bagnold_periodic_cohesif.c

Periodic free surface flow of a Plastic fluid
Code

Periodic free surface flow of a Plastic fluid

Problem

An example of 2D cohesive granular flow along an inclined plate (angle -\alpha) with a free surface is presented here. The configuration is periodic what is injected to the left comes from the right. On the bottom solid wall there is a no slip boundary condition, on the top fixed free surface a slip condition is imposed (and zero pressure). We use the centered Navier-Stokes solver with regularization for viscosity. This flow is a simple model for rock/gravels/sand wet avalanches along a slope.

It is a kind of Nusselt film (or half Poiseuille) non newtonian counterpart.

Equations for a granular cohesive fluid: \mu(I) rheology

The \mu(I) cohesive rheology supposes that the tangential and normal stress are proportional (GDR MiDi): \tau = \tau_c + \mu(I) P with a coefficient of proportionality that is a function of a single dimensionless number, called the inertial number I (square root of Savage Number…), and \tau_c the yield stress: I = \frac{d_g\dot \gamma}{\sqrt{p/\rho}} where d_g is the grain diameter and \dot\gamma reduces to \partial u/\partial y for an unidirectional flow (note that the final analytical result and numercial computations we present are a for a linearized case \mu=\mu_0+ a \frac{d \frac{\partial u}{\partial y}}{\sqrt{p/\rho}} with a=\Delta \mu/I_0).

To write this in tensorial form, \tau_{ij} is linked with D_{ij} the shear strain rate tensor (tenseur de taux de déformation) D_{ij}=(u_{i,j}+u_{j,i})/2, whose components in 2D are:

D_{11}=\frac{\partial u}{\partial x}, \text{ } D_{12} =\frac{1}{2}( \frac{\partial u}{\partial y}+ \frac{\partial v}{\partial x}), \text{ } D_{21} =D_{12} =\frac{1}{2}( \frac{\partial u}{\partial y}+ \frac{\partial v}{\partial x}),\text{ and }D_{22}=\frac{\partial v}{\partial y}

And where the second invariant is D_2=\sqrt{D_{ij}D_{ij}} hence D_2^2= D_{ij}D_{ij}= ( \frac{\partial u}{\partial x})^2 + (\frac{\partial v}{\partial y})^2 + \frac{1}{2}( \frac{\partial u}{\partial y}+ \frac{\partial v}{\partial x})^2

for a unidirectional flow \partial u/\partial y is \sqrt{2} D_2

The total stress is written in a pressure part and a dissipative part \sigma_{ij}= - p \delta_{ij} + \tau_{ij}

tangential stress is linked to the shear rate tensor by \tau_{ij} = \sqrt{2}(\tau_c +\mu(I) p)\frac{D_{ij}}{D_2} where the friction law compatible with the experiments and linearised for smalll I is: \mu(I)= \mu_0 + \frac{\Delta \mu}{ I_0} I the coefficients depend on the nature of the granular media \mu_0=0.38 \Delta \mu = 0.26 and I_0=0.3

Equivalent rheology

From this, one can exhibit an equivalent viscosity if we write: \tau_{ij} = 2 \mu_{eq} D_{ij} we thus define \mu_{eq}= \frac{\tau_c +\mu(I) p}{\sqrt{2} D_2 } This equivalent viscosity will be coded next (with a regularisation to avoid infinite viscosity at rest).

Exact solution in the proposed case

This problem admits an analytical solution called “Bagnold” profile in a steady case with no cohesion. We introduce the cohesion yield stress and obtain a kind of mofified Bingham fluid. This is a steady flow invariant in x. We look at an unidirectional flow, a pure shear flow u(y), v=0, so D_{11}=D_{22}=0 and D_{12}=D_{21}=(1/2)(\partial u/\partial y), this gives (mind square root of 2): D_2=\sqrt{D_{ij}D_{ij}} = \frac{1}{\sqrt{2}}\frac{\partial u}{\partial y}.

so that we have, as we wish: \tau_{12} = 2 (\tau_c +\mu(I) p) \frac{D_{12}}{ \sqrt{2}D_2} = \tau_c + \mu(I)p The component \tau_{12} is as we wanted \tau_c+\mu(I)p.

Equilibrium between pressure gradient and viscosity (writing \tau for a shorthand of \tau_{12}), the projection of the equations 0=\rho g \sin(\alpha) + \frac{\partial \tau}{\partial y} 0=-\rho g \sin(\alpha) -\frac{\partial p}{\partial y} as there is no stress and no pressure at the free surface y=h, the stress is \tau = \rho g \sin(\alpha) (h-y) and the pressure is p = \rho g \cos(\alpha) (h-y) the stress \tau and the pressure increase from the free surface where they are both 0. From those expresssions, the ratio of the \tau and p remains constant: \tau / p =\tan(\alpha), but by the Coulomb fricton \tau is \mu(I) p + \tau_c, hence: \mu(I) + \tau_c/p = \tan(\alpha) For sake of simplification we use here a linearised \mu(I) with a=\Delta \mu/I_0: \mu(I)=\mu_0+ a \frac{d \frac{\partial u}{\partial y}}{\sqrt{p/\rho}}, \text{ with equilibrium }\mu(I) = \tan(\alpha)-\tau_c/p hence the shear is (as I = d_g \frac{\partial u}{\partial y } /\sqrt{p/\rho} and p=\rho g \cos(\alpha) (h-y)) : \frac{\partial u}{\partial y }= (\frac{(\tan \alpha - \mu_0)\sqrt{g \cos \alpha}}{a d })\sqrt{ (h-y)} - \frac{\tau_c \sqrt{g \cos \alpha}}{a \rho g d \cos \alpha} \frac{1}{\sqrt{ (h-y)}} it is zero \frac{\partial u}{\partial y }=0 for y>Y, the yield height defined by Y= h -\frac{\tau_c}{\rho g \cos \alpha (\tan \alpha - \mu_0)}. We decompose the shear in two contributions, the Bagnold \frac{\partial u}{\partial y }\bigg\rvert_B one and the cohesive one: \frac{\partial u}{\partial y }\bigg\rvert_c as follows \frac{\partial u}{\partial y }= \frac{\partial u}{\partial y }\bigg\rvert_B + \frac{\partial u}{\partial y }\bigg\rvert_c with K_B =(\frac{(\tan \alpha - \mu_0)\sqrt{g \cos \alpha}}{a d }) and t_c=\frac{\tau_c }{\rho g h} ( \frac{\sqrt{g }}{d})\frac{1}{a \sqrt{\cos \alpha}}

The sum : \frac{\partial u}{\partial y }= K_B \sqrt{ (h-y)} - t_c \frac{1}{\sqrt{ (h-y)}}, with \frac{\partial u}{\partial y }=0 for y>Y with Y= h -\frac{t_c}{K_B} is integrated, for y<Y in: u=\frac{1}{3} \left(2 h^{3/2} K+2 \sqrt{h-y} (K_B y+3 t_c)-2 h K_B \sqrt{h-y}-6 \sqrt{h} t_c\right) or with t_c=K_B(h-Y) u=\frac{2}{3} \left(h^{3/2} K_B-h K_B \sqrt{h-y}+K_B y \sqrt{h-y}+3 t_c \sqrt{h-y}-3 \sqrt{h} t_c\right) with the plug flow velocity for y>Y: U=\frac{2}{3} \left(h^{3/2} K_B-3 \sqrt{h} t_c+2 t_c \sqrt{\frac{t}{K_B}}\right)

Code

just before, let us include the NS code, define the precision 2^N, define the angle 28.64 degrees

#include "navier-stokes/centered.h"
#define LEVEL 5
double mumax;
double dg;
# define alpha 0.6
# define tauc 0.125
    scalar mu_eq[],foo[];

now, come back to the velocity shear which is \frac{\partial u}{\partial y }= K_B \sqrt{ (h-y)} - t_c \frac{1}{\sqrt{ (h-y)}} with \frac{\partial u}{\partial y }=0 for y>Y with Y= h -\frac{t_c}{K} K_B =(\frac{(\tan \alpha - \mu_0)\sqrt{g \cos \alpha}}{a d }) and t_c= - \frac{\tau_c }{\rho g h} ( \frac{\sqrt{g }}{d})\frac{1}{a \sqrt{\cos \alpha}}

double dUbc( double y){
    double dU;
    double a = 0.26/.3;
    double h = 1;
    double KB = ((tan(alpha) - 0.38)* sqrt(cos(alpha))/(a*dg));
    double tc = tauc/(sqrt(cos(alpha))*(a*dg));
    double Y = h -tc/KB;
    dU=(y<Y? KB*sqrt(h - y) - tc/sqrt(h - y):0);
    return dU;
}

