Heavy trucks are mostly used for international transportations, with longer highways and long driving hours contributing to corresponding increases in the driver's fatigue that is related to accidents. Therefore, this study aims to improve the truck ride performance using multistage leaf springs and semi-active suspension for the driver seat. This analytical study describes the influence of the truck main suspensions on the performance indices analytically using MATLAB Simulink for different loading conditions in three case studies: fully laden articulated truck (case A), unladen truck (case B), and empty semi-trailer and a multistage leaf springs is considered after designing the main leaf spring stiffness based on particle swarm optimization (case C). This study exhibits a contribution based on the fact that changing the trailer cargo weight has considerable effects on the natural frequency of the vibration modes of the vehicle system, particularly for articulated carriage. Subsequently, the influence of the dynamic interaction of an articulated vehicle between the semitrailer and the tractor on its ride behavior has been investigated. The model has also predicted the effect of total trailer cargo on performance indices for 13 degrees of freedom model of a 6-axle articulated truck semi-trailer vehicle with a random road excitation. Additionally, a semi-active driver seat suspension based on skyhook strategy and seat passive suspension are compared in terms of the power spectral density and root mean square values. The results showed that the truck ride performance is improved significantly, and all the acceleration responses are suppressed dramatically when a designed multistage leaf spring suspension is considered in case C. The current analysis demonstrated that using specific and adjustable suspension parameters can positively enhance the riding behavior of the unladen vehicle. The results showed that the cab, tractor, and trailer acceleration improved by 22%, 21%, and 28%, respectively, which provides a comfort driving trip essentially for long distance traveling.

This study highlights the frictional behavior of the disc brake interfaces and microstructure of the friction layers (FL) and transferred materials (TM) on the tribopairs. The designed dynamometer was utilized to evaluate the frictional performance of the disc brake materials during dry sliding under different braking events (applied loads and speeds). The morphological properties of the brake lining and disc surfaces were examined by EPMA, FESEM, EDS, 3D surface profilometry and Raman spectroscopy to illustrate the FL, TM and wear mechanisms. The tribo-braking results exhibited that the friction coefficient (FC) increased with increasing the applied load. This is due to the fact that, at high applied loads, the actual contacting area of the friction pair is increased resulting in the growth of the secondary contact plateaus and flattening of primary contact plateaus that help the enrichment of the FL. Whilst, the initial stop-braking speeds did not hugely affect the FC. Eventually, the formation and destruction of the FL showed a significant role in controlling the brake friction behavior. The formation of the FL increases the frictional stabilization.

Misfire in spark-ignition engines is one of the major faults that affect the power produced by the engine and pollute the environment and may cause further engine damage. This paper presents an evaluation of an artificial neural network based performance system through three most popular training algorithms namely Gradient Descent, Lavenberg-Marquadt and Quasi-Newton to determine the misfire location. Misfire is simulated by removing ignition coil to that cylinder namely Cylinder 1,2,3,4 and Cylinders 1 and 2, 1 and 4 and 2 and 3 with three different conditions such as idle, 2000 rpm and 3000 rpm. The results showed that the Quasi-Newton is higher in recognition rate average of 98.19 % but it takes more time to train. The Lavenberg-Marquardt algorithm is also good with an average recognition rate of 96.09 % with the fastest performance than Quasi-Newton. The gradient descent algorithm requires the network size to be more complicated to perform well with least time and high recognition rate.

No abstract

No abstract