Close
0%
0%

3X0 - 3D Printable exoskeleton concept

A functional 3D printable full body exoskeleton - Passive concept development

Public Chat
Similar projects worth following
3X0 is basically a dream I had since my university years, so now that I got some time free during work I started to "give life" to the ideas I had in the past.
The aim is to develop a full body exoskeleton that could be used both for user enhancement or support of basic functions.
For this reason the design is developed in such a way that it would finally unload at the ground and not just redistribute the weight on other parts of the body.

The whole project is at concept level and the choice of fully 3D printed version is due to the fact that at the moment is developed on my own without availability of a workshop (so yeah, a laptop and a 3D printer).

Every component is accurately designed with some calculation background (such as the 3D printed springs) and many factors have been considered in order to achieve the proper results when 3D printing them.
If you want to follow and support the journey on IG, it's @nozzle_torino

Measurements & design approach

There are different ways of taking proper measurements before modeling the parts, the important is to avoid as much as possible to go by "trial-and-error" because well, no one wants to waste material and time when 3D printing.

In my case I went for photogrammetry, meaning that with about 50 pictures of my body I created a 3D mesh of myself and used it to model around it.

Of course, the mesh is not super accurate in details but all I cared about was the sizes and for this it's been perfect. No need of going to a 3D scanner which is way more expensive for example.

For some pieces this approach has been enough to achieve successful results at first print, for other no, obviously.

Apart from modeling the pieces, obviously some preliminary calculations have been done to determine a proper dimensioning and design for each piece. Given the current stage of the project, most of the calculations have been analytical (either handmade or with Excel tools) taking into account a proper safety factor.

A choice of good and reliable material brand providing accurate mechanical properties have been also very important in the process, as well as good print settings and the choice of color (yes pigments have a great impact on plastic’s mechanical properties, that’s why my whole exoskeleton is red 😉)


Arm piece

First piece to be sketched/designed has been the arm (bicep + forearm).

Took me a couple iteration to get to the right design having a proper comfort but the result was quite good.

I had a couple ideas for the elbow joint, but I found a simple offset ball joint to work properly for the aim with a good compromise of stability and freedom of movement.

The whole arm is constituted of a hollow frame in which resistance bands will be added in a pattern very similar to our muscle structure (of course much more simple than that) in order to aid arm flexion or extension, depending on the pattern chosen.

After this initial passive concept, nylon bands will be added in the frame and pulled by a high torque motor placed with appropriate path on spine or shoulders. Here is as example picture with a simple switch and a brushless 24V motor.

The reason I went initially for this concept is to avoid placing motors on the elbow, increasing the overall envelope and impacting the arm frame resistance and also penalizing the overall look. Drawbacks of this solution might be the high friction generated by the resistance bands path but, given the frame structure, an alternative design could be quickly implemented after initial full test.


Boots & legs

Boots are going to be the component taking the overall load from the full exoskeleton at the end, so I decided to add that bit of "push" to make it easier to walk. A bit of geometry, the right printing direction/settings for the different pieces and with a couple of gas springs I came up with this design which gives about 250N push at the heel.

Proportions have been accurately chosen in order to guarantee a good compromise between performance, push force and “walkability” and I decided to patent this whole piece since it’s one of the core ones of the whole concept.

Pivots for the gas springs can be moved and arranged properly (based on some geometric constraints equations) in order to change the maximum plantarflexion angle or the push force at the feel.Right after the foot, I designed the knee and leg piece in order to have the full lower limbs.

Knee piece uses similar gas springs to the one used for the boots. The overall design is very similar to commercial knee braces (with some features to increase resistance of the 3D printed piece) while the flexion/extension support is given by gas springs of 100N each, giving a total support at the knee of about 15Nm.

The choice of linear “actuation” (gas springs) has its reasons: mainly the easy availability (they are furniture gas springs which come...

Read more »

  • 1 × 3D Printing Filament (PLA+)
  • 1 × 608RD Roller bearings
  • 1 × Bolts and nuts in different sizes
  • 1 × Gas springs (about 100N)
  • 1 × Lattice resistance bands (thk. 0.8mm)

View all 8 components

  • Upper exoskeleton module parallel development

    Andrea Piccinno04/27/2022 at 20:36 0 comments

    Since I had basically ready the shoulder module from the full body exoskeleton (still to be completed), I started a quick parallel development of an upper limb solution, very similar to commercial exoskeleton.

