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

Tubes: Hollow vs Solid

  1. Mar 22, 2008 #1
    Why is the Hollow Hexagonal tube stronger than the Circular Hollow or Solid tubes? And why is hollow tubing stronger than solid tubing in general?

    Can someone please explain this to me in a simple way and with a link for further info if possible so i can read more on it. Would really appreciate it
     
  2. jcsd
  3. Mar 22, 2008 #2

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    For a given diameter, a solid tube is stronger than a hollow one. For a given mass, its the other way around. The extra mass in the middle doesn't add a whole lot of integrity, but it does add a whole lot of mass.

    Where did you get the idea that hexagonal tubing is stronger? Failures typically happen at edges; a circle doesn't have any.
     
  4. Mar 22, 2008 #3
    I tested some using an FEA package. I always get stronger results for hexagonal hollow tubes with same conditions as circular ones!?
     
  5. Mar 23, 2008 #4
    At the corners of the hexagon, you will have stress concentration. Hence the remark above that "Failures typically happen at edges".

    What kind of problem are you doing with FEA? Beam-bending?

    If it is beam-bending then usually the explanation is that one cross-section has higher moment of inertia (I), since the denominator for the deflection is always EI.
     
    Last edited: Mar 23, 2008
  6. Mar 23, 2008 #5
    The problem is a roll cage roll over simulation.. I have a cage (2- in circular and hexagonal tube structure), then i applied equal force on them (Basically same boundary conditions).
     
  7. Mar 23, 2008 #6
    Which equivalence (circumscribed or inscribed) did you use?
     
  8. Mar 23, 2008 #7
    Both had 10 diameter (well for hexagon it was 10 between opposite sides), and shelled with 2 for hollow.
     
  9. Mar 23, 2008 #8
    OK, that's a common way people do it. So, looking at that, can you make comparisons of: mass/unit length, moment of inertia, etc and draw any conclusion? When you say the hexagonal is stronger, what do you mean by stronger? What is your design intent?
     
  10. Mar 23, 2008 #9
    a hexagonal equivalent to the circular should have a diameter of 10 between edges not sides if that's the diameter of the cylinder
     
  11. Mar 23, 2008 #10
    Ok good, i was scared for a second lol Well the intent is to use FEA to determine strength during a roll over. I applied a force load on both in X and Y axsis to simulate the roll. I used surfaces only no edges.
     
  12. Mar 23, 2008 #11
    I am sorry, you are right. it was between edges
     
  13. Mar 23, 2008 #12
    so you have to apply the forces on the edges you should find that circular is stronger.the hexagon could be stronger if it is on the sides, but when you are testing for strength then you are accessing failure measurements to do that you must obtain minimum force to collapse.(on the edges)
    that is why circular is stronger because it has no edges and therefore no weak points
     
  14. Mar 23, 2008 #13
    Damn it , i have no time to go through all of those analysis again using edges. My paper is due this week..
    Stupid, but i didn't think a quick roll over would make it important to test the edges :(:(
     
  15. Mar 23, 2008 #14
    that is why i mentioned failure because in the roll over test the tubing is hit on all sides.and all the edges.
     
  16. Mar 23, 2008 #15
    I'm F***ed, this is my final year project ..
     
  17. Mar 24, 2008 #16
    not at all, this gives you something new to write about... your prof will be impressed, more so than if you hadn't asked the question in the first place! Showing an analysis with both edges and corners and talking about the stress concentrations that arise in corners will likely strengthen the report.
     
  18. Mar 24, 2008 #17
    You think so :) i can mentioned it at least to show that i know if i analyzed the edges results would be different
     
  19. Mar 24, 2008 #18
    I don't know your time limit, but you should be able to salvage something from this. There is value in incorrect solutions if you know what is incorrect. In the case of a roll bar, you might look at actual requirements (NHRA, for example). You would see that weight is critical for race cars. You might also run into the design requirement that force directions in crashes can never be predicted. You might look at the nodes (which are absolutely critical) and ask how difficult it is to weld them for various shapes. Finally, consider the ease of cold drawing round tubing (with the attendant increase in strength). Do a comparison between round and hex. You're not nearly the first engineering student to do this sort of thing; you may get a bad grade but you're not gonna get traded to the Performing Arts Department. Show that you've learned from it!

