Physics-based Validation Research Articles

Validating hybrid fire testing with full-physical twin experiments

Real fire incidents show that structural systems have specific fire-induced redundancies that make them perform better than classical structural fire analysis predicts. Properly validated hybrid fire testing has the potential to tap into such redundancies for designs. In this study, we conceived and implemented a comprehensive and novel physics-based validation strategy for hybrid fire testing. We developed a laboratory-scale thermo-mechanical proof-of-concept problem that is representative for structural systems exposed to compartment fires. We solved the proof-of-concept problem using our novel thermo-mechanical hybrid solution procedure (Schulthesset al.2020Comput. Struct.238, 106301 (doi:10.1016/j.compstruc.2020.106301)) and compared the resulting hybrid model response against the results from a full-physical laboratory set-up of the same proof-of-concept problem, which we call a full-physical twin experiment. Full-physical twin experiments constitute the top layer of a suite of benchmark tests any specific hybrid fire testing procedure must pass to qualify not only as theoretically accurate and correctly functioning but also (and more importantly) as physically sound and able to yield realistic simulation results. This paper delivers such a first-of-its-kind physics-based validation, and in doing so, the first comprehensively validated method for hybrid fire testing—a method that now allows fire test facilities to conduct meaningful full-scale hybrid fire tests of entire structural systems.

Reconstruction of Missing Gas, Oil, and Water Flow-Rate Data: A Unified Physics and Data-Based Approach

Summary An incomplete data set of flow rate and pressure is detrimental to reservoir management and operation. It has the potential to increase uncertainty and has the potential to unfavorably affect operational and managerial decisions. Such a data set might transpire because of failure in the flowmeters, pressure gauges, and/or unrecorded shut-in periods. This study proposes and evaluates unified physics and data-based analytics for “learning” the underlying behavior of a reservoir and reconstructing missing gas, oil, and water flow rates. The proposed workflow is evaluated using real field data obtained from a North Sea reservoir. Validation is done by using a whiteness test, a goodness-of-fit test, and a novel physics-based validation using material balance and pressure back-calculation. The outcome has shown the capability and flexibility of the selected machine-learning techniques in estimating the missing flow rate on the basis of pressure responses. The features that are extracted and expanded on the basis of physics have resulted in a high-fidelity model with less computation time.

3D Design Using Generative Adversarial Networks and Physics-Based Validation

Abstract The authors present a generative adversarial network (GAN) model that demonstrates how to generate 3D models in their native format so that they can be either evaluated using complex simulation environments or realized using methods such as additive manufacturing. Once initially trained, the GAN can create additional training data itself by generating new designs, evaluating them in a physics-based virtual environment, and adding the high performing ones to the training set. A case study involving a GAN model that is initially trained on 4045 3D aircraft models is used for demonstration, where a training data set that has been updated with GAN-generated and evaluated designs results in enhanced model generation, in both the geometric feasibility and performance of the designs. Z-tests on the performance scores of the generated aircraft models indicate a statistically significant improvement in the functionality of the generated models after three iterations of the training-evaluation process. In the case study, a number of techniques are explored to structure the generate-evaluate process in order to balance the need to generate feasible designs with the need for innovative designs.

Contact resistance extraction methods for short- and long-channel carbon nanotube field-effect transistors

Three different methods for the extraction of the contact resistance based on both the well-known transfer length method (TLM) and two variants of the Y-function method have been applied to simulation and experimental data of short- and long-channel CNTFETs. While for TLM special CNT test structures are mandatory, standard electrical device characteristics are sufficient for the Y-function methods. The methods have been applied to CNTFETs with low and high channel resistance. It turned out that the standard Y-function method fails to deliver the correct contact resistance in case of a relatively high channel resistance compared to the contact resistances. A physics-based validation is also given for the application of these methods based on applying traditional Si MOSFET theory to quasi-ballistic CNTFETs.

Recognizing jump patterns with physics-based validation in human moving trajectory

This paper presents an approach to recognize jump patterns in human moving trajectory, differentiating jump tracks from planar moving tracks. Since human moving trajectory is one of the most informative representations for content understanding and event detection, trajectory-based video analysis has been gaining popularity. However, a jump action typically leads to violent change in human moving trajectory, since the person suddenly leaves the original plane on which he/she has been moving. The abnormal tracks of the trajectory would influence the performance of trajectory-based video analysis. Hence, differentiating jump tracks from planar moving tracks is of vital importance, not to mention that jump actions typically imply significant events, especially in sports games. In this paper, volleyball videos are used as case study to demonstrate the effectiveness of our proposed jump pattern recognition approach. We derive player trajectory by head tracking, analyze the movement of each player, and recognize potential jump tracks in player trajectories based on two important characteristics: (1) jumps cause pulse-like tracks in the trajectory and (2) the extensions of such tracks go through the vanishing point of vertical lines in the scenes. Finally, the jump positions/heights are estimated, in addition to the planar moving trajectory of each player on the court ground. The experiments show that satisfactory results can be obtained with the proposed recognition scheme.

Direct numerical simulation of interfacial instabilities: A consistent, conservative, all-speed, sharp-interface method

We present a conservative and consistent numerical method for solving the Navier–Stokes equations in flow domains that may be separated by any number of material interfaces, at arbitrarily-high density/viscosity ratios and acoustic-impedance mismatches, subjected to strong shock waves and flow speeds that can range from highly supersonic to near-zero Mach numbers. A principal aim is prediction of interfacial instabilities under superposition of multiple potentially-active modes (Rayleigh–Taylor, Kelvin–Helmholtz, Richtmyer–Meshkov) as found for example with shock-driven, immersed fluid bodies (locally oblique shocks)—accordingly we emphasize fidelity supported by physics-based validation, including experiments. Consistency is achieved by satisfying the jump discontinuities at the interface within a conservative 2nd-order scheme that is coupled, in a conservative manner, to the bulk-fluid motions. The jump conditions are embedded into a Riemann problem, solved exactly to provide the pressures and velocities along the interface, which is tracked by a level set function to accuracy of O(Δx5,Δt4). Subgrid representation of the interface is achieved by allowing curvature of its constituent interfacial elements to obtain O(Δx3) accuracy in cut-cell volume, with attendant benefits in calculating cell- geometric features and interface curvature (O(Δx3)). Overall the computation converges at near-theoretical O(Δx2). Spurious-currents are down to machine error and there is no time-step restriction due to surface tension. Our method is built upon a quadtree-like adaptive mesh refinement infrastructure. When necessary, this is supplemented by body-fitted grids to enhance resolution of the gas dynamics, including flow separation, shear layers, slip lines, and critical layers. Comprehensive comparisons with exact solutions for the linearized Rayleigh–Taylor and Kelvin–Helmholtz problems demonstrate excellent performance. Sample simulations of liquid drops subjected to shock waves demonstrate for the first time ab initio numerical prediction of the key interfacial features and phenomena found in recent experimental and theoretical studies of this class of problems [T.G. Theofanous, Aerobreakup of Newtonian and viscoelastic liquids, Ann. Rev. Fluid Mech. 43 (2011) 661–690.].