Results
After assembling the prototype, it became apparent that with the current set-up, compliance wasn’t working as designed. Because the real-life belt tension was considerably higher than anticipated in the analyses, pretensioning bent the spring steel tensioners more than expected. As a result, the tensioners couldn’t bend as much to give compliance as designed, and many pucks could not hit their hard stops.
In addition, assembly proved to be particularly difficult in this iteration of the design, as the belt wound very tightly in the grooves of the plate. This, combined with the lack of room for the tensioners to give compliance, made it difficult to properly align the belt with the puck pulley stock. Many of the pucks weren’t in good alignment and couldn’t be corrected easily.
Thirdly, the lack of a moving tensioner between the cam and the first puck made it such that spare belt would “pile up” during the cam’s rotation.
Finally, imperfections in the timing belt’s advertised width led to it binding between some of the puck flanges.
Long-term cycling and thermal performance has yet to be tested, but will be investigated in the near future.
Project Description and Requirements
The primary goals of the project were to improve on the original sliding U-bar prototype focusing on three main objectives. The new prototype should:
be able to withstand and properly function in temperature extremes of -55°C and +175°C, and each DUT test site should be within 3°C of the set point.
operate upwards of 11 million cycles (back and forth rotation) without showing visible wear or requiring significant maintenance.
accurately and individually rotate each DUT by 90° with an error of no more than 1°.
Secondarily, the new prototype would:
incorporate a sensing mechanism to monitor the rotation of each puck to ensure that each rotated the correct amount.
use only pneumatics to provide the rotation force and/or not require modification to the existing E-chain used to power the shuttle underneath the kit plate.
consistently use metric fasteners and dimensions.
Motivating Ideas Behind Final Design
eliminate all sliding parts to reduce frictional wear.
eliminate all air pockets and mechanisms between the underside of the pucks and the hot plates and cold channels at the test sites to improve thermal conductivity.
Final Design
The design solution will consist of a pneumatically driven cam pulley connected via a timing belt to the eight pucks. The eight aluminum pucks will each contain a pocket for holding a DUT and span the height of the kit plate to provide the best thermal path between the shuttle and the DUT. Each puck will be held by a large, thin-section bearing near the top surface of the plate, be driven by a timing belt across its midsection and be fitted with a flange near the bottom to restrict its rotation. The pneumatic actuator will be attached to the kit plate by a bearing and to the cam pulley by another bearing with the intention of eliminating sliding components otherwise created by the linear actuation. Because each puck will be coupled through the timing belt to its neighbor, and because each puck must reach the correct angle independently, a pre-tensioned spring steel roller compliance mechanism will also be included. The spring steel will be designed to flex under the tension created by the cam pulley and actuator and provide
the timing belt with enough slack to allow its pair of pucks to rotate independently to a degree. A series of precision steel bars will be placed such that the flanges on each puck may collide with them while rotating to stop their motion at the correct angle in either direction. The Final Design page gives greater detail of the purpose of each design component as well as models and animations to illustrate the component behavior and interactions. The Quartus Engineering page gives greater detail of the project motivation and first prototype iteration and the Executive Summary page gives a concise explanation of the entire project.
Underside of the UCSD Design Team 9 final prototype