Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

What kind of pump would work for discharging water out of a small manned submersible?

  1. Oct 9, 2018 #1
    Hello, apologies for the crude drawing. I was wondering if it is possible to have a centrifugal pump that pulls in water and pumps it out down like in the drawn picture? I know usually its to displace water up. I always see centrifugal pumps with discharge oriented up like in the photo. If it cant be used like this what kind of pump can pump it down? This is a pump in a small manned submersible that is pulling water out of a tank and out into the outside to increase positive buoyancy. The pump has a Ns, specific speed, of 61 rpm.
  2. jcsd
  3. Oct 9, 2018 #2


    User Avatar
    Gold Member

    Can you describe the relative elevations of the pump, the water source, the water source level, and the where the pump will be discharging?
  4. Oct 9, 2018 #3


    User Avatar

    Include the maximum operating depth for the submarine.
  5. Oct 9, 2018 #4
    1000 Meters sea water
  6. Oct 9, 2018 #5


    User Avatar

    At that depth the required discharge pressure for the pump will be approximately 114 Bar.
    What is going to be the source of air required to replace the water volume being pumped out. Without an air source evacuating the water will create a vacuum in the tank; which can both starve the pump and create a risk of tank collapse due to the 114 Bar external water pressure stated above.
  7. Oct 9, 2018 #6
    Well we have a compressed air system to make sure there is never a vacuum in the ballast tank. How did you get 114 Bar. But Im having trouble picking a pump. Can the centrifugal pump simply be rotated so it discharges down?
  8. Oct 9, 2018 #7
    For Pressure at 1000 meters

    Pressure = rho x g x h
    = 1024kg/m^3 x 9.81m.s-1 x 1000m
    = 10045440 Pa
    = 10045.44 kPa
    = 100,45 Bar

    The pressure value that I'm getting is just a bit lower than what he got, but at the end of the day I suppose it depends on what value you use for the density of sea water. I've seen sources putting it at anything from 1024 kg/m^3 to 1050 kg/m^3 deep down in the depths, so I suppose it depends on where exactly you are. I wasn't able to get 114 Bar the way he did, but maybe he'll explain his calc at some point.

    As far as pointing a centrifugal pump downward, I have no idea. If you really have to you can just take the piping up and then down. I'd be very interested to hear why or why a pump can or cannot be mounted upside down. I don't imagine that it's a good idea to mount a self priming centrifugal upside down because the pump holds fluid in a chamber before it goes into the impellor and I think it will interfere with the operation of the self priming aspect of the pump if you mount it upside down.

    I also don't think that a vertical centrifugal should be mounted upside down because the weight of the impellor stack is supported at the top (as far as I know) and it is meant to hang. I think mounting it upside down will wreak havoc on the spacing between the impellors and the guide vanes, but I cannot think of any reason why a horisontal centrifugal pump cannot be mounted upside down.

    This is the kind of question you should ask the pump supplier. In general you should mount a pump in accordance with the pump suppliers guidelines. Warranties can be voided if you don't and specialist pumps are expensive. Any pump pushing 100 bar definitely will be expensive.
    I'm no expert, so if anybody wants to call anything that I've said rubbish, feel free. I'm listening intently.
    Last edited by a moderator: Oct 9, 2018
  9. Oct 9, 2018 #8


    User Avatar

    Staff: Mentor

    The short answer is that the orientation doesn't matter.

    However, from your response to the issue of pressure, it doesn't sound like you've even really begun to engineer this system. A sub that can go down 3,000m is a really serious and dangerous project. Are you qualified to be undertaking it? What is your experience level?
  10. Oct 9, 2018 #9
    1000m, Im a undergrad student im not actually building it!
  11. Oct 9, 2018 #10
    Goulds is a well known manufacturer of pumps. Search Goulds 3393 and Goulds VICR for two of their models that meet your pressure requirements. Read both the pump brochures and the installation & maintenance manual for those pumps on the Goulds web site. All of your questions will be answered.

    Note that high pressure centrifugal pumps have minimum and maximum flow rates. If you go outside those limits, internal pump parts (bearings, shafts, seals) get overloaded. If the flow is too low, the pump overheats. If that is the case, look at positive displacement pumps such as piston pumps.
  12. Oct 9, 2018 #11
    Thank you all for your answers. After talking to my professor it seems you're right. I was not supposed to use a centrifugal pump. I feel like such a dummie. What is a example of a piston where it similarly takes in water horizontally and discharges it down perpendicularly like the drawing?
    Last edited: Oct 9, 2018
  13. Oct 9, 2018 #12
    Google is your friend. Try search terms piston pump and pressure washer. But first define what range of flow rates meet your needs, the power required, and how you plan to drive the pump. It is not sufficient to say "with an electric motor", you need to specify the power, voltage, and current.
  14. Oct 9, 2018 #13


    User Avatar

    With regard to my earlier post stating a pressure 114 Bar, in US units the pressure of 1 ft of water depth is actually about .445 psi which, as you have determined results more accurately as 100 Bar. When working offshore .5 psi/ft is used as a walking around rule of thumb for quick estimates and that was the basis of my first post 114 Bar value; obviously, this practice has its limits for deeper water depths because even though it falls on the safe side, the amount of error becomes excessive.
  15. Oct 11, 2018 #14


    User Avatar
    Science Advisor

    Anything that pumps water out of the pressure hull will for safety reasons need to be a positive displacement pump, probably a piston pump with a minimum of three cylinders. That pump will need to be duplicated and each unit isolated with a stop valve.

    When not spinning, a centrifugal pump allows a free flow through the pump. You will need valves to control the flow when the pump is not running. When starting a centrifugal pump, the pump must be primed with water and all air must be vented. That is why the outlet is usually pointing upwards, to allow the air to escape.

    Your ballast tanks should be outside the pressure hull. Any air in those tanks that is exposed to external hydrostatic pressure will be compressed, which will increase the water ballast and so cause an uncontrolled plunge to greater depths. For that reason you need to avoid open flow centrifugal pumps and revert to positive displacement multi-cylinder piston pumps to transfer water ballast in or out of a ballast tank.

    Vacuum in a ballast tank will not be a problem since you are pumping liquid water from the lowest point in the tank. The water free space in the ballast tank will fill with CO2 rich air and water vapour when you pump the water out. There is dissolved gas in seawater that will fill any ullage space.

    It will be advantageous if you can adjust the gas pressure in the ballast tank to close to the operating depth as that reduces the pump energy needed to adjust or trim the ballast.
  16. Oct 14, 2018 #15
    So many good tips lol! Why exactly three piston pumps though?
  17. Oct 14, 2018 #16


    User Avatar
    Science Advisor

    A minimum of three pistons because the pressure on the pump crankshaft or equivalent is then balanced and turning the shaft has much less harmonic variation of torque during each turn.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted