Flood frequency and flood water depth were estimated using satellite remote sensing data for the development of flood hazard map and flood risk zoning map. Physiographic divisions, geological divisions, administrative districts, road, rail and drainage network digital data were used through GIS approach. Flood damaged area for each physiographic division, geologic division and administrative district were estimated. Finally, the most safe flood hazard map for flood countermeasure was constructed by considering interact effect of flood water depth and flood frequency simultaneously. Flood risk zoning map for each administrative district was developed, which provides the information about the development and preparedness of the administrative districts, road and rails, etc. on the priority basis against the flood damages.

View full abstract

It is indispensable that disaster information enables the appropriate action to be carried out for the benefit of the affected inhabitants in the case of a flood. Also, in disaster research, it is important to understand the information acquisition situation of the inhabitants in the event of such a disaster. In this study, we examine information acquisition under actual conditions, as gleaned from a questionnaire completed as part of a survey of the inhabitants of Koriyama-City at the end of August 1998. We also explore the relationship between information acquisition format and evacuation behavior. In addition, the information needs of the affected inhabitants under actual disaster conditions are discussed.

View full abstract

Recently, it becomes prosperous as for trial which activates public participation and cooperation for environmental preservation of the river. The purpose of this study is to make obvious some knowledge about the means of activation for improving the residents' interest and the intention of public participation for preservation of the familiar river in the basin. Therefore, the questionnaires was collected in the Sendai river basin and analyzedabout improvement of interest for the familiar river and the intention of public participation to activities. As a result, some new knowledge was made obvious which were the relations of consciousness and action in the basin, the main factors which affect interest, the notes when supply the opportunity for participation, and the others.

View full abstract

Each person in the household takes action based on there own role, when flood disaster happens. In particular, it is estimated that a householder takes the leadership on the action of other household person. He lets person of tender age and advanced age with first priority evacuate, and takes preservation behaviours of their possessions. There is, however, a high possibility of causing a number of victims, as he tends to delay or abandon evacuation because of these behaviours. In this study, we analyze that the allotments of household exist in any pattern and then have any influence on evacuation activity of each person in household, and the characteristics of household behaviour in flood disaster in Koriyama-City due to heavy rainfall at the end of August 1998.

View full abstract

Recently, in flooding-prone area, the flood control facilities have been provided positively, and the incident of flood disaster have decreased. So the opportunities that inhabitants experience the large-scale flood disaster decrease. Consequently, the disaster education and folklore are the chief opportunities for inhabitants to acquire the wisdom on mitigation of flood disaster damages. In this study, we examine the effects of disaster education and folklore of the past flood disaster on consciousness and human behavior for flood disaster. The effects that the disaster education and folklore make consciousness of inhabitants for flood disaster higher, and promote countermeasures and evacuation behavior of inhabitants against flood are confirmed.

View full abstract

This paper deals with flood mitigation measures by only one side dike on almost primitive rivers. If those rivers are left at existing conditions, the lands might be improved at random with irrigation channels and drainage channels constructed without any relation to flood mitigation, at which conditions it will take a lot of cost and time to execute effective flood mitigation works. Here, it is proposed to be beneficial to make a flood mitigation plan by one side dike at early stage development of land through two examples on the Apure River and the Chama River in Venezuela.

View full abstract

In developing countries, elevated roads on embankments that cross flood plains and approach roads to bridges which thrust out into river channels are often seen. Such structures interfere with flood flow, causing scouring around piers and soil embankments. One possible solution to this problem would be the construction of bridges completely over flood plains or rivers; however, this is not entirely realistic due to limited financial conditions in the affected countries.
When designing such roads and bridges, two facts must be taken into account: rivers are in their natural state (i, e., without flood control) and there is a high probability of floods occuring beyond a design flood discharge due to insufficient hydrological data. Design criteria in Japan cannot simply be applied to the local situation. In this paper, such problems actually occuring at individual locations are analyzed and feasible countermeasures are proposed from the point of view of river disaster prevention.

View full abstract

We evaluate an economic effect of potential flood risk which is defined by height difference between high water level and ground level with the two-regional residential location model. The model describes residential behavior of households which choose one of the two regions to maximize their utilities. First we get a parameter in a utility curve which shows a preference rate of the risk. Second we examine the risk-mitigation benefit under a variety of condition. Finally we discuss that a relationship of beneficiaries or a total benefit depends on an extent of influence which we set.

