摘要 : Existing nonisothermal thin-filament melt spinning models treat temperature as a one-dimensional (1D) quantity, varying only with axial distance. Such models accurately simulate processes in which the melt has high thermal conduct... 展开 Existing nonisothermal thin-filament melt spinning models treat temperature as a one-dimensional (1D) quantity, varying only with axial distance. Such models accurately simulate processes in which the melt has high thermal conductivity relative to the surface heat loss (i.e., small Biot number), or the ambient air temperature is near the melt temperature. However, in industrial melt spinning processes, where neither of these characteristics is true, a leading-order transverse temperature variation exists, no matter how slender the filament. Alternative models implement a hybrid 1D mass and momentum, 2D energy computation. Here we extend existing 1D models to include radial as well as axial resolution of temperature, thereby also providing a 2D resolution of stress, while retaining the computational advantages of 1D models. The model's predictions depend on the thermal conductivity of the melt, a property absent from existing 1D models. In addition to demonstrating the 2D resolution afforded by the new model, we also demonstrate significant differences between the new model and existing 1D models in predictions of fundamentally 1D quantities such as average axial stress, the quantity from which performance properties of the spun fiber are typically inferred. 收起