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

Analysis of Lifting Hook for Rolled Cans

  1. Oct 2, 2012 #1
    Hello All,

    I am brand new to this forum and have found information on here quite helpful in the past before I decided to join the forum, so hopefully some of you can help me with my current project.

    Currently I am trying to design and fabricate a lifting hook for rolled steel cans. The max. length of the cans we roll is 10' and the yard has requested hooks that can handle 15 short tons, 25 stons, and 30 stons.

    My main concern is how to analyze the hook for design. The hook will be fabricated from 2" thick, 36 ksi steel plate and will consist of a main plate with cheek plates for additional strength.

    Thus far I have analyzed the bending stress and shear at section B-B and tension at section A-A.

    What other stresses need to be calculated at section A-A? I am trying ot make it as simple as possible and intend to use a design factor of 1.5. Is simple beam theory applicable in this situation?

    I have attached a conceptual design.

    Any thoughts or concerns would be greatly appreciated.

    Regards.
     

    Attached Files:

  2. jcsd
  3. Oct 3, 2012 #2
    I worked in the rigging business for 25 years and never head of anyone designing their own hook. In many places, it is illegal or else they make it very expensive for you to do so legally.

    I normally start with WWW.thecrosbygroup.com. Once I find what I want, I do a Google search to see if anyone offers the same thing for a better deal. Yes, I know how to design a hook and have designed many lifting fixtures, sometimes for payloads costing billions, sometimes for bulk sand. But hooks of many different configurations are to cheap to simply buy out of a catalog.

    I can't open your file, but I suspect that one of many sorting hooks or lifting clamps would work well. Google those terms.
     
  4. Oct 3, 2012 #3
    Thanks for the response Pkruse, but why would it be illegal for us to design an in-house lifting device? This is basically a fabrication aid, similar to a company designing and fabricating their own spreader bars.

    My inquiries are in regards to analysis of this type of hook. I have not been able to find something similar to this design online and am curious if anyone else has designed/analyzed a similar piece of equipment and how to adequately calculate stresses experienced in the hook.

    I'm not sure why you are unable to view the attached pdf, any advice on how to post this drawing would be appreciated.

    Regards.
     
  5. Oct 3, 2012 #4
    Since your measurements are imperial I assume you are in the USA.

    I had no trouble opening your pdf.

    I do not know the US lifting regulations , but anyone designing lifting gear should know them for the country of operation.

    In the UK the regulations require frequent proof testing as well as or even instead of design calculations.

    I am concerned about the possibility of slipping as the prong does not go all the way through.

    Is it not possible to thread a conventional lifting bar through the corespace and secure the bar at both ends?

    If there is a safer way to do something, under UK health and safety, you would need a very good reason indeed to operate another way.

    Is your plant foreman simply trying to bypass such safety proceedures?
     
  6. Oct 3, 2012 #5
    Studiot, thank you for your response and yes I am located in the US. By no means are we trying to bypass any safety regulations, which is why thoroughness in the analysis of this hook is vital.

    I agree with the proof testing and intend to do so, but typically we would like something to work on paper prior to field experimentation. Design of this hook is required to meet all ASME, ASTM, AISC, BTH etc. design parameters applicable.

    I believe one of the major concerns here is in regards to the fatigue limit which needs to be accounted for when applying safety factors. My thought process is to determine a conservative lift frequency of this hook and basically place an expiration date on the hook prior to reaching the point where the hook has fatigued below the stress imposed. Typically with the design of our padeyes we impose a safety factor of 1.5-2, but slings and padeyes as industry standard are held to a design factor of 5:1 which I am considering applying here.

    Regards.
     
  7. Oct 3, 2012 #6
    Well, you found most the requirements. Don't forget OSHA 1910 and 1926, whichever is applicable. Keep in mind that case law and interpretations often sends you to 1926 when you think you should be in 1910. You will need a means of positive retention so the load does not slip off.

    I've designed similar rigging for very large stuff. I assume that a poorly engaged hook might be tip loaded, calculate the bending moment, and apply the SF of 5.

    Have you considered a 2-leg sling with sorting hooks? That would be the standard rigging solution.

    Edit: Fatigue is not a concern when using a SF this high, especially with the tip load assumption. That is part of the reason it is that high.
     
  8. Oct 3, 2012 #7
    Okay, great, thank Pkruse.

    I was already usig a tip load assumption when calculating bending moment but will have to integrate the safet factor of 5. Currently in the field they are using chains with sorting hooks, similar to what you suggested, but the issue in that lies with overhead clearance which is why this sort of solution has come about.

    I will definitely look into the OSHA regs, thanks for you help.

    Regards.
     
