TRO+-+Testing+the+mirror

The mirror is now suitably polished after 2708 mins (over 45 hours) of polishing to begin figuring.

The spreadsheet attached is a log of progress:

The following page is a log of figuring: TRO_30inch_figuring_log

__ Test stan __d
The image below shows the test stand bolted to the roof of the garage. It can be moved along the fixing beams to cater for changing focal length.

All Ronchi images below are taken with a 5 line/mm grating.



Figuring would begin with reducing the height of the central hill using an 11 inch star lap. Initially we used the machine, but we found that the tool had too much load and that the pitch on the tool was too hard. This was removed and some new medium pitch added. We began figuring by hand. We will in time use the overarm of the machine and use the counterweight system that it has to reduce tool pressure.

**Videos of figuring the mirror:-**

Reducing the high centre with the machine. Cerium and BM303 mix. 11 inch star lap. [|TRO_30inch_figuring_highcenter.avi]

Hand figuring with the 11 inch star lap [|TRO_30inch_figuring_highcenter2.avi]

Geoff hand figuring with the 6 inch star lap (16th December 2008) [|TRO_30inch_figuring_geoff_161208.avi]

Other video clips taken during the making of the mirror can be seen here:- TRO_30inch_figuring_vidoes



Figuring strokes
The word doc is here:



The graph below shows the wavefront errors after polishing and before any figuring work.





Above image is after 140mins of using the star lap by hand working on the center 80% and going edge to edge slightly off center with a 2 inch overhang. See figuring page for details. Some marks in the center of the mirror are marks from my fingures!.





Image above is after 265 minutes of figuring using the 11 inch star lap. We have had some problems with what looks like astigmatism. The Ronchi line bending when placed on the test stand. BUT, when we rotate the mirror the lines are not consistantly rotating with the glass. Sometimes it does not move with the glass, some times it moves a little. Each time the Ronchi lines look slightly different. We think it is one of a number of problems.

It does appear that the lines are more 'bent' after coming straight off the polishing lap. After a while they settle down. Also, shifting the mirror in the test stand has an effect too. It could be that the test stand is not doing it's job properly. I stiffened up the stand at the back and ensured that the front face was flat (to within 1mm). The carpet and bubble pack should do the rest. After all, the majority of the lead is radial through the strap. Maybe this is a source too? I changed the strap for stiff 15mm wide plastic packing strap. This has seemed to make things worse?

Image below shows the progress after 265 minutes of polishing. Not that much more has been taken out of the middle. We did put less pressure on the tool in fear of the tool flexing teh mirror too much. It is only 10mm thick in the middle.

Image below shows the knife edge readings after 373 minutes. You can see from the previouse graph that it has not changed a lot. We had done some work with the 11 inch star lap, again using Cerium on it's own, but the figure did not appear to change much when looking through teh tester. The lap would also stick and bind at times and it would be difficult to get smooth motion across the mirror.

We woundered if the cerium was doing it's work? If the bottle containing the cerium and water was not shaken, even for a few minutes, it would congeal in teh bottom of teh bottle. We think that not much cerium was actually getting onto the mirror and that it was mostly white water! We decided to add some BM304 (BM 304 = 7 micron, the Cerium we were using is 3 micron) to the bottle.

After only a shory period of work you could easily see the change when looking through the tester. The motion of the star lap was MUCH smoother and easier to control. We did not need to put much pressure on the lap. This will hopefully reduce the stress put into the glass. The BM404/Cerium mix did not appear to dull the mirror in any way. More work should give us a clearer idea of how well this mix will work.

One other small change was to remove the hard plastic packing strap on the tester and use 1 inch wide nylon strap. This reduced the stress on the edge of the glass. The bubble pack was reversed also so that the bubbles were facing teh carpet and not the back of the mirror.



Some images below using the 11 inch star lap. Center over center with a rounded end moving slow with cerium only. The other images below the top two are using the BM304/Cerium mix, were we moved the tool a little faster across teh mirror. Working through the center and at the 50% zone.





Above image is after 453 minutes of figuring: TRO_30inch_figuring_453mins_a.jpg. It was taken only a short time (20 mins maybe) after coming off the lap.
 * [[image:TRO_30inch_figuring_453mins_a.jpg width="551" height="469"]] || [[image:TRO_30inch_figuring_453mins_b.jpg width="508" height="457"]] ||

The image to the right (TRO_30inch_figuring_453mins_b.jpg) was taken 2 days later after NO more figuring was done.

The differerence is clear to be seen. It appears that the amount of work done is less. In fact, we think that the glass has re-bounded over time. Exactly how long is not sure.


 * [[image:TRO_30inch_figuring_583mins_a.jpg width="543" height="448"]] || [[image:TRO_30inch_figuring_583mins_b.jpg width="458" height="439"]] ||

Image above above left is after 583 minutes of figuring and was taken 20 minutes after coming off the lap. The image right was taken the following morning with no more work done. TRO_30inch_figuring_583mins_a.jpg.



The image above is after 819 minutes of figuring (image TRO_30inch_figuring_819mins_b.jpg). The image shows that the figure is now beginning to respond and the edge improve, yet the readings off the knife edge shows that hardly any work was done at all? We thought of a number of reasons for this:-


 * 1) The Cerium oxide is not working. Unlikely.
 * 2) Not enough tool pressure. Likely, but with that amount of work we should see some change.
 * 3) The glass is flexing with tool pressure and not much work is going on? This is unlikely.
 * 4) We are doing the wrong type of work. Likely! Why? Because I think we may be working the whole mirror and not concentrating on one area. If we are doing centre, then 50% then edge, then we are doing the whole mirror and taking glass off the whole surface at a similar rate??

So....lets concentrate again only on the centre 75% of the middle for an hour or so and see what happens!

Image above are after 889 mins of figuring. the image left was taken when teh mirror was jsut taken off the lap. The image right was taken the next morning after no more work was done. Again, there appears little difference from the previouse figuring session. Knife edge readings yet to be taken. The differnece between the two images above though shows again that the mirror changes a LOT over time after figuring.
 * [[image:TRO_30inch_figuring_889mins_a.jpg width="515" height="479"]] || [[image:TRO_30inch_figuring_889mins_b.jpg width="481" height="476"]] ||



Image above was taken after 1218 mins of figuring. The image was taken on the morning of the 5th December 2008 after the previouse nights figuring session.

You can see that thee is a change, but looking at the spreadsheet, the mirror has relax back yet again quite a lot after figuring.

The previouse 2 sessions before the 4th December showed NO change in the figure. We had been doing straight long strokes through the middle (3 inch overhand average) mixed with W strokes. We think that the mixture of strokes was the primary reason for seeing little improvemnet in the figure. That is each stroke type was cancelling out the previouse.

We decided to opt for straight strokes only. Initially 1 hour through the middle, followed by 30 mins 2 inches off centre. The Ronchi indicated that this time we had definately done work. But, until the mirror had time to settle, we did not know how much.



Image above shows the new 6 inch lap. This was made in an attempt to have closer control over the figure. We have had trouble reducing the middle zones. The 6 inch lap was used only for 15 minutes in the midddle. This had the desired effect of deepening the middle. BUt 15 minutes was too much. It also made the mirror rough in the middle and at the 50% zone. We then re-pressed the lap and reduced the pressure. In addition, we used oval strokes (see above image) rather than tangental strokes which appeared to rough the surface.

The image below shows that the overall figure is getting closer (4 waves approx) and the edge has improved.



Image above taken on the 16th December 2008. Geoff reducing the 60% zone with the 6 inch star lap. The larger 11 inch star lap would then be used to smooth the figure.

Geoff hand figuring with the 6 inch star lap (16th December 2008) [|TRO_30inch_figuring_geoff_161208.avi]

Image above is after 1728 mins of figuring: image reference: TRO_30inch_figuring_1728mins_a.jpg

The graph below shows the slow, but in the right direction, progress of figuring the mirror over the last few polising sessions. Compare the red line to the ones near the top of the page.



Image above is after 1758 mins of figuring: image reference: TRO_30inch_figuring_1758mins_a.jpg

The graph below shows that the last 30 minutes of figuring has really improved the overall correction. FigureXP reports a jump from 4.6 waaves to 2.8 waves.



Images below are after 1893 minutes of figuring. Figure XP states (on 4/1/2009) that the mirror is now 1.27 waves. Not much work is now required to complete.

But, after retesting on the 10/1/2009, the figure has relaxed slightly back to 2.1 waves. See the graphs below to compare the differences. This must be due to temperature.

Test data on the 4/1/2009 was taken 50 mins after coming off the lap.

We did another 70 minutes work on Sunday the 11th January 2007 using the 6 inch star lap (1963mins total figuring) mainly concentrated on the 225mm zone and the centre of the mirror also. The mirror is repsonding but slowly. Figure XP states now 1.7 waves (stable?). See the graph below. We could work the edge, as it is only the last few inches that si not throwing the light long enough. But we will probably leave the edge alone and work the rest of the mirror down to the edge.



Image above is after 2093 mins of figuring. After leaving for 2 hours the figure was 1 wave on the wave front.



Image above is after 2259 mins of figuring. After leaving for 2 hours the figure was ??? wave on the wave front. The figuring has been 2 steps forward and 1 3/4 steps back mostly. Somtimes going backwards. The middle zone not repsonding to the lap. We decided to use more pressure. Thsi roughened the middle and gave some shallow zones, but it had the effect of 'doing waht we wanted'.

Image above shows the 4 inch star lap being used on the 275mm zone near the end of figuring. The mirror is around 0.66 waves PV on the wavefornt at this time. Close up image of the 4 inch star lap



Image above is a ronchi image after 2349 minutes of figuring. There is some roughness in the middle, but the figure is getting close. Time for star testing.





Above graphs show the figure after 2454 minutes of figuring. The mirror is now getting smooother after using the 6 inch lap (soft pitch) and the 4 inch ot control the figure. See the spreadsheet for details.

Figure XP now suggests 0.411 PV waves on the wavefront. Star testing on the night of the 12th March 2009, showed that the images were much better, but that hardware mods are reuqired to the secondary kinematic mount as it was difficult to alighn. The Stars are smaller, but still not ideal. This could still be a collimation issue?

PROGRESS UPDATE 28/10/2010

We have hit a big snag! The mirror is just not performing well enough...:-(( Images are astigmatic and we cannot push the power up above a 40mm eyepiece without big problems. We can get good images in 'parts' of the field of view. But we need to understand what is going on before more costs are incured. We belive that it could be the strap? We have re-made the mirror support (27 point float) and that has improved images a little. But we will next follwo the advice below:-

Good article below. NOTE! item 3 on slings! One of ou=r major issues I think [] BUT...read this about sling missalignment also! Geoff....well spotted mate! [] And to top it all off. Check what this guy says about slings! He uses the piano wire method. We looked at this years ago and dismissed as MAD:- [] got above from links in:- []

So do do list:- More to follow..............
 * Check at zenith first NO edge support!
 * Then try the piano method at the edge
 * Then maybe silocone it?
 * Then last resort bell crank or pistons (lego technics type)?