Simulating IEC Fatigue Load Case 1.2 for Jacket Structure

Wind & Water General
1

Hello!
I am wanting to generate 12 load cases to apply to ABAQUS for my project as stated in another thread. However I am unsure about the writing of the input file when it comes to the Inflow wind.

If possible can someone validate my approach? I am using the k13 Deepwater Project Upwind Site data, table 59 to set up 12 load case files. The table provides the Turbulence intensity, peak spectral period, significant wave heights for a particular mean wind speed. I want to carry out an hour of each load case and as such have changed the simulated times in the files respectively. I also increased the time step to 30s to reduce to be less computationaly intensive in later analysis. I have used the site data for HydroDyn input files and now I am having trouble with the inflow file.

I am unsure as to how to obtain the binary .bts files that FAST needs to generate the inflow wind input file with this data.

Thanks,
Rory.

2

Dear Roy,

A time step of 30 s sounds too large for FAST model. See the following FAQ for guidance on selecting a time step for FAST v7: wind.nrel.gov/designcodes/FAQ.html#FASTTimeStep. (FAST v8 has some further numerical options, but the FAQ still applies to ElastoDyn and AeroDyn.)

TurbSim can be used to generate the turbulent wind data (*.bts) files needed by FAST: nwtc.nrel.gov/TurbSim.

Best regards,

3

Dear Jason,

I have been trying to perform a fatigue analysis on the Upwind jacket structure. Initially, I have made wind fields using Turbsim for each mean wind speed at hub height. Then I ran the FAST 17 times for 17 load cases given in the Upwind design basis with a period of 10 min for each of them. Now I have load time series at the end of joints (via OutAll option) within the jacket. Furthemore, I want to calculate the stresses using these loads but they are quite small (3.437 E+02, …)and they will not lead to such stress ranges that would be considered for fatigue analysis. In other words, FAST loads output units are N and N.m for axial load and moments respectively and when they are changed into MPa the results are even less than one MPa !

  • Apparent fixity length is 26 m.

Would you please tell me what is the problem with that ???

Best regards,

4

Dear Arsalan,

I’m not sure I can answer your question without knowing more about what you are doing. I would not expect SubDyn to output loads that are smaller than expected.

Best regards,

5

Dear Jason,

Really appreciate for your answer. What do you need to know precisely ? I really trapped in it.

Best regards,

6

Dear Arsalan,

Here are a couple clarifying questions:
[]What is the source of the Upwind jacket model you are running? Is this a model you maid or was it provided to you by someone else?
[]Does the SubDyn model you are using produce results you trust for any load case?

Best regards,

7

Dear Jason,

Here are my answers ;

  1. The basic model which I use is Test #21 in FAST v.8 package but with some modifications. Initially, I tried to find relevant mode shapes of the structure using BModes for apparent fixity length of 26 m and checked that with you. Besides, four new joints were defined in SubDyn joints list and subsequently in member list. Then I used the " Upwind Design Basis " presented by Fischer et al. in which there can be found 17 load cases including relevant Hs, Tp, and Vhub as well as their frequency of occurence. Moreover, for different cases both those are in idling time and operation time I have modified the pitch angels and rotor speed. Then I ran the FAST for simulation time of 10 min for each load case.

  2. Actually SubDyn outputs have the same range (#.##E+02 to #.##E+03) for members dynamics components of axial load and in-plane and out-of-plane moments for all 17 load cases !!!

Best regards,

8

Dear Jason,

It would be really appreciated if you can give me any advice about this problem.

Best regards,

9

Dear Arsalan,

When you say you output the dynamic components of the axial load do you mean you are outputting e.g. M1N1FMxe rather than M1N1FKxe i.e. the static (elastic) component? I can see the dynamic components being much smaller than the static components.

SubDyn can only output the static components (KUe) or dynamic components (MUedotdot), rather than the true reaction loads (which cannot be computed easily for a lattice structure with closed loops). However, the static loads are usually quite close to the true reaction loads, so, I’d recommend that you use e.g. M1N1FKxe rather than M1N1FMxe. I generally do not use the dynamic component. See the SubDyn Theory section (6) of the draft SubDyn User’s Guide and Theory Manual for more information: wind.nrel.gov/nwtc/docs/SubDyn_Manual.pdf.

Best regards,

10

Dear Jason,

I really appreciate for your cralifying answer. I found it really helpful. But in SunDyn theory manual on page 37, figure 5 represents the local coordinate system for members. Thus it seems the axial load in a member is e.g. M1N1FKze and for in-plane and out-of-plane bending moments they are M1N1MKxe and M1N1MKye respectively while you have mentioned M1N1FKxe for axial load.
Is my understanding incorrect ?

Best regards,

11

Dear Arsalan,

Yes, I agree that M1N1FKze is the axial load (not M1N1FKxe from my prior post) and M1N1MKxe and M1N1MKye are output-of-plane moments. In my prior post, I was simply using one of the load components as an example. What I wanted to emphasize is the third-from-last character should be a K instead of an M.

Best regards,

12

Dear Jason,

Your answer is clear and I am really grateful fot that.

Best regards,