Need Help Simulating Pole Vault Through FEA in ANSYS 2

[khorn06]

Hi, I am stressing over a class project I have to do. I got myself way in over my head with this one. I am trying to simulate pole vaulting through the use of FEA with ANSYS. My professor is new to the topic and has not been able to help me much. All I have to go off of, that I consider useful, is a research paper "SIMULATION OF ‘SMART’ POLE VAULTING" by Mats Ekevad and Bengt Lundberg. This paper is actually really cool because the authors have done exactly what I am trying to do. However, because I am new to this, I am left with very little to go off of in terms of creating the actual models and setting the proper loads and constraints. I was really hoping someone here has a better understanding than I do and could help me out. If there are tutorials that I can use to piece my project together or if someone has done anything similar. I am really stuck and have exhausted just about all other resources that I can think of.

The overall goal of my project is to accurately simulate pole vault as best I can in order to determine the optimal pole length for any given vaulter.

The gist of the paper:

The pole is considered to be a one-dimensional beam with transverse shear, large displacements, large strains, and rotary as well as transverse inertia. Pin joints connect the bottom end of the pole to the ground and the top end of the pole to the vaulter. The pole is modeled using 20 equally long two-node beam elements. The initial velocity of the pole is taken to be the same as that of the vaulter except at the box end which is simply supported. It has a variably thin-walled pip cross-section, with three layers of linearly elastic materials (glass fibre).

The vaulter model consists of 6 segments with five pin joints. The two legs of the vaulter are modeled as a single, two segment leg. The same is true for the two arms.

Here's an image of the initial stance of the pole vaulter

UPDATE: By placing a point mass located at the vaulter's center of gravity I was able to accurately mimic the pole's motion

Here's an image of essentially the goal of what I want my simulation to look it.

If you have absolutely no idea what pole vaulting is. Here's a quick link to give you a good idea on what pole vault entails from an engineering perspective.
Link

At this point in time I have close to nothing completed that gets me closer to my goal. I have modeled the geometry of the pole and the vaulter using simple line bodies in ANSYS DesignModeler. Instead of trying to create the three elastic layers of the pole, I used a single layer and am treating it as a uniform material with material properties similar to that of the composite. When I try to apply an sort of loads, the outcome does not even come close to what I was expecting. Right now it appears that when my pole is supposed to bend, it collapses in on itself and shrinks, then recoils back to its original shape. When I add in gravitational effects, it gets even worse. Everything just falls.

 

[btrueblood]

Hmm...I quit vaulting in Junior High after a nasty fall that cut my scalp open...but have some familiarity with it. At the time of pole plant, the vaulter (at least one that is really trying for altitude) is using his arms to apply a moment - i.e. the arm held lower on the pole is kept "stiff" to an extent, and this helps to put the pole into first-order buckling/bending. If you don't have an initial bend in the pole, or if your modeller is not set up to consider bending of the pole, then I don't think it will replicate the results shown in your graphic.

 

[gravityandinertia]

You need at least the following:

1. A non-linear dynamic study
2. A moment applied to the pole to bend it, then a compressive force from the vaulter running to further bend it.
3. I don't believe a pin connector of the pole to the vaulter is correct. If that was the case the vaulter's center of gravity would likely stay much more under the connection, they can/do obviously transmit a bending moment using their muscle, so a rigid connection is likely much more accurate.

Beyond that, I don't know enough about the situation or have enough time to dig deeper.

 

[IRstuff]

It seems to me that FEA is an incredibly difficult option to use for modeling this. The main reason is that the vault is a dynamically controlled operation, i.e., the vaulter changes their body position and configuration (call this body state function) as a function of where they are in the vault. Therefore, you can only simulate the case where the vaulter is both ideal and controlling their vault ideally, since that's the only case where the solution is relatively deterministic.

I can sort of see how you could possibly make the vaulter's body state function into a function of discretized time and use that in conjunction with an ideal pole motion, but it's going to be very non-trivial. Nevertheless, for any given length of pole, the vaulter would not necessarily use the same body state function, as the vaulter would adapt themselves to the length of the pole.

