Air Raid Siren

thingiverse

Human: I submit r3 versions for MakeItLoud\nThe sound level is too loud for iPhone microphone--it's over 94dB at 2 meters, also measuring 94dB at 1m. The microphone is surely swamped. You can add multiple sirens to double or quadruple the sound level easily (adding 3dB or 6dB more). But it gets harder to crank without binding up.\nThe printer is a little picky about printing quality. I suggest using 0.3 mm layers and medium quality. Some compensation for inaccurate printing has been built in (I use Afinia, which commonly prints 0.07mm too large). It might take a few tries to get it just right. If the .STL files make the crank too sloppy, let me know and I'll create new ones with lower 'undercut' when possible.\nDue to my error in reading the due date (thought it was Dec 16 until today), I will be short on try/fix iterations of what I'd like. My first misprint had the siren 16mm smaller than the crank, requiring me to drill new holes for it to attach. The second rev was too large for the contest rules by the same 16mm.\nTheory of operation: A squirrel cage fan (rotor) has 60 blades (50 in r1). The stator outside it has 59 blades (49 in r1). As the blades line up on each side, air flows and is blocked on the opposite side. This happens approximately 1500 times per second. The wavelength of 1500Hz is about one circumference (using approximations; might have the frequency slightly off). This corresponds to 2.5-3 cranks on the handle each second and 25-30 turns of the rotor each second.\nRev 2 has been added. It fixes a size mis-match between the siren and crank, designed for a higher frequency but lower crank speed (about 3/sec should produce a howling sound). Rev 2 eliminates the need for supports on the siren-rotor; nothing else requires supports. So this version is highly printable.\nRev 3 has been added. This revision only conforms to the 10x10x12.5cm print area rule for the MakeItLoud contest (I'm glad I noticed that in time!).\nPrint Settings\nPrinter Brand:\tAfinia\t\nPrinter: AFINIA H480 \t\nRafts:\tDoesn't Matter\t\nSupports:\tDoesn't Matter \t\nResolution: \t0.30mm layers; medium quality \t\nInfill: loose (medium)\t\nNotes:\tEdit 'Print = 0', to get each of the parts. (see notes on the same line).\nVariables have been created to account for poor print dimensions. You might need to adjust 'undercut' and 'backlash'.\nRafts are only needed for Siren-Rotor. I experienced warping on Siren-Stator, but it didn't harm anything.\n45-degree overhang is necessary for Siren-Rotor. The .scad file contains a variable to change this if required. Some drooping filament occurred, but wasn't a significant issue.\nI estimated the fastest spindle speed at 20 rev/sec (120rpm). That's using a 10X planetary gear set and taking into account two planets' rotation per turn of the output shaft (Crank.scad).\nA target frequency of 1kHz led to estimating 50 blades, resulting in a circumference of 0.363m or radius of r=58mm. I intended for it to circle the siren by mis-matching the blade count, using 49 for rotor and 48 for stator (instead of standard r-1 or r+1 relationship). Because of an error while copying numbers, I used a smaller circumference than required. Rev 2 switched to using 60 and 59 blades, achieving 1200Hz frequency at 20rev/sec.\nThe revised siren now resonates around 2000Hz instead of the intended 1000Hz (circumference being r=219mm is too short), making it difficult to reach without hitting a harmonic but capable of producing an ear-piercing howl.