FEM analysis of helical springs, with pitch angle, lenght , diameter

[hermatician]

hi all,

I have to analyse, by FEM, the stress and stroke change of spring, in compression fatigue.

please help me, how to make a cylindrical structure of spring in ABAQUS??

2. how to set pitch length, angle and the biggest question,

3. how to set pressure which is not exactly on the surface but in the direction of the axis of the spring.

please help me!!!!!!!!!!!

 

[MikeHalloran]

I don't know anything about Abaqus, but I know a little about springs.

The first thing you need to understand about helical springs is that the forces applied to the ends actually stress the wire in torsion. There is no 'pressure'. Maybe that will get you started.




Mike Halloran
Pembroke Pines, FL, USA

 

[hermatician]

hey mike, thanks for your answer, but I have grounded surface at the top and bottom of the spring and there i can have pressure, but ja its actually a different machine in which 300 springs are being tested together and i have been going through many dramas with these springs.
as you said you know abt springs can you tell me or send me details of setting the springs. i don know why its needed and though my springs are given set already its changing its length every time. can you explainwhy???

 

[MikeHalloran]

Most any text that goes into detail about spring design can tell you why 'setting' is necessary, in more detail than I can remember.
If you are involved in spring production, you need such a book on your desk. Spring manufacturers used to give them away; you can probably find a pdf on the web now.

What I do remember from many misadventures with springs is that setting becomes an issue when a design is not as conservative as I like them to be. If the end product doesn't have to fly, I can't think of a good reason to design a spring so close to its yield point that its free length changes in service.

That probably doesn't matter to you; it sounds like you have no authority to change the design, even if you wanted to. Perhaps we could help more if you could outline the dramas you have experienced in more detail. ... without giving the store away, of course.



Mike Halloran
Pembroke Pines, FL, USA

 

[hermatician]

well, i will first explain my task

I am working on very high cycle fatigue behavior of helical compression springs, right now I m involved in FEM modeling of it. I am not concerned with manufacturing. we have a special machine where we put around 576 springs of different rate/stress levels, and prestrain them to a point say 4mm(compression) and then give a cyclic loading of say 10 mm amplitude.

now the problem is this 10mm value is changing because of change in spring free length during the test. so we tested for this and saw that after setting twice and thrice its changing its length still.

my question is, as we are not going to the plastic deformation regime, why the spring is showing permanent deformation.

and about setting i read the materials available but i would like to know the mechanism why it happens that you give it a compression till solidification of coil and then you should not have change in length. n it is calculated say that if i set it to 10mm it wont change the length but if i do it to 8 mm will it?? or not?? and main question is why?

I hope i made myself clear.

 

[TVP]

I suggest you review the following references for behavior of springs, including methods for finite element modeling. The last link has a number of .pdf files that you can download, including an excellent paper by Prof. Kaiser on VHCF testing of springs.


FEM for Springs

http://www.jsse-web.jp/english/FEM_for_sp.html

Materials for Springs

http://www.jsse-web.jp/english/materials_for_sp.html

JSSE 60th Anniversary International Symposium Proceedings

http://www.jsse-web.jp/english/proceedings_list.html

 

[hermatician]

well Mike I am working with prof Kaiser and Prof Brita Pyttel

 

[MikeHalloran]

Well, I can answer one of your questions: The reason you preset to coil-bind is that >you just flat can't compress the spring any farther than that<.
... so that any portion of the spring that's going to yield locally will do so, and subsequent cycles will then not produce yield.

However, I'm a little confused by your procedure. It appears that you are not prestressing beyond the working stress range. So it should not come as a surprise that your springs are yielding.



Mike Halloran
Pembroke Pines, FL, USA

 

[hermatician]

yes exactly,I too did not understand that whats the point in compressing but not deforming plastically....... but they r saying that they do not want much change in length and accrding to them, in machine the displc is say 12mm so they r comressing to 11mm so they sud not have problem, but i think they will have as its not set, n so it will have loss in length....right

 

[MikeHalloran]

Loss of free length is kind of a red herring, because springs are installed with a preload in nearly all real applications. I.e., in Real Life, the only forces of concern are the force at installed height and the force at actuated height. The free length and the force at solid height are of interest for analysis, but don't matter to most machines.

If you test a real spring, measuring force vs. length from free to coil-bind, you will generate a graph that's a fairly straight line, but has distinct hooks at the ends.

The hook near coil-bind has to do with coils bottoming out and removing themselves from the (real) spring equation, in a manner that will repeat for one spring in one machine, but not otherwise.

The hook near free length has to do with spplying the load first at the very tip of the wire, and transitioning to a condition where there are effectively at least two forces applied, producing a twisting moment on the wire, all of which is dependent on the taper and precise geometry of the end finish of the wire and of the geometry of the mating machine, and which is again repeatable only for a given spring and a given machine, and perhaps not even then.

Which is kind of a long-winded way of saying that using free length as a reference point introduces a lot of uncertainty into any measurement you care to make.


Mike Halloran
Pembroke Pines, FL, USA