Project Log 5: let's list the objectives.

A project log for DIY Human-sized Mech

Mechs are not viable, nor cheap, so I will try to design and build one alone anyway.

FulanoDetailFulanoDetail 12/02/2022 at 16:252 Comments
Friday, 13:06, 02/12/2022 - I'm super procrastinating.

In this mess of thoughts and cool robots that I call "my mind", I did had some simple objectives that I already put some thought into it. Like this project being easy and cheap enough to build at home, unlike the carbon nanotube artificial muscles that need acetylene (a volatile, flammable and explosive gas) in a fricking furnace in order to be synthesised.

But let's lay everything in a plain and simple way:

  1. The skeleton, the actuators, the system and the power supply need to be simple and cheap enough in order to be bought locally or made at home.
  2. The full mech needs to be able to lift at least 1000 Kg of weight either on the hands or in the shoulders. This does not include its own weight and the weight of pilot, which can shift the weight demand to higher numbers. So let's keep with 1,5 to 2 tons in total.
  3. It is only possible to calculate the amount of muscles needed if there is a skeleton designed to support all this weight.
  4. After designing the skeleton, it will be time to design the actuators.
  5. And by measuring the amount of energy that it requires to lift a certain amount of weight, just then it will be possible to design the power supply.
  6. If the power supply can't power the entire body in a practical way, then it will be useless.
  7. If all the previous steps are done without trouble, it will be required to research and add the sensors to the design in order to be able to control it.
  8. After everything is designed, then it is time to build.

Although the best course of action is to continuously prototype and go back to the drawing board in order to continuously address the problems faced during the building process, I'm too broke to actually do that.

So I can only hope to do it in a single well planned shot in the sky that may or may not work.


dekutree64 wrote 12/03/2022 at 21:25 point

Two tricks to reduce the cost of the prototyping cycle: Work small, and own the equipment. You'll quickly run out of money if you have to pay other people to make parts for you, so most likely your first objective will be to obtain a CNC machine. If you have a suitable space for large machines, you could look for good deals on industrial surplus. My approach was to buy a 3D printer and use that and cheap Chinese parts to make a mill, and then use that to make better parts for itself. My machine is tiny, but I have very little space so I need to be able to move it out of the way to do anything else.

Design for full size, but build a scaled-down version so you can use a small CNC and material cost is insignificant. Hands-on work is essential to the design process. Most of my ideas turn out to be stupid, but after attempting to fabricate them I come up with something that really does work. When I do get to building full size, I'll be reasonably sure that it will work on the first shot.

As much as possible, use the same materials as the full size design. You can design robots for direct 3D printing, but it won't be scalable because everything has to be more chunky due to the low strength/stiffness of plastic, and you have to work around the low precision which results in a lot of different design decisions compared to machining. But one thing 3D printing is good for is making molds for carbon fiber parts. Lost PLA aluminum casting is fun too. Not as strong as machining from a solid block, but allows you to use cheap scrap aluminum and eliminates a lot of the machining work.

If you decide to use electric motor actuators, try to integrate the gearboxes into the skeleton as much as possible to reduce weight. But it would be very expensive to make actuators for dynamic walking of a 2 ton robot. And statically stable walking is boring :) Better to just use tank treads at that point. But you could build a pretty awesome mid-size prototype using modified hoverboard motors, which can often be had for cheap or free after the battery gets too old. And you get some hunks of aluminum for casting too :)

  Are you sure? yes | no

FulanoDetail wrote 12/03/2022 at 22:25 point

Thanks for the advices! I will sure try to make those. :)

  Are you sure? yes | no