Fused deposition modeling
3D printers consume plastic filament,
from which they are making the printouts. The filament typically comes on spools or reels,
which have to be held in place in a way to facilitate free, smooth rotation to unwind
the filament with minimal mechanical resistance.
There are many filament holder designs out there. None satisfying was found; they are usually too large, consume too much material to make and take too much time to print. A hybrid approach was chosen, with steel bolts and threaded rods for the large parts, with only the fairly small couplers for the custom-made parts.
All printing was done with 0.4mm layer thickness.
The spools come with a wide range of central holes. The most common inner diameters are
31-32 mm, 51-53 mm, and 74 mm. See the list here
.
A two-part spool holder was designed to facilitate clamping on both the 51 and 74 mm spools, the choice was based on the two available spools.
The clamps were designed around a 6282 ball bearing, with inner diameter of 8 mm, outer
diameter of 24 mm, and height of 8 mm. This facilitates smooth rotation around a 8 mm pin,
which can be a readily available M8 bolt. The top part of the hole is conical, to create
the narrowing to hold the bearing in place while keeping the overhang angle in check to avoid
need for support material.
Originally the clamps were intended to be a set of stacked disks. First printout attempt however failed due to failure to bed adhesion, and seizure of filament that forced abortion of the printing. The printing would also take over 1.5 hour and consume too much of filament with too little additional effect.
 Rendered spool clamp, old |  Rendered spool clamp, old |  Failed partial printout |
The decision was made to lighten up the design. Cylinders of unneeded material were removed, getting rid of two thirds of the material and resulting in about a half-hour print time.
 Rendered spool clamp |  Rendered spool clamp |  Rendered spool clamp |
 Spool holder |  Spool holder |  Spool holder |  Spool holder |
 Spool holder with bearing |  Spool holder with bearing |  Spool holder with bearing |  Spool holders |
 Spool holder in spool |  Spool holder in spool |  Spool holder in spool |  Spool holder in spool |
 Spool holder in spool |  Spool holder in spool |
This led to some mechanical weakening that was deemed acceptable. After clamping the spool with a wing nut, however, two pads on one of the clamps delaminated with a telltale crack sound. They however stood in place and maintained functionality, just were springing away.
 Delamination |
The repair involved opening the gaps and filling them with 5-minute epoxy. Remaining epoxy was smeared on the faces and edges of the other pads, to strengthen them.
A variant of the repair that was not used involved a countersunk-head M3 screw in each of the pads, which would hold the layers in compression and prevent delamination.
A possible improvement could be an integration of a hole for the bolt. The hole itself could act as a strengthening mechanism, due to its walls.
A new generation of the spool clamp was designed on a whim. The OpenSCAD file now supports arbitrary number of lightening holes (and pads) with presets for 3, 4, and 5. The parametrization is more complete now.
Holes for M3 screws were added. There are three possibilities - countersinks on the spool size (with nuts to be installed to the outer side), countersinks on the outer size (to rely on threads cut in the plastic), and no countersinks (to use flat heads and threads in plastic, or alternatively to rely just on the strengthening effect of the hole walls).
The M3 thread does not hold entirely well in the PLA, especially in shorter lengths. Using nuts may be more reliable. Be aware of the need to keep the spool-touching areas flat and use countersunk screws there.
 Rendered spool clamp, 3-pad |  Rendered spool clamp, 3-pad |  Rendered spool clamp, 3-pad |
 3pad clamp, fresh printout |  3pad clamp, fresh printout |  3pad clamp, fresh printout |
 3pad clamp, thread cutting |  3pad clamp, thread cutting |
 3pad clamp, screws |  3pad clamp, screws |  3pad clamp, screws |
 3pad clamp, bearing |  3pad clamp, bearing |
 3pad clamp, installed |  3pad clamp, installed |  3pad clamp, installed |
The stand needs long protruding parts for mechanical stabilization. The choices available included a length of steel 10x10 square tube, cut and welded or held with a coupler; steel bolts, held with a coupler; or threaded rods. The steel bolt variant was chosen. Long M10 bolts were chosen on the basis of easy availability; long M8 ones would be sufficient but there was not enough of them on hand.
As the load was asymmetrical, the legs were designed to protrude to left, right, and in 45 degree angles. Three-leg variant was considered but decided against due to possible lateral stability issues.
The shapes were designed from cylinders and bars. The holes were dimensioned for M10 thread, which was later manually cut with a tap.
 Rendered bottom coupler |  Rendered bottom coupler |
 Bottom coupler |  Bottom coupler |  Bottom coupler |  Bottom coupler |
 Bottom coupler |  Bottom coupler, thread tapping |  Bottom coupler, thread tapping |  Bottom assembly |
 Bottom assembly |  Bottom assembly |
The top part requires coupling of a vertical M10 threaded rod or a bolt to a horizontal M8 one. A simple assembly with two perpendicular, somewhat offset holes was designed. Layer orientation was chosen so in case of need a pair of large washers on the M10 rod can hold them in compression.
No thread was intended for the part, external nuts were used for facilitating compression of the part.
 Rendered top coupler |  Rendered top coupler |
 Top coupler |  Top coupler |  Top coupler |
The parts were assembled without a major problem. A minor delamination was observed at the bottom coupler during thread cutting, which did not impair functionality and was too small to photograph.
A length of M10 threaded rod was used for the vertical stand. It was cut about 3 centimeters too short, due to mistaking a M12 and M10 rods for M10 and M8. The M8 one turned out to not be in stock so a bolt had to be used.
The available M8 bolt was too long. A M10 coupling nut was used as a spacer. It turned out that this facilitates carrying. A carrying handle attached at this side may be benefical.
 Stand assembly |  Stand assembly |  Stand assembly |  Stand assembly |
 Stand assembly |  Stand assembly |  Stand assembly |
The spool was attached to the stand using a M8 wingnut. Both spools rotate freely. The clamp "teeth" are a few millimeters too small, so some reasonable clamping force is used to avoid the spool sliding over them and rattling during rotation. This is where the delamination (and subsequent epoxying) occured.
 Spool on the stand |  Spool on the stand |  Spool on the stand |  Spool on the stand |
 Spool on the stand |  Spool on the stand |  Spool on the stand |
The device functions flawlessly. The four legged stand is a little wobbly but this does not show during operation.
- Filament clamp, old non-lightened version
- Filament clamp, nonstrengthened, 4-pad
- Filament clamp, new strengthened version, parametric, preconfigured for 3-pads
- Bottom coupler
- Top coupler
- Filament clamp, nonstrengthened, 4-pads
- Filament clamp, strengthened, 3-pads
- Top coupler
- Bottom coupler