1. Limited time only! Sign up for a free 30min personal 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!

Materials for dynamic seal application

  1. Jun 15, 2010 #1
    I am about to talk to a prototype company and need some advice on O ring, friction and seals.

    I am creating a device that works on slight water pressure differential. I have a piston like structure that is cylindrical and there will be an o ring inside the cylinder. The gas/liquid that the o ring seal will operate under is water. What should be the material inside for the 2 mating surfaces (the wall of cylinder, and the o ring material?). The environment will be low pressure for the water (under 1000 PSI). I need as low friction as possible to allow the o ring to slide easily against the cylinder wall, but still need to keep the water out (a small bit of leakage is not a big deal).

    I was thinking maybe teflon for the cylinder wall, but I understand the o ring cannot really be made of teflon b/c it's not an elastomer. Can anyone make a suggestion??

  2. jcsd
  3. Jun 17, 2010 #2
    o-rings make very poor dynamic seals, and on top of that, water is one of the poorest lubes that you can have. (even worse if dirty) Often low differential pressures will not "load" the o-ring enough to seal completely, causing leakage. My suggestion is that you seriuosly should think about using the water as a seal, if possible, like a proving bell. Or consider a diaphram type set up

  4. Jun 17, 2010 #3
    Unfortunately the water is the dominant liquid/gas that is being manipulated. It was designed to function in and with water. I'm sorry but what do you mean by "proving bell" or a diaphram set up? Is there a website u can direct me too?
  5. Jun 18, 2010 #4
    For water applications EPDM rings tend to perform the best according my oring book. However this is not a job for an o-ring, movement will just shag it so it's useless.

    It's acutally fairly impossible to suggest a sealing method unless we know more about the application. Temperature, geometry, whats it's main function. If it's low movement, and you won't be actuating it often then something like Trelleborg's Variseal, which are PTFE (teflon basically) based seals with a spring inside toget the compression needed, although this isn't suitable if you are continually moving the piston.

    The next step would be a metal to metal seal, the tight fit will make moving the piston incredibly hard though.
    Last edited: Jun 18, 2010
  6. Jun 18, 2010 #5
    A proving bell is a device used for high accuracy measurements, It consists of a cylinder with one end open (bottom) and a "stand" that precisely inserts into the cylinder. The whole assembly sits in a bath of fluid, which forms the seal.
    look at this: figure 6 is a bell (this link is a good one on flow calcs, so the mods may want to add it to the resources "sticky")


  7. Jun 18, 2010 #6


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    What is the pressure differential across the seal?
    What velocity does it need to move at?
    What kind of duty cycle does it experience (ie: one stroke a day? 1000 RPM constantly?)
    What temperature is it exposed to?
  8. Jun 18, 2010 #7
    What is the total working pressure?
    Will high drag/friction cause the device to not work?

  9. Jun 18, 2010 #8
    I would think that lathe-cut teflon or epdm rings would be as efficient as ant as long as temps did not exceed the maximum working temps for the material and pressure was below 2000 psi. The next best material would be stainless steel and rubber - water actually is a good lubricant for this combination (the prop shafts of submarines use rubber bearings on stainles shafts). A rubber "D" ring in a square cut piston slot sealing against a stainless barrel.
  10. Jun 18, 2010 #9


    User Avatar
    Gold Member

    You might want to investigate Thordon seals. Very tough.
  11. Jun 21, 2010 #10
    Q Goest :

    Pressure differential - the device moves up and down based on the differential pressure of water (depth). Additional water is added to the cylinder to effect an incremental downward push on the cylinder but a valve is then opened to allow the cylinder (which has a buoyant empty chamber on the bottom) to be pushed back up by the greater buoyant pressure (cylinder deeper in the body of water). The seals on the outside of the cylinder prevents the water from penetrating higher up the cylinder walls but allows the cylinder to move up and down (hopefully with the lowest coefficient of friction).

    Velocity - the velocity is quite slow (i.e 3 to 8 feet per minute??)

    Duty - the cycle happens 50 times a day give or take (really low duty)

    Temperature -whatever outdoor environments it is situated at (maybe not North pole or antartica)

    dr dodge

    - total working pressure? - deepest depth of pressure is at 300 to 500 feet (or more if possible due to engineering specs - acutally the deeper the better pending on the seal strength but 500 is pretty good)

    drag/friction - the only drag / friction will be the seal - thus the lower the drag the better.
  12. Jun 22, 2010 #11


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    I'd suggest doing a bit of reading on O-rings. Check out the Parker O-ring handbook here:
    http://www.parker.com/literature/ORD 5700 Parker_O-Ring_Handbook.pdf

    O-rings will have higher friction and won't last as long as other types of seals such as rod seals used in hydraulic applications. You mentioned friction, which can be calculated using the equations found in the O-ring handbook.

    Also, consider using wear rings between the sliding parts to keep things aligned during operation. See the picture http://www.ahpseals.com/products/hydraulic.php".

    Here's my suggestion. Write down what you know about the seal. Saying it has x amount of water on one side isn't going to help - figure out what the pressure difference is and write down the number in MPa or psi. Write down what you know about it and make a sketch, then contact a seal manufacturer such as one of those listed above and let them help you select the proper seal for your application.
    Last edited by a moderator: Apr 25, 2017
  13. Jun 22, 2010 #12
    Q Goest - i really appreciate your time and diligence in helping me. Your advice on O rings and the use of "wear rings" is extremely helpful. I will definitely consult with a seal company here.

