the middle. There is still a very light haze in the central zones. I'll try two more hours of polishing and see what it brings on the tester.
010508	Polished 1:15 MOT After cleaning up there are a few more scratches, the turned up edge has only slightly lowered. What is the cause of the scratches, washing before the session with grits in the water? Polluted pitch? I need an airtight container for sure to store the lap, a plastic bag just won't do. Next step: SiC#320 (snif).
010510	Back to SiC#320 it is: only a few wets is enough to get the tool and mirror in good contact and get rid of the scratches. Did last wet TOT to get a better shape. I noticed that the wooden base of the tool has a slightly convex form, probably due to the shrink in the layer of plaster. So i ground it flat and revarnished before going to SiC#800.
010511	Put in 1:00 with SiC#800, last wet TOT and extended by wetting a few times. The filament of the lamp reflects at 20° already, and appears the same on every place of the mirror surface. It's probably a good sphere now. Next 0:30 with the SiC#1200 that's left over, 3 wets (second TOT), extending them as much as possible. Marker lines seem to wear first at the mirror edge, but are gone in a few minutes anyway. Reflection after cleaning up is at 30°, so good enough for polishing. Made a new lap, recycled the pitch from the old one: horrible mess. This time the pitch was too soft really, and the lap base too warm, so it flowed quickly all the way to the edge and further: more than horrible mess.
010512	Channeled the pitch lap, and trimmed the edges. Did 0:15 of preliminary polish. I also made a dust-free box, with separate floors for lap and mirror.
010514	Did 1:00 MOT and 0:15 TOT, two intermediate pressings. Cleaned the mirror only when done, and hope the lap will be OK for next time. The mirror appears to polish out evenly, the haze is decreasing equally over the surface. It seems that the surface is better spherical than last polishing attempt.
010516	Did 0:50 MOT and 0:15 TOT. Cleaned the mirror and inspected the surface: haze is getting vague now, evenly spread. There are a few light scratches again, possibly lumps of dried CeO. The list sort of confirms this, so i will go on with cleaning the lap also, take better care of cleanlyness, press warm before each session and don't let it dry out. otherwise, during this session the drag was very high but regular, indicating something close to spherical.
010521	Did 1:00 MOT, cleaned the lot very thoroughly before starting, brushing out any CeO remainders. Pressed new channels and did a long warm pressing. This helps, no new scratches! Need to use special means now to see the surface: a 500W lamp and a flocking-paper background. Initial foucault and ronchi show a ROC getting shorter towards the edge, and a very narrow TDE. Maybe i should do a bit more TOT now?
010523	Put in 1:00 TOT. Again, no additional scratches, the two smallest are hard to find now. Figure hasn't changed very much, but there was not enough time to let the mirror thermally equilibrate. Surface still shows up under the 500W light (i use this also for heating the lap, works great!). This shows how the lap is warmed up using a 500W light at appr. 50cm. The black pitch takes up much heat: if left under the lamp for too long the pitch is too soft for warm press but perfect for re-channeling! Barely visible surface under a heavy lighting. The scratches and rubbish you can see are mostly on the back. The mirror is resting on "velours paper", which is hairy black and has a very low reflectivity.
010528	Re-press channels. Thorough clean-up; soak in warm water loosens many CeO lumps. Warm press for 15 minutes causes very good contact from the start. Polished 0:45 TOT. Only one scratch left, and one more which is hardly visible. The surface is still visible with focused light.
010714	Heat tool and mirror under lamp, separated by CeO slurry. Hardly any pressing, shape feels OK from the start. Consistent drag doing short strokes, very heavy drag after 30 minutes, lengtened stroke to 1/6. Then Polished 1:00 TOT, 1/10 - 1/6 COC, slight W. Foucault and Ronchi say: almost a sphere with slight TUE. Haze seems almost gone. The total polishing time is about 6 hours now.
011105	It's been a long while, but the basement is clean again, and i have a brand new tester made from old floppy drives. See elsewhere on this site. Started with cleaning the mirror and tool in plenty warm water, and then warm press for 15 minutes. Used fresh CeO slurry. Polish 1:00 MOT <1/3 COC. Foucault shows a pronounced TUE and a depressed center. The lap seems to have recessed edges, making it too small. Probably also started with too warm mirror/tool. Smoothness is also pretty bad. Probably the new tester performs very well, and shows more detail than the old one. For next session: longer strokes, lower speed, longer press, TOT instead.
011107	Polished for 00:30, TOT, 1/4 W. Put extra pressure on overhanging side, ala Texereau. Unfortunately there were some interrupts, and i had to pack up early. Foucault and Ronchi show a somewhat decreased TUE and a shallow central dent. A light filament can still be focused on the surface. I saw on the web that TUE can be treated by pressing often, lengthening strokes and working TOT.
011108	Half of the channels were closed, so this session started with rechanneling. Tool in hot water, then under 500W lamp. Pressed channels with 3mm aluminium bar. Then warm press very well, to establish full contact. The polishing periods were short (10 minutes), doing 1/3 TOT wide W strokes. Polishing spells were interleaved with pressing for a few minutes. This is done in order to improve the turned up edge. Total time spent actually polishing is about 00:30. Ronchi now shows )( inside and () outside ROC, with acute hooks in opposite direction. If i'm correct this means that i have an oblate spheroid now, with a TDE of a few mm. Diffract seems to agree with this conclusion. The mirror surface can still be made out when the 500W is focused onto it. Next session i'll do the frequent pressing again, and use shorter strokes and narrow W.
011113	Used musquito screen for micro channels: only some shallow pits, pitch is too hard, maybe too cold. Did 0:45 TOT short COC anyway, press often. Foucault and Ronchi agree on oblate shape, and a few mm TDE. Dumped lap and mirror in hot water and re-pressed with screen: pits are deeper, especially in the center. Did 0:15 MOT slow long and wide W strokes. Ronchi shows less oblate now, but with hole in the center. Foucault agrees, surface looks smooth too. Visual inspection reveals still many pits, and a 5mm deep scratch (grr). This was caused by too much hurry for the last MOT spell.
011114	Heated lap + mirror in hot water and pressed 10 minutes with musquito screen and 10kg. There are better facets now, although still not very deep. Pressed 20 more minutes without screen and only 5kg. Polished for 1:00, interleaved with about 5 short pressings. Systematic TOT 1/3 narrow W strokes, rotate mirror before each pressing. Last 5 min MOT, same stroke. Quick test says still oblate.
011115	Warm press 5 minutes before + few times during polishing. TOT 1/3 COC, sometimes going to 1/3 offset chordal strokes to work the edges more. Spent 0:30. Edge seems to be less steep now, center hole still there. The lap is getting bad too, air bubbles rise to the surface and ruin it. In the holes that are left CeO piles up, i hope this won't scratch! (how hard is pitch itself?)
011120	Experiment: last week i painted the lap with turpentine, and let it simmer for the week. The pitch surface is now sticky at many places where i overdid it, but the surface is much softer in general. Pressing with musquito screen leaves disaster: pitch all over the place! Now clean the mirror with turps and apply plenty CeO slurry + detergent. Repeat short press and MOT stroke sequences, until after 0:15 the surface seems to be better. No more pitch on mirror so let's give it a try. Did TOT 1/3 COC for 1:00, and pressed often. Drag is extremely high, but consistent: the lap seems to conform very well! The surface shines like the glass itself. Quick test shows improved figure, but still TUE and center hole.
011122	Heat lap + mirror in water, press with fresh screen, and then some more: very nice facets. Started off MOT until assumed thermal equilibrium. Then off for 1:00 TOT 1/3 COC strokes, press often, rotate mirror at each pressing. Quick test shows again improved figure, visual surface inspection still reveals pits, but less numerous.
