5MW wind turbine blade root Fatigue parameters

[yc-shuai]

Hello everyone:

I would like to consult a question about mlife, in mlife I noticed that NSlopes recommends 3 and 10 respectively when calculating tower fatigue and blade root fatigue, but I did not find the LUlt corresponding to tower and blade root, I hope to get your help.

In addition, I have a question, the blade is a composite material,what is the logic of defining the SN curve in mlife, and which material is used to replace it？
Finally, I wish you all a happy life.

[jjonkman]

Dear @yc-shuai,

Regarding LUlt in MLife, this input has been discussed many times on our forum, e.g., see: https://forums.nrel.gov/t/mlife-user-defined-distribution/1970.

I agree with your concern regarding the SN slope for composite materials (given that it is composed of multiple materials), which is a limitation of the DEL approach. As an alternative to DEL, you could derive local material stress in a post-processing step of OpenFAST, upon which you could compute fatigue at the local material level.

Best regards,

[yc-shuai]

Dear @jjonkman ,

Thank you for your guidance, which has deepened my understanding of MLife. I’ve summarized the key points about LUIt in the discussion, hoping it benefits others:

1. Unit Consistency: The unit of LUIt must match that in the input files. For instance, if stress in the input files is given in pascals (Pa), LUIt should also be in pascals.

2. Determining LUIt: Ideally, LUIt is obtained through FEA. But if calculation isn’t possible, it can be defined by multiplying the maximum value of relevant loads (e.g., from MExtremes) by scaling factors like 1.25, 2.5, 5, 10, or 20.

3. Tower Considerations: For towers needing shear forces and moments in x, y, z directions as inputs, and assuming axial symmetry:

   • Shear force at tower base (x/y): MAX(SQRT(TwrBsFxt² + TwrBsFyt²))
   • Axial force at tower base (z): MAX(ABS(TwrBsFzt))
   • Bending moment at tower base (x/y): MAX(SQRT(TwrBsMxt² + TwrBsMyt²))
   • Torque at tower base (z): MAX(ABS(TwrBsMzt))

4. LUIt Equation for Thin-Walled Circular Sections:

   • Transverse shear: LUIt = tau_yield × A/2
   • Axial load: LUIt = sigma_yield × A
   • Pure bending: LUIt = sigma_yield × I/y
   • Torsion: LUIt = tau_yield × J/r

   Where:

   • D_o = outer diameter, D_i = inner diameter, r = D_o/2, A = (π/4)(D_o² - D_i²), I = (π/64)(D_o⁴ - D_i⁴), J = 2 × I = (π/32)(D_o⁴ - D_i⁴).

I also have some questions:

1. Stress Calculation Scaling Factor: I’m using stress for calculations. How should I determine the scaling factor? I noted a paper “Model Development and Loads Analysis of an Offshore Wind Turbine on a Tension Leg Platform, with a Comparison to Other Floating Turbine Concepts”suggesting a factor of 20, but you previously advised a factor slightly above 1.
2. MExtremes Alternative: I haven’t used MExtremes. Can its functionality be replaced by Excel’s Max and Min functions?
3. LUIt in Ansys for Blades: I have a composite blade model analyzed as a cantilever beam. To get LUIt from Ansys, should I use the stress at which the section fails, or is there another standard?
4. Regarding the alternative method you mentioned for DEL, which involves deriving material stresses during post-processing, could you please explain it in more detail?

Best regards,

[jjonkman]

Dear @yc-shuai,

Can you clarify what you are doing? In (1), you mention you are using stress; are you applying OpenFAST and deriving stress from sectional load in a post-processing step? E.g. for pure bending of the tower at its base, are you computing the maximum stress as sigma_max = MAX(SQRT(TwrBsMxt² + TwrBsMyt²)) * y / I where I is given as you stated and y = D_o/2?

Here are my responses assuming this is what you are doing:

1. In my example of pure bending of the tower at its base, you can compare sigma_max to sigma_yield directly.
2. MExtremes does more than just compute the min and max of time series. MExtremes also applies scales, offsets, and partial safety factors; calculates statics; locates peaks through peak-finding algorithm; and computes extreme event tables.
3. As in (1), if you are working directly with stress rather than sectional load, you can use the yield strength (e.g, sigma_yield in my example above for pure bending of the tower at its base) as LUlt.
4. Again, you can derive stress from sectional load, e.g., for sigma_max as I summarize above for pure bending of the tower at its base.

Best regards,

[yc-shuai]

Dear @jjonkman ，
Thank you for your response. Yes, your understanding is correct. My thoughts are as follows:

Use OpenFAST to calculate the loads at the tower or blade root sections, convert them into stress, identify the maximum stress value, and then multiply it by a scaling factor to obtain the LUIt.

Alternatively, directly obtain the LUIt through FEA.

Best regards,