    Finally, how did the FEA not show stress in the corners? Look at your restraints. Be critical of your test report. Your prof knows all this stuff, but remember this is education - show him you've been able to learn from your mistakes.
     
  20. Mar 24, 2008 #19
    First of all thanks for all this info, it is helpful :)

    Well it is my mistake. I think my boundary conditions were wrong or the package is not doing what i want it to do. After i finished most my analysis (This weekend). I was surprised by the results and i think i was wrong in the boundary conditions.
    Basically what i designed was a half roll cage (Attach/stitched parts together using the FEA package modeler) , but when i restrain the cage from it's sides i feel that i am wrong in doing so and should restrain every part individually (i didn't do that). Plus, the person who is suppose to help me with my project (doing PhD) is not sure himself and when i gave him my report he didn't say anything about the content.
    I didn't think i would be freaking out like this :P
     
  21. Mar 24, 2008 #20
  22. Mar 24, 2008 #21
    It's a little hard to see. What are the weights of the hex and square relative to the round? Is your mesh small enough to show edges? It looks like you fixed the ends and loaded the middle downward? If so, the square shows some obvious weld problems (assuming the model is realistic).
     
  23. Mar 24, 2008 #22
    First thanks for your continuous replies, i really appreciate it.
    I used the same material for all. I tried to make the circular, hexagonal and square the same size and length. Restrained the ends and applied same load to all.
    If i wanted to compare strength using this picture (Just assuming), which would be the strongest?
    My presumption was that the bigger the von mises stress figure, means the stronger the shape is..
     
  24. Mar 24, 2008 #23
    This thing has bounced my answer twice already. Hope this goes through.

    It's hard to see all the detail I'd like, but it looks like the hexagonal has the highest stress. Stress is bad. Stress leads to failure. The square tubing looks like it has the lowest stress, but what it does have is right there where your welds are gonna have to be. That's a problem. I'm also betting the square tubing is the heaviest of the 3.

    if I were doing the FEA, I would have set the three Von Mises scales up to be the same. That way, I'd get a quick color confirmation (love to see blue! hate to see red!). I would also set up a factor of safety calculation for calculated stress divided into maximum design stress. I would also, if I had CPU time, have run a parametric optimization for number of sides from 4 to say 12, looking for just that barely acceptable factor of safety and looking for minimum weight. I'd bet I'd get 12 which is just a round tube. Again, depending on time, I might have put a more realistic node in there and run that to see what kind of weld problems I might have; my gut tells me I'll need gussets.
     
  25. Mar 24, 2008 #24
    Thanks again, really..
    I wish i knew all of this in the beginning of the year when i was starting the report. I already did everything and im on the discussion part of the report which has to be handed in by next week, and find out all what i did so far was wrong! I am dead and stupid of course..
    I had 5 designs. I analyzed the 1st design with circular, hexagonal and square tubing with different material. I choose the structure with highest von mises thinking it would be the strongest, then i continued with that choosen structure (Hexagonal) with the rest of the 4 designs..
    And now i am in the discussion and find out that all that was wrong.. I don't think there is time to fix anything now :(:(
     
  26. Mar 24, 2008 #25
    Nah, you're not dead. Part of engineering is learning to think on your feet, so put the best spin on this you can and take your lumps. I've found the best way to deal with mistakes in engineering is: 1) never apologize; 2) always fix them if you can; 3) don't repeat them. Crikey, I worked with the engineer who designed the "Fremont Flyer" and he survived (that's the infamous subway train that was supposed to come to a smooth stop in front of a gaggle of government officials standing at the end of the line. Instead it came smartly past them about 30 mph and on into the parking lot. Truly an "Oh Crap!" moment)
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook