Coriolis Weather Tank
Coriolis Weather Tank is another exhibit for the Science Center of Singapore. This time it’s a brand new exhibit for us. Charles Sowers had prototyped it, and my role was to bring it to fruition as a robust, reliable exhibit.
Primary challenges:
Strong enough to withstand abuse. It needed to tolerate a person sitting on the top surface. Discourages sitting on the top surface and being spun my friends.
Easy to tune to air/water ratio.
Electrical power transfer from the floor to the spinning pump.
Brakes on high speed motion, but spins with low drag at low speeds.
Cylindrical shroud, per the environmental designer’s specification
Pinch hazard safe, while being tolerant of loose tolerances for assembly and fabrication
Features:
The HDPE basin is supported by a 3/8” aluminum stiffener. There are 4 pillars integrated into the frame that support the stiffener when a large load is applied and the stiffener flexes. There are Delrin feet that provide a soft, quiet surface should the load be applied when it’s already spinning. FEA simulation was done to assure that even the worst case situation of a guest sitting exactly between the nearest two pillars would not flex the aluminum or acrylic to the yield point.
Three position valves are incorporated for shutting off, or redirecting flow in the fluid circuit. These allow relatively easy access which allows an open reservoir air purge. Water flows through the system, entrains air bubbles as it goes, and exits into an open-top bucket. A booster pump pulls through a filter from the middle of the water in the bucket. This means bubbles and light debris float to the surface, and heavy debris falls to the bottom. The filter catches debris suspended in the middle volume the water. After some amount of re-circulation, the tank is effectively free of air.
Hollow main shaft with slip-ring installed in adapter in bottom allows electrical power to flow to the pump, which is mounted under the spinning platform.
An Eddy-Current brake concept was tried. It’s under-powered, so would need more magnets to provide a sufficient high speed braking effect. I have not yet tracked deceleration vs rpm, and would want to map that against the control of the same system with the magnets removed to see the difference to tune the effect some more. The exhibit works well, and spins freely enough at low speeds, and takes 10 seconds or so to slow down from very high speeds to functional speeds. This seems reasonable from a user perspective.
Uhg, this rolled steel shroud was a bit of a nightmare. The fabricator missed the mark a little, and my design really needed to be a little more tolerant to really do this gracefully. The circumference was off by ~ 1/8” each, and the height was off by ~3/16” on both the top and bottom sides. The issue of the slightly undersized circumference of the shroud was remedied with a special bracket system with cams and actively tensions the shroud as the bolts are tightened. The shroud was also a little tall, so forklift clearance needed some skids designed and installed on the bottom of the frame to support the load on the frame, rather than the shrouds. Likewise, the Delrin skids on the top needed to be stood off to allow them to contact the rotating stiffener before the shroud did. Corona Virus meant there was a significant lag between receiving the shroud and realizing the issues it had. They were solved, and the exhibit works just as well as i would have, but it was a bit of a headache.
A rotating ring provides something of a “labyrinth seal” that has a 1/8” clearance radially, but overlaps the gap between the shroud and the rotating top by >3/4”. This means the table can flex around with abuse and still not expose the shearing/pinch hazard.
FINAL VIDEOS
The final Exhibit is demonstrated below by Colin Wang, my friend and colleague at the Exploratorium. It’s really his exhibit, but i’m proud of my contribution to the x/y stage, control software.
These videos are intended for the SCS staff to review the final exhibit, address any last minute concerns, and demonstrate the maintenance associated with each exhibit.