1 Introduction

With the rapid worldwide development of the offshore oil industry,depending only on the floating crane to install large offshore platforms has become increasingly impractical.As such,developing ways to install these platforms safely and efficiently has become a focus of the offshore industry.Platform installation methods generally include the traditional lift installation and the float-over method.In offshore installation work,the traditional lift installation(Du 2007)cannot meet deck installation requirements and the associated cost becomes unreasonable.Float-over installation is a new method of topside/deck installations that has the concept of transporting the topside/deck using two groups of floaters connected at its opposite sides(Geba et al.,2017).The float-over method has become increasingly popular for the installation of large decks onto offshore platforms due to its great capacity and cost effectiveness(Sun et al.2012).Globally,the float-over method is widely used in offshore installation projects and is especially suitable for jacket platforms.Chen et al.(2017)describe an efficient methodology for analyzing the dynamics of float-over deck installation,particularly arising from the wave induced impacts between multiple structures.

Float-over installation represents a new method(Seij and Groot 2007)for integrated topside installations.In this installation,the topside is constructed as one large integrated module and is completely assembled on shore.Then,it is transported to the installation site by barge and is finally transferred from the transportation barge onto the supporting structure via barge ballasting.The key technologies are researched for float-over installation of large topside for a gas field in South China Sea,Such as the modification of transportation barge,deep water mooring system and mating between barge and topside(Fan et al.2013).The float-over system is a complex multi-body coupled system,in which the barge and topside interact through leg mating units(LMUs)and deck support units(DSUs),while being subjected to loads from the jacketand seabed by mooring lines,mating lines,and fenders.In addition,the mean draft and ballast of the barge and the dynamic characteristics of the floatover system continually change during the installation process,which further complicates the problem.

因此，咨询单位有必要积极研究BIM的有关技术，提高应用水平。各个专业需着手研究适合于本专业的、基于三维模型的技术；另一方面，还需考虑不同专业之间数据交换的可行性。可在Bentley、Dassault、Inventor、Autodesk Civil 3D 等平台上开展二次开发。

With the gradual recognition of the superiority of the floatover installation method compared with traditional lifting methods,researchers around the world are devoting increasing attention to the development and application of float-over installation technologies.Xia et al.(2005)develop a barge and offshore-platform-model test method to investigate the floatover installation of small platforms.Tahar et al.(2006)describe comparisons between numerical predictions and model test datas of the compliant tower in the float-over installation operation.Hamilton et al.(2008)study the importance of the dynamic characteristics of mooring systems and their environment and the distance between the barge and the jacket.A time-domain 3 Degrees of Freedom model by(Hu et al.2017)is developed to investigate nonlinear dynamics and impact loads during float-over installations,which generally involve multi-body interactions between wave-induced vessel motions and nonlinear constraint components.Researchers in China have also undertaken thorough exploration and research of the float-over method.Key technology of DP has been used by(Wang et al.2017)in the float-over and improved the efficiency and safety of offshore operation dramatically.

Xu et al.(2013)investigated wave drift forces on three barges arranged side by side in float-over installation.Liang et al.(2012)used the numerical analysis software MOSES to study the collision force between a barge fender system and jacket in the floatover installation process.Liao et al.(2014)studied barge movement performance during float-over installation as well as the stress and strain conditions of fender and buffer devices in a cold environment.Xu et al.(2015)investigated the hydrodynamic responses of three barges moored side-by-side in a float-over operation in the frequency and time domains.

In China,most of the key steps in the float-over installation process must be simulated and forecast prior to the installation of a float-over project.To this end,in this study,we used the hydrodynamic software Sesam to numerically simulate a floatover installation for an offshore platform based on a large-scale offshore float-over installation project to provide a reference for the actual installation and to ensure its safety and feasibility.

2 Computation Theory

2.1 Time-Domain Analysis Method

As shown in Table 4,the range of mutative errors is large for different operating conditions,but stay within the allowable range,e.g.,the maximum experimental value for B1 is 0.68 m and simulation value for A1 is 0.64 m.Due to positions A1 and B1 being on the same side of the barge and their very small errors compared to the experimental values,we can conclude that the simulation results agree well with the expected values,which means that the simulation values are effective and precise,Therefore,the maximum vertical displacement is less than 1.35 m,which also ensures the safety of the float-over installation loads-transfer process.