011126	Same routine as last time, 1:00 in total. Used LOTS of pressure slow speed (2 cm/sec) and less moisture. Last 0:15 the ensemble started squeeking, and the mirror surface looked almost dry and clean. Quicky F+R tests reveals better figure again. I can hardly find any pits under a loupe and 500W focused light. What you see on the surface is the scum leftover from the water, and possibly some CeO. Next time i might take longer before assuming thermal equilibrium: not doing so may cause TUE because the egde cools first. I also have to use distilled water for rinsing the mirror.

4:Final figuring

This section contains the log of the figuring process, finally picked up for the 250mm mirror to be completed for the Travelscope MkII. The mirror making process finally ends, about 7 years after starting it!

070509	Left image: After an attempt to bring down a high outer zone: bad TDE. The TDE is treated with one round of 1/4 TOT strokes, applying local pressure on the tool edge, working between 70 and 100% zones. The result in the right image is after 3 rounds of 12 directions, each direction 10-15 half strokes.
070511	A new attack is tried out: score the (left) edge of the lap and use this to wear down specific zones in the mirror. MOT, tool has fixed orientation, mirror turns a little every stroke.    Left shows the lap surface through the blank, right shows the polishing setup           The above are the results after four times three rounds and two more times six rounds. Strokes are 1/3 length, mostly up, slightly overhung to the left, fairly high local pressure on the left side. Nothing has changed much, really, maybe even worse ...
070512	Today the plan is to apply the TOT normal stroke: 1/3 long, 1/4 wide. This should at least show some change after 30 minutes or so. Possibly the strokes need to be lengthened and widened to lay more focus on the edge. Well, this is 45 minutes and three rounds of (longish) normal strokes later. The figure seems to be getting flatter now. Plan is to do some hours more and maybe enlarge the strokes to bring the edge down quicker.
070514	Today the stroke is TOT, 1/2 long, 1/3 wide. With the pressure applied, only about 20 full strokes per minute were possible. The session of three full rounds took about 1 hour. The figure seems to have improved more, the high zone has softened. But progress is so sloooow.
070515	One more hour of the same: Also the Foucault image shows a diffraction ring all around the edge, albeit a bit stronger on the side of the light source. The edge is looking good now, center is a bit high. Now I can either go on like this to work the edge further down slowly, or start MOT parabolization to deepen the center quickly.
070516	Started parabolization strokes to deepen the center, but after trimming the edge of the tool. On the left image the result after a short round of 15min, on the right a starting correction becomes visible after another 15 minutes. The roughness in the images is from air currents.
070518	Revived the lap, since the channels had closed. The channels were cut out with a V shaped piece of metal heated with a soldering iron. This provides fairly clean and deep channels, without touching the lap surface too much. The edges were cut back to match the mirror circumference. Working the lap back in contact has lowered the central 70%, since the next zone focuses short and then tapers off. I think the best way to go forward now is to get rid of this by flattening the edge, like before, and then continue parabolization. The edge still shows diffraction all around, so the apparent TDE is not too bad in reality.
070519	Attempt to bring back the high outer zone by TOT as before. Now the edge has turned down a bit more than anticipated. The diffraction is barely visible, so maybe it is best to correct the edge before going to parabolization...
070520	One session TOT 1/2x1/5 stroke, does not yield much result. Right image: Texereau fig42 method 6B to bring down the hill just inside the edge. Edge seems to become a little better, but the center lags behind. Next session I will try to deepen the center with MOT parabolization, concentrating on left and right decentered parts.
070521	Tried to WIDEN the stroke instead: TOT 1/3 long x 1/2wide, see result in Ronchi and Foucault. Drag is enormous, I need to fix the mirror/lap holder better. Otherwise not much difference, let's do one more go tomorrow, and the day after...deepen center.
