My task is to determine the size of piping needed in a system. In this system, hydrogen gas flows horizontally in a $\frac{3}{4}\,\mathrm{inch}$ pipe and into a vertical pipe of a larger diameter. My constraint is that the velocity of the hydrogen cannot exceed $1\,\mathrm{m/s}$ once it expands and goes up the vertical tube.

| (2)
|__________________ (1) <- Hydrogen Gas

The pressure in the horizontal tubing at 1 is $2\,\mathrm{psig}$ and the pressure in the vertical tubing 2 is $0\,\mathrm{psig}$ (open to atmosphere).

The 2 equations I attempted to use were:

  1. Bernoulli's equation:$$\frac{P_1}{\rho} + \frac{1}{2} V_1^2 + gh_1 = \frac{P_2}{\rho} + \frac{1}{2} V_2^2 + gh_2$$

  2. equation of continuity:$$\rho_1A_1V_1 = \rho_2A_2V_2$$

Using Bernoulli's equation, I receive a very large negative root or a velocity of about ~550m/s in section 1 which seems very ridiculous. Is there a better suited equation for this application? The goal is to determine the size of piping needed for section 2.

  • 2
    $\begingroup$ Psig is psi gauge, it is in reference to atmospheric pressure. Because the tube reads 2 psig, it is 2 psi above atmospheric pressure. $\endgroup$
    – AkzoNorman
    Sep 24, 2015 at 14:35
  • 2
    $\begingroup$ @AkzoNorman: Bernoulli's equation is only applicable to imcompressible fluids, not gases. Have a look at this: en.wikipedia.org/wiki/… $\endgroup$
    – Gert
    Sep 24, 2015 at 15:55
  • 1
    $\begingroup$ @Gert Most gases at low Mach numbers can be modeled as incompressible. Moving at 1 m/s is most decidedly low Mach and so it is not an issue to model this as an incompressible fluid (and note -- fluid means liquid, gas, or plasma). $\endgroup$
    – tpg2114
    Sep 24, 2015 at 15:56
  • $\begingroup$ @AkzoNorman Have you ever heard of choked nozzles? Your very large velocity would indicate that the incoming pipe has to have supersonic flow to meet your maximum velocity in the much larger pipe. This is pretty common in experimental setups. $\endgroup$
    – tpg2114
    Sep 24, 2015 at 15:58
  • $\begingroup$ I have heard of choked nozzles but have not done any real work or calculations with them. Is the large number I am getting the actual velocity needed for 1 m/s maximum velocity in the larger pipe or am I applying the wrong equation to this scenario? (I used Bernoulli's still because it was a very low velocity) $\endgroup$
    – AkzoNorman
    Sep 24, 2015 at 16:06

2 Answers 2


I used a numerical model to estimate flow parameters. Here are a few options for pipes of different diameters at the outlet. The problem is unsteady, the Bernoulli integral is obviously not preserved. The output velocity first increases, then reaches a maximum and decreases. It is not clear which speed should be limited. In the above figures, the flow velocity is normalized to the speed of sound at the input, so even the smallest output velocity is about 10 m / s. The animation shows how the modulus of the flow velocity changes with time.

Figure 1

Fig. 2 shows the modulus of the current velocity (left), the maximum output velocity depending on time and the distribution $p+\frac {1}{2}\rho \vec {v}^2$ at the last moment. It can be seen that the Bernoulli integral is not conserved due to the large density and pressure gradient in the thin pipe. Figure 2

And a couple of pictures that show the same thing, but the Reynolds number is increased 10 times. It can be seen that a vortex was formed in the thick pipe, which led to the blocking of part of the pipe.

Figure 3

Figure 4


You can try using a modified form of the Poiseuille equation.

$$u_2 = \frac{q_2}{A} = \frac{(p_1 - p_2)R^2}{8 \mu_{av} L} \frac{z_2(p_1+p_2)}{2z_{av}p_2}$$

where $\frac{z_2(p_1+p_2)}{2z_{av}p_2}$ is a correction factor that accounts for the gradual expansion of the gas when going from the high pressure inlet to the low pressure outlet. See this link for the compressibility correction factor derivation: http://oaktrust.library.tamu.edu/bitstream/handle/1969.1/ETD-TAMU-2010-12-8739/LING-DISSERTATION.pdf. view pages 26 through 34 (of 243).


Your Answer

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

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