The Goal

In Stanford’s course titled design for additive manufacturing our project spec read as follows:

“Design a part that can be printed in its assembled state. Including two or more connections in a mechanism whose tolerances have been fine tuned for the function of the design by using the ‘print-in-place’ method”.

Additionally, for my own personal constraint, I added that the product had to be wearable.

Initial Research

In my initial research, I found many print-in-place articulated animals which moved in an extremely satisfying way. I wanted to take this kind of motion and simplify the complexity while making the product more functional for everyday use.

Articulated dragons which inspired my connection mechanism

The human spine served as inspiration to mimic this articulated motion

Continued Research

As I ideated on how to both simplify the articulated animals while making them more functional, I decided I wanted to make something I could wear that could move. Below are some inspirations from this research.

Visual research on organic geometry and 3D printed wearables. Repetition and organic forms emerged as key pieces of my interest . Vitaly design (see above left) was a big inspiration for me in their use of bone-like geometries and sharp curves. Additionally 3D printed fashion helped me gain inspiration of what work had been done in this space and how linkages might work (see above right)

Early Ideation

My initial sketching was extremely broad. From my initial research, I knew I wanted to be able to wear my product, but I wasn’t yet sure where or how. Below are some pages of sketchwork to find out.

Continued Ideation

The additive manufacturing ability to print in place allowed me to print many preassembled linkages all at once (instead of having to manually assemble them which would have been time intensive), which is why I moved forward with a scarf / necklace of some kind. In this ideation I experimented with many different vertebrae inspired linkages (seen below).

First Prototype

Once I had settled on a first iteration of a linkage, I printed two stages of the link to test the ball-in-socket joint and note sizing.

One major takeaway from this first prototype was that linkages were very large–and while this was cool for a scarf, it was quite unfeasible for many lineages given print bed size constraints.

Another issue with this prototype was that the socket joint was fused together. This indicated a need to increase the gap between the links

Interface Prototype

Perfecting the ball-in-socket joint was my top priority in the early phase of prototyping.

In response to the size and joint concerns from prototype 1, I significantly shrunk down the linkages to make a necklace instead of a scarf. This size was promising for printing many links at once and I additionally increased the gap between the ball and socket.

CAD of joint prototype. In it you can see an increased ball and socket gap to decrease the risk of parts fusing together

The print successfully avoided fusing the parts together, but overcorrected such that the ball slipped through the socket

To correct my previous issue with the ball slipping through the socket, I decreased the gap between the two components. This marked a third trial for fitting links together.

In addition to fixing the fit, I increased the angle between shaft and opening to allow for increased range of motion for the links

This fit ended up working amazingly. The joints stayed together and rotated so smoothly I unintentionally created a fidget toy

Print Configuration

After finding the perfect fit, I assembled together pieces until they reached roughly 18in in length (enough to fit around my neck). Printing this though turned out to be much harder than I imagined.

Originally assumed printing this configuration would be simple and fast, however it did not fit on the print bed

This second configuration tried to utilize more dramatic angles, but also turned to be out of bounds

This final configuration was not ideal for supports, but did manage to accommodate a continuous loop. It however would take 16 hours to complete

Final Print Configuration

One unintended feature of my design was that the ball and socket joints were so well fit that they could snap in and out of place while not slipping out.

Upon discovering this, I pivoted to printing the design in strips to save supports and print time, while still capitalizing on the assembly benefits of printing most joints in place.

Final print layout featuring only four links to snap together–saving print time and assembly time with print in place

Final print bed with the finished rows. The print time ended up being 8 hours and 20 minutes

Below you can see me enjoying wearing this necklace after its completion. I ended up being very happy with the connection mechanism and visual appeal of the necklace!!

Me wearing Vertalace in May 2023.

I had so much fun making this project! So much so that I think 3D printed wearables (clothing and jewelry) might be something I continue working on in the future. That being said, there are a few things I would still love to tweak and improve in successive iterations of this project. Overall I spent roughly 15-20 hours on this project.

Secondly, I would have loved to find a way to expedite the process of removing the supports off the bottom of the print. My spheres were slightly larger than the main body of the necklace, which required a thin layer of support on the bottom of each link. Using water soluble supports would have helped greatly with this, or possibly shrinking the size of the sphere to be smaller than the thickness of the main body of each link.

Overall though, learning the power of print-in-place has felt so empowering and I am so excited to keep working with it! There is nothing like the perfect tolerance :)

Nearly a year later, I enrolled in Stanford’s Product Realization Course titled “Injection Molding”. After studying the design and manufacturing process throughout the quarter, our final assignment was to,

“Design an injection molded part that can interact in a geometrically meaningful way with multiple copies of itself”

Immediately, I thought of Vertalace.

Modular Clothing

Concept art for wearable Vertalace garment

This above concept of a Vertalace garment is one of many concepts that can be created from these design elements. It draws inspiration from drag art and could function as a garment for such artists.