070523	Parabolization to deepen the center first. One round of once left and right 4/5D, concentrate on lateral overhang to work on center.   Once more of the same.   One more round with double left and right, regular parabolization working the whole mirror. A flat part has formed in the middle now, but this can easily be fixed. Next will be to somewhat flatten the edge with some more TOT, or maybe first concentrate some more on deepening the inner half of the mirror.
070525	Two more 30min rounds parabolization,MOT. The center is deepening now, and an overall parabola is starting to form. It is now getting time to measure the amount of correction. The stroke seems reasonably under control now, still focusing a bit more on the lateral overhang parts. Cold press 15min before session, evenings seem to go better than daytime.
070527	The first order of magnitude measurement corroborates what you see in the last foucault image: a high central region. The plan is do do just one more round deepening the center with parabolization strokes accentuating the lateral overhang. After one short round deepening the center, followed by a few licks on the 80% zone. The figure has improved, but the lap had developed a small bump near the center, which apparently could not be pressed away in several hours, and left a sleek. This sleek is visible in the foucault image. Another short round concentrating on the 80% hill, not much effect. Next try a short 3/4x3/4 TOT to work on the edge.
070528	The TOT 3/4x3/4 does not have much effect, only that the edge seems to turn down. Next, try Texereau Fig.43 method C-II. The Tex local retouching results are promising, analysis has been done this time with a 4 zone mask (24, 63, 87, 105, 120mm edges). Also the tester has been modified to enable cornea-cutting views and hence more egal illumination.
070530	One more local retouching action has brought down a zone allright, but it's a bit lower radius than expected... The next will be to approach it a bit gentler, and do a wide TOT (1/2 wide, 1/3 long) as this worked before. The goal is to smoothen the high areas, especially the outer one, while keeping undercorrection. Then a final (?) MOT parabolization should do the trick.
070602	The foucault tester has been re-calibrated, i.e. the knife edge better aligned with the slit to give a more vertically balanced image. The central part is clearly undercorrected, and now shows the surface roughness quite well. The idea is to smooth this when deepening the center later, after the edge is corrected. Note that diffraction is still visible on the dark side. The first session wide TOT strokes, two full rounds. The edge is coming down and, as can be expected, the center is getting flatter. Tomorrow some more TOT is planned, to continue this until the edge becomes acceptable. It may be desirable though to use more pressure on the edge to flatten the persistent 80% ridge.
070603	Accented pressure on edge of tool, over 5-30mm inside edge, 1 round. Accented pressure on edge of tool, over 5-25mm inside edge, 1 round. Smooth with 1 round 1/2 wide TOT. The mirror is for the larger part almost spherical, at least much undercorrected. Readings therefore are not very accurate, but give an indication nevertheless. The edge still has a ridge, but it is moving outward. Do some more wide TOT, maybe applying some more pressure on the overhanging side of the tool.
070606	One full and one short round of 1/2 wide W, TOT. Measurements have not been taken, since the inaccuracy is becoming too much to be useful. Not much has changed since 070603, this asks for more rigourous measures: local pressure on the overhanging tool edge.   One round of local pressure on +5 to -15mm from edge. Then one round to blend it into the -10 to -30mm area. The latter was probably not necessary, since a familiar ridge is now forming. Do it again with a bit more overhang.   One round of local pressure on +15 to -15mm from edge. The turned down edge is now accompanied by a trench at the 80% zone, according expectation both would cancel out somewhat after a smoothing session with wide W TOT. Note, camera setting at 1/20 sec gives better images   One round of wide W TOT. The edge 5-10mm seems quite horribly turned down now, also the diffraction is gone. This calls for some dedicated TDE treatment, for example working the 10-30mm inside the edge, TOT or MOT short strokes.
070609	Two rounds of 1/2 TOT, 4/5 wide, lenghtening in the COC part and shortening at the edges. The expectation was that the edge region would overall lengthen ROC and thus reduce TDE, but not to much effect. Surface is getting smoother though. The foucault is taken outside ROC, so the edge zone is more pronounced than in the previous picture. The center is now showing some correction in the Ronchi image, indicating too wide strokes.   Two quick rounds of MOT accented 5-30mm withing edge. The effect seems to be more on the -30mm part than on the edge part. Probably some more TOT sessions would be better, only with less overhang.   As planned, three full rounds of 1/2x1/2 TOT, high pressure and drag. Center is getting flat, probably since it was not touched by the 1/2 D wide stroke. This is OK, but should travel towards the edge. Next could be to narrow the stroke even further and turn the edge more upward by avoiding too much sideways motion, or even avoid this altogether.   Two rounds of 1/2 TOT and COC with slight sideways deviations. TDE seems to be getting less now, hopefully the scratch is not really a scratch...
070610	After trimming lap channels and edge: Three rounds of 1/2 TOT and little lateral motion, a few cm max. The scratch is real, possibly the encounter of mirror surface with faucet that happened yesterday. The figure is quite good now, a few more sessions and the TDE will be gone.   Three more rounds of 1/2 TOT and little lateral motion. Some roughness is developing, probably due to little W. Center focuses long, posibly as long as the TDE now. The area 60%-90% has relative short ROC, and hence a ridge has formed just within the edge. Widening the stroke should take care of this. The plan is one more session before moving to parabolization, as 070525, only with a better starting point.
070611	Two rounds of 1/2 TOT, about 1/3 wide W. Edge has become slightly better again, but is still turned. The foucault looks better visually, and is taken at shorter ROC than previous. Maybe try a MOT session now, to smooth the center a bit. Not too wide strokes though, so the edge will also profit.
070612	One extended round of 1/2 MOT, maximum 1/4 wide W. The plan was to flatten the center and at the same time work away the TDE, which was moving very slowly in the previous TOT sessions. Something has probably gone wrong pressing or maybe in the first strokes, since there is now an excentric shallow region visible in both Ronchi and Foucault images.   One short round of MOT parabolization. The idea was to deepen the center somewhat, and get rid of the asymmetry. This has somewhat worked, but the edge has not been affected. To correct the edge entirely, it may be better to do some longer TOT sessions (some wide). The idea is that this will smoothen the center and at the same time lengthen overall ROC to catch up with the low edge. To detect how roughness moves about, the testing will from now on be done with the mirror in the same orientation.
070616	Two extended rounds of 1/2 TOT, maximum 1/4 wide W. Channels were trimmed before pressing. The central roughness seems to smooth out OK, but the edge is still turned down.   Four rounds of 1/3 TOT, COC with slight lateral deviation. The polishing speed was much higher than before, just to see what this would bring. Last round slow strokes again. The edge region must be worked, so the tool edge should pass over this area. Well, the edge has been worked and the figure is clearly oblate now. The images are lousy, mirror not yet stabilized. However, the plan for tomorrow is to work the center TOT, i.e. wide strokes and pressure on the tools edge when passing over the center regions.
070617	Three rounds of 1/3 TOT, almost full diameter W, longer strokes COC and shorter EOC. The COC strokes are traversed quickly, while the lateral strokes get accented pressure on contact side. The center slowly shortens ROC and gets smoother overall. There are still remains of the excentric hill left of center. Edge seems not to be affected much.   Two more rounds of 1/3 TOT, as above. Excentric hill is almost gone, low area to the lower right (which mirrors the high area on upper left side). Edge seems not to be affected much. Looking back, half may the edge was OK but went bad upon accented pressure rounds early june, and never recovered entirely. Just go on with TOT, like 070521. The edge needs rechamfering too, it is almost gone at some places (see the chip at 4:30).   After channeling the lap broke it in wit 1/3 TOT, 1/2 wide. The lap has now approximately 20x20mm squares instead of 50x50. On the left image also the new support can be seen, a 400x400mm floortile, and the synthetic chamois used as an egalizer. The idea is that the smaller squares will conform easier to the mirror and average its action a bit quicker than with large squares. The KE image already looks smoother.
