Furniture Design MIT 4.120 Final Project
To this day, I’m still rather baffled at how I got into this class in the first place. To give you some perspective, this class only accepts 12 students (predominantly architecture students) per semester and hence overprescribed. And as a student at the business school, my chances of getting into this design intensive class seemed to be close to none. Even though Chris assured everyone at the info session that everyone had equal chances of getting in, many who submitted their pitch about why they wanted to get into the class also dropped down rather impressive portfolios as well. At this point, I thought it was hopeless. But maybe it was the enthusiasm I showed by submitting some chair design sketches (although at the time, I hadn’t made anything substantial) or mentioning that my mom had been a carpenter back in the day, or perhaps everyone had just dropped the class for something else (aka class fomo). What ever the reason, this class was one of the best things I’ve ever done. And it taught me numerous life lessons as well:
- Nothing is more therapeutic than starting your day in the woodshop
- Furniture is interactive and weight bearing, so don’t forget about physics
- Do not use a hand router if a CNC router will do. You can see the scar from my terrible routing skills on the seat in later pictures.
- The Festool system is stupid good. Although, I’m sure some purists would disagree.
- Do not use spray adhesive in an enclosed area. Actually, don’t spray anything in an enclosed area, unless it’s intentional.
- Steam bending white oak requires at least 2 grown ass men
The requirement of the final project was to “design an original piece of furniture that will support 185 lbs at least 12 inches off ground.” I think 12 inches was added to deter students from handing in polished tree stumps.
With open-ended requirements, I went to work on designing a piece that represented my experience at MIT. I wanted to make a chair that represented rapid-prototyping, agility and living on the fine line between balance and chaos. Aesthetically I wanted each element to flow into each other seamlessly, yet structured enough to stand the test of time. At first glance, the piece may look like it’s only balanced on two curved legs, but it’s actually reinforced with a cross-beam underneath the seat. Another goal of mine was that I wanted to use as many tools in the shop as possible,so that I could transfer those skills to later projects. I’m happy to say that I achieved that with the jig saw, joiner, table saw, hand router, drill press, band saw, hand saw, and the CNC router.
As if my stubborn out-of-thin-air requirements were not enough, I wanted to make a chair. The one piece of furniture that Chris said not to make because of the difficulty of pretty much everything. You have to factor in posture, ergonomics, weight distribution, tilting and shifting body weights, etc. So since I didn’t want to compromise on the design, I pretty much picked the hardest wood available at the time, hoping that it would compensate for what ever miscalculations I may make.
The chair ended up being made entirely out of white oak (quercus alba). I chose white oak because it was a widely grown hardwood native to eastern North America, and hard enough to withstand the test of time, plus Chris found a bunch of planks at the lumber yard for a good price (win win!). White oak is also had the perfect properties for steam bending. Not only is it pliable when steamed, it also holds its shape for decades without much recoil.
The legs and arm rest are quarter inch thick steam bent laminates. The seat was designed in AutoCAD, rendered in Rhino 3D and machine cut on the CNC router. The beams that hold up the seat and the arm rest are hand carved and partially shaped with machine tools. If you’ve ever hand carved white oak, you’d know why machines were involved. The entire chair was then sanded down with P220 sand paper and finished with Danish oil. The entire process is outlined below.
I learned very quickly that chair design is quite the science. I had to do many profile sketches before landing on the final design. I wanted something simple and streamlined, yet versatile and functional. As you can see the physics just didn’t add up in this original design. The arm rests would’ve been hard to support with the seat back without contorting. It would’ve been fine to look at as an art piece, but terrible for long-term use.
This miniature foam model was cut on the band saw and pinned together with nails. To keep to the purity of the design, the chair was intended to be held together only by wood joints. Though, I had to settle for nails instead glue for holding the seat in place because I needed to disassemble the chair and ship it across the country.
