
Diaphragmless shock tubes have emerged as a promising alternative to conventional diaphragm-based systems, offering contamination-free operation, minimal manual intervention, and quick turnaround time between tests. This study investigates the shock formation and propagation dynamics in a diaphragmless shock tube actuated by a high-speed pneumatic valve. Experiments in a high-pressure diaphragmless shock tube with helium as the driver and argon as the test gas provided time-resolved valve motion data, capturing stroke length and opening speed of up to 28 m/s. The measured valve dynamics were incorporated into axisymmetric CFD simulations, which replicated the experimentally observed shock velocity profiles. Comparisons with conventional diaphragm-driven configuration simulations show that the diaphragmless mode requires up to ∼ 40 % higher driver pressure to achieve the same shock Mach number. The simulations further confirm that the gradual valve opening produces an extended shock acceleration region that results in a longer shock formation length and a shorter test gas length. Notably, the diaphragmless setup yielded lower peak temperatures in the post-shock region, which is potentially advantageous for reducing remote ignition effects in chemical kinetics experiments. To the knowledge of the authors, this study presents the first combined experimental and numerical framework that captures realistic valve dynamics and the full flow evolution in a diaphragmless shock tube, laying a foundation for accurate modeling and reliable operation of next-generation shock tube facilities.
The previously developed universal method for measuring non-stationary force and moment vectors using six-component strain-gauge balances by A.R. Gorbushin, A.E. Kozik, and E.N. Anokhina (Measurement 226, 114176 (2024)) is improved. The original method relies on a physical model, in which damped springs are substituted for sensitive elements of each component of the balance. The problem is solved for the general case, when the balance coordinate system is non-inertial—it moves with acceleration along six coordinates, namely, three linear and three angular ones. Initially, the method took the acceleration of the earth part of the balance into account. The improved method is based upon the same physical representation of the sensitive elements; however, the derived mathematical model is altered—it now considers the acceleration of the metric part of the balance instead of the earth part’s acceleration. The applicability of both methods has been compared in the case of drag measurement of a sharp 5^∘ half-angle cone in a Ludwieg tube at zero angle of attack. In the conducted tube run, the implemented modification reduced the dynamic oscillations that corrupt the measured mean drag coefficient by 2.7 times compared to the results obtained through the initial method.
Following the perforation of a target plate by an explosively formed projectile (EFP), the behind-armor debris (BAD) generated at the rear poses a severe threat to personnel and internal equipment. To accurately quantify this threat, the mass distribution of the debris is considered to be a critical indicator for evaluating the lethality of the BAD generated by the EFP. This paper establishes a neural network model designed to rapidly predict BAD mass distribution under various projectile–target interaction conditions. Flash X-ray radiography and a sandbox recovery system were employed to investigate debris characteristics, combined with a numerical simulation using the coupled FEM–SPH method. The simulated results showed strong agreement with experiments, with the error in debris cloud head expansion velocity consistently below 10 R^2 values exceeding 0.96 and MAPE remaining below 5
Terrain strongly affects blast wave propagation, but Mach stem evolution over sloped ground remains insufficiently understood. This study used numerical simulations, supported by experimental validation, to investigate TNT blast wave propagation over upward slopes, with emphasis on triple-point trajectories and the peak overpressure of ground-reflected waves. When the slope angle was sufficiently large, the wave reflected from the slope toe interacted with the existing Mach stem, causing it to detach from the slope surface and propagate above it. When Mach reflection was developed during uphill propagation, increasing the slope angle shifted the onset of Mach reflection farther downstream and reduced the triple-point height at a given horizontal distance. The peak overpressure shape factor exhibited an oscillatory variation closely associated with Mach stem evolution. A conventional artificial neural network and a physics-informed machine learning-aided framework were developed to predict the peak overpressure of ground-reflected waves under slope conditions. PIMLAF successfully captured the abrupt overpressure increase near the slope toe, outperformed the conventional ANN, and achieved a test set MAPE of 0.63
One-dimensional analysis shows that a 100-nm ITO film affixed to a MACOR substrate matches the thermal response of a 1.4- m platinum film–MACOR thin-film gauge typically used in impulse facilities. The analysis also provides an estimate of the required thickness for an insulating substrate such as MACOR .