SEL Thermal Card Guide

The goal of this project is to develop a card guide design that can be applied in SEL equipment that has a net thermal resistance less than 1 ⁰C/Watt between an aluminum tray (.05"-.1" thick) and the chassis (.05"-.1" thick). This involves creating an instrumented test stand for measuring heat transfer across the card guide. Results should be reconciled against an analytical model to assure accuracy. The recommended solution should meet SEL standards for vibration.

Deliverables

 * Experimental test set-up that can accommodate different guide configurations
 * Experiment design and test results that demonstrate which guide configurations are most effective
 * Comprehensive project documentation for easy future reference by client
 * Proof of Concept guide design

SolidWorks Flow Simulation
To determine a suitable card guide deign our group used SOLIDWORKS to model heat transfer rates with Flow Simulation. This was then calibrated to match the experimental setup, so we could have a verified test apparatus once the guide was manufactured. The testing aparatus was then simplified to a two plate study, so we could control temperature boundary condition and solve for varying surface area, thickness, contact pressure, and material properties. All these factors were varied in parametric studies to determine the point where thermal conduction is optimal.

Test Setup
For our test setup we have created a scale version of the thermal card guide.

Test Results
After completing a test with the new material and thermocouples, convective heat loss does seem to cause issues when conducting the test in open air. Using the equation: R_t=(T_1-T_2)/q_x =L/kA We were able to calculate resistance values in the model. With temperature values obtained from the test T1 = 54⁰C and T2 = 36⁰C (see image below). The power input to the system was qx = 7.068 W. Measuring the dimensions of the tray and distance between the thermocouples we found A = 82.656x10-6 m2 and L = 0.135 m. Using these values the expected resistance value was Rexpected = 12.098 ⁰C/W. We have concluded that the convection heat loss is not allowing us to acquire correct measurements in open air. Looking at the resistors for our power input, they have fins around the entire housing allowing for greater convective loss resulting in power that isn’t being sent into the tray.