New Pieces

With the idea to make patterns in order to create dresses, shirts, and more, I introduced two new connection pieces to introduce 2D shapes and increased potential for play (seen below).

Internal mating piece

External mating piece

Vertalace panels with connector pieces (1), and design pieces (2)

With two design elements featuring very similar geometries, I modeled various injection sites with blue polypropylene for the standard design element which I would then extrapolate to the other design elements

Option 1 proved to be the best option with equal fill times throughout the extremities of the part and ease of cutting the gate

Option 2 showed variable fill times between the left side versus the right side of the design element which was less ideal than option 1

Option 3 also had equal fill times but the gate location inside of the internal mate made it almost impossible to access for gate cutting

First Mold Design

After deciding this gate location, I designed and machined the first mold. This revealed some key issues that informed my later molds.

1. Machining Boundary: Tool center on boundary caused a small cusp outside of the part which was undesirable

2. Bottom Heights: Setting the bottom height too high prevented the ball end mills from fully machining the insert pockets. This made placing the inserts impossible and required re-machining

CAM Operations

To manufacture my final mold I developed a robust CAM strategy which would allow for minimal offsets and minimal time.

1. Face

2. Trace gates

3. Center drill

4. Through drill

5. Counterbore contour

6. Pocket

7. Face

8. Steep and shallow

Second Mold Design

After making the aforementioned adjustments to the mold I re-machined the mold to include both the design element and the internal mating piece.

To save materials I made sure to use the same mold blank as previously due to the cost and previous work done to prepare the countersunk holes.

For molding material, I chose to use the pre-dyed blue polypropylene to have a consistent fill color across many parts. I chose polypropylene for its ease of availability and its rigidity. I needed firm plastic that held its shape so that my joints functioned properly, while not being so rigid that press fits would not function.

First Parts

Material Temperature

Using the pre-dyed blue polypropylene resulted in some issues with sprues catching in the mold base. As a result I had to reduce the temperature to prevent excessive material dripping which then resulted in too cool of a temperature which resulted in rapid cooling deformation (see left). Calibrating the right material temperature was an ongoing process throughout the project.

Shutoff Calibration

Another issue that came up throughout the project was calibrating the fill volume and injection pressure needed to reduce flash around the ball bearings. There was a delicate balance between adequately filling the part and preventing overfill flash as seen right.

Tuning and Assembling

After tuning the temperature, fill volume, and injection pressure parameters, I had some high quality parts that I was able to begin assembling into prototype bracelet bands.

Successful internal mate and design element parts

Bracelet assembly test showing how a Vertalace sleeve might be constructed

Final Mold

After creating this new part, I re-machined my mold in order to add both the double mating piece and an additional design element piece. Adding the additional design element was primarily to create a higher proportion of design elements to connection pieces, as most designs require more design elements than connection elements.

Side by side final mold

Final mold loaded into Arburg machine

Inserted 0.25’’ ball bearings to create shutoff which allowed the formation of the internal ball joint.

3D Printed Assistance

To assist with rapid production, I created a 3D printed bearing remover to assist with removing bearings required to form internal cavities for my mold.

First prototype, guidance features were not robust enough to support pressing out shutoff bearings

Final 3D printed bearing remover, supports all part types by including two stamp shapes

The fixture uses 1/8’’ steel dowels to press ball bearings out of injection molded parts

Final Parts

With my final mold and 3D printed bearing remover complete, I was ready to enter production mode. Some process pictures can be seen below.

Notable Elements

Building 2D planes: introduction of two new connector pieces enables 2D chainmail like geometries which can be expanded to all kinds of clothing and jewelry applications

Easily repeatable: rapid production mode capabilities of injection molding machine allow for quick replicability and full body scale designs

The perfect fit: well tolerances mating features allow for rigid and stable connections between design elements while not being too difficult to press in to form intricate shapes

I really appreciated revisiting one of my favorite projects to date and manufacturing it with a process that was better suited to its application. Vertalace has been something I have been excited about for a long time and I even still would love to continue working on it in the future.

I appreciated seeing how 3D printing designs translated to injection molding. Thinking about intelligent CAM operations to avoid cusp lines, and considering undercuts and shutoff methods was super useful and interesting as I expect this translation between manufacturing methods to be important in my future when transitioning between prototyping and production.

A magic moment that I wanted to call out that really inspired me for this project moving forward occurred just  at Meet the Makers - the final mechanical engineering showcase at the end of each quarter.

A little girl came up to my parts and began snapping them together like legos. Seeing her serendipity and joy interacting with Vertalace made me feel so happy with how far this project has come. I design for moments like that.

Thus far this project has shown me how prototyping and continued effort on a project can yield incredible results. When I continue working on it I would love to further optimize the cycle time by creating even more efficient bearing removal tools, as the cycle time took much longer than I expected which slowed production down considerably. Additionally, I would love to create even larger family molds to produce parts even more efficiently because the parts required to create a human scale design piece are much larger than I originally envisioned.