S-N Curve - not even the manufacturer knew this one!

[Alexisonsmith]

Hey,

So I am currently trying to workout the fatigue characteristics of my Formula Student Space frame design, I am using Cold Drawn Seamless mild steel.

I have the following mechanical properties:
Re - 360 MPa
Rm - 450 MPa

I think these are the effective stress intensity range and the max or mean I am not sure stress.

How would I go about creating an S-N curve from this data? If you know you will be a lifesaver!

Thanks

[Alexisonsmith]

Hey,

So I am currently trying to workout the fatigue characteristics of my Formula Student Space frame design, I am using Cold Drawn Seamless mild steel.

I have the following mechanical properties:
Re - 360 MPa
Rm - 450 MPa

I think these are the effective stress intensity range and the max or mean I am not sure stress.

How would I go about creating an S-N curve from this data? If you know you will be a lifesaver!

Thanks

[Chris Allbee]

There really is no substitute for proper test data, but for an ESTIMATE....

Fatigue Strength Coefficient(sigma'_f): Su + 345 MPa
Fatigue Strength Exponent (b): -(1/6)log(sigma'_f/(0.5*Su))
Fatigue Ductility Coefficient (eps'_f): ln(100/(100-%RA))
Fatigue Ductility Exponent (c): Strong steels, c=-0.5 Ductile steels, c=-0.6
Transition Fatigue Life (2N_t): ln^-1(13.6-0.0185*BHN)
Cyclic Strain Hardening Exponent (n'): b/c
Cyclic Stress coefficient (K'): sigma'_f/((eps'_f)^n')

Su is ultimate tensile strength.

Information taken from: Fatigue Testing and Analysis. Yun-Li Lee, et al. 2005.

The equations to plug those coefficients into will be in your materials or mechanical design book. Or the internets.

Again, I have to stress that this will give you a ROUGH idea of the fatigue behavior of the material. My own experience has shown that the values given can be off plus or minus as much as 5x actual test values. Basically, there is a lot more uncertainty going this direction compared to getting actual fatigue testing data.

On a side note, its not uncommon for material suppliers to not have the fatigue data. It is very expensive to get and its use is surprisingly limited in the grand scheme of things.

Good luck.

[vandit]

Even if someone gives you a S-N curve of a tube, it just wont be useful for assuming the lifetime or safety of your complete space frame. In a simple language, every tube in your space frame is not experiencing the same loading behavior as the loading for a tube for which you have S-n curve. Moreover, stress concentration region in a single tube will be definitely different than stress conc. region in complete space frame.

The approach for determining the S-N curve is little different in such complex structure. The focus of fatigue design shifts from 'complete space frame' to 'stress concentration region(s)' in space frame. Instead of determining the S-N curve for whole space frame we now determine the S-N curve locally at the node where the high stresses are occuring. As prevalent in fatigue design, the S-N curve for this critical node would have to be determined from S-N curve of base material.

To determine our local S-N curve, 'Notch effect' is one of the most important factor which influences the material S-N curve (un-notched specimen data). Normally for a components with defined cross section (c.s) and notch geometry one can look up in design handbook and take care of notch effect. In our case the stress concentration region neither has definable c.s nor definable notch geometry so one will have to determine the 'stress gradient' in this region to take into account notch effect.

Once you know the 'stress gradient' and 'maximum stress' in stress concentration region it is possible to determine the local S-N curve at this region/node.

I am very much interested in doing this for you but need some more information. Can you send me a short email at goyal.vandit@gmail.com so that i can communicate with help of figures (which otherwise here whould take ages to approve)

[Demon Of Speed]

You really should have covered how to calculate fatigue stress in one of your classes. I know Shigley's Mechanical Engineering Design, covers it. Additionally, you can add in experimental data into the equations and get more accurate results (I have a paper somewhere of a now retired professor that when over how to do this).

http://www.supercars.net/pitlane/pics/180059/3356542d.jpg

[vandit]

>> Line 3: What surface finish value (Ra or Rz) did you consider for calculating this factor ? '0.507' is way too conservative for a Cold drawn extruded component.

>> Contradicting 7th and 12th lines: In 7th line you mentioned Axial loading and then in 12th line you considered factor for shear loading !!

>> Line 24: Stress concentration factor is due to notch and its geometry and certainly not due to 'unknowns in material properties'. For such 'unknowns' you already took material factor in 2nd line '0.506'.

In a seamless tube, i dont see an external notch anywhere.

[Demon Of Speed]

It has been a long wile since I did these calculations so...

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by vandit:
&gt;&gt; Line 3: What surface finish value (Ra or Rz) did you consider for calculating this factor ? '0.507' is way too conservative for a Cold drawn extruded component.

&gt;&gt; Contradicting 7th and 12th lines: In 7th line you mentioned Axial loading and then in 12th line you considered factor for shear loading !!

&gt;&gt; Line 24: Stress concentration factor is due to notch and its geometry and certainly not due to 'unknowns in material properties'. For such 'unknowns' you already took material factor in 2nd line '0.506'.

In a seamless tube, i dont see an external notch anywhere. </div></BLOCKQUOTE>

Line 3: If I remember correctly I was unsure at the time if I could get the size I wanted (something weird) in cold drawn, so I think that is the hot extruded factor.

Line 7 and 12: If I remember right the load is axial (indicating pin-pin loading on line 7), however because of how the part is made during the axial loading shear stress was induced (I think it was where/how is was welded).

Line 24: The stress concentration factor is for welding (as welds are considered stares concentrators).

Remember this was a sample calculation based on my part. For other parts things will change, and to know what factors to use, look at the book I recommended, or any other book that covers fatigue.