1. Not finding help here? Sign up for a free 30min tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Calculating velocity for water pressure

  1. Jan 27, 2015 #1
    1. The problem statement, all variables and given/known data
    upload_2015-1-27_22-7-7.png
    A tank of cross-sectional area ##A## contains a liquid having density ##\rho##. The liquid squirts freely from a small hole of area ##a## at a distance ##H## below the free surface of the liquid.
    If the liquid has no internal friction(viscosity), with what speed does it emerge?

    2. Relevant equations
    ##p = H\rho g ##
    ##F = \frac{\Delta (mv)}{\Delta t}##
    3. The attempt at a solution
    ##F = pa = aH\rho g ##
    Again, ## F = \frac{\Delta (mv)}{\Delta t} = v \cdot \frac{\Delta m}{\Delta t}## ; because, ##v## is constant;
    Now, ## \frac{\Delta m}{\Delta t} = \frac{a (\Delta x) \rho }{\Delta t} = a \rho v ## ; because, ##\frac{\Delta x}{\Delta t} = v ##
    So, ## aH\rho g = v \cdot a \rho v = a \rho v^2 ##
    Finally, ## v = \sqrt{Hg}##
    But, it doesn't match the answer.
     
    Last edited: Jan 27, 2015
  2. jcsd
  3. Jan 27, 2015 #2

    Orodruin

    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper
    Gold Member

    It seems you are randomly sticking equations together. If you are going to use force arguments you need to consider what mass the force is acting on, if that mass is changing, and whether or not its velocity is actually constant as you assume.

    Or you could take the easy way out and solve it using energy considerations instead.
     
  4. Jan 27, 2015 #3
    Could not understand what is the problem in my approach.
    OK. I calculated in this way:
    ##\frac {dP.E}{dt} = - \frac {d K.E}{dt} ## ; because, P.E. + K.E. = constant
    ##\frac {d}{dt} (mg \frac{h}{2}) = - \frac{d}{dt} (\frac{1}{2} mv^2)##; average height is ##\frac{h}{2}##
    ##\frac {d}{dt} (Ah\rho g \frac{h}{2}) = - \frac{1}{2} \frac{dm}{dt} v^2##; as, ##v## is constant;
    ##\frac {1}{2} A \rho g \frac {d}{dt} (h^2) = - \frac{1}{2} \frac{d}{dt}(- Ah \rho) v^2##
    ## g \cdot 2h \frac {dh}{dt} = \frac{dh}{dt} v^2##
    ## v = \sqrt{2gh}## ; which is the expected answer.
    But still couldn't understand what's wrong in my previous approach?
     
  5. Jan 27, 2015 #4

    Orodruin

    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper
    Gold Member

    As I said, you need to consider what volume the force is acting on. Before you have done that and specified your other assumptions, you really cannot tell if what you are doing makes sense.
     
  6. Jan 27, 2015 #5

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    The problem is that you are taking it as though the water accelerates instantaneously from 0 to v. That makes it like an inelastic collision, with about half the energy being lost. In practice, a much smaller portion of the energy is lost. The water accelerates more gradually from the area surrounding the hole from inside the tank. So assume conservation of energy instead (Bernoulli's principle).
    Of course, some energy is lost to turbulence, so that doesn't give exactly the right answer either. Turbulence can be reduced by fitting a Borda mouthpiece inside the tank around the hole.
     
  7. Jan 27, 2015 #6
    To expand on what haruspex was saying, the fluid flow is converging in the region approaching the outlet hole. This is similar to what happens in a converging nozzle, where the surrounding fluid forms a sort of converging channel for the fluid approaching the hole. For an analysis of incompressible inviscid flow in a nozzle, see post #22 in the following thread: https://www.physicsforums.com/threads/fluid-question.791594/page-2#post-4973184. It has the term you are talking about for the rate of change of momentum, but it also includes the retarding force resulting from the pressure of the converging wall on the fluid. This is how the factor of 1/2 comes into play. The analysis also shows that the Bernoulli equation gives the same result as your conservation of energy analysis, and correctly yields the factor of 1/2.

    Chet
     
  8. Jan 28, 2015 #7
    But, how the retarding force acts in my example ? Would you please explain this in more detail ?
     
  9. Jan 28, 2015 #8
    Your problem is more complicated geometrically than a gradually tapering nozzle; the outer streamlines provide the converging "wall" for the inner stream flow. To solve your problem requires solving the inviscid flow equations for the velocities, pressures, and streamlines approaching the outlet. These equations are 2D PDEs. However, along any streamline, even in this more complicated case, the Bernoulli equation applies. But, to implement the bernoulli equation, you need to know the magnitude of the velocity at a given location from the solution to the PDEs.