This gives the Bagnold’s solution if \tau_c=0 u = \frac{2}{3} \frac{I_\alpha}{d_g} \sqrt{g h^3 \cos(\alpha)} (1 - (1 - \frac{y}{h})^{3/2}) (note the numerical value for alpha=0.43, dg=0.04 U0 = 2.06631)

double Ub( double y){
    double U0=(sqrt(cos(alpha))*(-0.114 + 0.3*tan(alpha))/(dg*(0.64 - tan(alpha))));
    return ((1. - pow(1. - y, 1.5))*2./3.*U0) ;
}

This gives the Cohesive Bagnold’s solution if \tau_c \ne0 u = \frac{(2(h^{3/2})K_B-3\sqrt{h} t_c-h K_B \sqrt{(h-y)}+3 t_c\sqrt{h-y}+K_B \sqrt{h-y}y)))}{3}

U =\frac{2}{3} ((h^{3/2} K_B - 3 t_c \sqrt(h)+ 2 t_c \sqrt{(t_c/K_B)}))

double Ubc( double y){
    double a = 0.26/.3;
    double h = 1;
    double KB = ((tan(alpha) - 0.38)* sqrt(cos(alpha))/(a*dg));
    double tc = tauc/(sqrt(cos(alpha))*(a*dg));
    double Y = h -tc/KB;
    double U ,Uy;
    Uy=(2*(pow(h,1.5)*KB-3*sqrt(h)*tc-h*KB*sqrt(h-y)+3*tc*sqrt(h-y)+KB*sqrt(h-y)*y))/3.;
    U = 2*((pow(h,1.5)*KB - 3*tc *sqrt(h)+ 2*tc*sqrt(tc/KB)))/3.;
    U = (y<Y? Uy : U);
    return (U) ;
}

The domain is one unit long. 0<x<1 0<y<1

int main() {
  L0 = 1.;
  origin (0., 0);

Values of grain size

  dg = 0.04;

the regularisation value of viscosity

  mumax=1000;

Boundary conditions are periodic

    periodic (right);

slip at the top

    u.t[top] = neumann(0);
  //  uf.t[top] = neumann(0);
    u.n[top] = dirichlet(0);
 //   uf.n[top] = neumann(0);
    /* u.t[top] = neumann(0);
    u.n[top] = neumann(0);
    uf.n[top] = neumann(0);*/

no slip at the bottom

    u.n[bottom] = dirichlet(0);
   // uf.n[bottom] = dirichlet(0);
    u.t[bottom] = dirichlet(0);

note the pressure

    p[top] = dirichlet(0);
 //   pf[top] = dirichlet(0);
 // with NO BC it works!
   // p[bottom] = neumann(0);
   // pf[bottom] = neumann(0);
   // p[bottom] = neumann(cos(alpha));
   // pf[bottom] = neumann(cos(alpha));

the \Delta t_{max} should be enough small

  DT = 0.75;

because U=u(y) e_x. stokes true => no CFL condition

  stokes = true;
  run(); 
}

face vector muv[];

event init (t = 0) {

prepare viscosity

  mu = muv;

pressure gradient, gravity acceleartion mdpdx -\frac{\partial p}{\partial x} = \sin(alpha) -\frac{\partial p}{\partial y} = -\cos(alpha)

    const face vector mdpdx[] = {sin(alpha),-cos(alpha)};

  a = mdpdx;

Initialy at rest

  foreach() {
    u.x[] = Ub(y);
    u.y[] = 0;
    p[]=cos(alpha)*(1-y);
  }
}

We check the number of iterations of the Poisson and viscous problems.

//event logfile (i++)
// fprintf (stderr, "%d %g %d %d\n", i, t, mgp.i, mgu.i);

old value of the velocity is saved

scalar un[];
event init_un (i = 0) {
    foreach()
    un[] = u.x[];
}

so that when it does not more change we are converged

event conv (t += 1; i <= 100000) {
    double du = change (u.x, un);
    fprintf(stdout,"#time Ucal/Utheo Ucalc err \n %g %g %g %g \n",t,interpolate (u.x, L0/2, .999)/Ubc(1),interpolate (u.x, L0/2, .999),du);
    if (i > 0 && du < .50e-4) //was -6
        return 1; /* stop */
}

