It is widely known that the critical strain to ductile crack initiation depends on the stress triaxiality. And the ductile crack may occur at the strain concentration region in the structural member unless initial crack by large deformation. Such large deformations occurs during ship crash, seismic condition, etc. And then, strain rates are usually high. But, the strain rate effect to the ductile crack initiation is not known.
To study the effect on ductile crack initiation condition, the axisymmetrically notched bar tensile tests were conducted under various tensile speed for SM400B, SM490A, SM490B and SN490B steels. And critical strains to initiate ductile crack were evaluated. To estimate the stress - strain states during dynamic tensile test, the thermal mechanically coupled FE-analysis (MARC K7.2) were used.
The results showed the strain rate effect on the critical strain to initiate ductile crack is small. And the results also suggested that if the load is controlled by displacement, the higher yield ratio steel is advantageous for the ductile crack initiation.

View full abstract

The specimen size requirements in toughness qualification are discussed on the basis of statistical local fracture criterion approach. The Weibull stress criterion for cleavage fracture is applied to the constraint loss field at the crack tip described with the toughness scaling model (TSM). Proposed new model (WSSM) demonstrates that the TSM is a particular case fora material with a particular combination of the Weibull shape parameter, m and the strain hardening exponent, n. The correlation between TSM and WSSM were also confirmed with 2 dimensional and 3 dimensional finite element analysis. The 3 dimensional analysis with J_p-δ for the stress intensity suggested the less constraint loss effect rather than 2 dimensional analysis in spite of the strong dependency of the constraint loss effect on Weibull shape parameter, m. The 3 dimensional analysis also suggested that the thickness effect on toughness mainly resulted from the statistical volume effect.
The values of the hardening exponent, n and Weibull shape parameter, m, that are the dominant factors for the specimen size requirement suggested from WSSM, were investigated from the experimental results for the practical constructive steels. These results demonstrated that the specimen size requirements prescribed in ASTM could be too conservative.

View full abstract

From the viewpoint of structural safety, leakage of cargo from welded plate structures must be prevented. Fatigue crack initiation and propagation tests were conducted in order to investigate the fatigue crack behavior at plate intersections of outer plate and inner longitudinal web plate and to make clear the outer plate-thickness-penetration condition of fatigue crack from inner members, by using welded plate structure models. In the case of longitudinal structural model with large bracket, a crack initiated from the fillet weld of large bracket and propagated the longitudinal face member and the web plate and then reached the outer plate. The crack did not penetrate the outer plate and propagated along the fillet weld of the outer plate and the inner web plate. On the other hand, in the case of the model with small bracket, a crack from the fillet weld of small bracket penetrated the outer plate. The plate penetration condition was derived on the base of the maximum stress criterion and was verified experimentally. This condition takes account of both the incident crack angle and the primary stress ratio at plate intersections. In the latter case, shear stress was relatively small.

View full abstract

In order to clarify the mechanism of fatigue fracture, fatigue crack propagation tests using Cu single crystals, as a first step, are conducted. Six types of specimens, crack-crystal orientation relation is different in each, are used. And SEM observations of fracture surfaces are conducted. It is shown that fatigue crack in single crystal propagates microscopically on the active slip plane which is characteristic of each crystal system, not on the mechanical principal stress plane. This mechanism can not be explained by the Laird model, which is thought to be appropriate in continuum mechanics, and needs expansion of the Neumann model.
Relation between fatigue crack propagation rate and stress intensity factor range is also investigated. In single crystal, an active slip system interferes intricately with other ones, and fatigue crack propagation rates depend much on this mechanism. Therefore, ΔK, a macro-mechanical parameter, is not useful for explaining the phenomenon in such single crystals.

View full abstract

Analyses of fracture behavior in Cu single crystal are conducted. Changes of atoms near crack tip under mode I loading are simulated by means of Molecular Dynamics (MD) method. Results are compared with experiments of fatigue crack propagation in Cu single crystal. MD simulations can reproduce the phenomenon that fatigue crack in single crystal propagates by slipping only on the active slip plane which is characteristic of the crystal system, which is confirmed by SEM observation in the tests.
Changes of atoms under cyclic loading are also simulated. It is shown that MD simulation, even when 2-body potential function is used, can represent the irreversible changes of atoms (i. e., accumulation of damage), which is thought to be essential behavior in fatigue fracture.
From the results above, it can be said that MD method may be a useful tool for analyzing a microscopical mechanism of fracture, at least in metal.

View full abstract

The micromechanical hardening rules developed by Cuitino and Ortiz [Modelling Simul. Mater. Sci. Eng. 1 (1992) 225-263] and Repetto and Ortiz [Acta mater. 45, 6 (1997) 2577-2595] are modified in order to establish a new approach to the study of PSB formation process. A crystalline FEM theory which incorporates this modified hardening rule is presented. We explore the propriety of the presented hardening rule and FEM theory by comparing the calculated and measured cyclic stress-strain relations of Cu single crystal.
As results, the followings are found :
1) The presented hardening rule and FEM theory give a good account of the nature of cyclic stress-strain relation of real f.c.c. crystals qualitatively.
2) The calculation result shows that the secondary slip occurs at the moment Bauschinger energy parameter β_E shows its maximum value. PSB formation begins at about the same time. It is presumed that the occurrence of secondary slip is related to the formation of PSB.

View full abstract

In the safety design of marine systems, it has to justly give a matter over human life in the case of outbreak of a marine accident such as fires and collisions. It may be difficult to establish a safe evacuation system by the cause of uncertain behavioral patterns based on human factors with ambiguity. Besides, the psychological effect on behavior of evacuee that is terrified under the stimulus by accident, should be considered in the redundancy system for safe evacuation. Accordingly, the safe evacuation system with the intention of securing evacuation routes and reducing escape time, is desirably designed by grasping the personnel behavior with psychological intelligence process in an emergency of marine accident.
In this paper, the evacuation behaviors can be predicted by using the proposed walking model with individual capacity of visible length and avoidance process for obstacles. Further, the mathematical model of psychological intelligence process in an emergency is contrived herein, in which the evacuation movement is simulated in the consideration of the decrease of walking speed and the occurrence of self-isolation in psychological action according to the magnitudes of accident stimulus. By applying these models to common spaces in a dinning cruise ship, taking account of the human factors with psychological action is rather valued in the safe evacuation system and the validity of the proposed models can be confirmed.

View full abstract

Evacuation behaviors are series of actions such as recognition of disaster, initial response, judgment of situation, starting of evacuation and evacuation action. Success of evacuation seems to greatly depend on starting time of the evacuation and selection of escape route among various evacuation behaviors. In case of fire, failure of evacuation often caused by “delay of starting of escape” and “improper selection of escape route”. Therefore, it is extremely important for success of the evacuation that the order for starting evacuation at the early stage of the disaster and the guidance for the appropriate escape route. In other words, guidance of the evacuation by crew is significant for the success of the evacuation.
In the previous paper, we presented the layered ship's space model constructed with primary elements by spaces and nodes, the evacuee's model which manages behavior of each evacuee, and introduced the fire scenarios. In this paper, new model for starting time of the evacuation is introduced to simulate the situation of the evacuation based on the guidance by crew. Using these models, several evacuation simulations were carried out and evacuation situation of evacuees were estimated. The results of the simulation shows that the starting time of the evacuation, the initial location of evacuees, the smoke movement and the selection of the escape route significantly affect the situation of evacuation. Especially, it was understood that the starting time of evacuation has great influence on success of the evacuation.

View full abstract

Formal Safety Assessment (FSA) considered at IMO, is formal methodology for improving the transparency and accountability of decision related to amendment to the regulations. FSA consists of PSA (Probabilistic Safety Assessment) and cost and benefit assessment. Risk assessment is the most important and difficult part to realize FSA. This paper introduces a methodology for risk assessment and a system for risk assessment composed of several programs including disaster development simulation and evacuation simulation. Furthermore, the result of the trial application of the system, in which the effect of sprinkler systems, i.e., one of the Risk Control Options (RCOs), on the risk due to cabin fire is estimated. Though there are many defects and problems in the system and the methodology at present, i.e., discussed in this paper, the methodology is considered to contribute for reducing the arbitrariness at the decision related to rule making process.

View full abstract

It is important to design effective escape routes in ships and offshore structure which can provide smooth escape of passengers and crew without any congestion and confusion. After the tragedy of ro-ro passenger ferry “Estonia”, requirement on evaluation of effectiveness of escape routes in ro-ro passenger ferries in its early design stage has been added to SOLAS Convention.
This study has evaluated the effectiveness of an escape simulation program of grouping model of evacuation in comparison between the calculation results of evacuation by the simulation and results of actual evacuation trial carried out on a passenger ferry of 10, 000 tons using about 400 people. It has been found by the simulation and the trial that cascaded starting time of evacuation from each compartment may result in smoothness or congestion in escape routes. It has been also found that the simulation could understand the flow, congestion and finish of various case of evacuation.
Calculations by the simulation program was also carried out for various scenarios such as a case of fire in daytime and night on a type ship which has 777 passengers and 55 crew. It has been found, as the results of the simulation, that congestion points do not depend on casualty cases and that congestion may occur when condition in downstream of escape routes is unknown even evacuation is taken place as planned.

View full abstract

More than 80% of the material movements, except raw materials, between Japan and its trading partners are conducted today in the containerized mode. In recent years, the marine transportation sector has been undergoing very rapid and multifarious changes, due to the alliances between shipping companies, introduction of larger ships as a cost reduction measure and the application of integrated logistics to satisfy the needs of customers.
However, many shipping companies tend to avoid calling at Japan's ports due to the inadequate terminal facilities and other services, insufficient channel depth, inefficient handling and the high cost of port fees and dues.
In the first report, we presented an evaluation method of the container handling efficiency of terminals conducted by a Petri net model of the container terminal operations for numerical simulations. In this paper, a high-density system for integrated container handling at a next generation terminal is proposed for solving those defective points and a practical simulation is attempted using the evaluation method presented in the first report. Further evaluation on the possibility of a high-density system for integrated container handling is attempted by the evaluation-decision model based upon multi-criteria objectives.

View full abstract

Generally, there are some relations of the queue that are occurred while the maintenance work and the recovering work. Therefore, it is able to calculated increase or decrease situation that is waiting time, waiting number of cases and occurrence ratio by using a queue procession. Furthermore, it's gotten the optimal number of maintenance person by making the occurrence situation. Engine failure in a ship can be regarded as exponential arrival model in random failure period. Also it's conceived that the distribution in recovery time complies with random Poisson or Erlang distribution. Recovering work data are calculated with a repair rate from a man-hour in terms of the kind of maintenance and queue simulation model. It was constructed by using queue processions and simulation programs. Finally, it's found out the number of optimal maintenance personnel concretely by it.

View full abstract

Recently, the significance of ship inspection is recognized more than before because of the strong paying attention to environmental and safety problems. However, current ship inspection relies on the experience of the worker. Therefore, it is difficult to grasp and to improve the state of ship inspection.
In this paper, ship inspection is classified into three stages (Plan, Do and Check), and total supporting systems for ship inspection are discussed. Moreover, prototype system for inspection planning stage is developed. This system is called “Ship Inspection Simulator”, in which worker can execute virtual ship inspection, then, damage finding probability of each hull part and looking-off area in inspection are calculated automatically. In order to realize this, some new models and functions are introduced. These are “Inspection Position Model”, “Extract Function” and “Analysis Function”. “Inspection Position Model” expresses worker's view at each inspection position. “Extract Function” and “Analysis Function” are defined to calculate damage finding probability and looking-off area in each hull part by the use of product model and “Inspection Position Model”.
Some examples are shown at the end of this paper.

View full abstract

A simplified method of estimating the ultimate strength of ship bottom plating under combined transverse thrust and lateral pressure is presented. A series of elastic/elastoplastic large deflection analyses is performed to clarify the influence of lateral pressure on both elastic buckling and ultimate strengths of a continuous plate. Based on the obtained results, a set of formulae has been derived for the estimation of the ultimate strength. Two fundamental collapse modes are recognised for a continuous plate under combined transverse thrust and lateral pressure i.e. simply-supported mode and clamped mode. For the former collapse mode, the increase of ultimate strength as a result of the increase of elastic buckling strength under lateral pressure is taken into account. The effect of lateral pressure on the yielding of plate is considered for both collapse modes. The ultimate strength is estimated taking the lower strength between those obtained for the two assumed collapse modes. It has been found from a comparison of the predicted strengths with the FEM results that the proposed formula gives very accurate prediction of the ultimate strength of a continuous plate under combined transverse thrust and lateral pressure.

View full abstract

The authors presented a simplified method to estimate the ultimate strength of a continuous stiffened plate under uniaxial thrust in the 1 st report. In this method, a continuous stiffened plate is modelled as a double-span beam-column, and the influence of the interaction between adjacent spans and that of the interaction between panel and stiffener on the ultimate strength are taken into account. In the present report, the influence of welding residual stresses produced by the fillet welding along panel/web intersection and/or web/flange intersection is further considered to apply the proposed method to a real welded structure. As for the stiffened plate with flat-bar stiffener, the effective thickness of stiffener is modified to consider the panel/stiffener interactions more correctly. The ultimate strengths estimated by the improved method are compared with the results of elastoplastic large deflection FEM analyses and those of the existing collapse tests for continuous stiffened plates. The comparison with the existing design formula and other estimation method is also carried out. It has been found that the present method can estimate the ultimate strength which is in better agreement with the results of FEM and tested results than the existing methods.

View full abstract

Precision of a ship building block gives large influence on a number of works to load the block into a dockyard. So many research projects have been carried out, and measurements of weld deformation and investigations to estimate weld shrinkage have been promoted. However, the methods applicable to a curved block that has curved shell plates were limited.
In this paper, the calculation methods on weld deformation are reviewed, and a new method applicable to the curved block is proposed. The results of experiments, in which weld deformations were measured in two models, are compared with the calculated values using the proposed method. Both experimental and calculated values are in good agreement, and the availability of the method is confirmed.

View full abstract

Numerical analysis of potential distributions in consideration for influence of some parameters of seawater in floating structures under cathodic protection using sacrificial anodes is difficult, because of complicated relation between mechanism of corrosion and parameters of seawater.
This report has discussed with usefulness of numerical analysis with potential distribution using Neural Network Method in consideration for influence of parameters of seawater.
First, polarization characteristics of electrodes in consideration for influence of parameters of seawater.has been estimated by Neural Network method. Polarization characteristics of electrodes have been assumed to be linear.
Second, using presumed polarization characteristics as input value, a change of potential distributions has been estimated by Boundary Element Method year by year.
From these results, the necessity of anticorrosion design in consideration for influence of parameters of seawater have been confirmed.

View full abstract

Author proposed the new method of gaseous CO₂ dissolution in the ocean. The system is composed of the horizontal vessel installed in the proper depth of the sea and CO₂ supplying and activating system of the interface of both fluids. Dense water in which CO₂ molecules are dissolved will flow down toward the sea floor inducing the bulk flow in the ocean and being accompanied with the molecular diffusions. These phenomena were treated analytically in the turbulent field by means of the k-ε two-equation model, which was applied to the semi-infinitive domain in the ocean. The basic datum of CO₂ mass transfer flux at the interface and density change of CO₂ dissolved water were obtained by the experiments of various kinds of pressure tests and the eddy viscosity and eddy diffusion coefficients were derived from the above mentioned turbulent model. Analyses were carried out with the two-dimensional cylindrical coordinate system under the still water assumptions.

View full abstract

Predicting the spread of smoke from a marine fire in enclosed spaces is still one of the challenge to the functional design of safe and comfortable environments for marine systems. A distinguishing feature of the problem is that the temperature difference as well as the density variation is very large while the smoke flow induced by the heat release is much slower than the propagation of acoustic waves. Therefore, this flow should be considered as a low-Mach-number compressible flow. The generally used Boussinesq approximation, in which the density is assumed constant except for the buoyancy term, is limited to very small temperature difference, and is not an appropriate model for the problem.
In this paper, a numerical method is developed, in which the control equations are derived on low-Mach-number approximation, and the Favre averaged compressible k-ε model is used to solve the turbulent transportation coefficients. A two-dimensional thermal plume above a line source is simulated by the present method and the method based on Boussinesq approximation. By comparing the distribution of vertical velocity and temperature above the heat source to theoretical and experimental results, it is found that when the strength of the heat source is very small, the two models give almost the same results. On the other hand, for the case of large strength of the heat source, computation by the low-Mach-number model gives better results than by Boussinesq model.

View full abstract

The authors have been studying on a factory simulator that bases on petri nets and published the paper on the autumn in 1997. This paper is the 2nd report of the authors' research and presents the enhanced concept of the previous simulator.
In the authors previous paper, the job shop are represented by petri nets as a core model of the shipyard simulator. The previous factory simulator can work by inputting job information that are separately generated with the production plan support system. However, we can see that there are many jobs according to the products in a job shop and the many kinds of planning schedules in the actual production activity, such as the general schedules planning, a wide-range schedules planning, a mid-range schedules, the detailed schedules planning for workers and etc. It is a characteristic that the content of work is various and the kind of the detailed degree of jobs is various also, and we can recognize the necessity for executing the factory simulation.
In this paper, two petri net models for simulator are defined separately, one is for the jobs in the job shop and another is for the factory. Furthermore, the some ideas are introduced in order to use the above-mentioned two petri net models.

View full abstract

Virtual Factory (Factory simulator) is useful for examination of productivity. For shipyard, it is especially useful for leveling the work amount. In this paper, authors developed factory simulator of shipyard for long term scheduling. In this system, scheduling knowledge of shipyard is expressed and, the series production of LPG carriers can simulate in a very short time. The result of simulation is acceptable for actual use. And, it can simulate the rebuild of the shipyard.

View full abstract

Dimensional uncertainties of ship blocks increase in accordance with each stage of fabrication and assembly by the accumulation of errors, and cause the increase of labor cost at the erection stage. Therefore accuracy control is considered to be one of the most important problem in ship hull production.
There will be two kinds of accuracy control ; dimensional control in assembly stage and positional control in erection stage, while the former has been studied mainly in shipyard so far. Now CAD system and product model of CIM have been applied widely, and also efficient measurement technology has developed in shipyard, then it becomes easy to apply a positional control in erection stage.
This report presents a computer simulation method to put a block, which has dimensional errors slightly, on suitable position as designed using neural network technology. A pilot program of C using personal computer is developed and applied for a flat block as an example. The application showed a good result to get optimum position.

View full abstract