Finite element based thermal analysis of sealed electronic rack and validation

Electronic equipment with sealed enclosures are used in many applications where ruggedness and harsh environment are governing factors. However, the reliability of such equipment will depend on the effective heat rejection capability. Thus, an optimized thermal design is of paramount importance. In the present paper, thermal analysis of a sealed electronic rack is described. Finite Element code NISA is used for the analysis. The 3D - FLUID module of NISA is used for computations. The rack contains PCBs populated with electronic components of specified temperature ratings. The PCBs are arranged in vertical orientation with small gaps in between. The rack is divided into two compartments separated by a common partition having openings for cables from lower compartment to the upper one and also for airflow across the compartments. There are other heat sources like power supply and CPU in the rack. The enclosure is made of aluminum and is made airtight so that internal air cannot communicate with outside environment. The rack is not provided with any forced cooling internal or external. Validation of the computed results was done by conducting experiments on an actual rack. The internal temperatures were measured at specified locations and compared with the computed values of temperatures at corresponding locations. The electronic components on PCBs were rated for required performance at specified ambient temperature. However, the ambient temperature may not define the true thermal condition of the component. The rating of the components may be based on its surface temperature or some electrical characteristics. In the present analysis, the temperature gradient between surfaces of the selected component at the specified points in the surrounding ambient air inside the rack is computed and compared with corresponding measured temperature gradient. Temperature contours along two cut planes across higher temperature zones were extracted by post processing of computed temp- rature data. The temperature at these selected locations on the cut planes are compared with the measured temperature values.