摘要: The thermal response of fixtured parts in a batch-type brazing furnace can require numerous, time-consuming development runs before an acceptable furnace schedule or joint design is established. Powerful computational simulation t... 展开 The thermal response of fixtured parts in a batch-type brazing furnace can require numerous, time-consuming development runs before an acceptable furnace schedule or joint design is established. Powerful computational simulation tolls are being developed to minimize the required number of verification experiments, improve furnace throughput, and increase product yields. Typical furnace simulations are based on thermal, fluid flow, and structural codes that incorporate the fundamental physics of the brazing process. The use of massively parallel computing to predict furnace and joint-level responses is presented. Measured and computed data are compared. Temperature values are within 1-2% of the expected peak brazing temperature for different loading conditions. Sensitivity studies reveal that the thermal response is more sensitive to the thermal boundary conditions of the heating enclosure than variability in the materials data. Braze flow simulations predict fillet geometry and free surface joint defects. Dynamic wetting conditions, interfacial reactions, and solidification structure add a high degree of uncertainty to the flow results. 收起