Hi @Jason.Jonkman and @Ryan.Davies ,
Thank you for the information about natural frequencies.
I will check the OpenFAST module and forum, and I will get back to you if I have any issues. Does it include both wet and dry natural frequencies?
Additionally, I have connected the bodies using line elements.
TypeName Diam Mass/m EA BA/-zeta EI Cd Ca CdAx CaAx
(name) (m) (kg/m) (N) (N-s/-) (-) (-) (-) (-) (-)
nylon 0.18 20.6 2.03E5 -0.8 0 1.0 0.0 1.6 0.05
beam 0.25 100.0 1.5e12 -1.0 1e7 0.0 0.0 0.1 0.0
It provides the forces in Fx, Fy, and Fz, but I am unable to get the moments at the connections of the joints. If you have any suggestions please answer for this. I am simulating connections where the lines are set as ‘POINTS’: Fixed, Free, and Coupled. The initializations are successful, so that’s fine. I also tried changing the values of UnstrLen, NumSegs, and dtM to 1e-6. Of course, it takes a long time to run these simulations, sometimes up to a full day.
Thanks and Regards,
Soman.SS
Hi all,
Sorry by the mistake. The connections beam line elements are unsuccessful during initializations. It is fine.
Thanks and Regards,
Soman.SS
@Somu.Sakthivel,
If you are simulating a rigid beam as a line, then you need to attach it to the bodies with zero length rods. Points are a 3 DOF connection, meaning they do not pass moments to their parent object. The rod documentation describes how to use a zero length rod as a 6 DOF point:
“A special case exists if a Rod is specified with zero elements: in that case it is given zero length, and its end B coordinates in the input file are instead interpreted as vector components to describe its direction vector. This case is meant for making cantilever points of a line with bending stiffness. A fixed zero-length rod can be used to make a cantilever point of a power cable to the ground, a body, or a coupling point. A free zero-length rod can be used to join two different types of power cable segments, and it will pass moments between the cable segments without adding any mass or other characteristics.”
1 Like
Dear @Somu.Sakthivel,
Regarding your question on natural frequencies, full-system natural frequencies calculated through an OpenFAST linearization analysis would include the wet natural frequencies of the mooring lines.
Best regards,
1 Like
Hi @Jason.Jonkman
Thank you for your rapid responses.
Regards,
Soman.SS
Hi @Ryan.Davies ,
Thank you for your suggestion and explanation about rod element.
yes, i checked in the documents of model structure section but i have doubt about how it is. Now i have a clear idea. i will check it.
Regards,
Soman.SS
Hi, @Ryan.Davies,
Good afternoon,
I have prepared an input file for two modules, and I encountered an error: “Warning: extreme pitch moment from body-attached Rod 2.” Additionally, the simulation took hours to run. Q1. How can I reduce the simulation time? I tried lowering the dtM values, as well as adjusting unstrlen and NumSegs.
Q2. Initially, I want to know if the following file is correct for adding a rod between the modules.
Q3. Moreover, I need to use multiple cylinders with rods, and my intention is for the cylinders to be considered as Morison elements. Are the following inputs correct? Additionally, how can I define a zero-length rod? Can I set the UnstreLen value to zero?
Regards,
SomanSS
Hi @Somu.Sakthivel,
Q1: Runtime can be decreased by both increasing dtm (fewer time steps) and decreasing numsegs (lower resolution of dynamics). The extreme pitch moment you have is probably due to your input file configuration being strange (fixed rods).
Q2: Its hard to say what is the right configuration because I don’t know what you are trying o reproduce in MoorDyn. However, you have rods listed as fixed which means they will not move and are not attached to anything. They are “fixed” at their initial positions. I also noted that you had points 13-16 listed as fixed and you are not using them. These will also be “fixed” at their initial position. If you want rods attached to your bodies, then the attachment needs to be “body#”, and the positions need to be defined in the body’s reference frame (same for points attached to a body). If you want the rods free to move in all 6 degrees of freedom, they need to be called “free”. Your beam elements should be attached to the body via 0-length rods rather than points, so that the bending moments are accurately translated to the body.
Q3: Zero length rods have the same xyz positions for end A and end B, and they have numsegs = 0. For all intents and purposes they are a point that transfers bending moments through them. The one point to note is that for the 0-length rod to work when attaching two lines together is that one line needs to attach to end A and the other to end B.
1 Like
Hi @Ryan.Davies,
Thank you for your quick response,
i will check my inputs based on your clear information.
let me see.
Thanks, and Regards,
SomanSS
Hi @Ryan.Davies,
My intention to analysis Morsion element of floating cylinder (i.e body consider as a three cylinder) and connect the mooringlines.
Below I have created a cylinder like rod1 (Free) and in between rod I have connected using rod2 connection as a Coupled.
Also, it is possible to give massless beam and rod to connect. The objective is using Morison element forces on the moorling lines.
Following inputs are corrects:
----------------------- LINE TYPES --------------------------------------------------------------
TypeName Diam Mass/m EA BA/-zeta EI Cd Ca CdAx CaAx
(name) (m) (kg/m) (N) (N-s/-) (-) (-) (-) (-) (-)
chain 0.060 79.0 3.0744E5 -1.0 0 2.0 0.8 0.4 0.25
nylon 0.18 20.6 2.03E5 -0.8 0 1.0 0.0 1.6 0.05
---------------------- ROD TYPES ------------------------------------
TypeName Diam Mass/m Cd Ca CdEnd CaEnd
(name) (m) (kg/m) (-) (-) (-) (-)
rod1 3.0 7238.095 0.0 0.0 0.0 0.0
rod2 0.50 0.0 0.0 0.0 0.0 0.0
---------------------- RODS ----------------------------------------
ID RodType Attachment Xa Ya Za Xb Yb Zb NumSegs RodOutputs
(#) (name) (#/key) (m) (m) (m) (m) (m) (m) (-) (-)
1 rod1 Free 17.5 -10.10 -3.5 17.5 -10.10 0.0 3 pvUDtcsd
2 rod1 Free 0.0 20.20 -3.5 0.0 20.20 0.0 3 pvUDtcsd
3 rod1 Free -17.5 -10.10 -3.5 -17.5 -10.10 0.0 3 pvUDtcsd
4 rod2 Coupled 17.5 -10.10 0.0 0.0 20.20 0.0 3 pvUDtcsd
5 rod2 Coupled 0.0 20.20 0.0 -17.5 -10.10 0.0 3 pvUDtcsd
6 rod2 Coupled -17.5 -10.10 0.0 17.5 -10.10 0.0 3 pvUDtcsd
---------------------- POINTS ----------------------------------------------------------------------------------
ID Attachment X Y Z M Volume CdA Ca
(#) (word/ID) (m) (m) (m) (kg) (m^3) (m^2) (-)
1 Fixed 59.35 -82.59 -27.0 0 0 0 0
2 Fixed 0.0 102.33 -27.0 0 0 0 0
3 Fixed -59.35 -82.59 -27.0 0 0 0 0
4 Free -51.81 -69.83 -24.0 0 0 0 0
5 Free 0.0 86.71 -24.0 0 0 0 0
6 Free -51.81 -69.53 -24.0 0 0 0 0
7 Point 17.5 -10.10 -3.5 0 0 0 0
8 Point 0.0 20.20 -3.5 0 0 0 0
9 Point -17.5 -10.10 -3.5 0 0 0 0
Or I need to add bodies locate the position and connect the rod to the body.
If I have bodies, then I need to give the time history of body motion?
Thanks and regards,
SomanSS
@Somu.Sakthivel,
Before we go too far with this, do you have a reason specifically to be using MoorDyn to model the platform? This seems like it would be better suited to some of the other OpenFAST tools (SubDyn, ElastoDyn).
If you want to make the platform work in MoorDyn you need to make some adjustments. It appears that rod1 here are vertical rods, while rod2 are intended to connect those rods. This will not work, as rods cannot be connected to each other. Additionally, the attachment coupled means that rods motion is controlled by either OpenFAST or a user provided time series. The point attachment point is the same as free, so these mooring lines will just sink.
What you would need is a single body, with no mass, volume or drag coefficients, and then each of the rod1’s would be attached to that body, either free of coupled depending on if you want to provide motions or allow the motions to be calculated. With this, the three rods would move as a single rigid body. As the rod2’s have no hydro coefficients and no mass, you can get rid of them. Additionally, if you want the mooring lines attached to these rods, then their attachment should be rod ends and not points. Below is an example of what I mean, but you will need to adjust it for your case:
----------------------- LINE TYPES --------------------------------------------------------
TypeName Diam Mass/m EA BA/-zeta EI Cd Ca CdAx CaAx
(name) (m) (kg/m) (N) (N-s/-) (-) (-) (-) (-) (-)
chain 0.060 79.0 3.0744E5 -1.0 0 2.0 0.8 0.4 0.25
nylon 0.18 20.6 2.03E5 -0.8 0 1.0 0.0 1.6 0.05
---------------------- ROD TYPES ------------------------------------
TypeName Diam Mass/m Cd Ca CdEnd CaEnd
(name) (m) (kg/m) (-) (-) (-) (-)
rod1 3.0 7238.095 0.0 0.0 0.0 0.0
----------------------- BODIES ------------------------------------------------------
ID Attachment X0 Y0 Z0 r0 p0 y0 Mass CG* I* Volume CdA* Ca*
(#) (-) (m) (m) (m) (deg) (deg) (deg) (kg) (m) (kg-m^2) (m^3) (m^2) (-)
1 Free 0 0 0 0 0 0 0 0 0 0 0 0
---------------------- RODS ----------------------------------------
ID RodType Attachment Xa Ya Za Xb Yb Zb NumSegs RodOutputs
(#) (body1) (#/key) (m) (m) (m) (m) (m) (m) (-) (-)
1 rod1 body1 17.5 -10.10 -3.5 17.5 -10.10 0.0 3 pvUDtcsd
2 rod1 body1 0.0 20.20 -3.5 0.0 20.20 0.0 3 pvUDtcsd
3 rod1 body1 -17.5 -10.10 -3.5 -17.5 -10.10 0.0 3 pvUDtcsd
---------------------- POINTS -------------------------------------------------------------
ID Attachment X Y Z M Volume CdA Ca
(#) (word/ID) (m) (m) (m) (kg) (m^3) (m^2) (-)
1 Fixed 59.35 -82.59 -27.0 0 0 0 0
2 Fixed 0.0 102.33 -27.0 0 0 0 0
3 Fixed -59.35 -82.59 -27.0 0 0 0 0
4 Free -51.81 -69.83 -24.0 0 0 0 0
5 Free 0.0 86.71 -24.0 0 0 0 0
6 Free -51.81 -69.53 -24.0 0 0 0 0
---------------------- Lines ------------------------------------------------
ID LineType AttachA AttachB UnstrLen NumSegs LineOutputs
(#) (name) (#) (#) (m) (-) (-)
1 chain 1 4 ... ... ...
