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

Convert crank driven system to an air piston one

  1. Jun 14, 2012 #1
    Hi!, nice place you got yourselves here, so I'm a mechanical engineer working as a maintenance supervisor working for a dairy and food corporation specializing in UHT milk and milk derivatives (varietieshttps://www.physicsforums.com/Nexus/editor/menupop.png [Broken] of yogurt, cheese, butter, ice cream, etc.)

    Anyway we have a bunch of old Cherry Burrell fillin machines, mostly mechanic based system (little pneumatic-driven motion) so I'm on a mission to update this old ladies.

    My first challenge is to adapt the actual filling system driven by a crank mechanism to a pneumatic driven one, below is a rough powerpoint based illustration of the current mechanism
    fill+system.jpg

    I'm thinking something like this

    fill+system+alt.jpg
    I have no clue how to start this since all my mind is thinking is how to calculate the force needed to push the liquid thru the filter net, osmosis is an obscure subject to me.

    Any clue would be much appreciated
     
    Last edited by a moderator: May 6, 2017
  2. jcsd
  3. Jun 14, 2012 #2

    Danger

    User Avatar
    Gold Member

    Welcome to PF, Phonnovo.
    With apologies, I can't access your attachments. They're just little blue boxes with question marks in them. Perhaps they aren't Mac-compatible?
     
  4. Jun 14, 2012 #3
    I am on a PC and the little boxes have red squares in them?? Also inaccessable.
     
  5. Jun 14, 2012 #4
    I think it's fixed now.
     
  6. Jun 16, 2012 #5
    Question is What is the problem you wish to overcome with the mechanical system by converting to pneumatics?

    And, will the new system be as reliable, more reliable, less reliable than the previous?
    Here I am relating one system to the other as regards to a consistant metering and maintenance schedule. You might want to do an assessment on all possible failures and breakdowns and expected cycles of operation before commiting to a changeover.

    I am not sure if one can call pushing a fluid through a filter as being osmosis. You might want to a test rig setup and obtain a curve of force on piston (ie pressure on the fluid) versus fluid flow to get a better idea of what you are dealing with. Though something tells me that the filter is not a fine mesh with substantial pressure drop but rather a screen to catch extraneous foreign substances.
     
  7. Jun 16, 2012 #6

    Danger

    User Avatar
    Gold Member

    Now that I can see the pictures, one glaring question comes to mind; why bother with all of those linkages at all? Can't you just connect an air piston or bellows directly to your pump piston? You should be able to eliminate the need for leverage by simply choosing an effective pressure level.
    I have no idea what scale this thing is at, but I suspect that something like a bag from an automotive or transport truck suspension system would be strong enough.
     
  8. Jun 16, 2012 #7
    Thank you so much for taking the time to answer me, here I ellaborate a little bit more and try to (hopefully) answer your doubts, sorry for my broken english.

    1.) Reduce load on the main constant rotation shaft, and subsequent parts i.e. chains, sprockets, motor, bearings, etc, thus prolonging life of it.
    2.) Easier/cheaper maintenence, I'm attaching actual drawings of the machine, so you can see it's a lot more hassle to replace parts now i.e. shafts, bearings, cam, the bearing housing, interlocking air pistons, etc.
    3.) Independence, today if one of the cams or bearing needs corrective maintenence we need to dismantle the whole bottom part of the machine thus loosing money, if we do this we can just take 1 station out of the 6 nozzles and keep the machine working.

    and last but I think it's the main reason...
    3.) When we CIP we need to pass NaOH thru the filling nozzles but we do not need the conveyor chain to keep moving, and due the "anatomy" of the machine this can be done bypassing the filling mechanism off the main rotational shaft.

    The current system is not much reliable, and we do have another machine that fills the cartons with no cam or shaft, it's just a rotational servo, a belt and a worm drive, much less parts, and you know what that means for a mechanical system.

    I'm working on it but a need this precise data I'm gathering

    I was just playing smart for that osmosis calling, I'm really not sure how to get that pressure value other than punching a hole on the filling bowl and stick a manometer in it.

    Leverage is on my side since I cannot give the air piston all the air supply I want since air supply is an issue on the current state of our industrial services department

    I'm attaching some actual drawings of it so you can get a better idea, and you can relate better knowing that this machine fills cartons of 1lt.

    This is the front view of the filling system, it has 3 of this, the upper part of line 1 is ommitted on the pic but it's the same as left side.
    1.png

    This is the bottom part, you see, this moves 2 of the next, if the cam or the bearing needs replacement we need to stop production entirely on the machine.
    2.png

    This is the arm that drives the filling piston, if I can attach an air piston to this (independant or both lines) I can take it away if needed and the machine can keep working with the other 2 stations.
    3.png
     
    Last edited: Jun 16, 2012
  9. Jun 17, 2012 #8

    Danger

    User Avatar
    Gold Member

    I see.
    Hmmm...
    I can't help thinking that those "no carton—no fill" latches could be exploited toward your goal, but I obviously don't know enough about them to figure out how. Another idea is to put quick-disconnect joints on your pump rods. I can't determine what those (tie-rods?) under the "side 1 line" and "side 2 line" labels are connected to or why they're there.
    I'll study your diagrams some more and think on it for a while.
     
