Since I will not be able to attend the Monterey meetings, I just want to reiterate the position I have voiced in the past, that it is important that we start "freezing" sections of the draft standard so that we don't keep rehashing the same territory endless times. I hope that recommendations to this effect will come out of the Monteey meeting. I realize in saying this that fundamental to the ANSI way of developing standards is successive review at various levels from TF all the way to public review. Our task force is small enough that we should be able to reach concensus and move this document up the chain. I wish I could have been more involved and a more timely contributor. Thie draft is very good - the task force is getting close!
Perhaps the solution is to limit future reviews to show-stopper comments only. From my perspective, the following are 3 critical issues:
1. My reading of the draft is that we are "binning" the scratch and dig specifications, so that each scratch or dig number essentially is a 1-sided tolerance limit? I.e., for a scratch that appears worse than a #40 but not as bad as the #60, it is classified as a #60, for both acceptance to the limit and for accumulation calculations.
Do we need the same "binning" decision for both the visibility method and the dimensional method? MIL-F-48616, for example, includes minimum-relevant scratch and dig values that do not appear in the lists of maximum-relevant values, and the use of interferometers, microscopes, etc seems to imply an accuracy for dimensional specification inconsistent with binning.
If binning is the decision, is the designer limited to the choices in tables 1-4? If yes, this would clear up many of the comments that have been made by earlier reviewers. Wording in the paragraphs of section 3.1.2 that mention specifications not appearing in the tables should then be eliminated, because they cannot happen. There has obviously been a lot of effort made to clarify this, but to me, it is still not explicit.
Having said all of this, I also want to remark that the companies I have worked for have binned consistently.
2. Section 6.3 prohibits area imperfections on coated surfaces. This is a tightening of requirements beyond the three MIL specs said to be the models for OP1.002. MIL-PRF-13830B, MIL-F-48616, and MIL-C-48497 all allow area imperfections on coated surfaces in limited circumstances. This tightening will impede acceptance of our document.
3. The inclusion of 4X magnification in the definition of "Unaided Eye" in 9.1.1 is inconsistent with my experience, both at the companies I worked for and every customer with whom the issue came up. Our experience is that 4X magnification causes significant improvements in visibilith and corresponding increases in rejection rates. The available comparison artifacts cannot be relied on the scale the same way as scratches on optical surfaces, and the result is worse correlation.
The following comments either cover lesser technical issues or are editorial in nature:
3.1.1.3 replace "she" with "the engineer"
3.1.1.5 if "classified according to the grade levels" measn that classification can be only the values found in the tables, then many issues disappear. Anything judged to be less visible than a #10, for example, becomes a #10.
3.1.3.2 In the last paragraph, suggest adding to produce "they can be useful but are not required..."
3.1.4.1.3.1 suggest adding to produce "Only relevant scratches in accordance with table 3 are graded...". Also, the second sentence in comment 21 at the bottom of the page would be a useful addition to this clause.
3.1.4.1.3.3 "S" is in the legend but not used in the equation.
3.1.4.1.3.4 It is not necessary to reference Test 2 in this paragraph. Any smaller grade scratches that violate or come close to violating Test 2 will fail the equation here sooner.
3.1.4.1.4 What if the tolerance is "5"?
3.1.4.2.3 I suggest indication this is limited to the clear aperture.
3.1.4.2.3.3 I believe it requires a careful reading to understand what constitutes a tolerance grade (range from the previous tolerance grade, not the difference between the maximum and minimum relevant limits for the grade). An example might help (for grade C, any dig greated than 100 and not exceeding 200).
4.1 It would be useful to finally clarify that the 135 degree angle is measured as the angle containing the optical material and not one of the outside angles. I've lost track of how many times I have been asked which way to measure this by cusomers, suppliers, or our own inspectors.
4.3.2 After stoning, the resultant imperfection should meet either the chip spec or the dig spec (or both) depending on location.
4.4.4 and 4.4.5 add "(see also 4.2.1)".
7.2 to be consistent, add "when peering through the component".
8.1.6 To the end of the last sentence, add "when 3.1.3 specs apply".
9.1.1 Consistent with mu comments above, delete "or a simple microscope (magnifier or loupe) of 4x or less".
9.7.1 What does 2) mean?
9.7.6 "Chamfer" is a toleranced feature in 13830, 3.3.1, and in this draft in 4.1. Many companies avoid the term entirely because it is used inconsistently throughout the literature. I suggest staying consistent with 13830.
9.10.3 I disagree with this definition, and suggest it be the definition of "cement separation". "Feathering" is a special case, and tends to be very inflammatory. the illustration in our draft of cement separation is not an illustration of feathering, either.
9.11.2 I disagree with this definition of sleek. To me, sleeks are fundamentally narrow. I have seen a number of instances of wide (> .5mm) scratch-like imperfections on surfaces that have smooth profiles, especially on but not limited to plastic optics, and nobody would classify them as sleeks. This definition should be limited to very narrow imperfections.
9.12 I have seen cases where poor control has resulted in localized, scratch-like (or dig-like) coating blemishes where only some layers of a multilayer coating are missing. the result is typically a dramatic difference in color - a green "scratch" on a blue surface, or the like. These are consistently evaluated to the scratch and dig requirements for the coating. We have always classified these as a type of void. Also, the term "void" is a common term in the optics community, and really should be included in our document, possibly as an "also called" clause.
9.13.1 This definition of chip is too restrictive. Conchoidal shapes are only one type of chip and the wording seems to imply that chips are inherently conchoidal.
Annex C The first two paragraphs should be combined.