2 chain 2 5 ... ... ...
3 chain 3 6 ... ... ...
4 nylon 4 r1a ... ... ...
5 nylon 5 r1a ... ... ...
6 nylon 6 r1a ... ... ...
Hi @Ryan.Davies ,
Thank you for your quick responses.
Indeed, my intention is to model the platform not with external body forces like diffraction and radiation from external RAOs (InputFile: MDInputs file), but instead as three cylindrical elements connected with mooring lines. In this case, would the three cylinders behave as a single body in motion? Moreover, would it be able to calculate the body forces and mooring line tension?
Yes, I understand that I need to provide the body mass, Ixx, Iyy, and Izz coefficients. My question is: since we already specify the body mass in the file, how does it account for or calculate the rod mass? I have simply divided the mass per length. Alternatively, do I need to define the rod element as massless?
Thanks and regards,
SomanSS
@Somu.Sakthivel
I’m still not entirely sure if MoorDyn is the right tool for what you want to do with simulating platform motion. While it is capable of doing it, there are other parts of OpenFAST much better suited for it.
To answer your first question, yes the three cylinders will behave as a single body, and yes it will be able to calculate the mooring line tensions and the 6 DOF forces on the center of the body. If the three cylinders are the only structures in your body, then MoorDyn will automatically calculate the mass matrix of the body using the rod masses and you don’t need to provide body mass and inertia values. To find the rod mass, MoorDyn multiplies the mass per length by the defined rod length, 3.5 m in your case. The body mass and inertia inputs are for when you want to account for properties that are not already included in the mooring system design. If you make the rods massless then you need account for that mass in the body inputs.
Hi @Ryan.Davies,
Thank you for the valuable information.
I performed the analysis without considering the mass of the bodies (using only the rod mass) and then with the mass of the bodies included (no added rod mass). What differences should I expect in the analysis? Will the mooring tension remain the same, or will it be entirely different?
Our goal is to compute the mooring forces on multiple connected platforms, but we are neglecting the diffraction and radiation effects of the bodies (assuming the waves are not affected by the bodies).
I have snapshots of the input for both cases. Please let me know your suggestions.
Thanks and regards,
SomanSS
Hi @Somu.Sakthivel
In theory the two approaches should be equivalent in terms of rigid body kinematics, but when you don’t include the rods you won’t get any of the hydrodynamic effects that they provide. Additionally when you remove the rods, you need to compensate for that in the body by providing the correct volume, I cannot tell if you’ve done that. I would tend to think including the rods is a better approach, but it is dependent on the results you see and how you set it up.
If you are looking to simulate the effects on multiple platforms then I would suggest looking at other tools than just moordyn. In particular FAST.Farm has the ability to simulate the floating platform dynamics including shared moorings. For your case it should work well because FAST.Farm can be run without diffraction and radiation, and does not account for the platform wake impacts on separate floating platforms. Here is an example: Coupled loads analysis of a novel shared-mooring floating wind farm - ScienceDirect
Hi @Ryan.Davies ,
Thank you for your suggestion.
I have completed two cases: one with the body and dummy body using rod elements. Yes, the displacement of the submerged body volume is provided.
In the following plot: (i) Using the rod = no body values are given, and with the rod element,
(ii) No rod = body values are given.
a) first plot is tesnsion line1 (at bottom connection chain)
b) same mooring line1 ( at fairlead position nylon)
corresponding body forces in Y direction
body forces with Rod
Moreover, you mentioned that FAST.Farm has the ability to simulate floating platforms, including shared moorings. That sounds very promising, and I have reviewed the recent publication. Although our platform is not a wind turbine but rather a floating structure, should I disregard the wind turbine aspect?
Thanks and regards,
SomanS.S
Hi @Somu.Sakthivel,
The higher forces and tensions seen when adding rods makes sense as they add more realistic hydrodynamics that the body object doesn’t have. As for FAST.Farm with no wind, see this forum post: FAST.Farm without wind modelling setup
Hi @Ryan.Davies ,
Thank you for your continuous support.
I will review the models and get back to you if I have any questions.
Thanks and regards,
SomanSS