The single arm bit is a little troubling, since the two arms need to be separated sufficiently to provide a stable frame upon which the vaulter can leverage their body position. Given that, then a pin connection may be OK, since the arm movement is constrained by the arm lengths and body width through the shoulders. The top of page 3 here:

http://www.ida.upmc.fr/~hoepffner/research/perche.pdf

shows how much arm movement there is relative to the hand positions. Nevertheless, it would appear from that sequence that the pole might rotate a bit to accommodate the vaulter's body orientation in the last two time points of the diagram.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[khorn06]

Thank you for the replies everyone, even to IRstuff whose post was deleted. So overall, this project is too complicated for me to accomplish what i had hoped to accomplish. On the bright side, I have actually been able to mimic the correct pole motions with the bending, simply by placing a point mass located at he vaulter's center of gravity directly below his top hand grip. I think with what i have now, ill just talk to my professor about finding a more suitable goal.

 

[btrueblood]

Dunno, sounds like you have replicated the analysis you referenced. How much more work would it be to add some more complexity to the body model?

 

[khorn06]

Well that's kind of the problem that i am a little confused about. When i model the pole vaulter as 5 different limbs and give them the appropriate mass's then my pole does not bend and everything just falls to the ground. Only when i idealize the vaulter as a point mass then the pole bends and goes through the correct motions.

What I assume based on my results. When i create the point mass, ANSYS is idealizing the inertial effects of the body (that's from the documentation). So when i try to create the vaulter and segment out his limbs, I guess I am not applying any inertial effects?. I have tried applying an acceleration to the body but that yielded nothing different. I am applying moments to the joints as well. Maybe I am using wrong values for moments but changing those as well have yielded no differences.

and technically I am only half way to accomplishing my goal, with the point mass, my vaulter can't do any work on the pole so I can not get an estimate as to how high he will jump, but half way there is significantly better than nothing

 

[IRstuff]

It's possible at your 5-segment model isn't moving correctly. There's basically a bit of a pumping action at the instant the vaulter leaves the ground so that their CG moves to maximize the momentum transfer into the pole. It's also likely that your point mass model isn't correctly modeling the motion either, since the body's CG is moving relative to the pole and hand positions. Note also that the body scrunching in mid jump vs. end changes the angular momentum further

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[btrueblood]

Your lumped mass jumper is a part of the pole? I.e. it has rotational as well as linear intertia, and can transmit forces and moments? How are you applying moments to the joints - I thought these were pinned joints? Can the stick model transmit a moment to the end of the pole?

 

[ThomasH]

Hi

Since this is a school project I was a bit uncertain how to respond ;-). Not only due to the forum rules but also since I remembered reading about pole vaulting and FEA several years ago.

But since it is friday and the sun is shining .

Try this link:

https://www.diva-portal.org/smash/get/diva2:998982/FULLTEXT01.pdf

It leads you to a doctoral thesis on wood drying. Perhaps not what interrests you right now but but if you move to the end of the thesis the author has included his licenciate thesis with the title "On the mechanics of pole vaulting" and it includes articles about simulation. The information may be useful for your project.

Good Luck

Thomas

 

[SkipVought]

"But since it is friday and the sun is shining..."

...if you get a handle on pole vaulting analysis, could you take a wack at Hungarian vaulting? I know that the Holy Crown of Hungary was IN a vault at Fort Knox and was returned to the Hungarians by Jimmy Carter. So any other vaulting related analysis would be greatly appreciated. ;-)

Skip,
[sub]
Just traded in my OLD subtlety...
for a NUance![/sub]

 

[IceBreakerSours]

Watch the OpenSim channel for a free solution or the AnybodyTech channel on YouTube for a commercial solution. FEM is an expensive tool, in addition to the limitations mentioned previously. The downside is that motion is required. However, not too long ago, Anybody has developed a solution to get over this hurdle.

*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[BILL_mce]

I would advise just getting through the course with a decent grade. You could probably do a PhD on this. I would suggest you outline the challenge of the whole problem and the tackle some appropriate aspect of the problem and skip the really tricky bits. Like figure out the energy storage in the pole and if the jump is executed properly what is the maximu height the jumper might achieve and how this would be sensitive to approach speed, technique, pole construction and damping. You could do some NL fem analysis of the bending for stress and the energy stored as a function of the displacement, you could do an NL dynamic like a single DOF system accept you have the pole where the print should be. You can develop a decent understanding of the basics of pole vault dynamics without doing the completely "free" problem as IRStuff has pointed out will be mighty tricky.

Bill McEachern, P.Eng.
Redstone Six Consulting Ltd.

http://www.redstone6.com