AC arc faults are one of the most important causes of residential electrical wiring fires, which may produce extremely high temperatures and easily ignite surrounding combustible materials. The global interest in machine learning-based methods for arc fault diagnosis applications is increasing due to continuous challenges in efficiency and accuracy. In this paper, a temporal domain visualization convolutional neural network (TDV-CNN) methodology is proposed. The current transformer and high-speed data acquisition system are used to collect the current of a series of arc faults, then the signal is filtered by a digital filter and converted into a gray image in time sequence before being fed into TDV-CNN. Five different electric loads were selected for experimental validation with various signal characteristics, including vacuum cleaner, fluorescent lamp, dimmer, heater, and desktop computer. The experimental results confirm that the classification accuracy of the five loads’ work states in the ten categories could reach 98.7% or even higher by adjusting parameters perfectly. The methodology is believed to be reliable for series arc detection with relatively high accuracy and also has important potential applications in other fault diagnosis fields.

Both experimental and theoretical methods were proposed to assess the effects of adjacent, parallel, and symmetric exterior wall structures on the combustion and flame spreading characteristics of rigid polyurethane (PUR) foam insulation. During the combustion of PUR specimens, the flame leading edge was found to transfer from a unique inverted ‘W’ shape to an inverted ‘V’ during flame propagation. This phenomenon is attributed to edge effects related to boundary layer theory. The effects of the adjacent façade angle on flame spreading rate and flame height were shown to be nonlinear, as a result of the combined influences of heat transfer, radiation angle, and the chimney restriction effects. A critical angle around 90 degree with maximum thermal hazards outwards by parallel fire was observed and consistent with the mass loss rate and flame height tendencies. For narrow spacing configurations or angles (e.g., 60 and 90 degrees), phenomenological two-pass processing in conjunction showed that increased preheating lengths were associated with enhanced heat transfer. The results of this study have implications concerning the design of safe façade structures for high-rise buildings, and provide a better understanding of downward flame spreading over PUR.

The characteristics of a series direct current (DC) arc-fault including both electrical and thermal parameters were investigated based on an arc-fault simulator to provide references for multi-parameter electrical fire detection method. Tests on arc fault behavior with three different initial circuit voltages, resistances and arc gaps were conducted, respectively. The influences of circuit conditions on arc dynamic image, voltage, current or power were interpreted. Also, the temperature rises of electrode surface and ambient air were studied. The results showed that, first, significant variations of arc structure and light emitting were observed under different conditions. A thin outer burning layer of vapor generated from electrodes with orange light was found due to the extremely high arc temperature. Second, with the increasing electrode gap in discharging, the arc power was shown to have a non monotonic relationship with arc length for constant initial circuit voltage and resistance. Finally, the temperature rises of electrode surface caused by heat transfer from arc were found to be not sensitive with increasing arc length due to special heat transfer mechanism. In addition, temperature of ambient air showed a large gradient in radial direction of arc.

Concerning the issues regarding driving mileage reduction for electric vehicles (EVs) in cold climates, a heat pump system with low global warming potential refrigerant R290/R1234yf is employed as one of the promising solutions. Different from the widely used mobile refrigerant R134a, R290 and R1234yf are both flammable or explosive. The application of R290/R1234yf in the mobile heat pump system is hindered by unexpected refrigerant leakage with the existence of fire and explosion risk. In this study, the combustion characteristics of R290/R1234yf in a potential leakage process from an air-conditioning heat pump system for EVs were investigated. Firstly, thermodynamic behaviours of R290/R1234yf used in a typical heat pump system were analysed based on a special experimental facility designed for EVs. Then the leakage and combustion characteristics of R290/R1234yf including flame shape, temperature, radiation etc. were obtained by the experimental method under different initial temperature and mass flow rate conditions. It was found that R290/R1234yf leaked is difficult to ignite at low temperatures, while the blow-off phenomenon of the jet flame would occur at high temperature with high leakage mass flow rate. In addition, the results showed that combustion intensity would be enhanced by the leakage mass flow rate between 30 and 60°C. These results could provide guidance for fire detection and rescue system design for new energy vehicles.

TDLAS (tunable diode laser absorption spectroscopy) is an important gas analysis method that can be employed to obtain characteristic parameters non-invasively by the infrared absorption spectra of tracer molecules such as CH₄, H₂O and O₂. In this study, a portable H₂O-based TDLAS system with a dual optical path was developed with the aim of assessing the combustion characteristics of flammable gases. Firstly, a calculation method of gas characteristics including temperature and velocity combining absorption spectra and a HITRAN database was provided. Secondly, to calibrate and validate this TDLAS system precisely, a pressure vessel and a shock tube were introduced innovatively to generate static or steady flow fields with preset constant temperatures, pressures, or velocities. Static tests within environment pressures up to 2 MPa and steady flow field tests with temperatures up to 1600 K and flow velocities up to 950 m/s were performed for verification. It was proved that this system can provide an accurate values for high temperature and velocity gas flows. Finally, an experimental investigation of CH₄/air flames was conducted to test the effectiveness of the system when applied to small diffusion flames. This TDLAS system gave satisfactory flame temperature and velocity data owing to the dual optical path design and high frequency scanning, which compensated for scale effects and pulsation of the flame. This work demonstrates a valuable new approach to thermal hazard analysis in specific environments.