Automotive Science and Engineering

Experimental and Numerical Investigation of the Effects of Linear and Nonlinear Modeling on the Performance of a Polymer Ball Dual Mass Flywheel in Reducing Engine Torque Fluctuations

Saeed Mahjoub Moghadas

In the automotive industry, reducing torsional vibrations caused by combustion irregularities and internal inertial forces is essential for improving powertrain quality, NVH (Noise, Vibration, and Harshness), and component durability. This study introduces and analyzes an innovative dual mass flywheel (DMF) with polymer bearings, aimed at enhancing torsional damping and ensuring smoother torque transmission in six-cylinder gasoline engines. The DMF consists of two masses, low-stiffness springs, and polymer bearings, all thoroughly modeled. Both linear and nonlinear dynamic analyses were performed, incorporating factors such as centrifugal force, nonlinear friction, and viscous damping. Component parameters were obtained via modal tests and experimental measurements. To validate the model, cylinder pressure and torque data were collected using sensors and a dynamometer, then compared with simulation results. Results reveal that employing the DMF reduces output torque fluctuations by up to 89.9% and significantly increases torsional vibration damping without a notable weight increase. Compared with a single-mass flywheel of equal inertia, the DMF decreased torque fluctuation amplitude by 46%, 41%, and 38% at 2250, 3000, and 3750 rpm, respectively. Achieving similar damping with a single-mass flywheel would require over an 80% mass increase. The nonlinear model provided a close match to experimental data, with simulation error below 5%. Overall, this integrated approach demonstrates that the dual mass flywheel with polymer bearings improves dynamic engine performance, NVH, and powertrain durability. These findings support the development of lighter, more advanced powertrain systems for next-generation vehicles.

Improving Defect Detection in GDXray Castings via Inverse Problem-Based Deep Learning

Mansour Baghaeian

In metal casting, detecting defects like pores and cracks in X-ray images is crucial for product quality and safety. This study presents an advanced U-Net architecture for semantic segmentation of defects in the GDXray dataset, achieving superior accuracy. By formulating defect detection as an inverse problem reconstructing material density from X-ray projections the method integrates transfer learning, data augmentation, and Convolutional Block Attention Modules (CBAM) to address low contrast-to-noise ratios and limited data. Pretrained on synthetic Radon transform projections, the U-Net, enhanced with CBAM, sharpens focus on defect regions, improving boundary precision by 5%. Data augmentation, including rotations, flips, and noise injection, generates 5,000 synthetic images to overcome data scarcity. Experiments on 2,727 grayscale GDXray images demonstrate a mean Intersection over :union: (mIoU) of 0.85, a 15% improvement over baseline U-Net models, with 97.8% accuracy for pores and 94.5% for cracks. The inverse problem approach reduces false negatives by 12%, excelling in noisy conditions. Compared to methods like Mask R-CNN, this approach advances non-destructive evaluation (NDE) for casting applications, ensuring reliability and safety. Validated on laboratory X-ray data, the model offers a scalable solution for industrial defect detection. Future work will optimize computational efficiency and explore multi-modal data to enhance robustness.

Lateral Motion Control of In-Wheel Motor EVs via Optimal Energy-Efficient Control Allocation

Majid Majidi

In this paper, a multi-level hierarchical control method for enhancing electric vehicle (EV) stability with four independent in-wheel motors is proposed. In the high-level motion controller, a sliding mode controller is used to calculate the total desired force and yaw moment, and in the low-level control allocation, an optimal energy-efficient control allocation scheme is presented to provide optimally distributed torques for four in-wheel motors. Moreover, both handling performance and energy savings are investigated in this research and evaluated via a co-simulation approach using MATLAB/Simulink, and CarSim. With a torque distribution algorithm based on energy efficiency optimization, the EV is controlled with and without a controller in J-turn and lane change maneuvers. The simulation results show that the proposed torque control system and torque distribution algorithm can maintain stability, reduce energy consumption, and track the desired values of yaw rate and longitudinal velocity of the vehicle in the mentioned maneuvers.

Thermo-elastic stress assessment of exhaust manifold outlet dimensional deviations in a heavy-duty diesel engine

Mohammad Parhizkar Yaghoobi

Generally, parts whose geometric form in the final configuration, or before additional operations such as machining, does not require special dimensional accuracy are produced using casting methods. Producing parts with this method results in significant deviations in dimensions and geometric forms from the main designed geometric model. Due to economic considerations, scrapping such parts will result in energy and material waste and high costs. In this study, dimensional deviations at the exhaust manifold outlet and deviations from the defined geometric tolerance limits of the part during the repair process are identified as geometric non-conformities and investigated using computer tools. Considering the harsh operating conditions under which the engine is under full load, the temperature distribution is determined using computational fluid dynamics, and the thermo-elastic stress distribution is calculated using the finite element method, accounting for the loads applied to the structure. The main part model and the non-conforming models with upper and lower limits in geometric dimensions have been investigated with respect to the deviation in thermo-elastic stress relative to the reference value in the part with the nominal size. The results showed that, given the amount of stress changes in the desired area of parts with deviations from the main design, these parts are also usable and have a lifetime almost the same as a part produced with the nominal size.

A Numerical Investigation of Inter-Vehicle Spacing on Aerodynamic Drag Using a Two-Dimensional Ahmed Body Model

Amirhasan Kakaee

This study employs two-dimensional CFD simulations to analyze how rear slant angle and inter-vehicle spacing dictate aerodynamic drag in a tandem Ahmed body configuration. We systematically evaluated slant angles from 15° to 45° and longitudinal spacings from X/L = 0.1 to 0.5. The results delineate three distinct aerodynamic regimes for the trailing vehicle: a drafting zone at close distances (X/L=0.1) with significantly reduced drag, an interference zone (X/L=0.2-0.3) where drag peaks, and an independence zone (X/L>0.4) where vehicles behave aerodynamically isolated. Furthermore, the model successfully captures the critical drag rise as the slant angle surpasses 30°, a key flow transition. While the simulation over-predicts absolute drag values, which is an expected outcome of the 2D approach, it demonstrates high fidelity in capturing complex trends, providing foundational insights for optimizing vehicle platooning strategies.