  9. Oct 3, 2012 #8
    Fatigue should not be a concern in any case for this application since the material should be ductile and stress levels well below the indefinite fatigue limit.

    Other materials considerations mean that you will need to consider operating temperature and crack check any welds.

    You have not stated why there is no access to both sides of the roll.
     
  10. Oct 3, 2012 #9
    There is access to both sides, but the use of the hook would allow us to butt the rolls up against one another on the jigs for welding. These cans are typically used to fabricate piles, vessels, etc.

    I do not believe there is any issue with operating temperature and all welds and materials will be submitted to full NDT (UT, MT, etc.).
     
  11. Oct 3, 2012 #10

    Q_Goest

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Hi aburt, welcome to the board.
    For section A-A in the highly curved portion of the hook, straight beam theory is going to give you lower stresses than actual. You can do an FEA analysis on it or you can do it by hand. I've attached a section from my textbook* that goes over curved beams. You can also find similar analysis in Roark's (I have the 7'th edition and it starts on page 267, chapter 9).

    You also need to look at tear out of the lifting hole per BTH-1, para 3-3.3 and probably a few other things as well. Note that this is actually considered a "Below the Hook" lifting device.

    *Robert C. Juvinall, "Fundamentals of Machine Component Design" 1983
     

    Attached Files:

    Last edited: Oct 3, 2012
  12. Oct 3, 2012 #11
    Q_Goest,

    Thank you, this is exactly the type of information I was looking for.
     
  13. Oct 3, 2012 #12
    Is the section A-A the only section besides the lifting hole that needs to be analyzed for lifting of the cans? Does bending at section B-B even come into play?

    Also, if cheek plates are added as per the attached drawing, does this only effect the area (A) used in Equation (4.9)?

    Regards.
     
  14. Oct 4, 2012 #13

    Q_Goest

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    I can't tell from your drawing, but I suspect the rest has lower stress.
    Assuming the cross section at B-B is the same as A-A, stresses should be lower.
    Certainly it affects A but also e. Try going through the sample problems at the bottom of that page, starting with Sample Problem 4.1. If you can understand those, you should be able to solve for your hook.
     
  15. Oct 4, 2012 #14
    I've run into hook height problems many times, and you are going down the correct path to solve that problem. Make your lower beam long enough to extend all the way thru the cylinder. Put a small stop on the end so the cylinder can't slide off. I'd probably use I-beams for the top and bottom beams, and design a moment connection between them.

    Ductile materials do have a fatigue life, but in this case it will be infinite. Every jet engine has parts with a defined creep and fatigue life, and they count load cycles to determine when to replace those parts. But this is not a jet engine.
     
    Last edited: Oct 4, 2012
  16. Oct 4, 2012 #15
    Jet engines are not made of structural steel.

    There is a fatigue limit stress, below which fatigue does not occur (or the life is infinite if you prefer) for some materials. Structural steel is one such material.

    For other materials there are no lower limits. Many alloys (including other steels) fall into this category.
     
  17. Oct 4, 2012 #16
    Many gas turbines have a great deal of steel structure in them. Everything I said about jet engines applies to them.

    But another idea that might optimize the flow of work is to get rid of the lower beam and put a hook on one end similar to a sorting hook. Then put an over center clamp on the other end.
     
  18. Oct 4, 2012 #17
    I am not a hook designer I leave that to J.M. Barrie and others.

    :smile:

    There is a difference between 'structural steel' and a steel used in a structure. I doubt that much 'structural steel' is used in turbine manufacture but it is not my field so I am happy to bow to an expert there.
    What do you use grade 43 steel for in turbine manufacture?
     
  19. Oct 4, 2012 #18
    You won't see ASTM A-36 called out on many gas turbine drawings, but most of the steel alloys in the Boiler Code have found their way into one GT or another. Keep in mind that the super alloys in the turbine section all have properties similar to steel, except with regard to high temperature.
     
  20. Oct 4, 2012 #19

    AlephZero

    User Avatar
    Science Advisor
    Homework Helper

    You might be surprised how "agricultural" some STEAM turbine designs are.
     
  21. Oct 5, 2012 #20
    AlephZero is right. Steam turbines operate at much lower temperatures and typically don't have a high speed shaft like most gas turbines. So they use a lot of steel. They are designed typically for a longer fatigue life, but they do have creep and fatigue limits similar to gas turbines.

    I still find it amazing that it is common practice that in certain areas the stress is designed to be more than. 100% of yield. That always leads to a limited fatigue life.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Analysis of Lifting Hook for Rolled Cans
  1. Lifting Lugs Design (Replies: 2)

  2. Spiral scissor lift? (Replies: 3)

Loading...