[Wave_star]

大家来讨论一下双浮体的时域分析
我用AQWA-LINE做双浮体的RAO分析，双浮体之间没有连接，然后希望通过NAUT得到双浮体在频域下的响应，可是每次都会出现这样的warning和error：
**** WARNING **** HYDRODYNAMICALLY INTERACTING STRUCTURES HAVE MOVED FROM THE ANALYSIS
POSITION
MOVEMENT EXCEEDS 30%
----------------------------------------------------------------------------------------

**** WARNING **** HYDRODYNAMICALLY INTERACTING STRUCTURES HAVE MOVED SIGNIFICANTLY FROM
THE ANALYSIS POSITION
RESULTS ARE INACCURATE - MOVEMENT EXCEEDS 50%
-----------------------------------------------------------------------------------------

**** ERROR **** HYDRODYNAMICALLY INTERACTING STRUCTURES HAVE MOVED TOO FAR FROM ANALYSIS
POSITION
MOVEMENT EXCEEDS 75%
STOPPED AT TIME = 16.10

我现在能想到的办法就是用Constraints - Articulations Between Structures，限制两者之间的位移，可是这样又与原始结构不一致了，大家有什么好的建议没，希望大家能多多讨论

[candit]

把surge 和 sway 两个运动modes固定。不要加Articulations。

[Wave_star]

引用框（candit ＠ 01-04-2011 - 18:44）

把surge 和 sway 两个运动modes固定。不要加Articulations。

谢谢您的解答，请问有没有类似的文献可以参照，我看好多文献都是拿计算结果的RAO与试验比较，但都是比较surge motion和surge force，如果要考虑到两个浮体前后或是左右碰撞问题的话应该怎么考虑。
另外，我想计算两浮体在规则波入射情况下的瞬态效应，因为两者间距比较近防止它们碰撞，请问，AQWA—NAUT计算瞬态效应是准确的吗?

[shaopizi]

How does AQWA generate the incident wave in its time-domain simulations?
If is prescribed by for instance Stokes waves combined with a ramp funtion to allow for a gentle start, tune the period over which you applied the ramp funtion, you may get a solution with minmum drifting.

Drifting is common issue in time-domain solutions. There are several ways of minimize its effect. However, i am not sure to what extend the commercial softwares e.g. AQWA, allow users to control their simulations...

[candit]

引用框（shaopizi ＠ 01-04-2011 - 18:10）

How does AQWA generate the incident wave in its time-domain simulations?
If is prescribed by for instance Stokes waves combined with a ramp funtion to allow for a gentle start, tune the period over which you applied the ramp funtion, you may get a solution with minmum drifting.

Drifting is common issue in time-domain solutions. There are several ways of minimize its effect. However, i am not sure to what extend the commercial softwares e.g. AQWA, allow users to control their simulations...

However, my experience shows that the ramp function is not always effective for controlling the drift, of course the initial transient can be minimized with an approporite ramp function. In practice they use spring system or autopilot in the motion solver. I think one can find similar applications in seakeeping model tests carried out in ocean basins.

[Wave_star]

引用框（shaopizi ＠ 01-05-2011 - 01:10）

How does AQWA generate the incident wave in its time-domain simulations?
If is prescribed by for instance Stokes waves combined with a ramp funtion to allow for a gentle start, tune the period over which you applied the ramp funtion, you may get a solution with minmum drifting.

Drifting is common issue in time-domain solutions. There are several ways of minimize its effect. However, i am not sure to what extend the commercial softwares e.g. AQWA, allow users to control their simulations...

Thanks for you reply. But there are always such warning and errors if no constrains attached to the two bodies, even if I shut down the drift calculation in AQWA-LINE. And the transient effect is one of my research aspects, so could you please give me some advice about that.

[shaopizi]

The 'drifting' i refered is different from the 'drift' you have understood. 'Drifting' can happen in even LINEAR time-domain analysis. It is a consequence of the homogeneous solution of the body motion equations. One simple example is Newton's law for a single degree of freedom, i.e. m*a=F, the homogeneous of which is x=A+B*t. A and B are constant depending on the initial conditions. If B is nonzero, the body experience 'drifting' from its mean position and never come back... While 'drift' in offshore hydrodynamic is related to mean or low-frequency wave loads. That is my reason to you why you still have 'Drifting' even though you have swifted off the function of 'Drift' in AQWA.

In order to minimize the homogeneous solution of the body motion equations, we have typically 4 methods (maybe more) to handle this problem,
(1) You design a special initial condition (phase relative to incident wave) so that A and B identically become zero (or nearly zero). So called target-shooting technique in other fields. But in practice, it is not easy in general to define such an initial condition...
(2) You let the body go, and follow the body. This require a more advanced hydrodynamic model. One can either use time-domain green function or Rankine pane method with accelerated coordinate system fixed to the body.
(3) Add soft spring as we have done in time-domain seakeeping analysis for sway/yaw motions;
(4) Apply a very long period of ramp function to start the simulation very gentlely and smoothly. I have been using this technique for time-domain simulations of single bodies. And it has been working well...

