"etd-150" is basically a modified 4100H alloy steel with an ultimate stength of 150 ksi and yield of 140 ksi. It's made by LaSalle Steel and is widely used for shafting material.
I doubt very much that you will be able to find S-N data for this alloy. Since LaSalle Steel is a relatively small company, they don't have much in the way of research staff who perfom this type of testing. I would start by calling their customer service people.
I agree that the manufacturer should be your first resort. Users of the material may have developed S-N curves, but they are probably proprietary and unavailable. In the absence of specific data, I use the following:
Reference: Shigley, Joseph Edward, and Mischke, Charles R., Mechanical Engineering Design, Fifth Edition, McGraw-Hill Book Co., Inc., 1989, Section 7.
A reasonably accurate approximate S – N curve for any steel can be constructed using the following equations given in this reference.
Log Sf = Log a + b Log N = Log (1.62 Sut) + Log N-0.0851
a = (0.9 Sut)2 / Se = 1.62 Sut
b = -(Log (0.9 Sut / Se)) / 3 (= -0.0851 when Se = .5 Sut)
Se = 0.5 Ftu = Endurance Limit = Stress corresponding to “infinite” life of 1,000,000 or more cycles.
Sf = Stress corresponding to a fatigue life, N, of 1000 to 1,000,000 cycles inclusive.
These equations are based on rotating beam test data, and must be adjusted for stress ratios other than R = -1.
I would caution against using this equation "as-is" for this material. This equation works well for structural applications where the fatigue loading is constant amplitude, and for common structural alloys. ETD150 has elements added to aid in machinability (N, Se, Te, S) that greatly reduce the fatigue strength vs. conventional SAE 4142 (ETD is based on SAE 4142).
If your application does not see a lot of variable amplitude loading, then perhaps you can use the equation, but substitute a more conservative number for the endurance limit-- maybe 0.35*Ftu or 0.4*Ftu instead of 0.5*Ftu.
TVP,
Interesting caveat. Thank you for the additional information on this material.
I would normally use Goodman diagrams, cumulative damage methods, etc., to account for variable amplitude alternating stresses. The stresses of varying amplitude, varying stress ratio, etc, can be converted to equivalent constant amplitude stresses—always, of course, with the understanding that these methods are not precise, and must be used with care and judgement.
It sounds as if this material is not a typical steel, and should be derated as you suggest; but if that is the case, it seems to me that it should be derated for any fatigue analysis. Why are variable amplitude stresses relevant?
You are correct in your assumption that this material should be "derated" for any fatigue analysis-- constant amplitude or variable amplitude. I was trying to explain that the equation you provided is suitable for constant amplitude loading, not for variable amplitude. Obviously you understand cumulative damage, so as long as those methods are employed, conversion to an equivalent constant amplitude stress can be done.
TVP,
I see. I misunderstood. Thanks for the clarification. I think I should have expanded slightly on my statement about rotating beam data and adjustments for stress ratios not equal to 1.0. There is probably not enough emphasis on that, and a little guidance might also have been in order.
From your knowledge of this material, would you suggest that the 1,000 cycle point be lowered, i. e., slip the entire curve down by 20 - 30 percent, or leave the 1,000 cycle point at 0.9 x Su and steepen the slope between 1,000 and 1,000,000 cycles? That is, do you know if this material is more or less susceptible to HCF than to LCF?
I have a copy of the S/N curves for this material. Fatigue limit for unotched specimen is approximately 500 MPa. For notched specimen it is approximately 125-163 MPa.
Can you provide additional details on what stress ratio was used? Stress amplitude? Notch geometry/stress concentration? What is the source of this data?
TVP,
My reference is a photocopy from La Salle Steel's "Engineering Data Book". I've been searching for the original with no success, and unfortunately my photocopy cuts off all test specifics except for:
"Rotating-Bending Test method for notched and unnotched".
I am sure that a more complete data sheet is availble from one's local steel distributor.
Thank lcubed for your valuable post.
I wonder if you have equations of bending fatigue for predicting S-N Curve for any steel. I don't know if we can use these equations to draw S-N curve of any kind of steel for design use.Thank you!
Who can help me with S-N Curve equations for bending fatigue? Mr.LCUBED or any other friends!
Mr.LCUBED has supplied the S-N equations for rotating beam test, I doubt very much if anyone can supply S-N curve equations based on bending fatigue data for any steel or alloy.I can't find the Reference: Shigley, Joseph Edward, and Mischke, Charles R., Mechanical Engineering Design, Fifth Edition, McGraw-Hill Book Co., Inc., 1989, Section 7.
