BuoyantKE湍流模型中浮升力源项的alpha

[袁宝强]

这是FLUENT中k方程浮升力生成项
$$G_{b}=-g_{i}\frac{\mu_t}{\rho Pr_t}\frac{\partial\rho}{\partial x_i}$$
把$\mu_t =\rho C_\mu k^2 /\varepsilon$ 带入上式以后应该是
$$G_{b}=-C_\mu g_{i}\frac{1}{ Pr_t}\frac{k}{\varepsilon}\frac{\partial\rho}{\partial x_i} k$$
但是openfoam buoyantKEpsilon.C中的形式却是
$$G_{b}=-C_\mu g_{i}\alpha\frac{k}{\varepsilon}\frac{\partial\rho}{\partial x_i} k$$
openfoam这里$\alpha=kappa/cp$么？感觉这里这个源项怪怪的。

另外。
下面是k方程源项的源码

buoyantKEpsilon<BasicTurbulenceModel>::kSource() const
{
    const uniformDimensionedVectorField& g =
        this->mesh_.objectRegistry::template
        lookupObject<uniformDimensionedVectorField>("g");

    if (mag(g.value()) > small)
    {
        return -fvm::SuSp(Gcoef(), this->k_);
    }
    else
    {
        return kEpsilon<BasicTurbulenceModel>::kSource();
    }
}

这里k方程的源项不应该是

    if (mag(g.value()) > small)
    {
        return kEpsilon<BasicTurbulenceModel>::kSource()-fvm::SuSp(Gcoef(), this->k_)
    }

[李东岳]

@袁宝强
你看openfoam里面H文件写的，尤其是1/Prt

This implementation is based on the density rather than temperature gradient
    extending the applicability to systems in which the density gradient may be
    generated by variation of composition rather than temperature.  Further, the
    1/Prt coefficient is replaced by Cg to provide more general control of
    model.

除此之外，Fluent看起来像是单相流kEpsilon模型。有可能他们在多相流里面也是用的单相流模型。OpenFOAM目前都是多相流模型。如果把$\alpha$去掉，再考虑1/Prt 的变化，二者应该是一致的了

[袁宝强]

@李东岳 谢谢，我再认真读一下