This blog documents our innovative lamp project for the Spring 2010 Penn class: Arch746 Immersive Environments / MEAM516 Advanced Mechatronics in Reactive Spaces with instructors Simon Kim, Mark Yim, and Dan Rothenberg. The Zen Master Group includes Brad Baxley, Tyler Zembrodt, and Bartley Gillan.
Here is a video of the lamp's white LEDs showing off in conjunction with the lamp motion.
Here is a great video of the first reaction of everyone upon seeing the lamp linkages move from the motor.
During our final stages of assembly, we ran into a few problems. First, we realized our whole structure weighs a good amount and is only supported by two aluminum support rods. The central crankshaft rod, which weighs the most, is not supported. So in thinking how the crankshaft rod could be supported while rotating with the motor, we decided to extend the rod's end to a tapered point such that only a small point would be rotating on the acrylic base to minimize friction while adding support.
Second, our gear train design was off before we even secured it to the assembly. The gears were added to reduce the high speed of the motor and increase the torque to the rotating crankshaft, so that it would spin no faster than 1 rotation per second. The current tooth ratio is 10 to 38 which was calculated from an incorrect no-load motor speed specification. The gear train needs more stages to reduce the motor's very high speed.
Third, we're not confident our acrylic to metal press-fits for both gears are strong enough to keep the bond from breaking out under the torque. We're also still not sure whether the motor will be powerful enough to rotate the whole assembly as it weighs more than expected, and there are many parts in the assembly that are torqued and held together by some form of adhesive or press-fit.
Fourth, a couple of our holes in the base are slightly too large, which introduces play in the support rod's positioning and makes the structure less stable. Also a few of our acrylic support pieces, which link the two support rods to the central rotating crankshaft, have holes that need to be larger to ensure no drag on the crankshaft's rotation.
And since our Digikey LED order has yet to arrive, we are sure to face more challenges with the wiring of the lighting to the motor, assuming all of the aforementioned problems with the kinetic rotating assembly are solved.
After the materials were all fabricated and then adjusted as needed, we began to assemble the pieces, starting small. Loctite was used to glue aluminum pieces together, and nylon or acrylic washers were inserted everywhere a linkage piece would be moving against another to reduce friction.
The linkages took quite some time to assemble in the correct order so that no kinks were introduced when the horizontal rotational motion was applied. We next put the skin over the assembled 4-tier linkage system to get a feel of how the lamp would look illuminated.
After our basic materials were fabricated, some post-processing work needed to be done to cater them to our lamp design. The aluminum metal edges were filed down to eliminate burs that could catch. Also, each acrylic linkage and edge that would be pressing and pushing against the dragon skin was sanded down to reduce the edge sharpness. We learned from the first lamp prototype that this would be necessary so the acrylic doesn't rip the dragon skin.
From our kinetic linkage prototype, we learned that our crankshaft center rod needed to have stronger bonds so we chose to make the next version out of aluminum. This required us to visit the metal shop to fabricate each of our aluminum pieces to the thousandth of an inch. Below a part of the center cylindrical rod is being shortened on the lathe, but the mill and drill press were also heavily used.
We have four layers of linkages plus a lamp base and top that are all made of acrylic for its lightweight and easy to cut properties. We used the laser cutter to cut each acrylic piece.
We also cast a large sheet of dragon skin which is going to wrap around the exterior of the lamp and act as its skin. The drag skin we purchased requires 6 hours to set.
To change the desired speed of the motor for the lamp, we were required to build a motor timing and driving circuit. The circuit was prototyped with a variable power supply, 555 timer, potentiometer, and MOSFET. The circuit sends a fixed frequency PWM signal to the motor to vary the speed of the motor depending on the duty cycle, which is set by the potentiometer.
Once the circuit and motor were tested, the board was soldered with a 12V power supply and on-off switch added later.
Inspired by Theo Jansen's kinetic sculptures, we decided to have each of our platforms be a modified set of four-bar linkages. These horizontal kinetic linkage platforms are connected to a central rotating shaft where each linkage platform is offset in rotation from one another. As the shaft spins, the horizontal platforms translate and shift in an undulating manner. The Dragon Skin will then be stretched over these platforms and allow the shape of the lamp to shift and undulate. Our first prototype of two connected linkage platforms is pictured below.
1-9 of 9