Incremental Nonlinear Control for Aeroelastic Wing Load Alleviation and Flutter Suppression
Round 1
Reviewer 1 Report
The paper is very well written and the work done is indeed interesting. I have just a comment on reference [22] - Geertsen,J. "Development of a Gust Generator for a Low Speed Wind Tunnel": I wasn't able to find a printed version, even on ETH theses portal. Even though it's not the subject of the article, it's strongly related to the experiment setup, so I think interested readers should have at least the possibility to read about it in detail.
Author Response
Thank you very much for your appreciation!
Based on your comments, we have uploaded "Development of a Gust Generator for a Low Speed Wind Tunnel" to TU Repository.
It is now publicly available via https://repository.tudelft.nl/islandora/object/uuid%3A902ddcb9-3884-46eb-ae1b-b2df506b6e07?collection=education
and
http://resolver.tudelft.nl/uuid:902ddcb9-3884-46eb-ae1b-b2df506b6e07
We have put the link in the paper reference list.
Thanks for your consideration!
Reviewer 2 Report
Outstanding paper combining rigorous mathematical background with wind tunnel test results. Application is done for response-type problems (1-cosine gust) and stability-type problems (flutter). Very well presented and easy reading. Conclusions fully supported by the evidences shown in the paper. The list of references is comprehensive and very relevant.
Author Response
Thank you very much for your kind appreciation!
Reviewer 3 Report
-written in a fair to good English, well organized and easy to read
-the topic is interesting and the wind tunnel experiment is a clear contribute of the paper
-some aspects of the approach should be more clearly presented: namely, some figure presenting the variables of the model and a block diagram of the control loop would be most welcome
-the steady state control is not fully satisfactory and should be further commented
Further details:
-sec.2.1 a REF and a figure for the dynamic model (1) would be welcome
> despite the ref to [16] the presentation is confuse
> the description of parameters of (1) should be consistent: use ; separated sentences for all of them
> no viscous damping term appears in (1): is there any approximation here?
-although I am not familiar with Wagner/Kussner functions, it is strange that Phi(t) has 2 psi paraneters and Psi(t) has 2 phi parameters!
-what is b in Phi(t) ?
-in (2): the description of variables is incomplete, A,B,…W,W1,W2 are missing, whereas the lift and moment inputs were presented for (1)
-where is Phi used in (2)?
-in (10): what is ug ?
> is there a dimension error ?
-at the end of sec.2 or at the beginning of sec.3, a block diagram would help to clarify the control law implementation
-along with fig.1/2, a schematic drawing would help be welcome
> figures are not very explicit
-fig.4-6 should have more explicit titles
> is the non zero steady state satisfactory
> in fig.6 at the end there is a residual oscillation which is not expected
NA
Author Response
First of all, we would like to express our sincerest appreciation for your suggestions! The comments are addressed as follows:
Q: some aspects of the approach should be more clearly presented: namely, some figure presenting the variables of the model and a block diagram of the control loop would be most welcome
A: Thanks for your suggestion! Two figures have been added.
Q: the steady state control is not fully satisfactory and should be further commented
A: That is indeed the case. The steady-state oscillations are not very significant in 3, 3,5, 4, and 4,5Hz cases, but is present in the 5Hz case. This is because the excitation frequency 5 Hz is close to the pitch mode natural frequency 6.39Hz. Since this study focuses on the heave motion, then the uncontrolled pitch motion inertially couples with the control surface movement, leading to the high frequency and low magnitude oscillations in control deflections in figure 6 although the steady-state oscillations in heave have been damped out.
These discussions have been added to the revised manuscript.
Q: sec.2.1 a REF and a figure for the dynamic model (1) would be welcome
A: a reference and a figure have been added.
Q: despite the ref to [16] the presentation is confuse
A: The sentence has been revised to "Theodorsen’s unsteady aerodynamic theory is adopted [16]. "
Q: the description of parameters of (1) should be consistent: use ; separated sentences for all of them
A: This has been revised accordingly
Q: no viscous damping term appears in (1): is there any approximation here?
A: Yes, indeed, it is a common assumption in aeroelasticity. We have added the explanation.
Q:although I am not familiar with Wagner/Kussner functions, it is strange that Phi(t) has 2 psi paraneters and Psi(t) has 2 phi parameters!
A: Indeed, you are right. We have changed the notations in the two equations.
Q: what is b in Phi(t) ?
A: half-chord length. We have added an explanation and indicated this parameter in the system drawing.
Q: in (2): the description of variables is incomplete, A,B,…W,W1,W2 are missing, whereas the lift and moment inputs were presented for (1)
A: explanations on the matrices are added. The lift and moment comparisons in (1) and (2) have been added.
Q: where is Phi used in (2)?
A: Phi is used in the aerodynamic matrices of (2) using Wagner's function. We have added a reference.
Q: in (10): what is ug ?
A: that is the gust input velocity. We have added the notation.
Q: is there a dimension error ?
A: No, there is no model reduction process and dimension error.
Q: at the end of sec.2 or at the beginning of sec.3, a block diagram would help to clarify the control law implementation
A: Thanks for your suggestion! We have added a block diagram.
Q: along with fig.1/2, a schematic drawing would help be welcome
A: Thanks! A schematic drawing has been added.
Q: fig.4-6 should have more explicit titles
A: Indeed, more explicit titles have been added.
Q: is the non zero steady state satisfactory. in fig.6 at the end there is a residual oscillation which is not expected
A: That is indeed the case. The steady-state oscillations are not very significant in 3, 3,5, 4, and 4,5Hz cases, but is present in the 5Hz case. This is because the excitation frequency 5 Hz is close to the pitch mode natural frequency 6.39Hz. Since this study focuses on the heave motion, then the uncontrolled pitch motion inertially couples with the control surface movement, leading to the high frequency and low magnitude oscillations in control deflections in figure 6 although the steady-state oscillations in heave have been damped out.
We have added the discussions in the revised manuscript.
