Finite Element Heat Transfer Analysis of a Cylindrical Body in an Enclosed Volume Subjected to Solar Radiation

Abstract

In aerospace industry, temperature of structures protected in an enclosed volume that are subjected to solar radiation should be known for critical applications. For instance, temperature of solid rocket engines (SRE) before they are ignited is critical for their operations. Therefore, numerical and experimental methods have been used to determine temperature of the SREs. In this study, a novel methodology, which gives the accurate temperature data of bodies placed in enclosed volumes like SREs in terms of hours instead of performing computational fluid dynamics (CFD) for days, was developed. The study included transient uncoupled heat transfer analysis with finite element (FE) method to predict the temperature distribution on a hollow cylindrical body located in a rectangular prism volume and the FE results were compared with a field test. Hourly solar radiation on a horizontal surface and hourly temperature values were measured by pyranometer to obtain the inputs for FE heat transfer analysis. For the tilted surfaces, solar radiation values were calculated by using the data obtained from pyranometer measurements. Absorptivity of enclosed volume was taken into account to determine the actual heat flow in the areas exposed directly to sunlight. Also, ground properties were selected identical to test condition to represent proper ground reflected radiation. Thermal conductance between the inner surfaces of enclosed volume and outer surfaces of cylindrical body has been defined to enable heat transfer mechanism between two separate components. In consequence of having accurate thermal conduction interactions between the inner surfaces of the FE model, CFD calculations of natural convection which was taken place inside the enclosed volume was eliminated. When the FE analysis-test comparison was concluded, it was observed that the calculated temperature values were found to be close agreement with measured temperatures and maximum error was within 10%. Furthermore, computation time was significantly reduced.