Hi Paul,
I took the smaller dimensions with a caliper which only shows [mm].
No problem. My expectation is that the as-designed dimensions from DEC were in fractional Imperial units so in my reverse engineering I convert your measurements back to multiples of 1/16" as, I think, a reasonable basis for cross-checking sums and differences in the various directions.
in case you want to fabricate the panel for the BA11-F box, I can make a simple drawing (not CAD, I never learned that) with dimensions It will take a bit longer, because the drawing is done at work and I only work on Tuesday and Thursday. That will probably be more clear than just the dimensions given in my previous reply.
If my sums and differences above seem wrong to you then please do put together a simple drawing pointing out where I am misunderstanding!
I will try to check how snugly the panel sits against the module. The handles of the hex modules are visible, the maroon handles of the smaller module may be a bit tricky, but I can check how far the maroon handles "protrude", compared to the metal bar of the hex modules. As the cover is made of aluminium, you can push it in the middle, so I can get an idea how close the panel is to the hex modules, and thus the plastic handles of the maroon modules. Putting the cover on the box never did give me the impression that the cover "touches" the modules. The quarter-turn fasteners does not move the cover in any way "closer", because the (vertical) panel edges already rest on the metal box.
I think that you've already considered the practical consequences here, but see my analysis below. My attempt to come up with a dimension in that direction must be incorrect -- perhaps the reference module that I am using is a bit tall? I'll experiment more as this dimension affects how the upper/lower lips are designed given SendCutSend bend constraints.
Yes, the panel is almost symmetrical with a tolerance of say 0.5 mm (neglectible). Thus i can be mounted in one way or 180° rotated.
BTW, you wrote "I've fiddled a bit with working up a CAD model for the 11/20 covers" ... yes, when I bend a metal plate, I always worry whether I got it right. I know, you have to account for the thickness of the plate, but still ...
SendCutSend has detailed instructions and a calculator (
https://sendcutsend.com/bending-calculator/), starting from specific characteristics of the sheet material chosen which produces quite precise modified dimensions ... hopefully they are also
_accurate_ :-}. They specify +/- .005" tolerance on the cuts; for other tolerances see:
https://sendcutsend.com/faq/what-are-your-tolerances-for-bending/
For aluminum (
https://sendcutsend.com/materials/5052-aluminum/) I'd design using either the 0.063" (1.60mm) sheet, slightly thinner than your measurement, or 0.080" (2.03mm). Based on some other experimentation these two choices might actually price out the same or quite closely. All things considered I'm inclined towards the thicker sheet to improve rigidity where the fasteners are attached.
Their bending brake requires a minimum of 0.25" (6.35mm) on tabs, which is more than the 4.7mm = 0.185" that seems to be your measured dimension for the upper/lower lips. In my limited experience with bending brakes BITD I don't see how such short upper/lower lips would have been formed without an additional step to cut off excess lip. I'm therefore contemplating treating the upper/lower lips as being double-bent but with minimal flanges of 0.25" (6.35mm) -- like those at the front/rear but narrower -- that would reach towards the edge of the slot-numbering sticker.
One could bend the lips *up* rather than down, leaving an air gap but still creating the stiffener, and accepting that the stiffener will protrude outwards by 0.25" (6.35mm). AFAICS that doesn't result in a clearance conflict with the rack posts although it's a bit close; I measure that gap as 7/16" (0.4375"). Having only 3/16" of clearance is probably not the best design decision in this location.
Rather than a double-bend one could simply angle the minimum 0.25" (6.35mm) tab at, say, 30 or 45 degrees outwards ... which would be a bit less stiffening but IMO stiffness isn't a critical consideration here.
If getting the inside of the panel closer to the module handles is important then to me this seems to be the best approach to managing the minimum tab dimension requirement at the top/bottom edge of the panel. What do you think?
Or one could accept the slightly higher 1/4" vs. 3/16" dimension; I guess some "push in" distance is acceptable here (you state "
As the cover is made of aluminium, you can push it in the middle, so I can get an idea how close the panel is to the hex modules, and thus the plastic handles of the maroon modules.").
I can't see any real harm in a potentially extra 1/16" of clearance between the inside-panel and handle-tops. It's insufficient to allow for modules to become misaligned and it doesn't significantly affect the clearance to the rack posts.
AFAICS the closely-spaced double bend at the rear/front of the panel is allowable but I can't be sure of that without submitting a completed design and seeing whether errors are raised.
I assume that the panel is completely symmetrical.