隧道火灾不同CFD模型的对比分析及试验验证

Comparative Analysis and Experimental Verification of Different CFD Models for Tunnel Fire

为筛选出适用于隧道火灾模拟、科学经济的CFD(Computational Fluid Dynamics,计算流体动力学)模型,为复杂结构的隧道内发生火灾时的数值模拟提供科学参考,基于Heskestad羽流理论,提出传质体热源模型。使用FLUENT软件对比不同物理模型、燃烧模型等的计算结果,并开展缩尺试验进行验证。结果表明,不扩大计算域可能造成顶棚下方温度的不对称分布,而隧道两端扩大的计算域能够消除不合理的边界条件对隧道内温度场的影响。布辛涅斯克(Boussinesq)近似理论假设空气物性不变,但实际上空气的热扩散系数会随着温度的升高而提高,因此使用布辛涅斯克近似的原则得到的温度与实际差异较大。提出的传质体热源模型能够较准确地预测隧道顶棚下方的最大温度,案例的计算误差在10%以内。然而,该模型没有考虑壁面对流散热和空气卷吸的情况,因此在一定程度上高估了纵向温度分布。

To select a scientific and economic CFD model suitable for tunnel fire simulation and to provide a scientific reference for numerical simulation of complex tunnel fire,based on Zukoski plume theory,Air Entrainment Model and Mass Transfer-Heat Source Model were proposed.Comparing the calculation results of different physical models combustion models,etc.using FLUENT software,and conduct scale tests for verification.The results show that the non-expanded computational domain may cause asymmetrical distribution of temperature.Expanded computational domain can eliminate the influence of unreasonable boundary condition on temperature field.Buossinesq approximation assumes that the physical properties of air remain unchanged,but in fact,the thermal diffusivity of air increases with temperature,so temperature predicted by Buossinesq approximation is lower.Because the heat dissipation by wall is not considered,temperature predicted by using the air physical property fitting parameters is higher.It is hard to control heat release rate using eddy dissipation model.Mass transfer-volumetric heat source model can predict the maximum temperature under the tunnel ceiling well,and the case error is less than 10%,but the model does not consider the wall heat dissipation and air entrainment,and overestimates the longitudinal temperature distribution.