Seyed Hosseini

Microstructure characterization of white layer formed by hard turning and wire electric discharge machining in high carbon steel (AISI 52100)

White layers, formed during hard turning and wire electric discharge machining, were characterized by means of X-ray diffraction, scanning electron microscopy and transmission electron microscopy. Different cutting speeds and flank wear were utilized in order to obtain different thermally and/or plastically deformed white layer. Since the white layer after wire electric discharge machining is mainly thermally induced, it was used as a reference structure. In the investigation, both bainitic and martensitic structures were studied. With a constant flank wear of 0.175 mm the thickness of the white layer increased from 1.5 μm to 3 μm as the cutting speed was increased. In both processes the white layer were characterized by nanosized grains and surface tensile residual stresses. M3C carbides were observed in the hard turned white layer, indicating that the time and temperature needed for completely dissolving the carbides were not reached during cutting. For both materials the white layers formed by wire electric discharge machining consisted of ~ 30 vol. % of retained austenite. Observation regarding the volume fraction of the retained austenite in the white layer formed by hard turning for martensitic material showed an increase in the volume fraction of retained austenite from ~ 2 - 3 vol. % to ~ 6 vol. %, while this observation was not seen in the white layer formed in the bainitic material.

Adaptive forward collision warning algorithm for automotive applications

This paper proposes an adaptive collision warning algorithm (CWA) that supports the driver by issuing early warnings for collision avoidance without noticeably increasing the risk of false alarms in real traffic. This algorithm can also detect when an emergency intervention is necessary. Compared to existing CWAs, the proposed solution in this paper triggers alarms by solving a linear convex program using traffic data, roads constraints and bicycle model dynamics, and by incorporating an adaptive (speed-dependent) warning threshold. A collision threat is detected by determining feasible steering trajectories without altering the vehicles longitudinal velocity.