In the electronic controlled 4WD system, front and rear wheels are driven, respectively, directly and through an electronic controlled coupling by engine power. It realizes low fuel consumption and high vehicle dynamics performance keeping good fuel consumption for high performance vehicle. The new system consists of electronic controlled 4WD coupling and the electronic controlled limited slip differential adapted to the front axle. The improvement of running through and turning performances were confirmed by actual vehicle tests.

View full abstract

With exhaust control reinforcement, the performance enhancement of the air fuel ratio sensor is expected as one of the emission reduction methods. In this study, to respond to the future engine control system needs, early activation, highly precise, and reliable for the air fuel ratio sensor with the next generation sensor was developed, aiming to implement from 10MY US vehicles.

View full abstract

The project discussed in this paper focused on the frictional characteristics of clutch facing materials in order to elucidate the mechanism by which the torque capacity of the multiple plate clutches used in automatic transmissions (AT) varies with mode of use. The frictional characteristics of the facing materials are affected by the temperature, contact pressure, and sliding velocity of the friction surfaces. The temperature of the clutch plate, sliding velocity of the friction surfaces, and degree of misalignment of the gears transmitting torque were therefore measured in an AT during torque transmission. Next, a torque capacity prediction model, able to simulate the flow of torque transmission in the AT, was constructed on the basis of these measurement results. An examination of the friction force vectors of the clutch disks showed that the torque capacity of an AT operating at a constant temperature is affected by radial components of the friction force vectors on the friction surfaces. In addition, this model enable torque capacity in an operating AT to be predicted for the first time with an accuracy to within ±5% of measured values.

View full abstract

The relation between a plastic distortion and a thermal fatigue life was statistically analyzed using Weibull distribution and a least square method. As a result, in Weibull distribution, when there is much examination data, a probability life S-N diagram can be created, but in a least-squares method, the S-N diagram which is reliable with almost all the analyzed quality of the materials can be created. This value is in agreement with the 50% probability life line of Weibull distribution.

View full abstract

Natural wind influence on fuel consumption, f_c , for highway vehicles is studied for real wind modeling. The increase in fuel consumption, β_f, is related to wind speed, U, through the aerodynamic influence coefficient on f_c, K_A, which is tentatively proposed for a class of small vehicles (passenger cars) and a class of large vehicles (trucks), respectively. A head wind of U = 3 m/s, nearly equal to yearly averaged wind speed, results in a 7 - 16 % increase in f_c for a small vihecle (100km/h (27.8m/s)) and a 9 - 14% increase for a large vehicle (80km/h (22.2m/s)).

View full abstract

Currently, various studies on pedestrian protection are conducted aiming at the improvement, especially, of the reduction of head injury criteria (HIC). In order to make pedestrian dummy collision behavior more uniform, engineers adopt that the arms of dummies must be tied. In this study, the influence of the arms on pedestrian head injury at the time of vehicle collision was evaluated. It was examined that the position of the impact-side arm of the pedestrian dummy was particularly apt to influence HIC.

View full abstract

In this study, the dynamic crushing characteristics of Front Side Member are studied by using the FEM. It is found that in low velocity the initial deformation of the Front Side Member mainly caused by the bend in plastic hinge. However, in high velocity the front portion before the plastic hinge is initially crushed. Thus the first peak load increases with increasing the collision velocity as an effect of the mass of the portion before the plastic hinge. Also, a Spring Model is proposed to predict the load behavior of the Front Side Member.

View full abstract

A numerical model for predicting the spot weld fracture is developed using the maximum principal stress and the equivalent plastic strain as fracture criteria. This model is applied to crash analyses of hat shaped automotive members of ultra high strength steel. The computed results of energy absorption are very close to experimental measurements both in axial crash test and in 3-point bending test. The influence of the component configurations, such as nugget diameter and spot pitch, on the energy absorption is evaluated, and it is clarified that the different material combination is effective in the fracture prevention of spot weld.

View full abstract

A novel combustion model for large eddy simulation (LES) of gasoline engines is developed. Unlike the conventional models based on Reynolds averaged Navier-Stokes (RANS) models, the new model is originally described by means of fractal characteristics of flame front and universal flame speed expression in the subgrid scale (SGS). The validity of this fractal combustion model is examined through the experimental combustion period and cyclic variations. The computational results agree well with the experimental data.

View full abstract

Homogeneous Charge Compression Ignition (HCCI) combustion has attracted considerable interest as a new combustion concept for internal combustion engines. The HCCI combustion process has three issues to be resolved: ignition timing control, slower combustion, expanding stable operation region. In this study, an attempt was made to overcome these issues by EGR and dual fuels. Light absorption spectroscopic measurement was made in the combustion chamber in order to investigate low-temperature reactions in detail. As a result, the ignition timing can be controlled and the rapidity of combustion can be lessened by using a fuel having two components with different ignition characteristics.

View full abstract

Real-time estimation of soot mass loading on diesel particulate filters (DPF) is essential to assure perfect and fuel-efficient regeneration. Computer models based on the pressure drop across the DPF are installed in the engine control unit to estimate the soot mass loading, but this approach is still subject to uncertainty for the lack of resources available. This study is intended to develop a portable instrument that can make real-time measurements of soot mass loading under transient mode for calibrating the models. The instrument is based on the light scattering technique and the intensity of scattered light from soot particles is reduced to soot mass concentration using a calibration data obtained under steady engine operating conditions. It is demonstrated that the prototype instrument can measure the temporal variations of cumulative soot mass and filtration efficiency of a DPF under transient modes.

View full abstract

The trapping and regeneration of particulate matter (PM) viewed on the surface of diesel particulate catalyst-membrane filters (DPMFs) were investigated in detail using an all-in-focus optical microscope. To improve filtration efficiency, a SiC-DPMF with K-, Na-, and Cs-zeolite catalyst membranes was developed and successfully evaluated. Visualization results showed that trapped particulates were homogeneously distributed along the fine surface pores of the membrane's top surface. In the regeneration process, the particulates in contact with the DPMF might be oxidized first by the catalytic effect of zeolite particles and simultaneously pushed through by gas flow, resulting in continuous oxidation from the bottom of the soot cake. The activation energy of particulates on zeolite catalyst membranes was about 24 kJ/mol or 20% lower than that on a conventional SiC-DPF without catalyst.

View full abstract

This study tried to improve ignitability of a lean-combustion CNG engine. CNG was directly injected into the combustion chamber by a gas injector and the gas-jet was ignited by a spark plug. The combination of gas-jet combustion and two-stage injection; namely, first injection at early combustion stroke and second injection just before ignition, has a great effect to realize lean-combustion at equivalence ratio ranges from 0.3 to 0.9 with high thermal efficiency and less CO emission. It is important for two-stage injection and gas-jet combustion system to control over-lean mixture distribution caused by the first injection for reduction of THC emission, and also to avoid excess-rich mixture formation near spark plug that results in incomplete combustion.

View full abstract

The aim of this study is to experimentally clarify the effect of biodiesel-fuel properties on diesel combustion and exhaust emission characteristics. Three kinds of fatty acid methyl esters are blended to make test fuels of which the ignition quality, volatility are the same as of gas oil. The engine bench test results show that the initial heat-release due to premix combustion is suppressed by using FAME with low volatility compared to gas oil, even when they have same ignition qualities. Also, it is observed that gaseous hydrocarbons and soluble organic fraction in PM in the exhaust are closely related to volatility of FAME.

View full abstract

In order to know the strain distribution of suction valves of a piston compressor for automobile air-conditioner, a new thin-film strain sensor was developed. The sensor is approximately 5 micro-meter thick, and was deposited directly into the valve surface by sputter deposition. An original alloy which is not affected by the temperature and pressure change was used for the sensing part. Furthermore, the original form of the sensing part was developed to reduce its transverse sensitivity. This paper presents a newly developed thin-film strain sensor and measurement results of the strain distribution on a suction value at different rotation speed.

View full abstract

An original road sign recognition algorithm using Levenshtein distance was introduced in 2007 by the authors. A new recognition algorithm combined Levenshtein distance with the least squares method enables to improve the recognition ratio is proposed. While the original algorithm could not detect a road sign hidden by trees, the new algorithm based on a shape judgment can find the road sign more reliable. The recognition ratio has been improved by 20% compared with the original algorithm. The new algorithm is implemented on Xilinx Spartan-3 using Bach-C at 49.8 MHz and achieved 17.0 ms execution time.

View full abstract

In this paper, we propose a reference driving maneuver model for the evaluation of lane change behavior by using driving simulator, and we also employ the reference model to clarify unskilled driving behavior. This research aims to development on advanced driver assistance system and an evaluation system which assist drivers in improving their skills in lane change maneuver. In the driving data collection, the lane change scenarios have been reconstructed based on the actual driving data on a highway. The reference driving maneuver is synthesized to evaluate the driving behavior, and compared with driving data in near-miss cases and collision cases. Human errors resulting in failure of lane change were described, and their driving behaviors were quantitatively evaluated.

View full abstract

This paper addresses an analysis of driving behavior in right turn behavior from the collected driving database in urban area by using continuous-sensing drive recorder. The driving behavior evaluation from the viewpoint of safety is systematically conducted. The right turn behavior database is constructed to evaluate normal driving behavior and clarify relevant factors of human error at intersection. The pedestrian and bicycle cognitive behavior of individual driver during right turn at a fixed intersection is focused. The pedestrian/bicycle collision threat assessment in right turn behavior is conducted based on the predictive value of relative distance between the host vehicle and other moving objects at the intersection.

View full abstract

About half of vehicle-pedestrian accidents occur on non-crosswalks, such as when a pedestrian runs into the street. In this paper, using data obtained from driving recorders, we analyze and classify the pedestrian behavior when drivers experience near-accidents with pedestrians on non-crosswalks. We then model the pedestrian behavior for incorporation into our traffic simulator STREET which evaluates safety systems.

View full abstract

In this study, we developed comfortable sitting posture for assist or rear passengers. We specified the basic sitting posture parameters that greatly influenced passenger's comfort by using Design of Experiments method. The results showed leg posture and leg supporting conditions were important factors. The best comfortable leg posture and leg supporting conditions were validated through subjective evaluations.

View full abstract

The algorithm of inferring driver's intention of "restart" operation, resuming vehicle motion after a pause at an intersection with a stop sign, was developed, in order to apply to the crossing collision prevention system. To infer with insignificant delay, the state transition model of driving behavior around an intersection with a stop sign was utilized. The algorithm can estimate restart intention by following the states from the initial to the restart using driver operations and vehicle motions. The performance of the algorithm was tested against driving data of subjects in public roads. The test results are reported, with the description of the algorithm.

View full abstract

This paper investigated the impact of deregulating the presence of stop signs prior to entering railway crossings. To this end, we evaluated the probability that a driver would stop inside a railway crossing, thereby obstructing the tracks, when a lead vehicle suddenly stops after railway crossing and a stop sign is absent. We estimated the probability based on the driving behavior data collected in a driving simulator and on the field, which were similar to the evaluation methods described in the 1st report. Thirty drivers participated in the simulator experiment and were instructed to drive at very slow speeds (10, 20km/h) and at slow and middle speeds (30, 50km/h). We also investigated the influence of following two leading vehicles on the driver's vehicle movements. The results indicated that slower driving speed led to a higher probability of stopping inside the railway crossing, and the probability was higher when the vehicle in front of the leading vehicle did not slow down when entering the railway crossing.

View full abstract

In this study, the safety and comfort level of the personal mobility vehicles (PMVs) in the pedestrian flows are evaluated. In the experiment, we used the conventional bicycle, the small wheel bicycle, the compact dynamically stabilized parallel two-wheel vehicle and so on in the mock pedestrian flows. From the questionnaires, the evade difficulty of the pedestrians, the discomfort level etc. were evaluated by the drivers and the pedestrians in order to discuss the safety and comfort of the PMVs. It was shown that the compact dynamically stabilized vehicle had higher safety and comfort level in the crowded pedestrian flows.

View full abstract

The effects of fuel properties on injector deposits were investigated with a Direct Injection Spark Ignition (DISI) engine under the stoichiometric condition. The injected fuel flow ratio was decreased as the lower temperature of 90% distillation and the higher aromatic contents. However, the decrease of fuel flow caused by injector deposits was small. The test fuels properties are within the range of commercial gasoline on the Japanese market. The effects of 7% ETBE blending on the decrease fuel flow ratio could be negligible. The detergent additive on the market was effective for removing injector deposits of DISI engine.

View full abstract

We have been proposing to apply visible light communication to ITS vehicle-to-vehicle communication, using LED lights on vehicles as the transmitters and high-speed cameras on other vehicles as the receiver. Visible light communication is a wireless communication method using luminance, transmitting data by emitting light. Detection of the transmitter is important for conducting the communication. In this paper, we proposed the algorithm to detect transmitter on the moving vehicle and report some experiments.

View full abstract