I have a pipe with radius $R_1$ with a smaller pipe within (radius $R_2$). I want to find an expression for the flow through the pipe which flows in the annulus area. However, the known values are the dimensions of the pipe, including the $R_1$ and $R_2$, the length of the pipe (L), and the pressure at each side of the pipe ($p_1$ and $p_2$).

The fluid is fresh water at 10 degrees.

I tried to use Poisseulles law under the assumption of laminar flow: $Q = \frac{(p_1-p_2) \cdot \pi \cdot (R_1^4-R_2^4)}{8 \cdot \mu \cdot L}$

However, this yields an unrealistically high velocity, and flow rate, which I guess means that it is not laminar flow.

So to Calculate the flow in turbulent, I use Darcy-Weisbach : $Q = \Delta p = f \cdot \frac{L}{D} \cdot\frac{\rho}{2} \cdot v_{avg}^2$

Where I find the friction factor in a moody diagram from a Reynolds number where the velocity is neglected.

$Re = \frac{\rho \cdot v_{avg} \cdot D_h}{\mu} = \frac{\rho \cdot D_h}{\mu} \approx 18000 \rightarrow f \approx 0.03 $

I am not sure this is a good way to estimate the Reynolds number. Does anyone have a better way?

Furthermore, the opening at the end of the pipe is controlled by the pressure how can I include this? enter image description here

  • 1
    $\begingroup$ The velocity is not neglected in the Reynolds number. $\endgroup$
    – Solar Mike
    Commented Jan 18, 2023 at 6:21
  • $\begingroup$ No thats just to have a reference $\endgroup$
    – Nil
    Commented Jan 18, 2023 at 15:46

2 Answers 2


For the pressure drop in pipe with annular section, you need to consider hydraulic diameter $D_h$, which is 4 times the ratio of section area to wetted perimeter:

$$D_h = 4\cdot \frac{\pi\cdot \left(R_1^2-R_2^2\right)}{2\cdot \pi\cdot \left(R_1+R_2\right)} = 2\cdot \left(R_1-R_2\right)$$

If $p_2$ is the pressure behind the spring valve, you need to define the pressure just before it $p_{valve}$. When all the pressures are equal, the valve is probably closed with force $F_{closed}$. Now in order to achieve some flow, you need to open the valve, so $p_{valve}-p_2$ has to generate a force $F_p$ higher than $F_{closed}$. For this threshold, you can use static pressure analysis, i.e. $p_1 = p_{valve}$:

$$F_p = \pi\cdot R_1^2\cdot \left(p_{valve}-p_2\right) > F_{closed}$$

Pressure drop at the valve will surely depend on how much the valve is opened (distance $d$), which can be controlled by linear spring with stiffness $k$:

$$d = \frac{F}{k} = \frac{F_p-F_{closed}}{k} = \frac{\pi\cdot R_1^2\cdot \left(p_{valve}-p_2\right) - F_{closed}}{k} = \frac{\pi\cdot R_1^2\cdot \Delta p_{valve} - F_{closed}}{k}$$

The local pressure drop at the valve can be expressed using minor losses formula and it depends on fluid density $\rho$, fluid velocity $v_{valve}$ and a minor loss coefficient $\zeta$:

$$\Delta p_{valve} = \zeta\cdot \frac{\rho}{2}\cdot v_{valve}^2$$

If you express fluid velocity at the valve using volumetric flow $v_{valve} = \frac{Q}{d\cdot 2\cdot \pi\cdot R_1}$ and combine the 2 equations above, you can derive relationship between volumetric flow and opening distance:

$$Q = \sqrt{\frac{8\pi}{\rho\cdot \zeta}\cdot \left(k\cdot d^3+F_{closed}\cdot d^2\right)}$$

As you can see, there is no easy direct solution for this problem, so you will have to use some iterative method. I would recommend using bisection method, where the volumetric flow has to be between zero and $Q_{no\;valve}$, a value you would obtain if there was no valve at all from the pressure difference $p_2-p_1$.

The biggest unknown in this is the minor loss coefficient $\zeta$. Also, be aware about the fact that dynamic pressure of the flow can also contribute to the opening of the valve.


I believe this may be solved by using Bernoulli's Equation Simply expressed Q1=Q2 Inflow must equal outflow regardless of size and shape of the pipe.

The annulus sectional area is all that matters. $$\frac{{^𝑃}{1}} {\rho g}+\frac{{^{V^2}}{1}}{2g}=\frac{{^p}{2}} {\rho g}+\frac{(\frac{{^A} {1}}{{^A} {2}}×{^V} {1})^2} {2g}$$ I hope this this will help answer your question.

Ref:Douglas, J. F., et al. (2005). Fluid Mechanics, Pearson Education Limited.

  • $\begingroup$ How will you define flow from this? the velocities v1 and v2 must be identical right? $\endgroup$
    – Nil
    Commented Jan 18, 2023 at 15:45
  • $\begingroup$ Velocities do not have to be identical. $\endgroup$ Commented Jan 23, 2023 at 10:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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