Collaborative Design Project: Weeks 2, 3, & 4




Weeks 3 and 4 of this collaborative project were dedicated to prototyping and building the bench and bookshelf designs shown in my previous blogpost. Our team worked in the woodshop to CNC cut, sand, and assemble our designs. This blogpost breaks down our process and some of the successes and challenges we faced.


a) Week 3: Cutting the Maze Pattern Prototype

Before we cut out final designs, we prepared a 1ft x 1ft sample of the maze pattern for the CNC machine. 

We used a 0.72 in sheet of plywood for this sample. At this point in our project, we thought we would be using this thickness of wood for all of our parts in the final design.


It took about 1 hour for the CNC to cut this first sample. The cut took a bit longer than expected because the jog speed was slow between cuts and the drill began spinning too high up before reaching the surface. This was an easy fix though and the following cuts were much faster. Bryan also suggested to make the width of the channels equal to the width of the drill bit. This would allow the machine to create the channels in one pass per layer instead of multiple.

Nadia adjusted the grasshopper definition I made by enabling the center lines of the channels and disabling the offsets. Then, Nadia and Bryan used this updated definition for our following maze pattern cuts and it sped up the cutting time.





The pattern looked nice from a distance, but it had many broken sections. One of the reasons is that the school had purchased construction grade plywood for our class instead of finishing plywood. Another reason for the broken pieces is that the angle of the drill bit spiral cause the wood to splinter upwards. We were able to significantly reduce this problem in subsequent cuts by using better wood and a different drill bit.



.

b) Weeks 2-4: Preparing the bookshelf model for the CNC

By the end of week 2, our bookshelf model was 90% complete and we were ready to add tabs or holes to each part. We wanted to set up our model to have press-fit parts, like the bench. Two of my teammates offered to help our group by completing this step. This was the last step we needed to complete in order to start cutting. We initially estimated that adding tabs would take 1-3 days.

However, we faced multiple setbacks in this stage, and made approximately 15 versions until we were ready to begin cutting. Even though our model was mostly complete by week 2, it wasn't until week 4 that we were done the tabs and able to cut the pieces.


One of the reasons it took so long to complete this stage is that we made our model mostly in Rhino, and not Grasshopper. Every time we needed to make a slight adjustment in the wood thickness, tab size, or hole size, we had to basically reconstruct each part.  

At the start of this project, a few of us proposed making this project in Grasshopper but some of the team members felt uncomfortable with creating parametric designs since it was a new technique for them. So, we decided to make it in Rhino to make the team more comfortable as a whole. We thought using Rhino would help us to work collaboratively on the file since everyone knew how to use it. We also thought that it would save us time since it can be quicker to build designs in Rhino than in Grasshopper.

However, this decision backfired on us when we found out that we had to use a new wood thickness. Even though our model was complete, we had to manually adjust all of the parts to account for the new thickness. I learned a lot from this experience and I will probably work exclusively in Grasshopper whenever I make a designs that might require iterations.

The 3D file labelled with the proper dimensions
 
Independently of that decision to work in Rhino, our team also struggled to keep the model at the correct scale between versions. This was a problem since our model will require hardware of specific dimensions. If the 3D model changes size randomly, then the hardware will no longer be compatible and that work will have to be redone. Even though our team was aware of the hardware, and I had labelled the files with the specific dimensions of each component, I should have emphasized more strongly the importance of keeping the model at the correct scale. In the future, I will keep this experience in mind and double check with my team mates every time we work on designs together. 

Additionally, our team struggled to make the tabs and holes the proper dimensions. One of our team members struggled with this because their Rhino file was rounding the measurements in the dialogue box but not in the workspace. For example, the text box would say a measurement was 3/4 inches but it was actually 0.78 inches in the model.  This inconsistency was due to their unit settings in Rhino.



Since I had set up my Rhino units to be in the decimal format, I was immediately able to tell that the dimensions were incorrect each time my teammate sent me an updated file. For several days, I wasn't sure why the dimensions were still incorrect with each new version. It wasn't until Week 4, when I borrowed my team mate's laptop, that I realized their unit settings in Rhino were causing this confusion. I don't think this was anyone's fault, it was just an unfortunate mistake that prevented us from completing our project on time. 

