Authors' Response Submitted

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

Dear Editor,

Thank you for considering our work.

Comments from the reviewer have been addressed, see table below.
In the first column, the reviewer’s comment is reproduced, then
our response is provided, and in the last column, the actions taken
in the revised manuscript are summarized.

We are looking forward to your feedback on our revision.

With our best regards,

David Stamenov
Giuseppe Abbiati
Thomas Sauder

1
Response to reviewer 1

Reviewer’s comment Authors’ response Action


- The authors would like to thank the reviewer -
for a thorough review of our work.
[1] Common practice presently is to use poten- We should indeed have made it clearer that Text Modified
tial flow theory in combination with viscous potential theory alone fails at predicting some
drag - for large body systems. This sentence important response parameters. Empirical
seems contrary to common practice. correction methods that complement poten-
tial theory are the very subject of this paper.
[2] Perhaps qualify this as catenary mooring By using ”moored” here, the authors wanted -
systems since these attributes are not true for to emphasize that the study was targeting sta-
taut or semi-taut mooring approaches. tionary floating structures, as opposed to ships
or to fixed offshore structures. The class of
mooring system involved (catenary, semi-taut,
taut, ...) is not crucial to the problem ad-
dressed here.
[3] Suggest describing what the QTF is physi- This is done at the end of the sentence: -
cally representing - transfer function between ”which models the relationship between a bi-
what? chromatic wave with frequencies f1 and f2 and
the low frequency wave loading at |f1 − f2 |”
[4] The common practice when using poten- It is indeed common practice to rely on addi- -
tial flow is to add a viscous drag term - typ- tional terms that can be tuned to match model
ically from Morison’s equation, or to add a test results or CFD calculations. This is indi-
linear/quadratic damping matrix. This is the cated in the next sentence: ”data-driven ap-
most common approach by industry, so needs proaches are not new and have been used for
to be included in the approaches described. the purpose of transfer function estimation be-
You need to show why your approach is poten- fore”.
tially superior to this industry accepted prac- It is however also known that current prac-
tice. tice (potential flow + tuning of Morison coef-
ficients and/or additional damping matrices)
has its limitations. Examples are a strong sea-
state dependency of the coefficients, difficulty
to properly account wave/current interaction,
etc. The present research aims at providing
more general load models, which can alleviate
this issue.

2
Reviewer’s comment Authors’ response Action
[5] Do you mean continuous-time? - Corrected.
[6] Strictly speaking, this paper only consid- Thanks to the reviewer for pointing this out. Text Modified, accounting also for comment
ered the wave excitation on a fixed structure The current paper only considers the diffrac- [7].
if I’m not mistaken, not an actual moored tion problem, and not the radiation problem,
floating structure with motion and wave ra- which will be addressed at a later statge.
diation/hydrostatic loads.
[7] I think you need more description on how This comment is related to comment [6]. The Text Modified, accounting also for comment
this approach might be used with an actual modelling strategy is indeed not made clear [6].
moored structure that moves in waves. If I’m enough in the paper. The present work only
understanding correctly, you will have the hy- addresses the diffraction problem, and aims
drodynamic model be completely represented at establishing better wave load models. The
by transfer functions? How do you account influence of the floater response on the load
for the hydrostatics (separately), and the ra- (mixed radiation/diffraction problem) will be
diation components? included in the NARX model at a later stage,
Is the approach to develop a model between by augmenting it with the floater response (re-
the waves and associated response? If so, this cent history of the position and velocity)
requires that relationship to already have been
measured or modeled. Is the idea to take se-
lect load cases from either experiment or CFD
and derive a model that can be used for other
wave conditions?
[8] Wave excitation has an acausal dependence Good point from the reviewer. The fact that Clarification included.
on wave elevation. Eq. (1) only considers the NARX model works probably means that
the wave elevation in the past in the exoge- in the present case, the wave loading process is
nous part, yet the model still works. Does nearly causal? An option to ”tend to causal-
this mean the missing information normally ity” is to use the wave elevation at an up-
conveyed in future wave elevation is actually stream distance of the model as an input.
embedded in the auto-regressive part?
[9] Looks like the relationship you’re build- We are indeed building a relationship between More info provided in the text.
ing is between load and waves. How will you load and waves. The objective is to infer the
obtain this load information for training this loading from the response, as done in Sauder
model? 2021, Abbiati and Sauder 2021, etc..
[10] [11] transform [...] Ω2 Thanks for pointing out these typos Typos corrected.
[12] Would the goal of this approach be to do For the time being, it has been restricted to -
this same process for all 6 DOFs? one d.o.f but the method is generic and can be
applied to all 6 dofs indeed.

3
Reviewer’s comment Authors’ response Action
[13] What transfer function? LTF for surge Clarified at an earlier stage in the paper.
force?
[14] More description would be useful here. - -
What are the components of your model you
are using to produce the time series?
[15] Please specify either here or in the text - Clarification provided.
what the y-axis is.
[16] So, again is the process to build a scale As discussed in our response to comment [4], Clarified in the text, see also comment [4].
model test or run CFD for some select load the objective of our work is indeed to establish
cases, and use these to build a model that a data-driven load model (trained from set of
can be used to predict the response for other well-chosen experiments) that is more general
wave conditions? How applicable is the model than e.g Morison equation. By more general,
trained from one wave condition at predicting we mean, in particular, that it should perform
the response for a new wave climate? well under a wide range of sea-states.
[17] In its current form, the method establishes We agree with the reviewer that QTFs are Additional details provided about the context
the relation between wave elevation and wave depending on first-order response, and hence of the study.
excitation on the structure. cannot be estimated experimentally by fixing
With a moored structure that moves under the structure. The idea is to infer the load
wave excitation, the hydrodynamic load on from the response, and that the experimental
the structure is generally not available from setup can be tuned as done in Sauder (2021).
experiments. In this case, it’s not clear how
experimental data can be used to tune the
model.
It is possible to experimentally measure the
wave excitation on fixed structures, but the
QTFs derived this way in general cannot be
applied to a moving structure.
In this sense, the present method might actu-
ally work better with CFD simulations, where
the hydrodynamic loads on a moving structure
is readily available.

4
Reviewer’s comment Authors’ response Action
[18]How do you know then that this method For the time being, the main focus was on es- -
would work when those viscous loads are in- tablishing the NARX model and apply har-
cluded? monic probing to ensure that we could re-
trieve the original LTF (sanity check). Go-
ing to higher-order, adding viscous loads, and
see how this affects the linear, quadratic, and
higher-order transfer functions is the object of
further work.

You might also like