“我需要你纯净的眼神给我灵感”摄影家诚恳地说，“这是纯粹的艺术创作，艺术的，保证不用于任何商业用途。”

The diffraction and radiation potentials can be solved by Green’s function method.Solving for the radiation potential may result in added mass and damping,and we obtained the retardation function by the additional mass and damping in the frequency domain.By solving for the incident and diffraction potentials,we can obtain the first-order wave excitation forces and wedetermined the second-order average drift forces using the far-field integration method(Pessoa and Fonseca 2013).

From Fig.7,it can be seen that the vertical displacement experience curve for position A1 corresponds with periodic temporal characteristics,namely periodic sudden increases or decreases,which are related to periodic heave motions of thebarge.The maximum and mean values of the curve are,respectively,0.64 and 0.26 m.

where,Da(ω)is an asymptotic approximation of D(ω)at high frequencies.The hydrodynamic coefficients and the retardation function are established:

Fig.1 Hydrodynamic model of barge diagram

Fig.2 Layout of mooring lines

By taking the inverse Fourier transform of Eq.(3),the retardation function K(t)can be formulated as follows:

where,M is the generalized mass matrix for the floating body,A(ω)and D(ω)respectively indicate the frequency dependent added mass and dam ping coefficient.Xu et al.(2014)investigate hydrodynamic performance of two side-by-side barge and show that numerical calculation accurately predictes the resonance period despite a difference in the peak values.C is the restoring coefficient matrix, is the first-order wave force, is the second-order wave forces,F W I is the wind force,F CU is the flow force,F MO is the mooring force,and F FE is the fender collision force.K(t)is the retardation function(Chen et al.2014),which is important to calculate the time-domain equation,is called the retardation function and results from the memory effects of the fluid.

3 Num erical Modeling

The hydrodynamic model is calculated mainly using the hydrodynamic software SESAM used to solve the frequency dependent added mass and damping coefficient of the lowto high-frequency fluid and the first-order wave forces RAO and the second-order average drift forces RAO,which is based on the potential flow theory.The overall length of the barge is 215 m,and the widths of the bow and stern are,respectively,42 and 65 m.The molded depth of the barge is 14.25 m,and its draft is 8.2 m.The hydrodynamic model of the barge is established based on the specific parameters shown in Fig.1.

Fig.3 Schematic of float-over installation model

Table 1 Positions and numbers of LMUs and DSUs

Name Position Number LMU Top of jacket leg 8 DSU Barge deck 8

插秧节省秧苗费。人工手插秧需秧苗30-33盘/亩，机械插秧需秧苗18-23盘/亩，机械插秧省秧苗10-12盘/亩，每盘秧苗可卖2.50元，节资25-30元/亩；如按成本计算秧苗大约1.5元/盘，可节资15-18元/亩。节本效益合计为32.16-40.76元/亩。

As shown in Fig.2,there are eight mooring lines and sway fenders are set at the barge bow and amidships.Figure 3 shows a model of the float-over installation.

4.要善于处理自己的负面情绪，比如急躁，慌乱，胆怯等。可以通过与他人沟通的方式来解决。在沟通过程中如果出现类似的情绪时，可以深吸一口气缓一缓，让自己冷静下来。或者找借口先离开现场，等情绪平复了再重新进行沟通，切记贪功冒进。

As shown in Fig. 5,we set the sway fenders on the barge bowand amidships.The mooring lines have two parts—an anchor chain and a steel-w ire cable.We selected polyester lacing cord for the connecting lines.Table 2 lists the properties and scantlings of the fender and mooring lines and the connecting lines.

[13] Glaeser E L, Kahn M E., “Sprawl and Urban Growth”, Handbook of Regional and Urban Economics, 2003, pp. 2481-2527.

4 Calculation and Analysis

4.1 Second-Order Wave Forces

Fig.4 Layout of LMUs

Fig.5 Layout of sway fenders

Based on the three-dimensional potential flow theory,SESAM software is applied for getting second order wave drift forces.This paper is based on the far field integration method to calculate second-order average drift forces RAO.As shown in Fig.6,second-order average drift forces of heave,roll and pitch three degrees of freedom mean is zero,so simply ignored.As the frequency increased,its changing law is analyzed to determine the basic parameters for the next step,which was the time-domain coupling analysis of the topside movement,mooring forces,and sway fender collision forces.The incident angles of the wave are set at 0°,30°,60°,and 90°,and Fig.6 show s response amplitude operator(RAO)-curve plots of the second-order wave forces experienced by the barge.

老巴说：“请个送货的，三五天帮忙进一次，每次多进点。给他付点工钱。四强哥哥已经帮我找了人，把工钱谈定，就没有事了。你赶紧回你的学校，丢了几天课，也不是个事。”

As shown in Fig.6,second-order average drift forces of barge RAO curve diagrams are plotted.

1)As shown in Fig.6,the surge,sway,and yaw values of the second-order average drift forces are very small when the frequency is less than 0.5,and the barge can continue to steadily move forward.However,these second-order average drift forces quickly increase with increasing frequency.At present,barge use must be limited to prevent drift-motion impact during float-over installations.

机械设计制造的自动化程度早已成为衡量一个国家工业制造业发展状况的标准，对于机械设计制造而言，自动化程度越高能发挥的实效力量越大，也是评判企业能否跟上时代的依据。通俗来讲，自动化程度就是要将机械设计制造与计算机、互联网紧密结合起来，同时利用电子技术的优势为机械设计制造创造更好的输出空间。多种技术的融合能缩短实际工程工期，降低投资成本，还能将原有冷冰、死板的传统机械设计制造转换成柔性化、智能化的发展方式。总的来说，机械设计制造自动化的优势集中在以下几个方面：

2)As shown in the RAO curve for the sway of the barge second-order average drift forces(Fig.6b),as frequency increases,the surge and sway of the barge second-order average drift forces increase and then stabilize.

3)As shown in the RAO curve for the yaw of the barge second-order average drift forces,the yaw values mutate with frequency.At a frequency near 0.65 rad/s,the yaw values exhibit mutations and reach a maximum,which should be avoided in float-over installation projects.

As shown in Table 1,there are eight LMUs and eight DSUs.The LMUs are symmetrically installed at the top of the jacket legs.Similarly,DSUs are symmetrically set onto the deck of the barge.Figure 4 shows the layout of the LMUs.

Table 2 Properties and scantlings of fender and mooring lines and connecting lines

Name Diameter/mm Stiffness/N Allowable tension/kN Sway fender 150 2.22E+06 6820 Steel-wire cable 68 2.08E+08 1954 Anchor chain 76 5.20E+08 2513 Polyester lacing cord 120 9.37E+07 914

Fig.6 Second-order average drift forces of barge RAO curve.a Surge in the barge second-order average drift forces RAO curve.b Sway of the barge second-order average drift forces RAO curve.c Yaw of the barge second-order average drift forces RAO curve

It is conducted that our model tests in the Deepwater Offshore Basin of Shanghai Jiao Tong University at a scaling factor of 1:40.It is conducted that mooring tests and loading transfer test in irregular wave conditions applied the test results to the float-over installation project and found them to be consistent with actual conditions.It is conducted that this study in conditions consistent with the selected model test conditions to ensure reliability in our comparisons.

4.2 Topside Movement

4.2.1 Calculation Conditions

The ITTC wave spectrum is used to study the impact of a floatover installation project under different sea conditions.Table 3 shows the wave directions,significant wave heights,and peak periods.We used the NPDw ind spectrum,set the average wind speed to 10 m/s,and contrasted the wind and wave directions.However,the average velocity of the current is set to 0.6 m/s and the flow direction to the same as that of the waves.

4.2.2 Topside Movement

In the float-over installation loads-transfer process,the most critical step is preventing a collision between the underside of the topside pile tip with the LMU.So,in the float-over installation loads-transfer process,we must ensure that the topside vertical distance does not exceed 1.35 m.Table 4 shows the maximum values of the topside vertical displacements at different positions.

The real-time motion state of the barge must be known to ensure the smooth operation of the float-over installation.The time-domain analysis method is used to calculate the mooring cable tension,maximum deviation of the topside,and the fender collision force in the float-over installation process.Howere,frequency-domain hydrodynamic coefficients are calculated based on potential flow theory(Dai and Duan 2008),and are converted with the fast Fourier transform method instead of the direct time-domain method.Considering the effects of storm and flow and the coupling effect between devices,we expressed the floating body motion equations in the time domain(Cummins 1962)as follows:

我开始时不时地往奶茶店跑，我就坐在那个固定的位置上，要同样口味的奶茶。青木开始注意我了。心里掠过一丝甜蜜。

Table 3 Calculation condition

Position number Wave direction/(°) Significant wave height/m Peak period/s A1 90 0.75 7.72 A2 90 0.75 7.72 A3 270 0.75 7.72 A4 270 0.75 7.72 B1 90 0.75 7.72 B2 90 0.75 7.72 B3 270 0.75 7.72 B4 270 0.75 7.72

Table 4 Maximum vertical displacementof topside comparative results

Position number Simulation value/m Experimental value/m Error/%A1 0.64 0.63 1.6 A2 0.43 0.42 2.3 A3 0.41 0.42 2.4 A4 0.61 0.63 3.3 B1 0.60 0.68 13.3 B2 0.40 0.45 12.5 B3 0.39 0.41 5.1 B4 0.58 0.61 5.2

In the float-over installation loads transfer process,the most critical is to ensure to insert under the pile tip of topside safely and accurately into the leg butt-coupled device of the upper portion jacket legs.So,in the float-over installation loads transfer process,we must ensure that topside movement cannot exceed Snap radius 0.6 mof leg butt-coupled device LMU(Leg Mating Unit).This paper obtained Chunk offset distance from the original center in accordance with topside horizontal and vertical displacement.Table 5 shows maximum deviation of topside from the comparative results in different conditions.

The total weight of the topside is 26,251.8 tons,and the topside length and width are,respectively,107 and 77 m.The jacket structure has eight legs.The length of the structure over the water surface is 11 m,and underwater is 190.5 m.

As can be seen from Table 5,simulation values are within the permissible range and with the experimental error that is very small.When confronting longitudinal waves,it is a larger deviation of topside,because the waves flow along the longitudinal direction of barge makes a great influence on the barge longitudinal movement;when confronting transverse waves,it is a larger deviation of topside,because waves attack to the side of barge and result in a greater transverse movement.

Fig.7 Vertical displacement experience curve for position A1

Table 5 Maximum deviation of topside from the comparative results

Conditions Simulation value/m Experimental value/m Error/%1 0.52 0.50 4.0 2 0.35 0.33 6.1 3 0.33 0.34 2.9 4 0.41 0.43 4.9 5 0.51 0.48 6.2 6 0.57 0.54 5.3 7 0.36 0.38 5.6 8 0.31 0.33 6.5 9 0.35 0.37 5.4 10 0.49 0.52 5.8 11 0.46 0.44 4.3 12 0.49 0.51 4.1

4.3 Mooring Force

4.3.1 Calculation Conditions

To realize the selected model test conditions,Table 6 shows the working conditions of the wave direction,significant wave height,and peak period.

4.3.2 Mooring Force

To prevent the mooring cable tension from being too high,we analyzed the mooring cable tensions of the float-over installation process in different sea conditions.We set 1954 kN as the permissible mooring cable tension.Table 7 shows the mooring cable tension results for the float-over installation process.It can be seen from Table 7,the experimental values are generally larger than the simulation values,due to the fact that the principle dimensions of the model test were limited by the text pool and the mooring cable was cut off in the horizontal direction.We designed a set of equivalent mooring systems based on principles of static stiffness,but the experimental values remained too large overall.However,in the float-overinstallation process,due to the fact that the mooring cable tension is much less than the permissible mooring tension,the experimental values still have some reference value.

Table 6 Calculation condition

Conditions Wave Direction/(°)Significant wave height/m Peak period/s 1 90 1.5 9.12 2 90 1.25 8.72 3 0 1.0 8.26 4 0.75 7.72 5 0 0.5 7.02 0 6 0.75 7.72 7 270 0.75 9.12 18 270 0.75 7.72 0

Table 7 Maximum mooring cable tension results in float-over installation process

Mooring number Simulation value/kN Experimental value/kN Error/%MP1 391 426 9.0 MP2 436 487 11.7 MP3 383 424 10.7 MP4 494 559 13.2 MP5 490 568 15.9 MP6 397 451 13.6 MP7 406 448 10.3 MP8 413 457 10.7

4.4 Sway Fender Collision Force

4.4.1 Calculation Conditions

To realize the selected model test conditions,Table 8 shows the working conditions of wave direction,significant wave height,and peak period.

(1)分别在干预前、干预后采用Fugl-Meyer运动功能量表(Fugl-Meyer Assesment,FMA)对两组运动功能进行评估,分数越高表示患者运动能力越好。(2)采用蒙特利尔认知评估量表(Montreal Cognitive Assessment,MoCA)对两组患者干预前后认知功能进行评估,分数越高表示认知功能越好。

试验地设在贵阳中医学院5号温室大棚，大棚配置有自动喷水装置、控温系统、换气装置。选择土壤疏松、排水性好的地块作为苗床，苗床宽100 cm左右，采用凹畦的形式，畦深约20 cm。苗床上设置高60～70 cm的遮阳网支架。将经过5‰的高锰酸钾溶液消毒处理的基质平铺于苗床上，整平，基质厚度15～20 cm。按试验设计条件，分别将珍珠岩、蛭石、草炭土按比例铺于苗床。

4.4.2 Sway Fender Collision Force

In the float-over installation process,the main function of the barge sway fenders is to limit the transverse movement duringsurges of the barge,mooring position,and loading transfer process and to avoid a rigid collision between the barge and jacket.When sway fenders collide with the jacket,this is called a broadside collision.We set the designed load of the sway fenders to 6860 kN according to the device.Table 8 shows the collision forces of barge-side fenders generated in different conditions and different phases of the float-over installation process.Figure 8 shows the collision force experience curves of left amidships fender when the barge is moving forward.

Table 8 Calculation condition

Phases Sway fender Significant wave height/m Peak period/s Wave Direction/(°)Ship forward Right bow fender 0.75 7.72 90 Left bow fender 0.75 7.72 270 Right amidships fender 0.75 7.72 270 Docking Right bow fender 0.75 7.72 90 Left bow fender 0.75 7.72 90 Right amidships fender 0.75 7.72 90 Left amidships fender 0.75 7.72 270 Ship back Right bow fender 0.75 7.72 90 Left bow fender 0.75 7.72 270 Right amidships fender 0.75 7.72 270 Left amidships fender 0.75 7.72 90 Left amidships fender 0.75 7.72 270

Fig.8 Collision force experience curve of left amidships fender with barge moving forward

From Fig.8,we can see that the collision forces of the left amidships fender while the barge is moving forward correspond with the periodic temporal characteristics,namely the sudden periodic increases or decreases related to the periodic oscillation of the barge.The maximum and mean values are,respectively,6259 and 2256 kN.In this case,the actual installation should strive to prevent collision forces that are excessively large due to the sudden collision of the side fenders of the barge and the jacket.

Table 9 Sway fender collision forces

Phases Sway fender Simulation value/kN Experimental value/kN Error/%Ship forward 5932 5659 4.6 Leftamidships fender 6259 5813 7.1 Docking Right bow fender 4443 4343 2.3 Left bow fender 4130 3998 3.2 Right amidships fender Right bow fender 4877 4589 5.9 Left bow fender 4878 4607 5.6 Right amidships fender 5292 5024 5.1 Leftamidships fender 4846 4801 0.9 Ship back Right bow fender 1819 1820 0.05 Left bow fender 1966 1870 4.9 Right amidships fender 4815 4544 5.6 Leftamidships fender 4952 4898 1.1

As shown in Table 9,the simulation values are within the allowable range and are smaller than the experimental values,which is due to the use of cushions to simulate the fender in the model test,and the fact that the barge fender and jacket flex during collision.Due to limitations of the software,the simulation results indicate that the barge fender and the jacket experience a rigid collision.As such,the experimental values are generally smaller than the simulation values and have more reference value.However,the errors of the numerical simulation are much smaller than those in the model test results,which verify the accuracy and feasibility of the numerical simulation in the float-over installation process.This result also provides data and technical support for the actual project program.

5 Conclusions

In summary,based on three-dimensional potential flow theory,we selected an appropriate calculation method for calculating the hydrodynamic performance of a T-barge in the frequency domain.We numerically simulated the coupled motion of the offshore platform float-over installation process in the time domain and verified the results with experimental values.We conclude that the experimental values of the topside excursion distance are generally larger than the simulation values,which is due to the principle dimensions of the model test being limited by the text pool.The fender collision forces of the side shell correspond with periodic temporal characteristics,namely,sudden periodic increases or decreases related to the periodic oscillations of the barge.We compared our results with those of the model test to verify our calculations and numerical simulations.The simulation values obtained from different conditions fall within the allowable range,which shows that the safety and accuracy of float-over installation in similar conditions can be ensured.Furthermore,it can also be used in many practical marine Engineering application areas and is practical to deal with problems associated with calculating marine engineering simulation,thereby providing a basis for the development of design and operation guidelines for actual installations.

