THWACKINATOR: An expanding, hammer based, battle top with a ripcord and geared launcher system

No supportsOVERVIEW

It's pretty widely known that expanding the diameter of a spinning object lowers its speed. But, by having my arms expand while it is finishing its acceleration, I aim to allow it to spin up to a higher inertia than traditional launcher designs, and not lose the speed, by triggering the hammer expansion several rotations before the rotation acceleration finishes. I do not expect it to be the most elegant, the clear battle winner, or the prettiest. It's my attempt at a refined combination of those, but I do hope it holds its own.

Everything here is 100% 3D printed, including the launcher's axles and assembly pins. Zero metal, zero glue, zero hardware.

INITIAL DESIGN

This spinning top is a collection of different engineering mechanics. I did not design the individual elements myself. Rather, this is an adapted version of those, into a fresh combination, for this top. The spin out hammer arms were inspired by chain link carnival rides. The hinged launching mechanism by childhood toys. The little deflect square by simple billiard physics. And, thankfully, there is a ton of public documentation on tops: spinning, leverage, and shifting COM to work with for this, so I didn't have to re-derive any equations. I just calculated folded and deployed state by plugging in the numbers with a simple script, using those formulas.

MECHANISMS

1. Load folded. Arms folded in, radially. The little slots, in the top of the hammer pivot rods, lock into the gear in the launcher.

2. Rip. As you pull, the tapered rip cord first lifts the hammer gear, releasing them to expand. Then it lifts the whole hammer gear, knocking the hammer into the bottom of the launcher, sending it flying.

3. LOCKED BATTLE STANCE. Each arm hits the little jutting square, which my testing (actually way too many times I just flicked the hammer by hand) showed greatly increased its locking in, at the end there. The little lock channel can, and should, be improved with an exacto knife, but do not cut the little protrusion at the top.

4. 

   

   

5. Re-arm. Load the ripcord into the launcher first. That sets it back to launch position. Then, insert the spinner. Grip the launcher firmly with one hand, pinched from top and bottom, so that your hand is free of the expanding arms. Then, well, let it rip!

NUMBERS

Note: numbers are rough, done off the CAD model, with theoretical print density applied. I do not have a system to accurately measure the model perfectly, and doing a water bath to get the mass seemed overkill :)

DESIGN CHOICES

• Five hammers The same total rim mass split five ways trades a few deep hits for many shallow ones.

• Deploy is against spin direction, deliberately. Head on impacts torque each arm into a wall, not out of it.

• Strike faces sit square to motion at deploy.

• Blunt rim, no spikes. Sharp features chip PETG on the first exchange and add grabby recoil.

• The lock is stability, not just mechanism. Balance rides on five joints seating solidly. This is admittedly the weakest part of my design. And can fail on launch. Cleaning the locks with a sharp blade is highly encouraged; the right angle corner catches.

BUILDING PROCESS

It took about two weeks, and around 40 hours by my time tracker. But it was a great learning experience. I already had rough gear clearances from previous projects, as well as the pivot tolerances. Shaping the initial spinner model took ~12 hours. After that, it was about ~20 of refinement, test prints, shaping, redoing the math, for spinner and launcher. The launcher system took another ~6 hours, mainly the pivot system, to get to the refinement point. The gears were just expanded from ones I already used, same with the couplings and bushings. But the low stress, U bracket shaped tilt mechanism solves a lot of the tension issues you get with 3d prints, and that was brand new for this. And it definitely loses some energy with the force release mechanism, but I chose simplicity over further reduction.

Hammer Progression:

Different Attempts (Row 1: Locking design. 2: Supports. 3: Launcher. 4: Hammer sweep):

Printing

Two spinner pieces plus 14 launcher parts. Tested profile: PETG, 0.20 mm layers, 4 walls, 15% crosshatch infill.

• Disc and arms print as one piece, print in place, arms deployed, top face down. The two pieces under the hammer arm are already snap off supports, with proper spacing. The inside pivot area should be filled with support in your sliced model: supports enabled, type set to Normal (manual), and the included enforcer volume bounds exactly where it fills. See the labeled photo above. Snap the link supports off first, then clear the cavities, all before the first fold.

• Shaft prints upside down, support free. Square middle part slides into the disc, should fit very snug.

• Launcher: All parts should be posed for the print direction in the files. The round gear axles print lying down with a D-shaped head flat; they are static (parts spin on them), so the flat never touches a bearing surface. The two body plates capture the axles in blind pockets; fastening the halves together with the four printed push pins.

• Assembly and build order. There is a full doc with a detailed build order. But there is also a blown-up version of the assembled launcher in the previews at the top.

I tested the system many times. But I do apologize: I ran into time constraints, and I can not ensure this slices accurately, with the custom support markers, on all slicing systems. Included are base models as stls, and the custom supports models as 3mfs. If it does not slice cleanly for you, please send me any model issues you encounter. I am happy to spend some time adjusting this, even though the contest is over. Though, I can not guarantee, months down the line, that I will respond to messages right away.

AI DISCLAIMER

AI was involved in the making of this model. The design, layout, operations, and mechanics, are all mine. It was used for research, math, and adjusting parameters. I am happy to share more info, if requested.

Further AI Detail

The best way to describe my AI use is this page itself. I wrote every line, every word of what is shown here. Every comma and parenthesis is mine. Including arranging the pictures, renders, etc. But I did use a base simple template. And I did have AI give it both a formatting and grammar read not edit. Then I decided what to change, and changed that. But every nuance, from the spacing to the bold, is my decision, not AI. And the final reviewer of this submission was actually my father. He did his best to understand the jargon and provide grammar feedback. The same is true of the model, AI parameterized my exact described dimensions, parameters, into a script. That script allows very quick editing of parts of the model. Linked heights, radii, fillet modifiers, etc. That, yes, at times causes just as much headache, dealing with the script version, as it sounds like. But I find it outweighs the drawbacks, by being able to instantly change most features uniformly, once the base design is fleshed out.

Design has not changed. Updates:
July 8th. Clarified phrasing in instructions for which parts need support, and how. Simplified model upload list. (added 3mf with custom supports built in, instead of just separate stls). Added a split axle for easier assembly.

July 9th. Removed incorrect dates from picture. The hand drawings were labeled on last opened date, not first saved and created date.