A NASA technical bulletin, identified as TB 26-10, describes two instances of thick-film chip resistor body cracking observed during environmental acceptance testing of printed circuit board assemblies. Both failures occurred during thermal cycling that followed vibration testing, showing intermittent open circuits at hot dwell temperatures that recovered at room temperature. The resistors, of case size 2512, were installed near board edges on two identical boards, and cracks ran along the short axis, consistent with bending or tensile failure.

Failure analyses ruled out a defective production lot and fatigue as causes. The bulletin lists contributing factors instead: placement of large components near board edges with the long axis parallel to the edge, excessive solder reaching the top pads, board bow or twist, hard mounting to a rigid frame, and fastener torquing patterns that concentrate stress.

The recommendations include avoiding edge-parallel orientation of large ceramic components, performing finite element analysis for components placed near edges, controlling solder fillet height, inspecting solder joints, and running one to three thermal cycles with a hot-dwell functional test before vibration testing. The bulletin states that no direct root cause was identified.