i meet some problem when i study this software,if someonge if familiar with it ,may i make a communicatin with you !

regards.

thank you very much

I am using that software from time to time. What is your question?

i have give you a email . thank you very much for your feedback!

how can we perform a steady computation with the NaSt3DGP

thank you for your feedback

You run a long time evolution until the solution doesn't evolve. Then you will get a steady computation if it has a steady solution.

thank you very much !

but i saw the time step changed automatically ,how can i set a time step that doennot change .

regards

check the file named as navcalc.cxx in the src direction.

//####################### //## Time-Step Control ## //#######################

NS_REAL NavierCalc::TimeStep() {

int i,j,k,l,oldtimer=timer->Start(6,Par) ;

NS_REAL deltneu=S.deltmax ;

static NS_REAL xmin=1e15,ymin=1e15,zmin=1e15 ;

if (xmin==1e15 && ymin==1e15 && zmin==1e15) {

PIJKLOOP {xmin=min(xmin,DX[i]); ymin=min(ymin,DY[j]); zmin=min(zmin,DZ[k]);}

}

// local CFL-condition including volume forces

PIJKLOOP {

deltneu=min(deltneu, S.tfconv*2/(fabs(U[i][j][k])/DXM[i]+

sqrt( (U[i][j][k]*U[i][j][k])/(DXM[i]*DXM[i])+4*fabs(S.g[0])/(S.froude*DXM[i])) +1E-15));

deltneu=min(deltneu, S.tfconv*2/(fabs(V[i][j][k])/DYM[j]+

sqrt( (V[i][j][k]*V[i][j][k])/(DYM[j]*DYM[j])+4*fabs(S.g[1])/(S.froude*DYM[j])) +1E-15));

deltneu=min(deltneu, S.tfconv*2/(fabs(W[i][j][k])/DZM[k]+

sqrt( (W[i][j][k]*W[i][j][k])/(DZM[k]*DZM[k])+4*fabs(S.g[2])/(S.froude*DZM[k])) +1E-15));

}

// Viscosity

deltneu=min(deltneu,S.tfdiff*S.re*0.5/(1./(xmin*xmin)+1./(ymin*ymin)+1./(zmin*zmin)));

if(deltneu<1E-15) {

printf("Maybe ERROR in parameters: Time-Step = 0.0 ! \n");exit(1);

}

if(S.TimeDis==1) { // Compute new time step (8 steps smoothing through parabola for AB)

deltneu=(S.timesteps<9)?((1-((S.timesteps+1)*0.1-1)*((S.timesteps+1)*0.1-1))*deltneu):deltneu;

}

timer->Stop(oldtimer,Par);

return deltneu; }

Hi all,

The NaSt3DGP is a multiblock or single block solver?

thanks