13.3 Design Process

13.3 Design Process

CAD Of Full Robot (Sorry about the bad color choices)

image-20251214-054721.png
CAD for the Current Assembly used During Demo Day (Note the exterior linkages are slightly different)

 

Design Process

Concept Development

The design process began by identifying the key challenge of jumping systems: achieving high instantaneous power output. Rather than relying on a powerful actuator, we pursued a store-and-release approach where energy could be accumulated slowly and released rapidly.

Additionally, since we were looking for a robot that has the ability to last for a longer duration without battery, have rolling motion (implemented in the future), and be able to potentially carry small payloads, we would need a bigger chassis than most standard jumping robots.

An inverted slider-crank was selected due to its ability to generate large output forces near its singular configuration, which is well-suited for impulsive motions like jumping.

 

Abandoned Ideas

Salto Robot

Concept Release Design

Salto Robot

Concept Release Design

image-20251214-222606.png
We initially considered a similar leg design but found it would not function well in a spherical enclosure
image-20251214-224007.png
The spur gears on the top are required to couple the two four bars, effectively turning 2 doF into 1 doF (vertical motion). We didn’t choose this idea as the gears can easily slip loose meshing under the high loads of the mechanism

 

 

 

Initial iterations included doing a worm gear as a form of creating the torque necessary to spin the motor, coupled with a pulley mechanism. This idea was abandoned as we found out we can get optimal torque through sizing the diameter of the pulley axle(settled with 8mm) and the motor. The rough cad of the worm gear and 3d prints are shown below

 

 

Energy Storage and Release Strategy

Rubber bands were chosen as the energy storage medium due to their simplicity, availability, and ability to tolerate repeated loading. A DC motor driving a pulley was used to stretch the rubber bands gradually, minimizing instantaneous motor torque demands.

Robot1.png
As the mechanism compressed downward the rubber band stretches

To initiate the jump, a clutch mechanism actuated by a stepper motor disengaged the pulley system, allowing the rubber bands to rapidly contract and drive the slider-crank into extension.

Robot2.png
Intermittent Gear (Violet) connected to Red Gear, which is attached to a linkage attached to the motor linear rail (Slider Crank Mechanism)

 

Previous Iterations and Comparison to Current

Pulley Mechanism V2

Pulley Mechanism V1

Pulley Mechanism V2

Pulley Mechanism V1

IMG_1544.jpeg
Here the axles on which the clutch mechanism slides on was made of 3d printed clamps, but was redesigned as it fractured during testing
IMG_1545.jpeg
This was a previous iteration of the pulley mechanism, but was redesigned as it used 8mm shafts (too much weight), an the supports for the pulley would have failed

 

Jumping Mechanism Top Support V2

Jumping Support Top Support V1

Jumping Mechanism Top Support V2

Jumping Support Top Support V1

IMG_1548.jpeg
This iteration uses a 4mm shaft hole for the linear slider, while also only having the I-Beam structure on the bottom half of the design, making it still stiff for torsional loads

 

IMG_1547.jpeg
The first iteration of the top support used 8mm shafts which would have been too bottom heavy preventing an ideal jump. Additionally, i beam like design made it hard to mount items on top of this mechanism.

 

Motor to Pulley Coupler V2

Motor to Pulley Coupler V1

Motor to Pulley Coupler V2

Motor to Pulley Coupler V1

image-20251214-220202.png
Here we do a simple 3d printed coupler (slightly modified to use a metal d-shaft coupler with a 3d print, to the clutch mechanism with an axle

We did not notice any failure with this mechanism after at least 15 tests. 30mm bore bearings were use to reinforce it axially, as well as a 3d printed motor mount (this was structurally stable due to the large radius (more area to apply frictional clamping loads)

image-20251214-213237.png
CAD Design of a Potential Coupler Mechanism
IMG_1546.jpeg
We didn’t end up using a worm gear as it was a very bulky design (small 3d printed worm gears wouldn’t function well) and the gear teeth would stiff due to the torque applied. Additionally, we could get motors with ideal torque speed characteristic making the gear ratio unnecessary

Clutch Release Mechanism V2

Clutch Release Mechanism V1

Clutch Release Mechanism V2

Clutch Release Mechanism V1

image-20251214-221433.png
This design was better integrated to the Top Support and had more rigidity. Additionally a custom mount was designed for the stepper. The intermittent gear was used so the crank slider would disengage when the intermittent gear was not active, while reengaging to active the clutch

 

image-20251214-221906.png
The servo was abandoned compared to a stepper as it did not have 180 degree functionality, and CR servos lack the required torque and tracking capability compared to steppers. Additionally this design was not as stable and lacked adequate attachment to the top mount

 

 

 

 

 

Iteration and Refinement

Early iterations revealed issues with structural bending, asymmetric loading, and clutch reliability. These observations informed reinforcement of key members and adjustments to alignment, though manufacturing tolerances remained a limiting factor.

The final design represents a balance between mechanical complexity, manufacturability, and achievable performance.

 

Exterior Casing

The premise of this robot is to not be only able to jump, but also to be able to handle the impact from jumping with high distances. We decided to experiment with a combination of using PLA and TPU as a way to keep being able to deform to impact, while also maintaining enough rigidity to prevent damage of the components housed inside the spherical enclosure

CAD

Manufactured Part

CAD

Manufactured Part

image-20251214-053401.png
Bottom Leg Housing: The components labelled one, two and 3 are using TPU. As shown below. Number 1 especially needed to be TPU after the PETG version of it got damaged from a drop test. Additionally linkages 2 and 3 are straight TPU Linkages, which are in a state of bending as a way of increasing rigidity
image-20251214-054003.png
Exterior Shown on the Robot Below

Failure with PETG

image-20251214-054417.png
After Performing a drop Test (3ft) on the Robot this part fractured, no damaged occurred with the same drop test using a TPU Version

 

image-20251214-054928.png
Top Outer Housing: Each of the curved linkages are slightly bent to create a more rigid housing, for future prototypes we will likely use a combination of PLA and TPU

 

Failure with Initial Prototype

The TPU linkages were made with too small of a thickness, resulting in too much compliance, and failure to hold its structure

Robot5.jpeg

Lessons Learned with designing the Housing: TPU is a material that could be good for shock absorptions and building compliance in an otherwise rigid mechanism, good for handling impact. Overuse of TPU can result in unstable mechanism and enclosure, which can damage electronics or internal components. Good design choices need to be made when using such a material.