Implementation of the Bagnold cohesive viscosity

event bagnold(i++) {

Compute D_{11}=\frac{\partial u}{\partial x}, D_{12} =\frac{1}{2}( \frac{\partial u}{\partial y}+ \frac{\partial v}{\partial x}), D_{21} =D_{12} =\frac{1}{2}( \frac{\partial u}{\partial y}+ \frac{\partial v}{\partial x}), D_{22}=\frac{\partial v}{\partial y}

And where the second invariant is D_2=\sqrt{D_{ij}D_{ij}} hence D_2^2= D_{ij}D_{ij}= D_{11}D_{11} + D_{12}D_{21} + D_{21}D_{12} + D_{22}D_{22}

with I = d_g \sqrt{2} D_2 /\sqrt{p/\rho}, and as \mu(I)= \mu_0 + (\Delta \mu) I/(I_0 + I)

the equivalent viscosity is \mu_{eq}= \frac{\tau_c+\mu(I) p}{\sqrt{2} D_2 } the final one is the min of of \mu_{eq} and a large \mu_{max}, then the fluid flows always, it is not a solid, but a very viscous fluid.

    foreach() {
        mu_eq[] =    mumax;
        if (p[] > 0.) {
        double D2 = 0.,In,muI;
        foreach_dimension() {
            double dxx = u.x[1,0] - u.x[-1,0];
            double dxy = (u.x[0,1] - u.x[0,-1] + u.y[1,0] - u.y[-1,0])/2.;
            D2 += sq(dxx) + sq(dxy);
        }
        if (D2 > 0.) {
            D2 = sqrt(D2)/(2.*Delta) ;
            In = sqrt(2.)*dg*D2/sqrt(fabs(p[]));
            muI = .38 + (.26)*In/(.3 + In);
            muI = .38 + (.26)*In/(.3 );
            foo[] = D2;
            mu_eq[] =  min( (tauc + muI*fabs(p[]))/(sqrt(2.)*D2) , mumax );}
        else{
         mu_eq[] =  mumax;
          }
    }
    }
    boundary ({mu_eq});
    foreach_face() {
        muv.x[] = (mu_eq[] + mu_eq[-1,0])/2.;
    }
    boundary ((scalar *){muv});
}

Save profiles

event profiles (t += 1)
{
    FILE * fp = fopen("xprof", "w");
    scalar shear[];
    foreach()
    shear[] = (u.x[0,1] - u.x[0,-1])/(2.*Delta);
    boundary ({shear});
    for (double y = 0.; y < 1.0; y += 1./pow(2.,LEVEL))
        fprintf (fp, "%g %g %g %g %g %g %g %g %g \n", y, interpolate (u.x, L0/2, y), interpolate (shear, L0/2, y),
                  Ubc(y), cos(alpha)*(1-y),interpolate (p, L0/2, y),
                 interpolate (mu_eq, L0/2, y), interpolate (foo, L0/2, y),dUbc(y) );
    fclose (fp);
}

We adapt according to the error on the velocity field.

event adapt (i++) {
 // adapt_wavelet ({u}, (double[]){3e-3,3e-3}, 8, 6);
}

Results and plots

To run the program

 qcc -g -O3 -o bagnold_periodic_cohesif bagnold_periodic_cohesif.c -lm
 ./bagnold_periodic_cohesif
 
 lldb bagnold_periodic
 
 make bagnold_periodic_cohesif.tst;make bagnold_periodic_cohesif/plots;
 make bagnold_periodic_cohesif.c.html;open bagnold_periodic_cohesif.c.html;

bug from MacOS ~~~bash cp bagnold_periodic_cohesif.c.html XXX.c.html

sed -i - ‘s/\)//g’ XXX.c.html;sed -i - ‘s/\(//g’ XXX.c.html sed -i - ‘s/\[//g’ XXX.c.html sed -i - ‘s/\]//g’ XXX.c.html; mv XXX.c.html bagnold_periodic_cohesif.c.html open bagnold_periodic_cohesif.c.html


Plots of the velocity,  $\tau$ and $p$, the two last are linear as expected:

 
<div id="plot0">
~~~ {.bash}
 set xlabel "y"
 set xlabel "U, tau, p"
 p'xprof' u 1:2 t'U computed' ,''u 1:($7*$3) t'tau comp.','' u 1:6 t'p'

We check that (\partial u/ \partial y) is \sqrt{2}D_2 and that the computed gradient of velocity is the Bagnold one.