How does Step Calculation works

[miriasmata]

Hi,

I am new using MoorDyn and I was following the example with the 3 mooring lines attached to a floating body.

My main question is how the Step calculation works:

In that example, we estimate the system for 5 time steps by moving the body at 0.5 m/s in the positive x direction.

What is not clear to me is that for each time step, the values of x (position) and xd (velocity) are not updated. From my understanding, it always receives as input the vectors from the initial position.

I would like to understand this function as I would like to use MoorDyn in Python as a standalone version, and force the system by including waves and currents.

Thank you for the help

[sanguinariojoe]

Hey Marisol! Welcome! x and xd are inputs, not outputs. They will not be updated by MoorDyn. In your simple example you must do something like x += xd * dt (If they are not numpy arrays you must use an inline iterator or something like) The underlying idea is that your code is responsible to feed MoorDyn with the motion pf the fairleads, while MoorDyn is responsible to report back the forces. You have some examples of staggered couplings on the closed issues, even with ac Runge-Kutta 4 scheme: #315 Let us know, please

…

On Fri, 19 Sept 2025, 05:22 Marisol Irías Mata, ***@***.***> wrote: *miriasmata* created an issue (FloatingArrayDesign/MoorDyn#328) <#328> Hi, I am new using MoorDyn and I was following the example with the 3 mooring lines attached to a floating body. My main question is how the Step calculation works: image.png (view on web) <https://github.com/user-attachments/assets/4faea0a6-f20c-4491-921a-e612054a7b19> In that example, we estimate the system for 5 time steps by moving the body at 0.5 m/s in the positive x direction. What is not clear to me is that for each time step, the values of x (position) and xd (velocity) are not updated. From my understanding, it always receives as input the vectors from the initial position. I would like to understand this function as I would like to use MoorDyn in Python as a standalone version, and force the system by including waves and currents. Thank you for the help — Reply to this email directly, view it on GitHub <#328>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMXKKBJ3R5DVEC434Y2UFL3TNZHXAVCNFSM6AAAAACG5T7KOWVHI2DSMVQWIX3LMV43ASLTON2WKOZTGQZTENJYGI3DINY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[miriasmata]

Thank you for the fast response.

I understand now how the Step function works.

Then my next question would be if MoorDyn also computes this resulting position. From the input file I believe I can get as output the position of each node of the line by adding a p:

Could I used this value as initial condition for the following time step?

My end goal would be to use the Water Kinematic functions available in MoorDyn-C. Can I force the model with WaveKin=3 or with Currents=2 and use the Step function with the position that MoorDyn computes?

I know MoorDyn can be coupled to other models, but I wanted to add the hydrodynamics using these available functions.

Thanks,
Marisol

[sanguinariojoe]

Short answer: No, MoorDyn is meant to be coupled with other solvers. Long answer: Well, you can, but the model is definitely too simple for any real practical application. Just replace the "coupled" points by "free" ones (set them a mass, a volume, and drag and added mass coeffs). Then you can call step() passing empty arrays to x and xd.

…

On Fri, 19 Sept 2025, 11:00 Marisol Irías Mata, ***@***.***> wrote: *miriasmata* left a comment (FloatingArrayDesign/MoorDyn#328) <#328 (comment)> Thank you for the fast response. I understand now how the Step function works. Then my next question would be if MoorDyn also computes this resulting position. From the input file I believe I can get as output the position of each node of the line by adding a p: image.png (view on web) <https://github.com/user-attachments/assets/e701b30b-7a8b-4d76-9096-eab85a18074a> Could I used this value as initial condition for the following time step? My end goal would be to use the Water Kinematic functions available in MoorDyn-C. Can I force the model with WaveKin=3 or with Currents=2 and use the Step function with the position that MoorDyn computes? I know MoorDyn can be coupled to other models, but I wanted to add the hydrodynamics using these available functions. Thanks, Marisol — Reply to this email directly, view it on GitHub <#328 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMXKKDM5EMCLDSCXKJBQ3L3TPAZRAVCNFSM6AAAAACG5T7KOWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGMJRGMYTCMBWGE> . You are receiving this because you commented.Message ID: ***@***.***>

[ZahidUllah33400]

Hello,
I also need some clarification regarding the input of the moordyn_step and will appreciate it. I have compiled the MoorDyn v2 using visual studio for its use with floating platform in Matlab. Now the the moordyn_step requires the position x, and xd (velocity) as input and returns the fairlead forces. As far as i understand the x is the initial fairlead position obtained from the Lines input file and xd is the velocity at at any timestep. The moordyn_step uses the velocity to update the the fairlead position internally and then compute the fairlead forces. I need clarification on the following points:

1. Does input position vector x remains the same (which is obtained from input Line) through out the simulation time and only velocity at each time step is provided in input which is then used to internally the position vector?
2. If i use my platform motion and updated the fairlead positions i.e., x vector, and input the updated x vector at each time step and input the velocity vector xd as zeros. Will the moordyn_step will give me accurate fairlead forces or it is not designed like that?
3. if the point 2 above is not possible. Does this mean that i have to compute the translational velocity of each fairlead position using the platform degrees of freedom (surge, sway, heave, roll, pitch, yaw) and provide that translational velocity vector as xd while keeping the x same as initial fairlead position (from input Lines.txt). Is this the right approach for coupling MoorDyn (matlab)?
   It is my first time to use MoorDyn, therefore, i will highly appreciate any helpful comments.
   Best,
   Zahid

[RyanDavies19]

Hi @ZahidUllah33400,

To answer your questions:

