- This topic has 17 replies, 3 voices, and was last updated 8 years, 8 months ago by
.
-
Posts
-
Hi everyone
I open this new thread to talk about the new project I have in mind to achieve.
The real motivation and proper that pushed me towards this project is the want me to challenge themselves in the creation of a secondary convex for a type Cassegrain telescope. The primary objective then is precisely to learn and verify my ability to realize this type of optical.
Despite the motivation that prompted me to face this new path was mainly the creation of a secondary, it seemed sensible, first of all, impress a complete project for a type Cassegrain telescope, in order to have the precise dimensional characteristics of the mirrors and thus have very specific results to be achieved.The dimensional calculations were performed using the ATMOS software and are reported below:
The choice of realizing a Cassegrain telescope is due to the fact that once the secondary realized I can, if in the future we do not have the time to realize the primary, buy one having conical constant k = -1 (then a common parabolic) having a focal length equal to 900mm. In this way, regardless of the diameter of the primary I could to have a working telescope (although maybe not optimized with respect to the obstruction).
While in the case acquisissi the ability to work a secondary convex and had the time to realize also the primary, I would steer in the construction of a Ritchey Chretien with the same specifications of the current project, in which they would change only the conical constants of the mirrors. I could then proceed in the realization of the primary and then retouch the secondary until it is brought to the shape necessary to properly mate with the primary.
The results of calculations suggest a secondary having a diameter equal 104.6 mm having a curvature radius equal to 881.25mm.
As it is suggested in the literature, for the realization of a convex secondary, you should work two mirrors having a (the mirror) a diameter equal to or slightly higher than indicated by calculations and an even slightly larger (tool) to minimize the problems easily occur at the edge.I am done with a waterjet cutting machine, from a 15mm thick slab, two discs of 115mm diameter, this because I have always enjoyed working two mirrors of equal size and cut because the disk still is over 10 mm of diameter greater than the value suggested by the formulas, granting me therefore a good margin to keep under control the problems at the edge.
who 10 millimeters more in the diameter of the secondary anyway, will not to unnecessarily increase the obstruction, because, as can be seen from the calculations in the previous post, the diameter of the baffle (Secondary) will be darkened 133 mm.Here are just mirrors before starting to work on them:
Great Mirco ! good luck for this new adventure, I'll be the first to follow it with interest, as you can imagine !
let two little tips on the parameters of the configuration Atmos:
1- if the decrease of 15mm fire extraction and ports in 260 mm, you are “simplify” the radius of curvature of the secondary and the primary-secondary distance, becoming measures without decimals.
2-if you stay on the Cassegrain “pure” the 30 mmm CPL there are excessive, the correct field is much smaller, I would suggest around 12 mm. In this way you will have a secondary smaller and an obstruction of the lower hood.Also it considers that a f9 is a right for a focal RC but it is too “thrust” for a pure Cassegrain, the coma would be important just out from the optical axis, but you can be rectified by reducing the diameter of the primary which does not alter, as you rightly said, the project parameters, but it allows you to have an overall focal least F12.
Also greetings from me.
I will learn something along the very demanding and this intriguing work.
I'm glad to know that each of you have a clear idea that they show the work in perspective.Hi Massimo, thanks great advice…
Answering a bit’ I would say to your clipboard:
1) It is true you're right, the numbers would be whole and not with decimals, but from a practical point of view having to do with a full measure or with decimals it does not change anything.2) I agree! however, keep in mind that my ultimate dream would be to get to R-C, then earthen these as values. If necessary to reduce the full light field could intervene on the internal baffles of the primary lens hood. Calculations in over, if imponessi a full range of light 15 mm, It would result in a lens hood of the secondary 115 mm, so big as the current mirror.
The actual machining I started the other day with chordal past and carborundum grit 80.
Slowly the excavation has begun to be generated with an estimated depth in about 3 tenths of a millimeter per hour.But what I have to dig around?
