Railcore II Kinematic Bed Mounting

thingiverse

##########\r\n(Update July 12th 2019)\r\nI've made some updates to this design, including pre-tensioning using springs stretched between the heads of screws on the platform-side and the yoke-side of the mount. The new files are prefixed MK2.\r\n\r\nThe ball mount is different for the left and right sides, to accommodate the cables for the heated bed on the right side.\r\n\r\nI've chosen springs that will be quite forgiving, but I used one of these on each left-hand mount:\r\nhttps://www.mcmaster.com/9044K194\r\nAnd two of these on the right hand mount:\r\nhttps://www.mcmaster.com/9044K178\r\n\r\nI included holes in the ball mounts, to allow you to thread long M3 screws through to stiffen the plastic. This seems to work quite well, but you can probably get by without them.\r\n\r\nThe photos and renderings should make it obvious how to put this together. Post in the comments or message me if you have any questions.\r\n###########\r\n\r\nI designed this kinematic coupling system to overcome the risk of z-axis binding on the Railcore printer.\r\n\r\nBOM (for Maxwell configuration with spring pretension):\r\nThe printed parts\r\n3x M3x30mm screws for the yoke mounts\r\n7x M3x ~16mm for the spring mounts\r\n6x M3 nuts & optional washers\r\n3x 8mm Steel ball bearings (Amazon link below)\r\n6x 8mm long, 6mm OD M3 threaded sleeve rods (Amazon link below)\r\n2x McMaster Carr Spring 9044K194\r\n2x McMaster Carr Spring 9044K178\r\n\r\nOptional stiffening screws:\r\n2x M3x40 For right-hand bed mount\r\n4x M3x25 For left-hand bed mounts\r\n\r\nThe kinematic coupling avoids over constraint of the bed and allows the z-axis to move freely regardless of any misalignment in the z-rails.\r\n\r\nUsing the Kelvin kinematic coupling arrangement allows the majority of positional control to come from the front left rail, which is easiest to align. The back left rail will have reduced effect on bed position, and the right hand rail will have almost no impact on the bed position, other than keeping it level.\r\n\r\nIt's also possible to use the same parts to generate a Maxwell coupling, which will average out the alignments of each rail. I don't see much benefit from that in this application, unless vibration from the printer being passed through to the bed proves to be a problem. The Kelvin configuration might be more susceptible to hard accelerations in the y-axis, compared to the Maxwell.\r\n\r\nThis is actually a modified version of the Kelvin coupling, since I have used two crossed cylinders for the single point contact position, rather than a ball and a flat. This has no effect on the functionality, but allows us to use only two types of component.\r\n\r\nThis is still a work in progress, I've run several prints without problems, but my Railcore is new so I don't have much before and after data, and I'd really appreciate feedback.\r\n\r\nA word of warning, with the current design, the bed is just resting on the mounts, a significant lateral force could knock the bed off, and cause damage. Take extra care when removing prints, or working on the printer. I suggest lifting the bed off and resting it in the bottom of the printer when not in use. You'll also need to reduce your max z-height by ~10mm.\r\n\r\nAlso note, the upper parts show how it goes together, the rods can be screwed in, the balls just pop-fit for me.\r\n\r\nParts:\r\nUxcell a16050300ux0563 M3 304 Stainless Steel Threaded Sleeve Rod Bar Stud Round Connector Nuts 10pcs https://www.amazon.com/dp/B01HSTUZW8/ref=cm_sw_r_cp_apa_i_AwRgDbDY8JVKA\r\n\r\n100 8Mm Chrome Steel Bearing Balls https://www.amazon.com/dp/B008M3584K/ref=cm_sw_r_cp_apa_i_LiVgDbXHMNW6C\r\n\r\nReading on kinematic couplings:\r\nhttps://en.wikipedia.org/wiki/Kinematic_coupling\r\n\r\nhttps://www.precisionballs.com/KINEMATIC_ENCYCLOPEDIA.php#k1