[jiahui_93]

hi, i'm cfd newbie. May i know why is it most cases using car as stationery object and air as moving object instead of car moving in stationery air as real case?
If i want to observe how the airflow and pressure flow behave over time when the car travel from start to end of tunnel. Is it acceptable if i set the car wall as moving wall, start of tunnel as inlet velocity, end of tunnel as outlet pressure, other regions of air in tunnel as zero gradient, and tunnel wall as stationery no slip wall?
Sorry for my silly question but I really need help in this matter.
Thanks for help =)

[FMDenaro]

Originally Posted by jiahui_93

hi, i'm cfd newbie. May i know why is it most cases using car as stationery object and air as moving object instead of car moving in stationery air as real case?
If i want to observe how the airflow and pressure flow behave over time when the car travel from start to end of tunnel. Is it acceptable if i set the car wall as moving wall, start of tunnel as inlet velocity, end of tunnel as outlet pressure, other regions of air in tunnel as zero gradient, and tunnel wall as stationery no slip wall?
Sorry for my silly question but I really need help in this matter.
Thanks for help =)

Galileian invariance is supposed to let you consider a fixed object in a moving fluid. In the case of the car, the difference appears in the BL at the ground but in some special wind tunnel there is care for that.
In the case of simulation of a car in a tunnel you could consider the car fixed and the walls of the tunnel moving.
However, if you are able to simulate the complete movement of the geometry of the car over a computational grid in the tunnel that would be ok.

[jiahui_93]

Originally Posted by FMDenaro

Galileian invariance is supposed to let you consider a fixed object in a moving fluid. In the case of the car, the difference appears in the BL at the ground but in some special wind tunnel there is care for that.
In the case of simulation of a car in a tunnel you could consider the car fixed and the walls of the tunnel moving.
However, if you are able to simulate the complete movement of the geometry of the car over a computational grid in the tunnel that would be ok.

Thanks for your reply For the case car is moving vertically in a narrow tunnel with 20 m/s, i set BC as following and get time step continuity error. I tried using movingWallVelocity uniform (0 20 0) and reduced the error to ^-15 to ^-20, but the result appeared to be wrong which shows no airflow throughout the tunnel. May I get advice on what should i do to solve this? thanks

internal field, U= uniform ( 0 20 0), P= uniform 0,
inlet, U= uniform ( 0 20 0), P=zeroGradient
outlet, U = zeroGradient, P= totalPressure P0=0
tunnelWall, U=movingWallVelocity uniform (0 0 0), P=zeroGradient
carWall, U= noSlip, P=zeroGradient
and zeroGradient for all other turbulent fields (eg: nut, k, epsilon)

Courant Number mean: 0.335262 max: 2.83168
smoothSolver:  Solving for Ux, Initial residual = 0.694223, Final residual = 9.76037e-06, No Iterations 102
smoothSolver:  Solving for Uy, Initial residual = 0.74074, Final residual = 9.45336e-06, No Iterations 107
smoothSolver:  Solving for Uz, Initial residual = 0.814794, Final residual = 9.14118e-06, No Iterations 111
GAMG:  Solving for p, Initial residual = 0.735763, Final residual = 0.0681959, No Iterations 10
time step continuity errors : sum local = 0.0811912, global = -0.0341982, cumulative = -0.0266287
GAMG:  Solving for p, Initial residual = 0.234852, Final residual = 9.16891e-07, No Iterations 115
time step continuity errors : sum local = 2.3507e-06, global = 6.11098e-07, cumulative = -0.026628