  10. Jun 17, 2012 #9
    All they do is activate the filling system, a clutch like mechanism, the follower/cam be still moving alongside the shaft but the no-carton no-fill will retract a shaft shot-pin that's connected to the ball joint attached to the tie rod that goes up to the arm fill system.

    Those serve as stabilizer for the system, it just add some constrains to the bearing housing

    Thanks very much, remember my main goal here is to determine how much air pressure an air cylinder needs to move the fill arms without the crank mechanism.
     
  11. Jun 17, 2012 #10

    Danger

    User Avatar
    Gold Member

    Right... that had actually slipped my mind. :redface:
    It will ultimately depend upon the surface area of your air piston. To start with, though, you need to either measure or calculate how much force is required at the lever input stage.
     
  12. Jun 17, 2012 #11

    Q_Goest

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    You could determine that by calculating the power input from the existing mechanism. Take the total energy required to run the mechanism over some period of time (power x time = energy). That energy goes into moving the piston. The piston force times length over the same time period is equated to the input energy. In other words, you have energy input from the crank mechanism over some time period which equals the force on the piston times the distance that the piston moves over the same time period; assuming the force is constant (ie: for the piston, E = F*L where E is energy, F is force, L is distance piston moves over the time period in question).

    I doubt the force on the piston is exactly constant but it's probably close. To be conservative, assume the force is higher than that constant value by some amount. Your engineering judgement comes in at that point. You have to take into account other factors involved that affect the force on the piston.

    Regardless, I suspect the bearing mechanism is a best way to go. Pneumatic cylinders have seals that wear and they will typically wear out much faster than bearings. A properly designed and maintained bearing system should last for tens of years in an industrial setting. Perhaps the existing bearing package just needs to be made more maintenance friendly or improved on to get it to last longer. How long does it run before it requires new bearings?
     
  13. Jun 17, 2012 #12
    Errr.. That's the whole point of the thread, how to get that input force needed. ;D

    The thing is that crank also moves another sub-systems, thru cams, actually there are 5 cams on that shaft, I need to get rid of the 3 belonging to the filling stations.

    I doubt the force on the piston is exactly constant but it's probably close. To be conservative, assume the force is higher than that constant value by some amount. Your engineering judgement comes in at that point. You have to take into account other factors involved that affect the force on the piston.

    Due the nature of the current system I'm guess-timating changing an air cylinder and take it to replace it's sealing it's 1000x faster than replacing the equivalent contraption that is the current mechanical system.

    I appreciate all your input on my question, but I need focus on how to measure the energy needed to actuate the filling arm to pass liquid through the mesh into the cartons.

    Once again thanks for taking the time...
     
  14. Jun 19, 2012 #13

    Q_Goest

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Regarding the force and energy required to move your piston, two thoughts:

    1) Can you not test it by attaching a pressure gage to the cylinder? Drill and tap a hole, insert gage? That would be the most accurate I suppose.

    2) Another option would be to remove the cylinder and filter arrangement, and do a flow test on it. To measure flow, filling a known volume in a specific time is one very easy and accurate method. To determine pressure, you could either maintain a specific head level in the cylinder or pressurize it with your liquid product and measure the pressure at the inlet. Once you have a relationship for dP and Q (flow) you can use the standard valve equations to extrapolate to other flows and pressures.
     
  15. Jun 19, 2012 #14

    Mech_Engineer

    User Avatar
    Science Advisor
    Gold Member

    There's some good avice being given regarding the implementation of a pneumatic piston, but I have to ask: why pneumatic at all? It seems to me you could more easily implement a stepper or servo motor driven piston, which would be capable of finer incremental adjustment. If driven by a stepper motor, control would be relatively easy using off-the-shelf hardware, and probably be easier to implement than a pneumatic system.

    Something like this Parker ET series electric cylinder:
    https://www.motionusa.com/sites/default/files/imagecache/product_full/ET_ParkerMotion.gif
    Link: Parker Cylinders (Pneumatic, Hydraulic, Electric, etc.)
     
  16. Jun 19, 2012 #15
    This is what I'm doing currently. (Y)


    Glad you point that out, some of our machines have servos to control the filling, this is Proposal N°1, the pneumatic system is proposal N°2, I gotta have all my fronts covered before I pitch these to my boss.

    Once again thank you all.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Convert crank driven system to an air piston one
Loading...