The depth to be achieved with the excavation it can be easily obtained using the following formulas:The table shows in the first column the formulas for calculating the radius of curvature of the surface (R) in case you know the depth of excavation (h) and the radius (r) which are placed 3 the legs spherometer, while in the second column shows the formulas for calculating the excavation depth (h) in case you know the radius of curvature of the surface (R) and the radius (r) which are placed 3 the legs spherometer, ray which can be understood as the total of the mirror if you want to know the total depth of excavation.
The table is divided in more than two rows, the first contains the exact formulas, while in the second the approximate formulas (h being generally much smaller than R).In the specific case the total depth of excavation turns out to be 1.845mm (using the exact formula) with a gap of less than 2 thousandths of a millimeter compared to the value that would be obtained using the approximate formula. Since the sensitivity of my measures will not let me in the least to get to the accuracy of the millimeter it is clear that use the approximate formula (easier) than the exact one is no problem.
At the moment I arrived to a 1.26mm arrow which is about the 60% of the total. The intention is to proceed with the grain 80 until the 90-95% the total depth and then move gradually to other grains to achieve the exact value, reversing in case the position between tool and mirror in the case to join him before finishing the sequence of abrasives.
…
From a theoretical point of view I agree, nothing changes, but I think in practice it is easier ( at least psychologically ) measuring a reference radius of curvature of 870 mm rather than 881,25. Similarly, being the Cassegrain ( in all its variants ) very susceptible to minor variations also on the distance between the mirrors, a measure “shoal” this distance 610 mm , (Mirror mirror on the center center) without decimal, It could help to position them more easily… but maybe it's just my “limit”, I know that you have a lot more engineers “feeling” me with numbers and are not intimidated certainly a few hundredths to be measured
ah, I had not noticed the next message…
As you can see it digging with Atmos, for simulating the concave mirror one newton with the same focal length and reading the “sagitta” reported
Eye to consider only the effective diameter net chamfer, works ( if not considered ) would increase the final depth.Then occurs in the vicinity of the final radius of curvature is best done as usual, by measuring the reflection on wet mirror.
Ah here, I did not know, Now I try immediately with the software…
You certainly have to consider the chamfer, if for measuring the excavation use a picture straight bar that leans from edge to edge. However, I do not I pose the problem having my spherometer legs arranged on a circumference of 45 mm radius, then smaller than the size of the mirror.Keep in mind that if I had the tools that allowed me to place the mirrors with an accuracy of 1 tenth of a millimeter would not change anything I put them in 600 mm distance between them or to 600.3 so always with the precision of 1 I could tenth of a millimeter to position them…
After another couple of hours of work have fallen to the share of 1.59 or 86 mm % of the total.
Little is it's time to move to the next grain ...
Worktop:
The machining of a convex mirror is not very different from the machining method of a parabolic primary mirror, simply invert role of what were previously considered to be the mirror and what was considered the tool. Now the disc of the two who will become convex mirror, while what will become concave will be the tool.
What will be the change of shape verification mode because all the classic optical test as Ronchi and Foucault can not be used on this type of surface (as convex).
What you can do, will analyze the interference fringes using an interferometer of the Newtonian type, placing one on the other, separated by a space light, mirror and tool. This will allow to trace the differences existing between the two surfaces.It will be necessary then, in this processing, polishing and bring in tolerance before the tool concave surface, which it can be tested with the classical methods and will become the reference surface, and then also the mirror surface that can be verified by comparing it with that of the tool reference.
The entire verification process and the construction of the interferometer will explain later to hand hand that the processing will proceed.
Hi everyone, work is proceeding… With grana 80 I got to 92% of the total arrow and are then passed to use the grain 220. The latest measurements indicate that they are in proximity to the achievement of total arrow, so that now, the difference between the measurement and the reference value to be reached is of the order of 1-2 hundredths of a millimeter, ie the order of magnitude of the measurement sensitivity of the comparator.
