After the previous hypothesis about Mini Linear Actuators, there have been a few comments regarding another possibility of how they could work. Also, there was a question about how could such a part be assembled.
First of all, this is how I think the mLA (the first proposal, shown on the previously mentioned blog post) could be assembled.
The bracket (red) part could be composed of two sub-parts, which are joined together after the axle/sleeve (green) part is inserted.
The green part would also need to be composed of two sub-parts, to allow the piston (blue) part to be inserted between them.
Finally, the piston (blue) part could be composed of the screw and the pin hole; these two would need to be separate sub-parts again to allow insertion inside the green part.
I found this a bit difficult to explain, and you most probably found it a bit difficult to understand, either. So, if you look at the cutaway image I posted earlier, you'll notice the narrowing at the tip of the green sleeve. This prevents the blue piston from coming off when extending too much, but also requires all these sub-parts to be constructed and assembled.
So, what I thought would be a simple and cheap way to construct mA's turned out to be rather complicated... therefore, let's see the other alternative that was suggested and I modelled. This one is much closer to how "regular" LA's work.
Comparing with the first mLA speculation, instead of the cylinder part being composed of two "layers", the outermost one spinning when the mLA extends or retracts, this proposal has three layers:
An external (in red), static one, fixed to the bracket. This allows the mLA to be comfortably held while it is spinning;
An intermediate (in blue) one, which constitutes the mLA's piston. It has an internal thread, which mates with the innermost layer.
An inner (in green) one, which spins and has on one end an axle (or maybe axle hole) and a screw on the other. This screw is responsible for extending and retracting the piston.
While this design is probably more fragile than the first proposal and therefore quite possibly unable to handle anything but the lightest loads, it is not only not more complicated (the assembly process would be similar to the one above for the previous suggestion) but also it's a lot more realistic.
At first I thought the first speculation was more likely, but after considering what you suggested on the other post, I too am more inclined towards this second possibility! I can hardly wait for next year, when all doubts will be cleared...