In the absence of surface tension, two incompressible viscous fluids are conned between a pair of parallel flat plates $y = -h$ and $y = h$. The region $-h<y<0$ is occupied with a fluid of density $\rho_1$ and viscosity $\mu_1$. The region $0<y<h$ is occupied with a fluid of density $\rho_2$ and viscosity $\mu_2$. A steady fully developed flow is set up with the lower plate $(y = -h)$ fixed and the upper plate $(y = h)$ moving with a constant velocity $U_0$ parallel to the x-axis. Prove that the velocity distribution in the two fluids is: $$\boldsymbol{u} = u(y)\underline{\hat\imath}\quad\text{where}\quad u(y) =\left\{ \begin{array}{11} \frac{\mu_2}{(\mu_1 + \mu_2)}U_0(1+\frac{y}{h})\quad-h\le y\le0\quad(1)\\ \frac{\mu_1}{(\mu_1 + \mu_2)}U_0(\frac{\mu_2}{\mu_1}+\frac{y}{h})\quad\,\,\,\,0\le y\le h\quad(2)\\ \end{array} \right.$$

I write the problem out: $$\boldsymbol{u} = u(y)\underline{\hat\imath} =\left\{ \begin{array}{11} u_2(y)\underline{\hat\imath}\quad\,\,\,\,\,\,\,0\le y\le h\quad(1)\\ u_2(y)\underline{\hat\imath}\quad-h\le y\le 0\quad(1)\\ \end{array} \right.$$ I have the boundary conditions: $$y=-h,\quad u=0\\ y=h,\quad u=U_0$$ I am not given a pressure gradient so I assume $p=p_0$. I check continuity holds: $$\nabla\cdot\boldsymbol{u} = \frac{\partial u}{\partial x} = 0.$$ I then use N-S equations: $$\frac{\partial u}{\partial t} + (\boldsymbol{u}\cdot\nabla)\boldsymbol{u} = -\frac{1}{\rho}\frac{\partial p}{\partial x} + \boldsymbol{X} + \nu\nabla^2u$$ The body force $\boldsymbol{X}=0$ and $\frac{\partial u}{\partial t}=0$ because steady flow. This then comes down to: $$u\frac{\partial u}{\partial x} + v\frac{\partial u}{\partial y} = -\frac{1}{\rho}\frac{\partial p}{\partial x} + \nu\left(\frac{\partial^2 u}{\partial x^2} + \frac{\partial^2 u}{\partial y^2}\right)$$ Here, both $\frac{\partial u}{\partial x}$ and $v$ are $0$ so the LHS $=0$. And I think that the pressure is $p=p_0$ so that makes $\frac{\partial p}{\partial x} = 0$ leaving: $$\frac{\mu_2}{\rho_2}\frac{\partial^2 u_2}{\partial y^2} = 0,\quad\frac{\mu_1}{\rho_1}\frac{\partial^2 u_1}{\partial y^2} = 0$$


For fully developed, steady, unidirectional flow, the velocity component $u$ does not depend on $x$ and $v=0$. The flow is driven by the motion of the upper plate, so you can assume the pressure is constant. In this case, the Navier-Stokes equations reduce to

$$\frac{\partial^2u_1}{\partial y^2} = 0, \,\,\,\frac{\partial^2u_2}{\partial y^2} = 0$$

where $u = u_1$ for $-h \leqslant y \leqslant 0$ and $u = u_2$ for $0 \leqslant y \leqslant h$. Integrating twice we obtain,

$$u_1 = Ay +B, \,\,\, u_2 = Cy +D.$$

The natural boundary conditions are used to determine the constants $A,\, B,\, C,\, D$.

We require no-slip at solid surfaces

$$\tag{1}u_1(-h) = 0,$$ $$\tag{2}u_2(h) = U_0,$$

and continuity of velocity and stress at the interface:

$$\tag{3}u_1(0) = u_2(0), $$ $$\tag{4}\mu_1 \left.\frac{\partial u_1}{\partial y}\right|_{y=0} = \mu_2 \left.\frac{\partial u_2}{\partial y}\right|_{y=0}$$

Applying these conditions we get

$$\qquad \,\,\,\,\,\,(1) \implies -Ah + B = 0\\ \qquad \,\,\,\,\,\,(2) \implies Ch + D = U_0 \\ (3) \implies B = D \\ \quad \,\,\,\,\,(4) \implies \mu_1A = \mu_2 C$$

From (1) and (3) we get $D = B = Ah$, and from (4) we get $C = \mu_1A/\mu_2$ leading to

$$u_1 = Ay + Ah =Ah\left(1 + \frac{y}{h}\right), \\ u_2 = \frac{A\mu_1}{\mu_2}y + Ah =\frac{\mu_1}{\mu_2}Ah\left( \frac{\mu_2}{\mu_1} + \frac{y}{h}\right) $$

From (4) we get $U_0 = Ch +D = (\mu_1/\mu_2)Ah +Ah = Ah(1 + \mu_1/\mu_2)$ leading to

$$Ah = \frac{\mu_2 U_0}{\mu_1 + \mu_2},$$


$$u_1 = \frac{\mu_2 U_0}{\mu_1 + \mu_2}\left(1 + \frac{y}{h}\right), \\ u_2 = \frac{\mu_1 U_0}{\mu_1 + \mu_2}\left( \frac{\mu_2}{\mu_1} + \frac{y}{h}\right)$$

(At the start, a clue that inertial terms and the pressure gradient are negligible is the absence of the densities $\rho_1$ and $\rho_2$ in the answer given to you.)

| cite | improve this answer | |
  • $\begingroup$ in you're first line you have what i have but without $\mu_2/\rho_2$ etc. Why is this? $\endgroup$ – MRT Apr 15 '18 at 17:59
  • $\begingroup$ using the conditions I have: $$0 = -Ah + B,\quad U_0 = Ch + D$$ But how do I determine these? $\endgroup$ – MRT Apr 15 '18 at 18:04
  • $\begingroup$ Divide both sides by the nonzero constant $\mu_2/\rho_2$ and you get what I have. There are four boundary conditions including the two at $y=0$ to solve for the four constants. Does that make it clear or should I add some more. $\endgroup$ – RRL Apr 15 '18 at 18:18
  • $\begingroup$ $\frac{\mu}{\rho} \frac{\partial^2u}{\partial y^2} = 0 \implies \frac{\partial^2u}{\partial y^2} = 0$. I finished for you. Were you really unable to do the algebra for the constants or was there some other confusion? $\endgroup$ – RRL Apr 15 '18 at 19:03
  • $\begingroup$ I do just get lost in the algebra and can't see the subtle connections. Like after you have those 4 equations, seeing that to get $Ay$ I just need to do $Ah\cdot \frac {y}{h}$ can just go right over my head. And then going back to $U_0$ to find the fraction equal to $Ah$ I just can't see it sometimes. But you've helped clarify me that what I was thinking was good. I think and try for so long and end up questioning myself with how can this not be the right way? And it is but I'm not reading inbetween the lines $\endgroup$ – MRT Apr 15 '18 at 19:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.