Analysis Driver Design for Electronic Systems
Recent advances in electronic hardware have been a boon to users of the devices but have made engineering systems that contain new components more difficult. The most challenging aspect of this is that more and more heat is created and at the same time there is always less space to handle the excess heat. This paper presents how Analysis Driven Design (ADD) can be used to quickly and accurately solve these problems with some discussions on approach followed by some specific case studies. The foundation of ADD for thermal problem solving is choosing the proper tools and developing methodologies for their proper application by applying the following process: 1) define the design questions, 2) determine the proper level of simulation, 3) choose the proper simulation tools, 4) develop a simulation plan, 5) carry out the simulation and 6) implement information from simulation. The tools chosen vary from hand calcs to complex spreadsheets to full-blown commercial Finite Element or Finite Difference packages. Another approach that is heavily relied upon is the use of parametric representations to allow for fast iterations and sensitivity studies. For parameterization of geometry, parametric CAD tools are used or the geometry is modeled in the analysis package using the parametric capabilities found within. In addition, the same parametric capability is used with scripting to parameterize loading, constraints and material properties. Some case study examples to be discussed include a cooling system designed for a telecommunication device. A coupled fluid/thermal analysis was run within a single package and parameterization was used to allow for quick and accurate design iterations. The resulting solution was then applied to the system and recent testing has shown it to be very successful. Another case study example: how transient thermal analysis was used to solve a cracking problem caused by repeated heating of a high power diode used in an optical communication package. This involved the coupling of thermal and structural simulation to minimize the stresses caused by the repeated cycling of the diode.