I am grateful for this experience because it will make me better prepared for future team projects. I will make sure to:
  • Double check the dimensions of each updated 3D file before working on it
  • Emphatically communicate the purpose of various design elements to team members. What may be self-explanatory to some may not be clear to others (e.g. the importance of maintaining the dimensions of a model equal between versions). 
  • Ensure that the software settings are the same for all team members (e.g. unit settings in Rhino)
  • Use Grasshopper as much as possible for team projects or projects that will require many revisions.
  • Ensure at least one team member is exclusively responsible for supervising the project. Each person in our group was either designing, building, or documenting the project but no one was exclusively supervising the project. Some of us who were designing and building tried to supervise at the same time but it generated an overwhelming amount of work. If we would have had better supervision, we may have avoided many mistakes.
  • Establish our team's expectations for the final product at the beginning of the work process. Some of us had different expectations for the final outcome but we didn't realize until the middle of the process.
Although we had many setbacks, I enjoyed working with my team because they were all hardworking people who tried their best to make this project work. We still got a lot done given the timeframe we had to complete the project. We were able to assemble the entire bench and most of the bookshelf. Our goal is to complete the bookshelf next semester and paint it along with the bench.


c) Week 4: Cutting and Building the Bench

Our team cut the pieces for the bench first since it had less parts than the bookshelf and would be easier to assemble. It was really satisfying to see the pieces in their intended scale after working so long on the computer. Like the bookshelf model, we had to adjust our bench model since the wood thickness was now 0.696in instead of the original 0.720in. The new wood had a higher quality than the one we'd used for the prototype.


First plywood thickness vs. new plywood thickness



 

The parts were designed to press fit together so they had to be cut at very specific dimensions. In our case, since the CNC is very precise, we made the size of the tabs equal to the size of the holes.

However, when we first tried to assemble the bench, the pieces did not fit together. We double checked the measurements of our 3D file and they were correct. 

So, we decided to add a tolerance by making the tabs and holes slightly different sizes. We didn't recut the entire bench but only parts of it. 

When we tried to assemble it a second time, the fit was still too tight so some of the team members sanded down the pieces until they fit together.


The side pieces of the bench were stacked, glued together, and clamped down. 

Three layers of plywood were used for each side. 

The seat section is meant to slot into a long hole at the top of each side piece. 

After the pieces were dry enough, the edges were sanded down to level the three layers and remove the glue residue. 

We ended up cutting the seat and maze pattern twice to make them fit better with the side pieces of the bench. 

The first maze cut for the bench (not the prototype) had channels that were too deep. 

We made the channel depth shallower in the second version and it looked much better and it would probably be easier to clean in the long run.

The seat was also glued together in layers like the side pieces of the bench from earlier. We used two layers of plywood for the seat.




 

We were able to successfully assemble the entire bench, glue it, and leave it drying over the weekend.


d) Week 4: Assembling the Bookshelf


Chloe & Chandler beginning to assemble the bookshelf




 

Even though we had incorporated the tolerances into the tab system, some pieces did not fit as intended. A few team member sanded the tabs and holes until they fit together properly. Then, the shelf was assembled.




.



The team members working on assembling the shelf noticed that the sides of the bookshelf were not mirrored versions of each other, but rather duplicates. I wish I would have checked this file before it got sent to the CNC because we were very close to finishing the main structure of bookshelf. If it hadn't have been for this setback, we might have been able to have a free-standing structure by the deadline. This moment reminded us of the importance of double-checking and having a supervisor.

Fortunately, we were able to complete our bench piece and appreciate it without the clamps.


.





 

I look forward to picking this project back up next semester and getting to work with my team mates again! 

I am thankful for all the work my teammates and our instructor Bryan put into this project. It was great to work collaboratively for the first time and get to know each other better.

We have learned a lot from this first attempt, and we will feel refreshed and well prepared for when we meet again in the fall :)



Comments

Popular Posts