View full abstract

he Ninoarate bank of the Hyakken river is a historic civil engineering structure which was built more than three hundred years ago. Under the original river restoration plan, the entire Ninoarate bank structure was scheduled to be removed. However, due to the historic significance of this precious hydraulic structure, the possibility of its preservation has been re-examined in recent years. In this paper, the possibility of preserving the Ninoarate bank is studied by the flow analysis based on MacCormack Scheme. Numerical results show that it is feasible to preserve a part of the Ninoarate bank at the actual location, though it may be difficult to preserve the whole structure.

View full abstract

This paper discusses the concurrent characteristics of the maximum storm surge and rainfall during a typhoon by analyzing the records of hourly rainfall, tide level and storm surge caused by typhoons. Using the bivariate exponential distribution theory, the joint distribution function of storm surge height and rainfall is analytically obtained. The diagram of return period is shown in order to evaluate the past main typhoon events.
We also investigate how to determine the design rainfall depth necessary in planning the drainage for a tidal river in case that the tide gates are closed against the storm surge and the precipitated water has to be artificially drained to the bay by pumping, etc.

View full abstract

To provide an assessmento f the effects of the year-to-yearf luctuationi n precipitationa nd forest growth on the low flow of a headwater watershed, 61-year precipitation and discharge data in the Shirasaka Watershed in Japan were analyzed. The minimum flow in winter tended to be greater than that in summer in the 1930s and 1940s, whereas the minimum flow in summer tended to be greater than that in winter from 1950 to 1990. The precipitation factors, which have a reasonable degree of correlation with the annual low flow, were different in summer and in winter. Vegetation growth has occurred for 60 years in this watershed and the differences in trends between low flow and precipitation factors may be due to the effect of vegetation growth.

View full abstract

There are some historical quantitative data on water level of Lake Biwa in Edo Era. Those are grouped into four types: 1) relative comparison of water level at peak between extreme floods, 2) observation records of water level using scaled posts, 3) inundation depth records of house, garden or rice field and 4) records on daily variation of flood level. In this paper, we picked up and collected historical records on water level of Lake Biwa from old documents, and attempted to represent daily water level data in Biwako Standard Level (B.S.L.) for three historical extreme floods at theend of Edo Era. Moreover, we reconstructed the water level at peak for other historical floods for about 150 years by estimating the standard level of Feudal Zeze Clan.

View full abstract

The inflow into dam reservoir is estimated from the change of hydrostatic water level. It is difficult to obtain the estimation of inflow enough to use for practical dam control using moving average process which is adopted as smoothing method for water level data by almost dam reservoir. The aim of this study is to develop a new estimation method of hydrostatic water level and of inflow into dam reservoir. In this paper, the new smoothing method of water level, which is constituted of the combination of notch filter (NF) and low-pass filter (LPF). In the new method, NF and LPF is used for eliminating a seiche and other random noise, respectively. In the case of Kanayama Dam Reservoir, it would be able to obtain the estimated inflow, in which the power of seiche is cut off enough to use for practical dam control if it is possible to admit 20 minutes time lag of flood peak in the estimated inflow.

View full abstract

It is important for the precaution activity of inhabitants to know how much rain will cause the urban inundation in their city. We discussed in our previous paper the possibility of prediction in case of Ube city using Tank model and point rain prediction method. The Tank structure, Tank parameters and critical occurrence level of residual water in Tank were determined to separate clearly the inundation occurrence cases and the non-occurrence cases in the past rains. We call the critical occurrence level D line. In this paper, we examine how the Tank water level reaches the criterion level (D line) in case of Ube city, inputting the design rain into Tank instead of real 10minite rainfall data.

View full abstract

We have started observations of rainfall amount and water stages at Sabo dams in some devastated basins and a natural basin which was chosen as a basin for comparative studies. This paper points out another role of Sabo dams after describing some characteristics of runoff and sediment yield from devastated basins and a forested basin. The obtained results are as follows:
1) The ratio of runoff/rainfall during June to November of the devastated basin exceeds that of the forested basin. These phenomena may be occurred by high intensity of evapo-transpiration in the forested basin and high storage capacity of water in group of Sabo dams constructed in the devastated basin.
2) The peak specific discharge from the forested basin exceeds that of from the devastated basin for large flood. These phenomena may be also occurred by storage effect (namely, flood regulation effect) of group of Sabo dams constructed in the devastated basin.

View full abstract

A severe flood disaster was occurred at the Yosasa River, a tributary of the Naka River, located in the northeast of Tochigi Prefecture due to heavy rainfall on August 27, 1998. The amount of precipitation of 1, 229 mm was recorded at the Nasu meteorological observatory, the Meteorological Agency, from Aug. 26 to Aug. 31. Owing to the discharge of a large amount of driftwood as well as bank erosion and river bed variation, many houses and bridges were damaged. In this paper, a channel variation of the Yosasa River was discussed by comparing the aerial photos of before and after the disaster. A Flood runoff analysis by using a tank model was done and one dimensional numerical simulation was carried out to examine the characteristics of the river bed variation of the Yosasa River.

View full abstract

The present study addresses the procedure to predict the time dependent of suspended material loads from discharge hydrographs by using of a newly proposed turbidity function. The turbidity, which is relatively easily measured even during big floods, is a convenient index to understand transport mechanisms of suspended material load. The turbidity has different characteristics in the rising and falling limbs of flood hydrographs. This hysteresis effect in the relationship between turbidity and discharge is modeled by a nonlinear function in the present study. On the basis of simulation results, the turbidity function appears to have effectual methods for predicting suspended material loads.

View full abstract

Yosasa-river of northern Totigi located eastern part of Japan suffered the biggest flood disaster by the strong rain fall, August 26, 1998. The natural bank of Yosasa-river has special characteristics for lateral erosion, and therefore, such peerless phenomena occurred, that floodwater seriously broke down the river banks and the width of channel became three to five times as wide as the original width. Moreover, some new channels were formed. On these occasions many sands, gravels and trees were yielded and flowed out.
In this paper, some characteristics of channel changes and their causes in Yosasa-river were discussed basing on the results of field survey.

View full abstract

The concentration of suspended solids was measured at the Mukawa Bridge located 2.6km upstream of the Mukawa River mouth, and at the Hobetsu Bridge located 40.7km upstream of the river mouth. The lateral diffusion of suspended solids was measured by comparing the concentration of suspended solid at the center of river course and shores. Suspended solids are transported toward the center from the shore in the early flood stage and afterward move to the shore from the center. Observations of sediments on the floodplain were conducted before and after flood at the Mukawa River to understand the relationship between suspended solids and sediment on the floodplain. The scouring of the sediment on the floodplain and the deposition of the particles, which have the same size as suspended solids on the floodplain, was verified. From these investigations, it is shown that suspended solids and the sediment are interchanged during floods.

View full abstract

Fine sand deposition on a floodplain during the Abukuma river flood in 1998 is studied both observationally and theoretically. It is found from the observation that most of the deposition consists of fine sand of about 0.1 mm, which is commonly refer to washload that is not included in bed material. The observed data shows that the deposition is maximized around 10 m inside of the floodplain along the boundary between the main channel and the floodplain. This phenomenon is found to be well explained by the theory of suspended sediment transport. The predicted value of the magnitude of the deposition agrees well with the observed value with the use of suspended sediment concentration obtained in another observation.

View full abstract

We first analyzed the reason of error accompanied in the CT-type laser concentration meter, to conclude that pulse-like error signals are caused by laser light which passes through the space between fine spray particles of tracer without decay. A technique is proposed to eliminate these error signals by application of wavelet transform.
The wavelet transform technique is further applied to the inverse problem of estimating the concentration field from the information of decay rates of laser light at receivers.

View full abstract

A new super-resolution PIV is proposed. The proposed super-resolution method consists of the standard correlation PIV, the Particle Mask Correlation method (PMC method), and Kalman filter/_χ²-test method (KC method). The present super-resolution PIV is referred as to the “super-resolution KC method”.
Performance of the super resolution KC method is examined by the Monte-Calro simulation. The super-resolution KC method is also applied to the measurement of the two kinds of flow fields, of which images are captured by the Double-Pulse/Single-Image method and Single-Pulse/Double-Image method, respectively.

View full abstract

A CT-type laser concentration meter for measuring 2-dimensional profile by solution of the inverse problem, has been proposed and developed (Hino, et al., 1997, 1998). In these papers, the ‘virtual-load’ method has been applied to solve a system of simultaneous linear equations (A-w=R), and then numerical simulation and laboratory experiments have confirmed the usefulness of the methodby using a stepping motor for scanning laser beams.
In this paper, optical scanning and signal processing systems are improved for a practical high-speed CT concentration meter, also based on the virtual-load method. A prototype of the meter, using laser light sheets of 12 emitter set and photodiode array of 12 receiver set without mechanical moving parts, is developed and confirmed to measure 2D concentration profiles in approximately real time (order of ms).

View full abstract

Large scale particle image velocimetry (LSPIV) is an efficient and powerful tool to measure river surface flow. In LSPIV oblique angled video images taken from a river bank or a bridge are used to measure two-dimensional surface velocity distributions. In the original LSPIV algorithm, water surface is assumed to be in a horizontal plane. This as-sumption holds true for a mild-slope channel, however, it can be a cause of error in the image transformation when a steep flow is viewed from an oblique angle. Hence, in order to generalize the LSPIV, a water surface equation is in-troduced to the image correction algorithm, which allows the image analysis on a tilted surface plane. In addition, a new formula for verifying velocity resolution is derived using an image transformation equation. As an application of LSPIV, large scale recirculating flows between the groins installed in the middle reach of the Nagara River were measured with the aid of non-toxic tracer particles. An improved LSPIV was found to be quite useful for analyzing surface flow with a reasonable accuracy.

View full abstract

Three-dimensional measurements in an open-channel flow were conducted by making use of a 6-beams laser Doppler anemometer (LDA). All three components of instantaneous velocities were firstly measured in the cross section of an open-channel flow. The Reynolds stress term in the vorticity equation was succesefully measured with high accuracy. It was found that the Reynolds-stress term and the generation term in the vorticity equation balance with each other. The diffusion terms in the energy equation (∂(uuv)/∂y and ∂(uuw)/∂z) were also investigated in the cross section of open-channel flow and it was found that the distributions of diffusion terms are complicated.

View full abstract

The continuous measurement method of swface velocity distributions during a flood flow in the river was developed by using Particle Image Velocimetty (PIV), in which video images of the oblique angle recorded at the high-building facing at the river were transformed into non-distoited ones. The development of the digital-video processing makes it possible by the reasonable cost. It was successfully examined in the Yodo River Study. The 2-D distributions of the velocity vectors, the vorticity, the turbulent intensity, and the Reynolds stress at the water surface were obtained. The characteristics of their distributions were disclosed in view of the relationship between the depth, channel geometry, hydraulic structures and channel alignmentand the basic characteristics of twbulence of the flood flow were also discussed in view of the channel geometry.

View full abstract

In this study, an inundation flow model is developed using the unstructured meshes, which can divide the computational area into arbitrary shapes of meshes. This model is applied to the basin which was inundated in 1947 due to the levee breakage in Tone River and the numerical results is compared with those of the Cartesian coordinate. Furthermore, this model is used to analyze the effects of small rivers and continuous dikes in the basin. An advanced inundation flow model which has an advantage of considering the influences of linear structures is developed.

View full abstract

The aim of this study is to estimate influence of sewerage system on overland flood flow. First, in comparison with the experimental results on water level of overland flood flow due to dam break, the accuracy of numerical model for overland flood flow is evaluated and validity of moving critical depth is found out. Next, numerical simulation for three cases (the case without sewerage system, the case with sewerage system, the case with sewerage system except discharge form pomp) is made, and then the characteristics of numerical simulation model considered with sewerage system for overland flood flow is obtained. In the heavy Flooding like this analysis condition, flood water spurt from manhole. From this study, it is found that numerical modelconsidered with sewerage system for overland flood flow is important to construct of evacuation system.

View full abstract

An numerical method for one-dimensional shallow water equations is proposed. The method is based on Flux Difference Splitting (FDS) with Roe's numerical flux. Source and sink terms are up-winded to solve the system of flow equations in a conservation laws with source and sink terms.
The applicability of the method is examined by comparing numerical results with sample problems and existing experimental data that bed slope or bed friction is presented, and with experimental data of one-dimensional dam-break flood wave propagation on dry bed that both bed slope and bed friction are presented.

View full abstract

The mathematical model which can simulate inundation process in urban area due to heavy rainfall has been newly developed. The network model in which streets, rivers andchannels are treated as a channel network is used. The discharge hydrographs obtained by the runoff analysis in the surrounding mountainous area are imposed as the upstream boundary conditions. A simplified sewerage model is also incorporated. The model is applied to Kyoto city area. The computed results show a good agreement with the record of inundation depth distribution in 1935. The flood risk of the present Kyoto city is also discussed.

View full abstract

A mathematical model is newly developed for inundating flow over a region in which a river network is extended. A 1-D mathematical model is applied to rivers, and the model is transformed into characteristics form to obtain a finite difference scheme. A finite leap-frog difference scheme for a horizontally 2-D model is applied to analyze the overland flow and inundation processes on the flood plain. Two schemes of 1-D and 2-D models are combined at several connection points by means of an overflow discharge formulae. The model is applied to the north-western Mekong delta in Vietnam, and characteristics of the inundation flow are discussed.

View full abstract

In a smooth-bed open channel flow of water at low Reynolds number, with flow depth 10 cm and width 50cm, the near bed vortex structure was visualized by injecting fluorescein dye from a slit on the channel bottom, and illuminating the flow by a light sheet tilted upstream of a cross-stream plane by 30°. Viewing from downstream reveals mushroom-shaped patterns of dye, inferred to be counter-rotating pairs of streamwise vortices. Newly appearing mushroom patterns are strongly inclined with respect to the vertical, and seem to alternate between oblique left and oblique right. Analysis of flow in a crossstream plane by PIV tends to confirm many of these data and subjective impressions. Both instantaneous vorticity distributions and two-point correlations of vorticity indicate preferred directions in the cross-stream plane at about 45° from the vertical.

View full abstract

Experimental studies were conducted on turbulent structures in open channel flows with adverse pressure gradients. A laser Doppler anemometer (LDA) was used to obtain velocity data, and we succeeded in acquiring velocity profiles in the viscous-sublayer. Accordingly, the friction velocity was precisely calculated with the ‘viscous-sublayer’ law. Although the friction velocity is important to analyze turbulence characteristics, most studies have not employed this method due to restrictinns of measurement techniqucs and anemometers. Using this accurate friction velocity, this study examined the opinions that other researches have presented about open channel flows with adverse pressure gradients. It was found that von Karman constant is universal as pressure-gradient parameter changes. On the other hand, the integral constant in the log-law becomes smaller than that of uniform flow. The results also showed that the log-wake law correctly represents velocity profiles in the equilibrium region.

View full abstract

Flux-difference splitting scheme of Roe for free-surface flow simulations is combined with the Preissmann slot to simulate flows in a closed conduit wherein the flow may change from free-surface to pressurized flow and vice-versa. The model can simulate conduits with uniform cross-sections of arbitrary shape, with bed slope and bed friction. The model is verified against available experimental data on freesurface-pressurized flow. Thereafter, the model is tested against some exacting sample problems. It is demonstrated that the model yields very reasonable results in all the cases considered. A sensitivity analysis is performed for the size of the slot and useful conclusions are drawn from the study for the simulation of free-surface-pressurized flows.

View full abstract

In order to clarify variation process of large-scale fluvial geomorphology caused by extreme floods, flow characteristics are studied experimentally and numerically in an open channel with both an expansion and a contraction, just like an elliptic plan-form reach. In this research, two large horizontal circulations spread between both side-walls and the main flow, and small scale eddies appear both sides of the main flow and migrate downstream along it. This flow pattern is simulated well by a numerical model based on 2-D shallow water equations. In turn, fluctuations of water surface levels and flow velocities produced by the numerical model are compared with those measured in the experiments and are proved to have some physical significance, such as seiche and/or shear instability.

View full abstract

The behavior of organized motion in the mixing layer between the main-channel and the concave is analyzed using a wavelet transform. Experimental data of the velocity are obtained by a particle image velocimetry. At first, we successfully detects the characteristics of the large-scale vortices along the mixing layer, such as the spatial scale, frequency, advection velocity and occurrence interval by using the continuous wavelet transform of the velocities. Then, these characteristics of the vortices are examined with respect to the difference of concave depths. The joint probability distribution of the scale and the frequency shows that the large-scale vortices are well developed along the mixing layer in the deep concave rather than in the shallow one. These results strongly suggest that the wavelet transform is a useful tool for analyzing the coherent structure of this flow.

View full abstract

This paper presents experimental results carried out in a laboratory channel to investigate the effect of bed liquefaction on local scour development around circular bridge pier under sudden water pressure variation. Many hydraulic structures collapse due to scouring under sudden attack of flood flows or storm waves. During such a natural conditions, the water level rises abnormally above its initial level and then decreases to its original level. This phenomenon creates abrupt change of water pressure variationnear the structures and cause to failure due to scouring. Considering such a real occurrence in the hydraulic engineering field, this paper deals with the distribution of pore water pressure, effective stresses, velocity distribution and local scour around a circular bridge pier. The results show that the sediment bedis liquefied by an increase of excess pore water pressure under sudden water pressure variation and that the equilibrium local scour increases considerably than that of clear-water steady flow at the bridge pier.

View full abstract

Many researchers have investigated the friction factor of stepped channel flows, but general agreement has not been obtained. Because a method for evaluating a supercritical depth of the stepped channel flow might not be established. Further, the hydraulic condition required to form a quasi-uniform flow on the steppedchannel has not been clarified. This paper presents the flow resistance of skimming flows, which is characterized by the formation of the corner eddy for each step. In order to estimate the equilibrium condition of skimming flows, the subcritical depth of the jump immediately below the step channel end has been utilized. Two methods for evaluating a clear water depth of skimming flows have been developed for a wide range of channel slopes. Under a quasi-uniform flow, the friction factor of skimming flows has been clarified systematically.

View full abstract

In the design of bank protection and calculation of flood inundation, it is important to clarify the resistance of large roughness and the structure of flow in the shallow water flow. We have developeda two-dimensionanlu mericalm odel that considerst he shape and arrangement of large roughness. In this paper, measured hydrodynamic forces acting on submersible large roughness and averaged velocity distribution around roughness element are presented. The resistance of shallow water flow over submerged large roughness is examined using the equivalent roughness k_s. Additionallyt, his paper compares one-dimensionanl numerical simulation using the equivalent roughness k_s with two-dimensional numerical simulation.

View full abstract

This paper presents the results of an experimental study of flow behavior around submerged dikes on rigid bed. The studies of flow behavior including water surface profiles and velocity profiles around a single dike with different cases were carried out with steady uniform flow on smooth rigid bed. It is observed that, the disturbance occurred in the water surface profiles and the velocity profiles for a short distance upstream the dike and for a longer distance downstream the dike. The flow patterns around fullysubmerged dike and just-submerged dike had a different phenomenon such as: A rise in the water surface level upstream the dike and a drop in the water surface level downstream the dike. The separation flow zones were different in each case depending on the kind of the dike.

View full abstract

An investigation of side-weirs, used as a diversion structures in rivers and canals, is presented. Most commonly, the side-weirs are aligned parallel to the direction of flow in the parent canal, giving an off-take angle, θ of 90 degrees. However, this study investigates performance of oblique side, weirs, that off-takes at an angle less than 90 degrees. Discharge coefficients for oblique weirs are obtained theoretically and experimentally as functions of off-take angle, θ, Fronde number in the parent channel and weir geometry. It is found that a high value of discharge coefficient can be maintained when θ lies between 60and 80 degrees. A side weir with θ equal to 70 degrees is selected for tiuther study with improved frontal apron and approach wall. It is concluded that the value of discharge coefficient can he further increased for a wide range of high Fronde numbers through these improvements.

View full abstract

First, a storm water runoff simulation model of surcharged flow with masses of air at manholes in urban sewer pipe systems is proposed. In this model, the compressibility of the air is considered.
Second, the model is applied to runoff experiments in a hydraulic model of storm sewer pipes with manholes, and the air-pressure hydrographs simulated by the simulation model are compared to the observed ones.
As a result, it is demonstrated that the simulation model proposed here can simulate very well the temporal variations in air-pressure at manholes.

View full abstract

The Gaussian solution for the advection equation in open channel flow was modified with displacements caused by the secondary flow introduced. The solution was compared with experimental data and was in good agreement. There are two concentration peaks for a shallow compound channel flow and the solution reproduces it nicely. A stream function was introduced to predict secondary flow. The solution with the stream function was found to be in good agreement with the experimental data.

View full abstract

This paper concerns a multiblock-parallel computation technique for three-dimensional turbulent flows, which is advantageous to deal with complicated boundary shapes and to improve computational efficiency as well. A computational domain is decomposed into multiple sub-blocks and their geometries are represented by curvilinear coordinates. The multiple blocks are connected on their surfaces withoutoverlap, so that the grids near the shared surfaces can be easily generated even when the three-dimensional connected regions have complex geometries. The spatial interpolation for some of the variables are performed with a cubic spline function, which prevents the first-order numerical error arising between sub-blocks. As a result of the parallel computations for laminar flows with a moving wall and pressure gradients, it was confirmed that the computational speed is about twice when using four workstations with 130, 000 computational nodes.

View full abstract

In the stratified closed water systems such as bays, lakes, and river mouths flowing into sea, the unique phenomena caused by the density differences are observed. Numerical simulation is useful for investigating such phenomena and three-dimentional numerical models are required when it have occured under the effects of changeable wind direction and complicated ground shape. In this paper, a three-dimensional numerical model which employs LES and CIP methods is proposed for the purpose of calculating incompressible density current. The model is closs-checked by the results of an experimental density flow in a flume, in which the movement of density front and internal seiche can be observed. The applicability of such a process is confirmed by the calculated results of miglation of density front, vortexes behind it, diffusion of inner density boundary, and period of internal seiche.

View full abstract

A Lagrangian solver of the Navier-Stokes equation, or the MPS method, is extended by installing the sub-particle-scale (SPS) turbulence model. The momentum equation for the particle-scale (PS) flow is derived through the filtering operation of the Navier-Stokes equation. The effect of the SPS turbulence is found as the Reynolds stress terms in the PS momentum equation. The Reynolds stress terms are described by introducing the Smagorinsky model, which is frequently used in Eulerian LES. The MPS method has the advantage in the calculation of the free-surface flow against the Eulerian one, because it is free from the numerical diffusion due to the discretization of advection terms. While, in this paper, the MPS method with the SPS turbulence model is applied to the turbulent mixing layer, or the jet, to show its performance on the simulation of the turbulent flow with a simple structure. The mean velocity profile is compared with the previous experiments. The unsteady behavior of the jet and the mixing process at the interface of the jet and the still water are displayed as the snapshots of the particle motion.

View full abstract

LES simulations have been performed using different finite difference schemes derived by a general procedure for discretising the convective terms to evaluate their importance in the results of thesimulation. Two typical subgrid-scale stress models have also been examined. The flow past a square cylinder is taken as a benchmark test case of general flow around a bluff body. The code was initially validated at different low Reynolds number and then calculation results are compared with experimental data. Dynamic Smagorinsky model is seen to improve the solution marginally. Upwind-biased scheme of appropriately high order has been found to give satisfactory results.

View full abstract

It is important to consider the secondary currents induced by the stream line curvature to predict the river channel processes. In view of the refinements of the model proposed by Kalkwijk·De Vriend, the depth averaged 2-D flow equations including the development and attenuation of secondary currents along a stream line, which are considered by Ikeda et al. and Parker et al. analytically, are studied in the generalized curvilinear coordinate system. The depth averaged equation of the strength of secondary currents along a stream line is derived by using the Engelund model, and then is transformed into the curvilinear coordinates. It is shown that the experiments by Rozovskii can be predicted well by using the model equations

View full abstract

A Dimet Numerical Simulation method using a regular grid under a generalized curvilinear coordinate system is presented which can simulate flows in complex geometry. In this paper, at first a one-dimensional linear convection problem was dealt to test the accuracy and the results strongly suggested that the way of defining a grid with certain types of unequal spacing cause substantial phase rrors in numerical solutions and the cause is the metric discontinuity. Next, a regular grid under a generalized coordinate system was used to analyze free surface turbulent flow with Reynolds number 150 defined by friction velocity. It was shown that this method enabled sufficiently stable calculation. In addition, a comparison with DNS of open channel flow using a staggered grid verified that this simulation gave fairy satisfactory properties with respect to higher-degree statistic values related to turbulent characteristics

View full abstract