    The aim will be, depending on the case, to support the user in keeping the arms either in front or up the head therefore relieving the shoulders.

    In this design, contrarily to the full body one, the spine will be still (from above the sacrum) since it it intended for safety of the user. It still allows forward leaning but keeping the upper back straight.

    Range of motion is full and support is provided by gas springs (design 1) or elastic bands (design 2).

    Following video shows a preview of the design 1.

    Many features are still missing here, as for example a cover bracket for the rods which will allow the user to carry a backpack without feeling it on the shoulders but rather on the hip (see picture below).

    As soon as the design 2 is ready, I will start some testing by recording muscle activity with and without the exoskeleton.

    In order to do this I setup a very simple EMG project with Arduino that basically requires the following components:

    Arduino UNO
    AD8226-based EMG sensor
    3 electrode pads
    A 3-lead connecting cable
    2 9V batteries
    Connecting Wires & Jumper wires

    In this way I should be able to ready muscle activity in order to compare it in the standard case and with the exoskeleton on.

    Updates about the Arduino code and the signal post processing code will follow as it's all work in progress.

  • Standing test #1

    Andrea Piccinno04/21/2022 at 20:14 0 comments

    One of the aim I had in mind from the beginning of the project, was for the exo to be able to stand on its own in a static position.

    This is not a must in general for exoskeletons, particularly passive ones, but I consider it a "nice to have" since any unbalance of the exo on its own, would need to be compensated by the user. The fact that this type of exoskeletons is mostly lightweight makes this not a big deal, however, as a personal challenge, I tried to achieve this, at least in a static position, since dynamically would be impossible for a passive system.

    Having an exoskeleton standing on its own would mean that the center of gravity of the frame is located properly and that all forcing elements are preloaded correctly therefore trying to bring the user to that static equilibrium position.

    Here is a first test I made with the whole body (arms still missing) achieving a good result since the exoskeleton already stands on it's own placing the feet correctly. It is still not a stable solution since the hip torsion springs are still missing (this would give stability to the position) but in this way some features of the design can be highlighted.

    For example the fact that the legs can rotate around z independently from each other (notice the left leg way more rotated than the right one). Or for example the fact that the core gas springs compensate the weight of the shoulders/arms which otherwise would cause the exoskeleton to learn forward creating a torque around the spine pivots.

    Following preview video of the exo standing will clarify all the described aspects.

  • Shoulders update

    Andrea Piccinno04/13/2022 at 20:56 0 comments

    Finally both shoulders have been printed and are ready with some updates (not only visual).

    In particular the attachment points for the gas springs has been shifted on both axis in order to have the following results:

    - Gas springs are preloaded at rest condition (this compensates the initial sticking point at actuation)

    - Arm sliding axis at rest condition is inclined 10° updwards.

    Both gas springs guarantee a growing reaction force the lower the arm is, therefore a bigger torque at the shoulder joint. They could be eventually replaced by tension springs in order to keep the hands pointing upwards.

    This specific shoulder module of the exoskeleton is going to be used and adapted separately in order to create a second upper limb exoskeleton. I am going to work soon on this and the main modification is going to be the spine, since in the case of a single upper limb exo that component has no need to being bulky and super felxible.

    Meanwhile, here is a preview of how both shoulders look like now (one scapula is still missing but I'll print that soon).

  • Achilles' tendon spring assembly

    Andrea Piccinno04/05/2022 at 21:23 0 comments

    Both springs have been finally assembled on the Achilles tendon area.

    They have been printed with different materials (both have a safety factor considered in calculation to account the material difference) and the new one (left foot) is definitely stronger. Additionally the salt annealing treatment enhances its fatigue life to the limit.

    Functionality of both spring have been tested and they correctly guarantee the return of the leg in vertical position after the squat movement. Here is a preview video of both spring, both on the legs alone and on body.

