Hexagonal Mirror Arrays (focused on a single point in space)

thingiverse

CAUTION: MAY CAUSE EYE DAMAGE OR FIRE. USE AT YOUR OWN RISK. ALSO THE FOCAL LENGTH IS LONGER THAN ORIGINALLY ANTICIPATED DUE TO SCALING AND A CURRENTLY UNKNOWN REASON (UNTESTED BUT PERHAPS UP TO TWICE AS LONG AS INDICATED...) Update 5: 12/15/2021 WARNING: I have not tested this design but be aware that the focal length was originally computed at 200mm but then the stl file was scaled up to 150% scale to fit 1" mirrors so the focal length then became 300mm due to the scaling. After testing a different design, the focal length may actually even be closer to around 400mm (I haven't tested this particular design or looked into exactly why the focal length is longer than was anticipated). Just be aware that the focal length is longer, perhaps approximately double) the 200mm originally anticipated! DO NOT DAMAGE YOUR EYES! Even just the scattered concentrated light from an object at the focal point with only a few mirrors could be damaging to your eyes! Update 4: 12/12/2021 There is a different large array I made here that I have been focused on: https://www.thingiverse.com/thing:5162272 Don't burn yourself or damage your vision as using this with mirrors is VERY DANGEROUS! Keep your eyes away from the focal point to avoid blindness and be careful that it is stored out of sunlight. USE AT YOUR OWN RISK! Update 3: 12/11/2021 I revisited the Jupyter notebook code to optimize for printing and to change the focus to 200mm. I also got rid of the 10.8 degree angle I introduced earlier. The center mirror is no longer recessed lower than the others as well. The latest file to fit directly on an Ender 3/5 bed which has a focal point 200mm or approximately 8 inches above the center of the array: "scaled_hex_prism_37_grid_single_focus_200mm.stl" except it is 300mm because of STL scaling or 150% and likely even higher than that due to another currently unknown reason. I also posted a larger 169 mirror array that could be divided to fit on an Ender 3/5 bed or print directly if you can print ~393mmx393mm. The stl is called: "scaled_hex_prism_169_grid_single_focus_200mm.stl" I generated these files from Jupyter notebook and then scaled to 150% in Fusion 360 to fit one inch hexagonal mirrors (measured point to point). I did cut the 169 mirror array into a square which has 69 full mirrors to maximize the print area on an Ender 3/5 print bed. The file is called: scaled_square_69_grid_single_focus_200mm.stl . I created it by chopping up "scaled_hex_prism_169_grid_single_focus_200mm.stl" (+/-103.924mm from center) I also added a 61 mirror array which is just slightly too large to fit on an Ender 3/Ender5 bed so I cut two mirrors off the tips of the hexagonal array to make it fit, turning it into a 59 mirror array. The area where the mirrors were could also be used for mounting tabs. These files are called "hex_prism_61_grid_single_focus_200mm.stl" and "hex_prism_59_grid_single_focus_200mm.stl" If I was going to print one next, I would have likely printed "hex_prism_59_grid_single_focus_200mm.stl" although I played around and removed the center mirror so there is now a 58 mirror version and also a 52 mirror version with the seven center mirrors removed (getting carried away) some with tabs for mounting and some without. hex_prism_52_tabs_no_center.stl hex_prism_58_tabs_no_center_fixed.stl hex_prism_58_tabs_no_center.stl hex_prism_58_no_center.stl hex_prism_59_no_tabs.stl So I will need to decide which one to print. I think I might print the 58 mirror version without the center mirror and with tabs for mounting. "hex_prism_58_tabs_no_center.stl". I like the idea of running an ~8" pipe up through the center of the array to the focal point to mount something to heat on as well as help with alignment as the sun will shine through the long hollow metal pipe if it is exactly aligned or perhaps a rod that would cast a shadow if it was not properly aligned. I also added "hex_prism_1.stl" and "hex_prism_7_200mm.stl" which could be added into the ring optionally. The 7 mirrors together could even be printed at a lower layer height for better mirror position accuracy without printing the entire design at the lower layer height (saving print time). There is no tolerance built into the 1 or 7 mirror files so you may want to scale to something like 99% so it fits inside the ring. When the array is aligned. these inner mirrors will be in the shadow of the object in the focal point so perhaps it makes some sense to get rid of them to save print time (or even shift the focal point from above the array by revisiting the code). There is also a hex_prism_7_200mm.stl which could be printed at a lower layer height and still leave a hole in the center. -- Update 2: 12/10/2021 I printed a sliver of the wedge (to cut down print time) with only eight mirrors at 0.28 layer height and 20% infill. I superglued eight mirrors without worrying about alignment and tested as the sun was setting. I confirmed that the focal point is 80mm (3.14"), not 800 mm as one of my pics incorrectly indicate. I also set it up at 10.8 degrees directly under a recessed light. It seems to focus! I called the file "wedge_8_mirror_sliver_test.stl". I had the most trouble with the end mirror furthest away from the focal point and are the most angled. From this experiment, I ended up creating "Mirrored_Triangle_22_80mm" which is what I will print next. There are only 22 mirrors but should be able to be combined with up to 5 more triangles into a larger 110 mirror array (update: I'm not certain yet but there may be an error if you want to assemble multiple triangles, as there could be base perimeter interference on adjacent triangle pieces since each triangle perimeter has a full width base I need to consider this possible issue and how best to fix any base interference with adjacent pieces of a larger array) -- Update 1: 12/9/2021: I realized that the scale of the models that focus on a single point that are generated via the Jupyter Notebook design are too small for the 1" mirrors (unlike the heart stl which appear to be scaled correctly for 1" diagonal hex mirrors in the Jupyter notebook). I verified using Fusion 360 that these single focal point 3D models must be scaled up to 150% in the X,Y, and Z axes to match the scale of the heart stl which uses 1" diagonal hex mirrors. -- These stl files are generated from the amazing project located at: https://github.com/bencbartlett/3D-printed-mirror-array using the author's Jupyter notebook design. I have not printed anything yet myself. BEWARE: VERY DANGEROUS AND VERY EASY TO DAMAGE YOUR EYES!