[duguyoudi]

Dear all,


When I tried to simulate urban boundary layer, I found that the imposed velocity fluctuations decrease sharply near the inlet. At the inlet, groovyBC was used:

inlet
    {
        type            groovyBC;
        variables  
                        (
                          "Uref=3.654;"
                          "zref=0.2;"
                          "alpha=0.31;"
                          "zh=pos().z/zref;"
                          "Uxm=Uref*pow(zh,alpha);"
                          "a1=-0.09365;"
                          "b1=0.1448;"
                          "c1=0.06564;"
                          "a2=0.8404;"
                          "b2=0.4765;"
                          "c2=0.5241;"
                          "a3=0.3152;"
                          "b3=0.1186;"
                          "c3=0.1319;"
                          "deltau=a1*exp(-pow((pos().z-b1)/c1,2)) + a2*exp(-pow((pos().z-b2)/c2,2)) + a3*exp(-pow((pos().z-b3)/c3,2));"
                        );
        valueExpression "vector(Uxm+randNormal()*deltau, 0, 0)";

    }

Please check the attached figures for velocity fluctuations at the inlet (x = -1.1, figure 1.1), and two locations near the inlet (x=-1.08, figure 1.08), (x=-1.02, figure 1.02). The x range of the block is from -1.1 to 1.9.


Does anyone have some ideas about this?


Thanks,


Leo