The images below refer to the measurement made on 'tool to the left (0.865 mm) , and the mirror on the right (0.860 mm).
Using the appropriate formulas is obtained:
R = tool 885 mm
Mirror R = 883 mmAt this point be able to precisely control the radius of curvature of the surface only with the comparator is no longer sufficient, necessary to verify directly the radius of curvature going to illuminate the mirror (after having wet the surface) and locate the point where it focuses the return laser beams, with the aid of a sheet of paper to detect the luminous diskette.
The method is simple, and it is very well exposed in this article written by Henry:
Determining the focal work in progress: a useful method in the phases of grinding
With this method I measured a value of the radius of curvature of 889 mm, or about 8 mm more than the value to be reached, which is great because I should be able to shave that amount with the rest of the grits ...
See you soon bye…
Wow, fast work ! This will be very interesting “logbook”, I do not believe there is a similar network in the processing of the secondary, or in any case is a rarity
One difference between the two rays is estimated that there is , In fact, it has to consider that the two surfaces are in contact via the interposition of abrasive which creates a small thickness, but not zero. With the transition to the next trouble this difference should diminish until it becomes almost zero in the process of polishing.
Ciao, thanks to Massimo. It is the work goes faster, because in these weeks are at home awaiting the start of the new work, then took advantage…
Say you are anxious to get to the polishing, because only then, with the first Ronchi test, I can realize if I worked well with machining with full diameter, Otherwise it'll have to go back to abrasives…
Check the difference between the radii of curvature of the mirror and tool, It was something that intrigued me too, but in the end I came to the conclusion that with measurement tools I have available I can not measure them accurately and then make a judgment…the radii of curvature that you see in the previous post are derived from measurements made, which I measure every time I dance more or less a penny (the value that I put is what I think is the average of all those found). The point is a hundredth of a millimeter of difference in the arrow to measure impact 10 mm on the radius of curvature. And being the difference between the radii of curvature of the two surfaces (I expect to find) less than the minimum difference that the sensitivity of the comparator allows me to estimate I can not give an answer to my curiosity…
Hi everyone, after finishing the session with grain 80 They moved gradually to the grits 220 400 600 ed 800…And at this point it is time to polishing:
First I made the two full-diameter tools into pitch, one for the mirror, and one for the concave glass, I have to say being able to peck the right hardness at the first attempt ...
During processing just you feel that the glass bites the pitch and forming the white froth cerium oxide and is heard that ssssssss sssssss typical of when things are going the right way ...As stated from the beginning, My intention was to be able to do even polishing machine with full-diameter tools but after a short polishing session, just enough to start pondering the surface, I made the first Ronchi test set noting that the configuration of the machine could not get a nice spherical surface. So I did several tests, but I've never found one that could generate a nice smooth surface. I noticed that it would be necessary to use a smaller tool type 75-85% the diameter in order to obtain a good ball (as I got to experience in working with 400 f4.8).
Anyway I made a good portion of polishing machine and now I am continuing, until the gloss, Hand with classic races 1/3 COC in order to fix the shape and get to the ball.Tomorrow if I have a bit 'of time I hope to post some pictures of the tests.
To build and test a convex optical surface is necessary to proceed in a somewhat 'different from how we proceed to realize for example the primary of a newton. The entire process for manufacturing, for reasons of time, no I describe it here, especially since I have already done magnificently in this article Giulio:
In summary then, it will be necessary to give the tool concave surface of the same curvature and conic constant of the mirror that you intend to accomplish, as this will serve as a reference against which will be evaluated that the convex mirror, through interference fringes with an interferometer of the Newtonian type.
This type of interferometer is extremely simple to implement and it will describe the construction later, as soon as I have all the necessary components. The actually fringes can see already putting just under a neon lamp, but it serves a type of monochromatic illumination for not having interference fringes of different colors, for this it is necessary to build the equipment specially.
- You must be logged in to reply to this topic.