Often times in addition to regular project work, surprises emerge that require immediate attention. Recently, a sudden issue came up with one of my tasks regarding a bonding issue. Essentially, the bond between a thermal spreader plate and a panel did not cure to specifications. The original bond was supposed to be a uniformly distributed 15 mil layer between the plates. However, upon inspection post curing, there were sections of the thermal spreader that could be depressed (essentially bubbling off surface), indicating that the bond did not set correctly and that there were voids. The concern was that because the bond wasn’t uniform as intended, the added thermal resistance would cause the electronics mounted to the thermal spreader plate to overheat. The urgency was that the panel (which had the imperfectly bonded plate affixed to its surface) was due to be delivered the next day and it was an issue of whether to accept the plate as is, or return it for re-work.
The hardware supplier had said that based on their thermal model, even with a dry bond (no interface material), once the electronic boxes were all bolted down to the panel, sandwiching the thermal spreader, the electronics would still be within their AFTs. However, the project management wanted someone on the JPL side to validate this. Unfortunately, my model of the panel is still in development, so it could not be used on such short notice. In order to have some validation of the hardware suppliers’ analysis, a simple ‘back of the envelope’ calculation was done. This was based on the bolt pattern of the hardware as well as the proposed bonding of the two plates. The calculation yielded a predicted change in temperature between the two plates based on the thermal resistance between them. To facilitate this calculation, it was entered into a spreadsheet so various parameters could be changed yielding quick results. By checking a range of assumed bonds (ex: complete bond, 75% bond/25%void, 50% bond/50% void, etc) even with a 10% bond and 90% void, the change in temperature across the two plates was small enough to allow for sufficient margin for the electronics boxes AFT requirements. Based on this confirmation of the hardware supplier’s analysis, the panel was accepted. When my model of the panel is complete, this study will be further verified to ensure the safety of the electronics boxes.