    I will definitely browse through the Parker O ring link. At this point with my protoype which is a small (2 feet high , 1feet wide cylinder) representation and will only be submerged up to 1 to 2 feet - I was thinking either teflon surface for the cylinder and a perhaps the spring energized teflon seal you pointed out in the "OmniSeal". At this point, my thought process is that with the slow speed, room temperature, low water pressure environment of the prototype, the aformnetioned material combination would be best to give me the lowest friction possible (sites mention teflon and teflon have a coefficient of 0.04). Do you think that is a realistic guess?
    Last edited by a moderator: Apr 25, 2017
  14. Jun 22, 2010 #13


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    The coefficient of friction for Teflon drops as the contact stress increases. At very low contact stress (think a block of Teflon sitting on a table) the coefficient of friction may be as high as 0.2 to 0.4. As the contact stress increases to a substantial portion of the yield strength, the coefficient of friction drops off by an order of magnitude (ie: 0.02 to 0.04). Teflon U-cup seals are intended to run in this range.

    Another big factor is the pressurized area of the seal. O-rings, Teflon U-cup seals and other pressure energized seals have a pressurized area that exerts a normal force on the sealing surface. The larger that pressurized area, the higher the normal force, and the higher the frictional force. So you want to look for seals with a smaller cross section if you want to minimize seal load. Check the Parker handbook for equations for O-rings. Note however, that as you go smaller in cross section, the seals generally become less forgiving of tolerances, contamination, scratches on the sealing surface, etc... so going small has its drawbacks. They also will wear out faster since they have less material to wear away.

    For Teflon U-cup seals, there's another factor affecting friction, and that's the spring load. The above mentioned manufacturers can often tailer the spring load to minimize friction.
  15. Jun 22, 2010 #14


    User Avatar
    Gold Member

    OP, have you looked up Thordon seal materials? They make water-lubricated bearings for hydro-dam turbines and marine tail-shafts, rudders, etc. You may be surprised at how sophisticated their bearings are, despite their simplicity.
  16. Jun 23, 2010 #15
    Q Goest - thanks for the advice again. I hear what you are saying about the smaller pressurized area but I think for my small prototype I can get away with it with no real problems for the time frame i need. One thing I am a little unsure of in what you said - how does an increase in contact stress drop they friction coefficient? I was under the assumption that the larger contact stress would result in more friction because the mating surfaces would be pushed harder against each other?

    turbo-1 - Thanks for the Thordon Seal reference. This is the first i've read about seals using water as lubricants which is extremely useful. It will be extremely relevant if i go past my prototype for a large scale operation. Thanks very much :)
  17. Jun 23, 2010 #16


    User Avatar
    Gold Member

    You're welcome. When I was selling seals and bearings for an industrial supply company I targeted hydro-dams for turbine-shaft bearing/seal applications. Since the materials are intended to be lubricated by water, they are a perfect fit. No leaking polluting lubricants to deal with. Thordon materials can be lubricated by salt water as well as fresh water, making them a perfect solution for ships that navigate both.
    Last edited: Jun 23, 2010
  18. Jul 8, 2010 #17
    Q Goest, Turbo-1 ..... this may sound kind of dumb, but the prototype firm I feel is not super experienced in seals and their dynamics. Should I have them make the cylinder first, then get the engineer seal firm to fit the seal, or should I pre-order the seal at a specifed diameter and then have the prototype firm make the cylinder to fit the seal?
  19. Jul 8, 2010 #18


    User Avatar
    Gold Member

    Either way would work, but you may save money by doing the latter. If you can get an off-the-shelf seal in a standard dimension from Thompson-Gordon or another manufacturer, that might save you some bucks and some time. You're going to have to get the prototype cylinder machined anyway, and there's noting be be lost by specifying dimensions that match your off-the-shelf seal.

    In the case of my large hydro-dam seals, all the seals had to be machined to match the existing equipment, much of which was 50-60 years old and all dimensioned in inches and fractions thereof.
  20. Jul 8, 2010 #19
    Just look at a sealing company's catalogue, it not only has the parts list but the dimensions needed for the grooves for them to fit in.

    You'll have to work out tolerances and surface finishes for yourself though. There will be a reccomended squeeze and overlap. The run-in's are also sometimes specified in the catalogues.

    It may be a prototype but everything you are doing is pretty standard stuff really, sealing wise that is.
  21. Jul 8, 2010 #20
    Yes, doing the latter was kind of logical but my thought process also pondered that the engineer seal firm would be more precise and professional in how they measure and fit things. Thus the prototype firm can make a cylinder specified for a certain diameter but it would kinda screw up slightly and have a slight variance BUT then that's ok b/c the seal company would compensate by taking the cylinder, measure it in their own way, and make the custom seal to fit it perfectly with regards to the normal friction forces between the seal and the cylinder walls. I guess it would make sense to take one of their catalogue b/c they don't have to custom make the size, but i was wondering if prototype firms were just not that experienced with making something so sensitive.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook