Glykas and Das (2001) calculated the energy dissipation on the bow structure during a "head-on" collision with a rigid body, using finite element analysis. Wang et al. (2002) presented an investigation of the longitudinal strength of ships with damages due to grounding or collision accidents. Scope of analysis The objective of the analysis was to compare behavior of three types of barges: (i) OBP-500W, being a barge operated by the Polish owner ODRATRANS, further referred to as barge A and two innovative barges devel-oped in Project INBAT – (ii) the design developed by the Design Office for Inland Navigation NAVICEN- Framesno. 7, 10 TRUM, referred to as barge B and (iii) the design de-veloped by a team in the Technical University of Szczecin, referred to as barge C. More detailed descrip-tion of the barges is given by Guesnet et al. (2004). Barge A is a typical barge being currently operated. It is a steel barge characterized by shell structure reinforced with numerous girders and stiffeners at the bottom and sides. Thickness of the plates of the outer shell and the cargo deck is 6 mm. In the regions exposed to increased loading the sides are reinforced by additional stiffeners. The barge has the explicit fore part – raised, with fore deck, equipped with the bulkhead, side decks and the aft part. A firm transverse beam is situated at the midship section. Main particulars of the barge are: length overall 45.13 m, breadth overall 8.988 m, depth 1.7 m, design draught 1.6 m, frame spacing 0 - 80 0.5 m, frame spac-ing 80 -91 0.4 m – Fig. 1. Fig. 3: Cross-section of barge C Authors believe, that the application of nonlinear, fast transient application of FEA, when the structure model is isomorphic with physical structure, permits to achieve accurate results in the analysis of the loss of structural integrity during ship collision.. As the barge structure is relatively simple and small in size, a detailed computa-tional model may be elaborated relatively easily and the necessary simplifications do not seriously influence the accuracy of the results. Computations have been done using the PAM-CRASHTM computer code which is a part of the PAM-SYSTEMTM, developed by the ESI Group. It is the computer code for analysis of destruction of technical objects by the finite element method employing the explicit time integration scheme. This approach pro-duces the best results for solving the dynamic problems including contact. Arbitrary type of s structure can be modelled using plate, shell, solid and beam elements. Functional persity allows to take into account various physical phenomena occurring during the collision of the barges. Fig. 1: Cross-section of barge A Barge B is a concept based on a relatively dense double bottom grid made of thin plates. The structure is stabi-lized by foam filling the double bottom space. The cargo deck is reinforced using the laminate plates with increased thickness of the plating directly exposed to the cargo in the expected larger pressures – Fig. 2. The problem of the stability of the numerical solution is overcome assuming appropriate time step according to Eq. 1 ρ= ∆Eltminmin (1) where lmin is the geometry discretization size (size of element). Fine meshing - small size of the finite ele-ments – thus leads to numerous time steps.. Even so small step size can be advantageous since high resolu-tion in the time domain can be important in accurately capturing the non-linear behaviour of system. PAM-CRASHTM user may model extreme non-linear behaviour under high transient load by simulation of elasto-plastic behaviour of material with strain rate, damage effects and rupture any type of material. When the software is used, considered models can be prepared using structural parts with solid, shell and beam elements as well as rigid parts. Validated material models include common engineering materials such as metals, composites, highly compressible polyurethane foams and others. Special essential features in mechani-cal structures can be easily modelled – behaviour of rivets and welding, which may be critical for collapse modes, may also be taken into analysis Fig. 2: Cross-section of barge B Barge C is an all-steel barge made using the typical for shipbuilding flat plates and innovative structural ele-ments - I-coreTM panels developed by MeyerWerft, being elements with internal structure (Roland and Metschkow, 2002) – Fig. 3. They are composed of two plates connected by densely situated vertical ribs. A contact model is the important feature of PAM-CRASHTM. It may be accounted for with the series of search algorithms. The friction between areas of contact are taken into account. The integration scheme stability is controlled with Cou-rant condition to avoid the numerical instabilities and time step collapse. Also subcycling algorithm may be used to increase the calculation efficiency through ex-ploitation advantage of local mesh density. Finite element modeling and computational procedure Models of barges Computational models were built using shell, solid, beam and rigid elements. Contact in collision was de-tected by the algorithms of PAM-CRASHTM. The model of contact includes friction. Finite element model was built using the 3D structural documentation. Longitudinal symmetry of the barges was utilised. The side of the barge which is in collision is modelled as a deformable structure using 3- or 4-noded shell elements and beam elements including the transverse beam. In the case of analysis of collision, the opposite side of the barge is modelled: (i) in the case of the struck barge using the shell rigid elements to allow to account the influence of water inertia effect while the barge moves perpendicularly to the direction of move-ment along current direction – as described in Section “Model of water”; (ii) in the case of striking barge using the rigid body elements having the mass equal to a half of the lightweight and a half of the actual deadweight. When grounding is analysed, the opposite side of the barge is modelled using the rigid body elements having the mass equal to a half of the lightweight and a half of the actual deadweight. Welds and minor elements which do not influence the structural strength are not modelled. All flanges of floors, girders and stiffeners were modelled using beam elements. To obtain the effect of loading acting on the inner sides of the cargo hold, 内河运输船舶碰撞与搁浅的数值模拟英文文献和中文翻译(3):http://www.751com.cn/fanyi/lunwen_63533.html