070618	Three rounds of 1/3 TOT, almost 1/2 W, to see how the new lap works. Still a slight trace of roughness present, visible in the middle left area where the edge shadow is somewhat thinner. This can however also be an artifact of the picture, what could be made in the lower region of the slit image, since the lower part is darker than the upper part over the complete width. Bad image anyway... The stroke has deepened the outer 30% or so, but the edge again lags behind. It would probably be better again to widen the stroke as a few days before.
070620	Three rounds of MOT parabolization (left) and one more (right). Correction is setting in, while the edge remains about the same. Using thick CeO, half stroke per second and low pressure, the lap seems to work in a gentle and controlled way. A few more sessions to go before measuring the curve. The edge remains hopeless, maybe a leftover from grinding?   Two more rounds of MOT parabolization. Correction is proceeding slowly, and so is the central roughness. Plan some 1/3 TOT to see if edge improves from (partly) corrected situation rather than oblate.
070623	One round 1/3 TOT, 1/4 wide. Foucault image is moved. Center seems to be flattening, hence the ege looks worse. Try another MOT parabolization session next.   One round MOT parabolization (4/5 x 4/5). Foucault image is moved again. Correction is extended, but the center has a flat spot. One more quick MOT concentrating on central (i.e. overhang) part of W.   One round MOT, accent on center, short TOT smoothing. The smoothing is done by dragging the lap-edge from 50% before to 20% beyond center, and back, in several positions around the work. The central flat has got worse, more accented, too much concentration on center part. Some wide TOT sessions may be needed to correct this. The lap surface is crusted a bit, it may be necessary to make a new lap altogether.
070624	Lap foundry Scraped off old lap New lap after first session Made a new lap, the old lap was worn out. It had developed a hard crust of pitch with embedded Cerium Oxyde, which possibly caused undeerministic behaviour. Scrape off top 2mm, warm the pitch on a hotplate, warm the mirror in hot soapy water. When the pitch is almost ready, heat and melt the remaining pitch on the old lap, and then pour new pitch on top of this (adding about 3mm). Press mirror covered in CeO onto lap, and leave it there. When cooled down, trim pitch along the mirror edge. Cut grooves with a soldering iron and V-shaped cutterblade. This time there is no bevel around the lap, so trimming must take place every session. The idea is that it might have a beneficial effect on the mirror edge.   Three short rounds 1/3 TOT, COC up to max 1/4 W. The lap has broken in nicely. After three rounds it even started squeeking. The edge is actually getting better, some roughness added probably due to the new lap. The roughness will presumably disappear, the edge should improvemore and the shape will become sherical again after some more TOT. Possibly widen the W to get more action in the central region.
070626	One quick round TOT normal, one slow round 1/3 TOT 1/2 wide shorter strokes left and right, one last quick round TOT normal followed by smoothing strokes. The center has improved, not much change otherwise. Plan is to do some more sessions normal TOT, and widen/lengthen strokes every now and then. Part of the apparent roughness is caused by thermals: the garage had cooled down to under 15° when pictures were taken.
070627	One hour TOT normal strokes. Shape improves only slightly, after an hour workout... Still lots of thermals in especially the foucault image, probably worse when close to spherical. Have to widen the strokes for a while to attach the 50% zone a bit more.   45 minutes, mostly 1/3 TOT 1/2 wide, last two rounds normal stroke. The progress is very veeeeeery slow, the new lap seems to act really slowly. Perhaps the strokes are in balance as it is, and the next try should be a normal strokes MOT round to see how it acts that way.
070630	One hour and a half in total, starting TOT normal strokes, then MOT normal strokes, then MOT edge treatment as the 070511 session, finally 1/3 TOT COC. The last part ended with a screeech when a bit of pitch came loose, but there is no visible scratch. Goal was to attack the edge specifically, and it seems to have improved slightly. Some roughness is introduced but with a few TOT normal rounds this should disappear.