随着城市的不断发展，市政工程建设也在如火如荼进行中，每天都有数以万计的市政工程在施工，以满足人们对基础设施建设的实际需要。市政工程是城市发展中不可缺少的物质基础，在建设的过程中只有保证了施工质量，才能提高市政工程的使用价值，其社会效益和经济效益都能实现，后期保养和维护难度也会减小。

References

Chen MS,Eatock Taylor R,Choo YS(2014)Time domain modeling of a dynamic impact oscillator under wave excitations.Ocean Eng 76:40-51.https://doi.org/10.1016/j.oceaneng.2013.10.004

Chen MS,Eatock Taylor R,Choo YS(2017)Investigation of the complex dynamics of float-over deck installation based on a coupled heave-roll-pitch impact model.Ocean Eng 137:262-275.https://doi.org/10.1016/j.oceaneng.2017.04.007

Cummins W.E,1962.The impulse response function and ship motions.Schiffstechnik,(09),101-109

Dai YS,Duan WY(2008)Potential flow theory of ship motion in waves.National Defense Industry Press,China,Beijing

Du XY(2007)Searching for float-over installation of Nanpu35-2 CEP moudule.China Offshore Platform 22(04):39-43

Fan M,Yi C,Bai XP et al(2013)Research and application of key technologies for float-over installation of large topside in China.China Offshore Oil Gas 25(06):98-100

Geba KA,Welaya YMA,Lehet HW et al(2017)The hydrodynamic performance of a novel float-over installation.Ocean Eng 133:116-132.https://doi.org/10.1016/j.oceaneng.2017.01.034

Hamilton J,French R,Penman AD,2008.Topsides and jacket modeling for float-over installation design.Offshore Technology Conference Paper,5-8 may,Houston,Texas,USA,DOI:https://doi.org/10.4043/19227-MS

Hu ZH,Li X,Zhao WH,Wu X(2017)Nonlinear dynamics and impact load in float-over installation.Appl Ocean Res65:60-78.https://doi.org/10.1016/j.apor.2017.03.013

Liang X X,Zhang Y G,He M,et al,2012.Application of MOSES software to offshore installation analysis for large jackets.Shipbuilding of China,Xiamen,53(S2),362-371

Pessoa J,Fonseca N(2013)Investigation of depth effects on the wave exciting low frequency drift forces different approximation methods.Appl Ocean Res 42:182-199.https://doi.org/10.1016/j.apor.2013.06.003

Seij M,Groot HD,2007.State of the art in float-overs.Offshore Technology Conference,30 April-3 May,Houston,Texas,U.S.A.DOI:https://doi.org/10.4043/19072-MS

Sun L,Eatock Taylor R,Choo YS(2012)Muli-body dynamic analysis of float-over installations.Ocean Eng 51:1-15.https://doi.org/10.1016/j.oceaneng.2012.05.017

Tahar A,Halkyard J,Steen A,Finn L(2006)Float-over installation method comprehensive comparision between numerical and model test results.J Offshore Mech Arctic Eng 128(3):257-262.https://doi.org/10.1115/1.2199556

Wang B,Yang XL,Zhang GL et al(2017)Key technologies of DP floatover installation and corresponding feasibility analysis in the East China Sea.Shipbuilding of China 58:162-169.https://doi.org/10.3969/j.issn.1000-4882.2017.01.018

Xia J,Hayne S,Macfarlane G et al(2005)Investigation into float-over installations of minimal platforms by hydrodynamic model testing.ASME Int Conf Offshore Mech Arctic Eng 62(3):193-215.https://doi.org/10.1115/OMAE2005-67092

Xu X,Yang J M,Li X,et al(2013)Wave drift forces on three barges arranged side by side in float-over installation.Proc.32th Int.Conf.on Offshore Mechanics and Arctic Engineering(ASME),Nantes,France.DOI:https://doi.org/10.1115/OMAE2013-10737

Xu X,Yang JM,Li X,Xu LY(2014)Hydrodynamic performance study of two side-by-side barges.Ship Offshore Structure 9(5):475-488.https://doi.org/10.1080/17445302.2014.889368

Xu X,Yang JM,Li X,Xu LY(2015)Time-domain simulation for coupled motions of three barges moored side-by-side in float-over operation.China Ocean Eng 29:155-168.https://doi.org/10.1007/s13344-015-0012-4