Cyclic plasticity for fatigue evaluation

[gurmeet2003]

I have an application for fatigue evaluation where the stress goes above yield. I have briefly surveyed the literature both FEA and fatigue. It appears that I need to specify cyclic stress-strain properties. The materials are AISI 4140 (UTS 150000 psi) and steel UTS 80000 psi. If any one could throw some light on these material properties and and where I can get them I will be highly obliged.

Thanks,

Gurmeet

 

[rb1957]

the short answer is not much, take a paperclip and plastically cycle 10 times, you should have a broken paperclip.

the other part of the question is what stress to you have ? the nominal gross section section, the stress at the root of a stress concentration ?

AR-MMPDS-01 (or Mil Hdbk 5) has some s/n data for 4140, which clearly shows the material going locally plastic (Kt*max. cyclic stress > ftu).

i think you have an FE result on the surface of a stress concentration. you can estimate a Kt based on a far field stress (a stress in your model that you think is unaffected by the notch). or you can use the s/n data to infer a local linear stress (Kt*stress), use several different s/n curves (several different Kts), to compare with your local linear stress peak.

 

[gurmeet2003]

rb1957,

The high stress is in a fillet. It is an area of stress concentration. The fillet is inside a bolted joint (on the bolt itself). The stress in the fillet due to bolt preload is above yield (by FEA and theoratical calculations). A varying external load is applied to the joint. A part of this load transmits to the bolt and is source of fatigue.

Gurmeet

 

[rb1957]

if you're dealing with a bolt, the cyclic load in the bolt is much reduced by the bolt preload; for details google "fatgiue in preloaded bolts", you should get lots of hits.

it'll work out something like this ...
the cyclic load applied to the joint is less than the pre-load, so the load in the bolt cycles between the preload (with no external load applied) up to preload + a %age of the applied load, say 5% maybe 10% ... this means that the joint gaps when the load applied is 105% or 110% of the preload which is the real world result, as opposed to the ideal world where the joint gaps when the external load equals the preload.

i am surprised that your bolt is going plastic under just preload ...
what'll happen at the top end of the scatter band for preload, unless you control the preload carefully (say PLI washers) the preload in the bolt will be between 67% and 133% of your design preload ... both extremes have a significant effect on fatigue life.

 

[israelkk]

A local plasticity on the fillet is basically a crack from the fracture mechanic approach to fatigue analysis. If you can estimate the crack size you may use the crack propagation theory of fracture mechanics to estimate the number of cycles until the crack becomes critical.

 

[gurmeet2003]

rb1957,

The bolt is really a part we call a piston rod. The fillet is designed to be the weak location so that it will fail on overload. The preload is applied by a hydraulic tightening tool and therefore is fairly well controlled.

Thanks,
Gurmeet

 

[prost]

Cyclic stress strain is a 'nuther animal; it seems to be mentioned rarely in texts, and appears to be a complex subject, if no reason other than each material seems to be behave differently in cyclic loading. Often people run a uniaxial stress coupon with a cyclic load, record the stress-strain in time. You can imagine that these stress strain curves would vary in time, until they reach some kind of 'equilibrium' or steady state. With some materials that appear to give you a steady state cyclic response (sounds oxymoronic, I know), you can fit the stress strain curves with a power law such as Ramberg Osgood. Other materials seem to never quite settle down to steady state.

To learn more about how do use this data, consult the SAE book on fatigue, Fatigue Design Handbook, AE-10.

If you surveyed the literature, I am sure that you noticed that there are a myriad of ways to deal with fatigue in a structure that don't involve specifying cyclic stress strain properties; crack initiation and crack propagation methods being the most oft used.