After much contemplation, I decided on this profile because I liked the aesthetics of the curves and how it gave the illusion of soft lines, yet be incredibly strong structurally (or so I thought). After Chris saw the model, he advised me to redesign the seat back to better distribute the weight of the arm rest. A logical suggestion since there was nothing counteracting the forward flex of the back rest (duh!). I’m sure you can also see why this design totally didn’t make sense physically. Chris also didn’t think the legs were practical, which was true, but I ended up keeping them because they looked nice :)
The wooden seat back in the background was actually a piece that I drew inspiration from. It was a piece that someone had made but never finished. I was actually able to use the formwork that this piece had originally bent around.
Originally, I wanted everything to be precision perfect, but for the sake of time, I had to toss some of it out the window. The picture above is how I wanted the back legs to align, but they ended up being a few centimeters apart. And the fact that I went through shoulder surgery weeks before the final projects were due, which made working with tools slightly more difficult.
Steam bending wood is a pretty intensive process that takes several days a) make formwork b) steam wood for at least 1.5hrs c) clamp steamed wood and let dry for at least 24hrs. Depending on how difficult you want your life to be, you can either try to bend a solid piece of wood (harder to bend, but shorter processing time), or slice the wood into thin laminates (easier to bend, with longer processing time. I decided on the later, because the curves in the legs were too curved to render a solid wood bend. Bending solid wood at those angles would likely cause the wood to break, crack, or splinter. So with this process in mind, I had to add a few more days to laminate and glue the wood.
Before steam bending anything, you have to make the formwork that the wood would bend against because wood don’t bend themselves. So to make the legs, I drew them at full scale on drafting paper. Then with the help of some strong spray adhesives (remember lesson #5?), I stuck them to a large board of MDF (Medium-density fibreboard). After that, I cut 6 identical pieces out with a jigsaw, glued them together with gorilla glue and refined the edges using a band saw. This formwork is probably still at the woodshop, hopefully I can go retrieve it one of these days.
Then, I cut my wood plank (post joiner) on the table saw length wise to create 6 quarter inch laminates; one of which served as protection from the clamps. After steaming the laminates for 1.5 hours, so that the wood is nice and malleable, and with the help of Chris and a nice classmate, we clamped the unglued laminates to the form work and let it sit for a few days to let it dry from the steam.
After this initial steaming, the laminates are able to hold about 80% of its new shape, which makes glueing and bending much easier. You can see below, how many clamps we had to use to make sure that everything was being forced in the right direction.
After finishing 2 legs, the arm rest felt like a breeze.
Here’s what the legs looked like after much sanding. As you can see they’re held together by a mortice and tendon joint with a horizontal beam. This one is actually a mock beam glue together to show where the joints will go and to ensure that the legs are also balanced properly.
Here’s a picture of my seat undergoing some pretty bad workmanship. I wanted to take off enough with what ever tools I had (i.e., chisel and hand router) so that I can shape the seat by hand, but it ended up being ridiculously inaccurate. And I had to accept the fact that 2 months in an wood shop did not make me an expert craftsman.
So, I opted for 21st century technology called AutoCAD and Rhino 3D, then let the CNC router do all the work. Oh the marvels of technology. This was the first drill bit taking out most of the mass.
This is the second drill bit smoothing up the lines. It was so awesome to watch.
After the seat was cut out, I took it to the mortiser to cut out where I wanted all the tenons to be.
I cut all the tenons using a combination of hand and machine tools. I’m kind of in love with the band saw because the table saw still scares me every time.
All the edges were softened with a file and sand paper. Lets just say that ‘blood, sweat, and tears’ is an understatement. Even a week after the project was done, my hands still stung when I washed it with soap.
This is the final design, albeit I still have to round the supporting columns. It quite literally fits me perfectly. Side note, I think there’s a lot of room to build a business on made-to-measure furniture. This is my first piece of furniture and it definitely won’t be my last.
Here it is after a nice coat of Danish oil.
Originally published at Lucy Zhao.