070704	3/4 hours 1/3 TOT mostly COC, sometimes longer strokes. The edge is becoming better now, even turning up slightly. The roughness needs to be taken out still, with wider strokes, en-passant taking down the center area too.   3/4 hours 1/3-1/2 TOT 1/2 wide, sometimes longer and wider strokes. The surface is much smoother now, but the edge has turned down again. The plan is to start MOT sessions anyway, first not too wide in order to flatten the part just inside the edge, and then wider and longer to correct the center.
070802	Picked up the work after nearly a month with 30min <1/3 TOT, COC and small W. The idea is to get into the routine again, and let the lap and glass mate before doing anything else. As can be seen, the edge is less turned, but the slight W causes the outer zones to roll up. There is some roughness in the surface, which I will try to remove with slowing down and lowering the pressure even further.   20min <1/3 TOT, COC. Not much is changing, really: probably the last session was too short. Pics are bad too. The COC sessions seem to reduce the TDE a little though.
070803	45min <1/3 TOT, COC (four rounds). The surface seems somewhat smoother and the TDE compensated with a slightly rolled up outer zone. The trench that is formed this way probably indicates the more active outer ring of the lap surface. Flattening out the rolled up part requires MOT work, which might as well be parabolization from the start.
070804	A few rounds 1/2 MOT, COC, very slow strokes. The surface has become overall smoother yet, although now the stroke directions become visible. The edge is still turned down and a bit rough, it remains problematic. Also the images duffer from the garage heating up, causing air currents.   Five rounds of MOT parabolization, approximately 80% strokes. Correction is clearly setting in. Surface roughness is increased significantly, although this may also be an artifact of the exaggerated contrast: the lighting is so low that photoshopping (auto contrast and despeckle) is needed to produce a usable image. The cause may be a too watery lap causing a very high, rubbery drag. Need to use more CeO to cope with this, and to make the action a bit faster and more reliable.
070807	One round of parabolization, using a richer CeO slurry. A hole has developed in the center, due to too much lateral overhang (up to 90%). Also there is a spiral shape in it, indicating that the single round changes its action while. Next session has to be narrower laterally, and it would be better to distribute the directions of one round more randomly (i.e. skip directions and make three 1/3 rounds).   1Rnd 1/3MOT, 1/2W to push down ridge around hole. 6halfRnd 1/4TOT, 4/5W to bring down outer 80% The hole is now reduced, average correction still too much although less than before: best conic is now -1.2, Strehl >0.8. The 50% zones need some more work (to decrease zone 1 to zone 2 correction) and then probably the outer zones too. Best to use TOT with edge over zone 2.
070811	After 6 half rounds TOT 2/3W, very short strokes lengthening over 50-70% zone. The difference from zones 1 - 2 and 2 - 3 is too high, while the difference from zone 3 - 4 is too low. Also overall correction is too much. This means that zones 2 and especially 3 need to focus shorter, and hence should be worked on. Ultimately also the edge must focus a bit shorter to improve overall correction. The plan is to start with zone 2 and 3, then zone 4 and finish with some final corrections of the inner zones.
070812	After 2 half rounds TOT, accentuated pressure on zone 2 and some on zone 3. As can be seen in the first foucault test analysis, the treatment seems to have been the right one and the overall correction is good. However, the lower diagram shows the analysis of later measurements during the evening with a less favourable figure. Also, the 90° rotated measures slightly differently (correction -1.1) than the 0° set, indicating mild astigmatism. Next morning, a double check of the foucault readings indicates again a better figure. This suggests there is some influence of ambient lighting conditions on the perception of the shadows. In any case there seems to remain some final tuning to do. The plan is to repeat the treatment of accented pressure over zone 2 one more time.
070813	After 2 half rounds TOT, accentuated pressure on zone 2. Nothing has changed much, really. A new measurement will be taken tomorrow, daytime. If the results are not worse, the mirror is done.