  • Achille's tendon spring

    Andrea Piccinno04/02/2022 at 20:48 2 comments

    Since left foot is getting printed, I also made ready a second tension spring to be placed in the Achille's tendon area. Differently from the previous one, this has been salt annealed in order to boost its mechanical properties and therefore be able to have a higher fatigue life @ working max deflection.

    Here you can find a short video with the main steps about the annealing process and below its detailed description.

    What is salt annealing?
    It's a thermal treatment of 3D prints that turns the inner structure of prints from amorphous to crystalline and therefore highly increases its mechanical properties (UTS almost doubles).

    How to do it (in this case for PLA+)
    - Print your part with 100% infill (the part shall not have inner voids)
    - Remove supports and perfectly clean and sand the surface (don't leave strings or other imperfections)
    - Completely cover the part in fine salt (sand or talc would work too, the finer the better) and make sure salt goes in all voids. Once covered, slightly press the salt and make it a compact block. This is an important step since the function of the salt is to keep the piece on place avoiding deformations when the piece softens.
    - Preheat an electric over at about 100°C (above glass transition temperature but below melting point of your material)
    - Leave the piece in over for 1 hour (control the salt temperature is stabilized)
    - Remove from over and let cool down to room temperature
    - Enjoy your strong piece


  • Leg movements

    Andrea Piccinno03/30/2022 at 20:35 0 comments

    While I wait for the second foot to be ready, I tried wearing both legs and trying to take some steps in order to check if there is any weird clash or issue.

    So far walking seems ok and also side abduction movements are quire smooth. Of course feel while make the overall structure stiffer since ankle has no degree of freedom about z for the moment.

    Here is a short preview of both legs connected to the waist (I had to hold it since there is not yet attachment to the body).

  • Gait simulation #1

    Andrea Piccinno03/28/2022 at 20:46 0 comments

    As I started to print the left half of the lower body, soon it will come the moment of the "heavy" tests, such as walking or full body loading.

    As I normally don't like to leave things to fate, also for this aspect I setup some simulations. Specifically here is a gait simulation. In order to keep the model light, only some of the degrees of freedom have been used, reason why on some points the movements might look stiff.

    The overall aim was to check anomalous clashes or constrained movements and overall everything seemed to be ok. Of course physical test will have the last word. For the moment here is a preview of one of the simulations.

  • Push-up test (arm only)

    Andrea Piccinno03/23/2022 at 21:39 0 comments

    I got many questions about the arm piece regarding overall flexibility. 

    In particular, it the overall range of motion and degreed of freedom would be enough to perform a whole push-up. I have simulated this beforehand on Fusion360 and results seemed to be ok.

    However real test has always the last word, so here is a push-up test wearing the whole arm.

    Movements are quite smooth and also support is provided by the rear tendons which aid arm extension.

    I might updated the wrist upper piece design in order to reduce contact with the hand which might feel disturbing after continuous usage. Tests with the whole exoskeleton will follow.

  • Wrist ready - wearability test #1

    Andrea Piccinno03/21/2022 at 21:24 3 comments

    After 2 iterations for proper wrist pivot location, finally I achieved a good compromise.

    Overall, the wrist has 2 degreed of freedom, one dependent from the other, which make it quite comfortable and allow pretty much all main movements.

    First rotation is given by a custom made bearing all around the wrist/forearm, second rotation is about 2 608RS bearings placed on the sides of the real wrist.

    The whole piece is jointed to the articulated hand. Next step will be a movement /wearability test with the rest of the exoskeleton and then hand actuation.

  • Full wrist bearing test #1

    Andrea Piccinno03/17/2022 at 21:22 0 comments

    Finally printed both pieces for the wrist bearing. It is constituted by 2 similar pieces placed symmetrically in order to compensate unbalanced loads. Each piece has 2 rails of steel ball bearings plus a centering  feature. Once assembled movement is smooth and wrist pivot seems to be placed in the right point.

    Now there will be the attachment with the hand piece to double check that everything is in the right place and proportions. Meanwhile here is a quick preview of the bearing design and functionality.

View all 11 project logs

Enjoy this project?

Share

Discussions

Peter Walsh wrote 04/28/2022 at 18:27 point

Saw your project on HAD - congratulations. Very nice project.

