General Mechanical

General Mechanical

What output parameter would be the best to reduce likelihood of buckling?

    • zjuv9021
      Subscriber

      All,


      I'm wondering what output parameter I can try to minimize/maximize to reduce the likelihood of buckling to occur with this hyperelastic tubing:



      2 Geoemetrical configurations would be to reduce the length of the tube before it enters the Rigid Joint, and to change (reduce/increase) the distance between when it hits the top surface.


       


      What output parameter could I utilize that would ultimately reduce the likelihood of buckling to occur? 


      Kind Regards,


      Zach

    • jj77
      Subscriber

      There quite a few. 


       


      Euler buckling of column (comes first to mind), have dependencies on Young's mod., cross section properties (Ixx, Iyy,..), and the length of the column.


       


      Similar dependencies are for local buckling of tubes (thickness comes in there also). Here also the radius to thickness ratio, plays a role, say for small ratios local buckling is not likely.


       


      So yes, one could start with the length,and/or other parameters as mentioned above (depending if they can be changed or not),and do a study.

    • zjuv9021
      Subscriber

      Thank you for your reply.


      If I was saying that I could not control the geometrical parameters current-state, and wanted to look at DoE #1 & 2 listed in the above graphic (i.e. the distance between the fixed support on the left and the start of the outside tubing, or the distance between the tubing and when it strikes the upper surface upon displacement), what could I use as an output parameters to try to optimize via response surface method to reduce the likelihood of buckling to occur? 


      Regards,


      Zach

    • jj77
      Subscriber

      Well just try i out. If you might make say doe 1 smaller that should perhaps hopeful influence the onset buckling assuming it is of Euler type (global flexural buckling, and not local).


       


      Just have a go, this is the exciting part.

    • peteroznewman
      Subscriber

      Zach,


      The force-time plot showed two negative slopes, meaning there were two buckling events. You seemed to pay attention to the second negative slope and ignore the first negative slope. Is the first buckling event desirable and the second one not?  Does the first buckling event stop flow in the tube, which is the desirable behavior, while the second buckling event degrades the system in some way? Please elaborate on the basic function of the system as a whole. That will help us think about an output response to track that is meaningful to the basic function.


      Regards,
      Peter

    • zjuv9021
      Subscriber

      Thank you Peter,


      Neither buckling events are desirable, and in both cases would stop flow within the tubing, which is another undesirable behavior of the system. I would like to run some form of a Response Surface/DoE with the above, altering the distance between the fixed support (far left of tubing), and the distance between the top of the tubing before it touches the surface above in the +z direction.


      Hypothetically speaking, if the distance between the fixed support and the rigid tubing that is bent is essentially zero, the displacement would not be able to occur as there is no "slack" or room for the hyperelastic tubing to buckle. In essence, I would like to know how this 'slack' or free tubing affects the likelihood of this buckling phenomena, and what geometrical restrictions minimize this from occurring. 


      Could I simply try looking at reducing the Maximum Stress of the process?


      Kind Regards,


      Zach

    • peteroznewman
      Subscriber

      Zach,


      You haven't said what the basic function of the device is. If you don't advance the tube, it won't buckle. That is the obvious answer. What is the utility of advancing the tube? How do you measure that utility?  It seems that you want the tube to make contact with a surface that is some distance away. So the other obvious answer that you mention above is that the closer that surface is to the tube, the less you have to advance the tube for it to make contact and therefore the less likely it is to buckle on the way to touching. However, if you continue to advance the tube after touching, you will again cause a buckle, but at a much higher force.


      Kind regards,
      Peter

    • zjuv9021
      Subscriber

      The function is to provide drug, this is an implanted catheter, the upper surface mimics the skin, whereas the rigid tubing is essentially fascia that the catheter is implanted through. The displacement or advancing of the tube is to mimic patient movement.


      Thank you,


      Zach

    • peteroznewman
      Subscriber

      I accept that patient movement could advance the far end of the tube through the fascia. I don't understand why the other end of the tube is fixed. It that fixed to an implanted device?


      You say the tube will push against the skin when advanced, but the skin is flexible, while the model is pushing the tube against a rigid surface.  I expect the model would be more realistic if the surface that represents the skin was flexible.


      Regards,
      Peter

    • zjuv9021
      Subscriber

      The other end of the tube actually goes through an anchor which is sutured to the fascial plane.... This anchor is made up of an ETA treated Silicone. The catheter does not slide much while inside the anchor, therefore, I had made the assumption that the 'fixed support' is due to the fact that the movement of the catheter within the anchor is quite minimal in comparison to patient movement.


      Do you know of a good start to model skin?


      Kind Regards,


      Zach

    • peteroznewman
      Subscriber

      I can find journal articles that may be relevant.


      March 2018 Review and constitutive model.


      A finite element model of skin deformation


      I don't have access to these articles, but if you do, there may be something useful.

Viewing 10 reply threads
  • You must be logged in to reply to this topic.