[candit]

引用框（shaopizi ＠ 01-05-2011 - 09:49）

The 'drifting' i refered is different from the 'drift' you have understood. 'Drifting' can happen in even LINEAR time-domain analysis. It is a consequence of the homogeneous solution of the body motion equations. One simple example is Newton's law for a single degree of freedom, i.e. m*a=F, the homogeneous of which is x=A+B*t. A and B are constant depending on the initial conditions. If B is nonzero, the body experience 'drifting' from its mean position and never come back... While 'drift' in offshore hydrodynamic is related to mean or low-frequency wave loads. That is my reason to you why you still have 'Drifting' even though you have swifted off the function of 'Drift' in AQWA.

In order to minimize the homogeneous solution of the body motion equations, we have typically 4 methods (maybe more) to handle this problem,
(1) You design a special initial condition (phase relative to incident wave) so that A and B identically become zero (or nearly zero). So called target-shooting technique in other fields. But in practice, it is not easy in general to define such an initial condition...
(2) You let the body go, and follow the body. This require a more advanced hydrodynamic model. One can either use time-domain green function or Rankine pane method with accelerated coordinate system fixed to the body.
(3) Add soft spring as we have done in time-domain seakeeping analysis for sway/yaw motions;
(4) Apply a very long period of ramp function to start the simulation very gentlely and smoothly. I have been using this technique for time-domain simulations of single bodies. And it has been working well...

Nice comment!
Concerning the long-period ramp function method:
It works well with the oblique following seas as well?

[shaopizi]

I did not try it on oblique following sea of ships. You can test it in you program. :)
What I did was for a vertical cylinder without restoring mechanism in the horizontal plane. You can find it in Section 8.3 of my thesis.

[candit]

引用框（shaopizi ＠ 01-05-2011 - 11:26）

I did not try it on oblique following sea of ships. You can test it in you program. :)
What I did was for a vertical cylinder without restoring mechanism in the horizontal plane. You can find it in Section 8.3 of my thesis.

OK, thank you ;)

[Wave_star]

引用框（shaopizi ＠ 01-05-2011 - 16:49）

The 'drifting' i refered is different from the 'drift' you have understood. 'Drifting' can happen in even LINEAR time-domain analysis. It is a consequence of the homogeneous solution of the body motion equations. One simple example is Newton's law for a single degree of freedom, i.e. m*a=F, the homogeneous of which is x=A+B*t. A and B are constant depending on the initial conditions. If B is nonzero, the body experience 'drifting' from its mean position and never come back... While 'drift' in offshore hydrodynamic is related to mean or low-frequency wave loads. That is my reason to you why you still have 'Drifting' even though you have swifted off the function of 'Drift' in AQWA.

In order to minimize the homogeneous solution of the body motion equations, we have typically 4 methods (maybe more) to handle this problem,
(1) You design a special initial condition (phase relative to incident wave) so that A and B identically become zero (or nearly zero). So called target-shooting technique in other fields. But in practice, it is not easy in general to define such an initial condition...
(2) You let the body go, and follow the body. This require a more advanced hydrodynamic model. One can either use time-domain green function or Rankine pane method with accelerated coordinate system fixed to the body.
(3) Add soft spring as we have done in time-domain seakeeping analysis for sway/yaw motions;
(4) Apply a very long period of ramp function to start the simulation very gentlely and smoothly. I have been using this technique for time-domain simulations of single bodies. And it has been working well...

Thanks for your advice,it worked.

[ansys_zhf]

第一点：aqwa求解两浮体相互作用，请使用hydi卡片
第二点：aqwa模拟碰撞请用fender单元
第三点：自己去琢磨。

[ansys_zhf]

还有第四点中国人交流不要用洋文。

[ytlee]

5. ansys_zhf should make no demand.

[cmm]

引用框（Wave_star ＠ 01-04-2011 - 18:04）

大家来讨论一下双浮体的时域分析
我用AQWA-LINE做双浮体的RAO分析，双浮体之间没有连接，然后希望通过NAUT得到双浮体在频域下的响应，可是每次都会出现这样的warning和error：
**** WARNING **** HYDRODYNAMICALLY INTERACTING STRUCTURES HAVE MOVED FROM THE ANALYSIS
POSITION
MOVEMENT EXCEEDS 30%
----------------------------------------------------------------------------------------

**** WARNING **** HYDRODYNAMICALLY INTERACTING STRUCTURES HAVE MOVED SIGNIFICANTLY FROM
THE ANALYSIS POSITION
RESULTS ARE INACCURATE - MOVEMENT EXCEEDS 50%
-----------------------------------------------------------------------------------------

**** ERROR **** HYDRODYNAMICALLY INTERACTING STRUCTURES HAVE MOVED TOO FAR FROM ANALYSIS
POSITION
MOVEMENT EXCEEDS 75%
STOPPED AT TIME = 16.10

我现在能想到的办法就是用Constraints - Articulations Between Structures，限制两者之间的位移，可是这样又与原始结构不一致了，大家有什么好的建议没，希望大家能多多讨论

你好 我跟你的情况一模一样，我在NAUT里算两条船的时候，给一条船一个初始速度，结果这条船就沉了，不知道什么原因，晕啊