Early Rider training wheels bracket

Allows installing cheap training wheels  (aliexpress) to Early Rider 14 and possibly larger models. 

The brackets mount on the frame pipes near the rear wheel. You will also need:

• 4 x M4 hex nuts
• 4 x M4 screws
  • Minimum screw length is 30mm, but if you use 35 mm or longer ones you can use the screw to force the captive nuts into their pockets instead of pressing them in, and they don't protrude in an obstructive way. 
  • I used ISO-7380-2 screws (BHCS with builtin washer), but I feel the force needed for installing the nuts was close to socket limits, and I would have used DIN 912 (SHCS) + a washer if I've done it again. 
• 2 x pair of “internal” + “external” bracket  - one each side of the bike. 
  • Brackets have been designed to be printed with the large flat face facing the bed (part with captive nuts - the pocket gets a bridge - don't use supports). 
  • Print 100% solid. I used PETG for its conformity, and under assumption that my bracket modelling isn't perfect. 
• 2 x spacer - its function is to go in the cutout in the bracket, and provide support for the washer. You want the washer to equally press the metal and the spacer - fine  tune the spacer thickness if necessary.

To choose the angle - I found 12.5 mm worked ok, albeit the angle on both sides of the bike was slightly different. Since this is a welded frame, not precision machined metal*, yours is likely going to be different. The practical way to choose the angle that I found was to first install the eyeballed angle bracket, lay the bike on the side, and use digital protractor app with a phone referencing the wheel plane while being pointed in a constant-ish direction.  The photo shows bracket with 10 mm angle, which was a bit too little.

I attached parametric CAD  so you can tune the angle yourself. Note the spacer is only modelled in the straight.f3d, but it is the same in both files.

The wheels are raised a bit from the ground. This was partially due to long have they came, and honestly I am not sure about it. To a degree I consider this as a feature, as the rider has some extent of freedom in balancing the bike. The design lacks adjustability in this regard. To make it adjustable I would likely get wheels for bigger bike and drill equally spaced holes in them.

They tend to do the job, but note the disclaimer - the user is still very careful and has low mileage with the bike, so they didn't get much beating yet. That said, I tried to design them in such a way that the major forces go perpendicular to the layer lines, and that tensile stress is counteracted by the compression resistance on the opposite side of the bracket  - or at least so I think. I.e. the screws do most of the holding, and with M4 bolts holding power being in range of several kilonewtons, I am personally happy with the risk and seeing the results was super satisfactory. 

*) Note, there is a machine surface towards the the rear wheel is actually machined and it was another point I considered to attaching the wheels. I concluded this was too invasive, harder to manage the forces, and would also result in the wheels being slightly lower.