Dear developer,
For a monopile-supported wind turbine model, could elastodyn tower reference point be located at difference location from monopile subdyn interface point?
Suppose for 15 MW wind turbine model downloaded from github, the tower base is 15 m above mean sea level, and then elastodyn tower reference point z coordinate is set 15m. However, the subdyn monopile interface point is set at joint at -10m below mean sea level. Is it acceptable?
Based on the openfast documentation, it seems that there is no need to make the same location for elastodyn tower reference point and subdyn monopile interface point, as they are assumed to be rigidly connected. However, making them the same location may contribute to a better convergency performance. Many thanks if you could correct me.
I agree with your assessment. It is generally recommend to locate the ElastoDyn platform reference point at the same location of the SubDyn interface joint, but this is not strictly required within OpenFAST. Any offset between these points in ElastoDyn and SubDyn will be handled by the internal spatial mesh-to-mesh handling within the OpenFAST glue code, where offsets between points are treated via rigid-body kinematics (that is, rotation of the platform reference point in ElastoDyn passed to SubDyn would introduce translation of the SubDyn interface joint based on the offset arm and forces at the SubDyn interface joint passed to ElastoDyn would introduce a moment at the ElastoDyn platform reference point based on the offset moment arm).
I may want to confirm how subdyn calculates the member property.
Below code is from SD_FEM.f90 of openfast v3.5.2.
if (eType==idMemberBeamCirc) then
E = Init%PropsB(P1, 2) ! TODO E2
G = Init%PropsB(P1, 3) ! TODO G2
rho = Init%PropsB(P1, 4) ! TODO rho2
D1 = Init%PropsB(P1, 5)
t1 = Init%PropsB(P1, 6)
D2 = Init%PropsB(P2, 5)
t2 = Init%PropsB(P2, 6)
r1 = 0.25*(D1 + D2)
t = 0.5*(t1+t2)
if ( EqualRealNos(t, 0.0_ReKi) ) then
r2 = 0
else
r2 = r1 - t
endif
A = Pi_D*(r1*r1-r2*r2)
Ixx = 0.25*Pi_D*(r1**4-r2**4)
Iyy = Ixx
Jzz = 2.0*Ixx
For tube member, its elastic modulus, shear modulus and density currently only depend on PropSet at the starting end of the member. And the diameter and thickness of the member are the average values of the counterparts at two ends of this member.
That’s correct. Tapered members (where diameter and thickness change from joint to joint) in SubDyn are considered as a series of stepped uniform beams, where there is one step for each division of the member (for NDiv > 1), which is why the SubDyn documentation recommends setting NDiv > 1 for tapered members.