    Chet
     
  10. Jan 28, 2015 #9
    Would you please explain what was wrong in this approach? I still cannot find out this. You said about 'retarding force' in post #6. But, I couldn't understand how it is applied in this example.
     
  11. Jan 28, 2015 #10
    There are a couple of things wrong with your approach, and yet you still get the correct answer to within a factor of 2.

    First of all, you are assuming that all the "action" is taking place right at the exit cross section, where the pressure suddenly drops from P to atmospheric and the velocity suddenly increases from zero to the exit velocity. This would be equivalent to an impulsive force at the exit cross section (like in a collision), which is not the case. In the real physical situation, the pressure in the tank is decreasing gradually as the fluid approaches the exit cross section, and its velocity is increasing gradually. So assuming that the pressure P (far from the exit) times the cross sectional area at the exit A is the force that accelerates the fluid is incorrect.

    Secondly, in the real physical situation, the fluid is not traveling in a horizontal line toward the exit. On the contrary. It is essentially converging radially toward the exit cross section, as if the exit cross section were a spherical point sink, with the surfaces of constant pressure being nearly hemispherical. Only within a distance on the order of about 1 exit cross section diameter from the exit do the streamlines straighten out and become horizontal. I hope you can see how this is very different from what you assumed.

    Chet
     
  12. Jan 28, 2015 #11
    Probably, I have understood what you meant. Thanks for this explanation.
    But I couldn't understand this point I mentioned above.
    upload_2015-1-28_21-14-25.png
    Let the velocity in the tank be ##v_1## and the velocity at the hole be ##v## ;
    So, change of mass in the tank per second ##= \rho Av_1## ; where, ##A## is the cross-sectional area of the tank.
    And, the amount of water flow per second at the hole is ## = \rho av## ; where, ##a## is the cross-sectional area of the hole.
    So, it becomes,##\rho Av_1 = \rho av##
    Hence, ## v_1 = \frac{a}{A} v ##
    So, I thought , the speed of the water in the tank would be constant (which is equal to ##v_1##) and the speed of water at the hole would also be constant (## = v##). And I thought, as soon as the water reaches at the height of the hole, it suddenly changes its velocity (which is in the horizontal direction).
    Would you please clarify this point ?
     
  13. Jan 28, 2015 #12

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    Imagine a speck of dust in the middle of the water. It won't descend vertically at a steady speed until it reaches the level of the hole, then suddenly go horizontally at exit speed. Rather, it will gradually curve around towards the exit as it approaches the hole level, getting faster as it does so. Admittedly, most of the acceleration is close to the exit, but it is not a sudden jump. Correspondingly, the pressure will reduce somewhat as it approaches the hole. This makes it hard to apply conservation of momentum.
     
  14. Jan 28, 2015 #13
    But , the vertical component of the velocity of water in the tank, I think, will be constant , and will be ##v_1 = \frac {a}{A} v##, as I calculated in my earlier post.
     
  15. Jan 28, 2015 #14

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    I didn't say the vertical component would change (though it must eventually in order to exit horizontally). I said it would gradually acquire a horizontal component.
     
  16. Jan 28, 2015 #15
    So, does the vertical component of the velocity change to 0 at the hole all at once?
     
  17. Jan 29, 2015 #16

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    Surely not. Water flow will be smooth within the tank.
     
  18. Jan 29, 2015 #17
    OK. Then the vertical velocity at a certain height should differ from point to point. But, I have confusion where the vertical velocity will be greater, I think , it will be at the point nearer to the hole.
     
  19. Jan 29, 2015 #18
    This is not what happens. To get an idea of how the parcels of fluid in the tank move, imagine a very long room full of people (from wall to wall) with an exit door on the long wall not very far from the corner. The short wall far from the exit door begins to move toward the people, and the people realizing that they might get squished begin to file toward the door. The wall moves in such a way that the number of people per unit area in the room remains constant. In the region close to the door, all the people will be converging radially toward the door. Even beyond the door, in the part of the room toward the non-moving short wall, people will be moving toward the door (with a velocity component opposite to that of the moving wall). So, aside from the people right next to the moving wall and rigid walls, all the people in the room will have velocity components both parallel and perpendicular to that of the moving wall, and some of their parallel components will even be opposite to that of the moving wall.

    Chet
     
  20. Jan 29, 2015 #19
    This was helpful to understand the matter. So, I think the velocity will be (roughly) something like the picture below:
    upload_2015-1-29_19-57-53.png
    And I conclude, there is no easy way to calculate the exit velocity using the formula : force = rate of change of momentum with time.
    It's better to use the energy conservation law for this problem.
    Thanks for your explanation.
     
  21. Jan 29, 2015 #20
    Yes. Yes. Yes.

    Chet
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted