S16-14 Handbook of Steel Construction CL26 Fatigue

[KB4444]

For Clause 26 of the Handbook it is covering the design requirements for Fatigue. I am not sure for my situation if this clause should include the stress range or if it is just the max stress value. 26.3.1 states stress ranges that are completely in compression are not considered, my case has the top and bottom flange alternating between causing the question.

For my situation I have a machine sitting on beams. The machine applies a dynamic load that can be +/- in direction. The load is large enough that it causes the top and bottom flange to alternate from tension to compression. It is provided that the machine operates at 4200 RPM so I have completed the calculation using the number of cycles expected in its lifetime for Fsr. Fsr was equal to 5, which is lower then Fsrt = 165 MPa from table 10. Therefore, I am using Fsr = 165 MPa (26.3.2 states that Fsr >= Fsrt).

When I am comparing the load induced fatigue I have a maximum stress while the force is acting in the positive direction and I have another maximum negative stress when the force is acting in the negative direction. The former is 95MPa and the latter is 85MPa.

Should I only be considering the higher of the two stresses (95MPa) to the 165 MPa? If so it passes. If I need to combine the stresses for a stress range because the flanges are not strictly in compression, then the condition fails as it is 180MPa. This doesn't seem right because the stress on the beam does not ever reach the endurance limit stress of 157.5 MPa (0.35 Tensile Stress or 0.35 x 450MPa) and the endurance limit accounts for an infinite number of loaded cycles.

 

[skewl]

I believe you will have to take your stress range as fsr = 95 - (-85) = 180 MPa if it goes into tension.

For your consideration, the Canadian bridge code CSA S6-19 states:

10.17.2.1 Calculation of stress range
The stress range for load-induced fatigue shall be calculated using ordinary elastic analysis and the principles of mechanics of materials. A more sophisticated analysis shall be required only in cases not covered in Table 10.7, such as major access holes and cutouts. Because the stress range shall be the algebraic difference between the maximum stress and minimum stress at a given location, only the stresses due to live load shall be considered.
At locations where the stresses resulting from the permanent loads are compressive, load-induced fatigue shall be disregarded when the compressive stress is at least twice the maximum tensile live load stress.

 

[KB4444]

Thanks for the response kewli.

Incorporating the full stress range leads into the last part that I had wrote then. Not sure how this Claus incorporates the endurance limit of steel. If the beam can only have the force applied in one direction at a time (resulting in stresses of 95 MPa or 85 MPa) then the beam is under the endurance limit at all times (Fendurance = 157.5 MPa). Wouldn't that conflict since the endurance limit could have that stress applied an infinite number of times without any concern but the Claus is stating it fails?

 

[skewl]

My understanding of this is a little iffy but I'd like to imagine when the member is in the fully compressed state (at -85 MPa), it experiences 180 MPa of tensile stress to bring it to 95 MPa. If the member never goes into the positive stress range (i.e. always in compression), is okay because the fatigue cracks "never" grow. But the moment it goes into the tension range, fatigue cracks can grow.

 

[SandwichEngine]

I been taught that you have to check the full 180 MPa because we're talking about stress range fluctuations. You have to check the entire stress range, positive to negative, not just positive to zero.

 

[SandwichEngine]

Posted too soon. The newest edition of the AISC manual makes it clear that you have to check the entire range.