Will you be making the design files available?

  Are you sure? yes | no

Joshua Felix wrote 04/18/2022 at 14:11 point

Not bad. What's missing is the headset that reads brainwaves and translates those waves into signals that can be used to expand and contract pneumatic (or hydraulic for harder materials like aluminum) pistons with the help of a control board. The point would be to enhance your human strength by a certain percentage. I'd do this using a custom board equipped with some GPIOs, but I truly have no idea if that'll be feasible. 

  Are you sure? yes | no

Andrea Piccinno wrote 04/27/2022 at 13:36 point

RIght, that's definetly a future implementation quite comples. So far the main part of the exo is passive, since I am at the point of building the frame. I'll start to implement the active part on a derivated upper module where I am adding some myosensors to monitor and control the activity with the exo. I'll be posting updates soon. Moreover, given the material used, additional motors would just increase the stress on the frame given the increased weight. But later on, upgrading the materials, situation will change.

  Are you sure? yes | no

Manuel Corrales wrote 03/18/2022 at 23:38 point

Hi Andrea!

Amazing project!!

Do you plan to share the project someday?
I have a family member who suffers from poliomyelitis and it would help him in his daily life with the mobility of his legs.
Congratulations on this great project!

  Are you sure? yes | no

Andrea Piccinno wrote 03/19/2022 at 08:34 point

Thank you Manuel. Yes I'll be sharing it when it's done and I come to a first whole reliable design. For the moment I shared some minor pieces on which I'm quite confident already, like the fingers for the hand piece

  Are you sure? yes | no

Manuel Corrales wrote 03/19/2022 at 09:34 point

Perfect!
I'll be looking forward to the moment you share the lower limb parts!

Thanks!

  Are you sure? yes | no

Andrei Bazavan wrote 03/16/2022 at 20:19 point

Amazing !!

  Are you sure? yes | no

Andrea Piccinno wrote 03/16/2022 at 20:32 point

Thank you!

  Are you sure? yes | no

happyday.mjohnson wrote 03/16/2022 at 16:06 point

I wonder if this can be used to help my 92 year old father.  He is very unstable on his feet.  GREAT GREAT WORK!

  Are you sure? yes | no

Andrea Piccinno wrote 03/16/2022 at 20:13 point

I'd love to bring the project to that level. Now it's a very first proof of concept with its limits but I want to bring it in that direction with the right development.

  Are you sure? yes | no

happyday.mjohnson wrote 03/16/2022 at 20:30 point

Does it make sense for folks like us to print out, get software, try it out and contribute?

  Are you sure? yes | no

Andrea Piccinno wrote 03/16/2022 at 21:10 point

@happyday.mjohnson definitely! that's what I'd like to do at community level, starting with some specific modules of it. The more feedback, the better. Probably I'd start from the lower limbs since they seem the ones that could bring the most benefit. I'm almost at the end of the design of this first concept, so as soon as I'll be done I will start that testing phase and I'll be posting about it so people who are interested can take part to it.

  Are you sure? yes | no

Ninjalicious wrote 01/21/2022 at 18:41 point

Love this. Keep it up Tony Stark.

  Are you sure? yes | no

Ninjalicious wrote 01/21/2022 at 18:42 point

Let me know if you need welded or metal parts for this, I'm happy to collaborate.

  Are you sure? yes | no

Andrea Piccinno wrote 01/21/2022 at 19:06 point

Thanks man! For sure in next version I'll keep you posted 😉

  Are you sure? yes | no

dekutree64 wrote 01/18/2022 at 18:28 point

Cool project :) Looks like a great place to use lost PLA casting to turn your 3D printed parts into aluminum. Use natural PLA so it burns away completely with no pigment powder residue left in the mold. Bearing holes can be made slightly oversize and then install them with epoxy. Make shaft holes slightly undersize and drill them out.

  Are you sure? yes | no

Andrea Piccinno wrote 01/18/2022 at 23:07 point

That's a great idea for the next version. Now being just a concept I'm going fully 3D printed just to check the overall kinematics but soon I'll probably use lost casting especially on the most as stressed components. Gotta hurry up

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates