Calculate Elasto_tower.dat with WISDEM

Wind & Water Structural Analysis
1

Hi everybody,

I am using WISDEM to calculate TwFAM1Sh(2-6) and TwFAM2Sh(2-6), which is based on Frame3DD. Hence I create a new topic. Calculate TwFAM1Sh(2-6) and TwFAM2Sh(2-6) is based on BModes.

  1. Below is shown in weis/aeroelasticse/openmdao_openfast.py. Is it the right way to transform the result from WISDEM to elastodyn_tower.dat?

z_tow = twr_elev
z_sec, _ = util.nodal2sectional(z_tow)
sec_loc = (z_sec - z_sec[0]) / (z_sec[-1] - z_sec[0])
fst_vt[‘ElastoDynTower’][‘NTwInpSt’] = len(sec_loc)
fst_vt[‘ElastoDynTower’][‘HtFract’] = sec_loc
fst_vt[‘ElastoDynTower’][‘TMassDen’] = inputs[‘mass_den’]
fst_vt[‘ElastoDynTower’][‘TwFAStif’] = inputs[‘foreaft_stff’]
fst_vt[‘ElastoDynTower’][‘TwSSStif’] = inputs[‘sideside_stff’]
for fass in [‘fore_aft’,‘side_side’]:
for idir in [0,1]:
if not np.any(inputs[f’{fass}_modes’][idir,:]):
logger.warning(f’WARNING: {fass} tower shape coefficients are zero which will cause errors in using ElastoDyn’)
fst_vt[‘ElastoDynTower’][‘TwFAM1Sh’] = inputs[‘fore_aft_modes’][0, :] / np.sum(inputs[‘fore_aft_modes’][0, :])
fst_vt[‘ElastoDynTower’][‘TwSSM1Sh’] = inputs[‘side_side_modes’][0, :] / np.sum(inputs[‘side_side_modes’][0, :])

  1. will the result from WISDEM the same as BModes?
  2. The inconsistence of mass between WISDEM and ElastoDyn.sum: in my case, tower mass in wisdem = 1514888kg; but in openfast, tower mass = 1497943.250ky.

Thank you for your help in advance!

Best regards,

2

Hi,

over the years we’ve transitioned away from BModes into WEIS/Frame3dd, and as far as we know both codes generate accurate ElastoDyn input files. The mass inconsistency of 1.1% is likely from the different meshing and I wouldn’t be too concerned about it.

1 Like
3

I’m not familiar with WEIS. But in case it is useful: ElastoDyn interpolates the input definition from the user and creates equal-length beams that are straight (not tapered).

1 Like
4

Thank you, Pietro.
One follow-up quesion. I am using AQWA to calculate the hydrodynamic coefficients for OpenFAST, should I include mass, I and CoG of all components or just the floater platform? If mass，CoG and I of tower is needed, is it accuracy to use the data from WISDEM?

5

I’m not familiar with AQWA, but I think the idea should be the same. When computing the potential-flow hydrodynamic coefficients, assuming you are only considering first-order coefficients, the mass and moments of inertia of the system are not relevant. HydroDyn only uses the wave excitation, added mass, and wave radiation damping coefficients, not motion RAOs. For CoG, you need to set it to whatever reference point you are using in AQWA to effectively remove the self-weight contributions to hydrostatic stiffness. This is because OpenFAST adds gravity force and moment internally, and it is important to avoid double-counting self-weight in the hydrostatic stiffness matrix. If you also want to compute second-order hydrodynamic coefficients, you will need to include the mass, moments of inertia, and CoG of the entire system into AQWA, so your first-order RAOs from AQWA match the full-system behavior as close as possible. There will be approximation involved in this step because it is not possible to capture the full time-domain physics of OpenFAST in the frequency domain.

1 Like