1. No, the input needs to be updated at every timestep. MoorDyn uses the position and velocity provided by an external program to calculate the force in response to those kinematics. I would suggest this video: https://www.youtube.com/watch?v=FqW7Xpl_VNk and this Jupyter notebook as a way to get a better sense of how this coupling works: https://github.com/FloatingArrayDesign/MoorDyn/blob/dev/example/MoorDyn_standalone_demo.ipynb

2. See above. You need to provide the position and velocity to the step function and it will return the response force to that motion (while also calculating the line dynamics internally)

3. See the Jupyter notebook linked above. It has two equivalent examples of using 3 coupled points (i.e. three fairlead positions) and using 1 coupled body (i.e. the platform degrees of freedom)

[ZahidUllah33400]

Thank you @RyanDavies19 for the instant response. I watched the demo and went through the two examples, which explained most of the question i had in mind. Since i am coupling the MoorDyn with a floating body where the equation of motion will be in solved in MATLAB and will provide the kinematics, So, i will be going with input similar to "3-lines with Body Simple Example". In the input file i will be defining the fairlead poisiton with respect body reference frame. The point i want to clarify is that during the simulation, the x and xd are 6 dof motion and velocties vectors of the body about the origin of body reference frame (which in my case will be provided by the floating platform solver). The moordyn will compute the fairlead position from these platform x and xd vectors internally. Am i getting it correctly?
Thanks once again.

Best
Zahid

[sanguinariojoe]

Yes Zahid, it will do that. Just a hint: Since you are carrying out a staggered coupling, you are responsible of granting that both models, MATLAB and MoorDyn, do not drift away. Along this line, I strongly recommend to do not use directly the velocity from your MATLAB model, but recompute that as matlab_x_end = matlab_x + dt * matlab_xd; moordyn_xd = (matlab_x_end - moordyn_x)/dt; You can always get "moordyn_x" calling to MoorDyn_GetBodyPos()

…

On Tue, 25 Nov 2025, 10:38 Zahid, ***@***.***> wrote: *ZahidUllah33400* left a comment (FloatingArrayDesign/MoorDyn#328) <#328 (comment)> Thank you @RyanDavies19 <https://github.com/RyanDavies19> for the instant response. I watched the demo and went through the two examples, which explained most of the question i had in mind. Since i am coupling the MoorDyn with a floating body where the equation of motion will be in solved in MATLAB and will provide the kinematics, So, i will be going with input similar to "3-lines with Body Simple Example". In the input file i will be defining the fairlead poisiton with respect body reference frame. The point i want to clarify is that during the simulation, the x and xd are 6 dof motion and velocties vectors of the body about the origin of body reference frame (which in my case will be provided by the floating platform solver). The moordyn will compute the fairlead position from these platform x and xd vectors internally. Am i getting it correctly? Thanks once again. Best Zahid — Reply to this email directly, view it on GitHub <#328 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMXKKCYWOBMXC6K6IWJQPD36QPRLAVCNFSM6AAAAACG5T7KOWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTKNZUGY4DONRRGA> . You are receiving this because you commented.Message ID: ***@***.***>

[ZahidUllah33400]

@sanguinariojoe Thank you for kind suggestion. i did not fully understand the final remarks. I have attached a helping figure and with reference to that figure i have explained step-wise procedure that i will follow.

1. Define the fairlead position with respect to body reference frame in the input file.
2. At ti, the matlab will compute 6 dof motions (x_matlab) and 6 dof velocities (xd_matlab) of the point o (which is origin of body reference frame) shown in Figure.
3. The x_matlab and xd_matlab will be provided as input to MoorDyn which will return the fairlead forces.
4. These forces will be transferred to matlab solver for computation of x_matlab and xd_matlab for next time step.

Looking forward to your suggestions regarding that procedure. Thanks

[sanguinariojoe]

Yep, but which time integrator do you use on your matlab model. To avoid getting swamped with those details, just simply compute matlab_x_end = matlab_x + dt * matlab_xd Then you can ask MoorDyn the state of the body: body = MoorDynM_GetBody(system, 1) moordyn_x, useless_var = MoorDynM_GetBodyState(body) And now you can compute the correct speed and apply it: moordyn_xd = (matlab_x_end - moordyn_x) / dt Forces = MoorDynM_Step(system, x, xd, t, dt)

…

On Tue, 25 Nov 2025, 15:09 Zahid, ***@***.***> wrote: *ZahidUllah33400* left a comment (FloatingArrayDesign/MoorDyn#328) <#328 (comment)> Body_mooring.jpg (view on web) <https://github.com/user-attachments/assets/deaae49a-9b6b-4212-bad3-0369d8659d4e> @sanguinariojoe <https://github.com/sanguinariojoe> Thank you for kind suggestion. i did not fully understand the final remarks. I have attached a helping figure and with reference to that figure i have explained step-wise procedure that i will follow. 1. Define the fairlead position with respect to body reference frame in the input file. 2. At ti, the matlab will compute 6 dof motions (x_matlab) and 6 dof velocities (xd_matlab) of the point o (which is origin of body reference frame) shown in Figure. 3. The x_matlab and xd_matlab will be provided as input to MoorDyn which will return the fairlead forces. 4. These forces will be transferred to matlab solver for computation of x_matlab and xd_matlab for next time step. Looking forward to your suggestions regarding that procedure. Thanks — Reply to this email directly, view it on GitHub <#328 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMXKKGI7GV5RBSTC4B7B5L36RPJPAVCNFSM6AAAAACG5T7KOWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTKNZVHA2DCMBZHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[ZahidUllah33400]

@sanguinariojoe
I am using RK4 integrator. I highly appreciate your guidance. i ran a tried test example on MoorDyn with a floating body. The returning force vector is 6 by 1, which i assume are the net restoring force and restoring moment of all the mooring lines about the body reference frame.

Best,
Zahid