Full-field Distributions Research Articles

Full-field expansion and damage detection from sparse measurements using physics-informed variational autoencoders

Recent advancements in machine learning and artificial neural network algorithms have created new opportunities for expanding sparse measurements to full-field representations. However, these methods often face accuracy challenges when used for condition monitoring of physical systems. For example, traditional autoencoders can capture complex phenomena and their underlying physics in the latent space, but due to their frequentist nature, they lack generative and damage detection capabilities. To address these challenges, a novel physics-informed variational autoencoder (PI-VAE) network is proposed for expanding sparse measurements to full-field representations while also detecting damage. The effectiveness of the proposed PI-VAE network is evaluated through analytical and experimental studies on a metal plate under thermal excitation with embedded defects of various sizes and types. In the analytical studies using finite-element model data, the PI-VAE accurately expanded full-field temperature distributions and identified the dimensions of cracks, spalling, and hole-like defects with errors smaller than 5%. When tested with experimental data, the PI-VAE network maintained robust performance, detecting damage with errors smaller than 6%, despite being trained on undamaged data only. These findings demonstrate the PI-VAE’s potential as a reliable tool for full-field expansion and damage detection in structural health monitoring and nondestructive evaluation, even when limited sensors and datasets are available.

Full-field temperature prediction in tunnel fires using limited monitored ceiling flow temperature data with transformer-based deep learning models

In practical tunnel scenarios, full-field coverage of sensors is impractical and costly. During a tunnel fire, the available information is constrained and localized, making the prediction of full-field smoke temperature distribution becoming a noteworthy challenge. This study proposes a transformer-based deep learning model to predict full-field smoke temperature distributions during fire incidents in real-time using limited temporal data from the sensors installed in localized regions below the ceiling, considering heat release rate of the fire source is unknown. The results indicate that proposed approach can predict the longitudinal temperature distribution throughout the tunnel with a length of 750 m by leveraging temperature data from limited sensors within a monitoring length of 210 m. It can further predict the vertical temperature profiles, and eventually estimate the full-field temperature distribution within the tunnel. The transformer model achieved R2 of 0.95 and 0.87 for longitudinal and vertical temperature distribution predictions, respectively. Under the influence of the self-attention mechanism, the transformer model has an advantage over the long short-term memory model in capturing global information, enhancing the accuracy of longitudinal temperature distribution predictions by 18.8 %. This study significantly contributes to effective emergency response and rescue strategies during tunnel fire incidents.

New high sensitivity approach for a quantitative visualization of the continuous whole-field stress evolution in digital photoelasticity

Digital photoelasticity is a widely used technique for experimental stress analysis on mechanical models subjected to loading conditions, due to its ability to reveal the stress information through color fringe patterns and consequently demodulate it computationally to achieve a quantitative evaluation. In this process, some limitations persist, such as, expertise dependence, experimental repeatability, sensitivity to load increments, and weak performance under experimental conditions, especially in time-dependent scenarios. In response to these weaknesses, the method “Quantitative Visualization of the Continuous Whole-Field Stress Evolution” introduced the ability to perform a complete demodulation of the fringe patterns generated in time-dependent applications, achieving high performance compared to conventional photoelasticity methods. This paper proposes a new highly sensitive approach for a quantitative visualization of the continuous whole-field stress evolution in digital photoelasticity by incorporating an inverse signal process and a multi-stage strategy. The algorithm aims to ensure the performance under high fringe order conditions while exhibiting high sensitivity in cases where the maximum stress obtained is less than one fringe order. The results presented validate its effectiveness in demodulating full-field stress distributions, especially in scenarios where the loading condition induces fringe orders less than one, thus contributing to the broader applicability of digital photoelasticity.

Full-field measurement of residual stress in single-crystal diamond substrates based on Mueller matrix microscopy

In this work, a method has been proposed for fast-measuring full-field residual stress of diamond substrates based on Mueller Matrix (MM) microscopy. The measurement system features a polarization camera and a fixed compensator at 45° orientation, and real-time birefringence imaging can be performed. The magnitude and orientation of residual stress have been characterized by the stress-optical law. Optical-grade synthetic diamond substrates after laser-cutting and grinding have been measured. After laser-cutting samples exhibit a central region where stress orientation is shifted by −90° from the x-axis. After double-sided fine grinding, the average stress value is relatively low at 0.094 GPa, while single-sided fine grinding yields average stress of 0.142 Gpa, whereas double-sided rough grinding leads to 0.157 Gpa. Full-field stress distributions obtained from Raman images show consistent patterns. The research indicates that the MM-based method is a promising technique for detecting diamond substrate birefringence and characterizing its stress variation during fabrication.

Aerothermal performance improvement by cross and X ribbed stripes

The incentive to save energy in heat exchangers has led to enhanced thermal performance in ducted flows, achieved via inserts designed to increase the heat transfer efficiency and efficacy. We introduce a novel approach using cross and X stripe inserts, roughened by 90° and 45° ribs, to enhance the structural integrity and heat transfer rate of the duct. Additionally, the incorporation of edged notches aims to improve aerothermal performance by directing flow traction away from the recirculating cell behind a rib, thereby reducing the drag. Utilizing the infrared thermography method, this research quantifies the full-field Nusselt number distributions for 34 enhanced ducts at Reynolds numbers 5,000–25,000. The outcomes of flow simulations elucidate the underlying mechanisms of aerothermal performance enhancement afforded by these inserts. These enhancements result from the combined effects of increased turbulence intensity, fluid mixing, and rib-induced multicellular vortices, which improve the synergy between fluid velocity and temperature gradient vectors, thereby optimizing heat transfer. Quantitatively, the Nusselt numbers (friction factors) for ducts with cross or X ribbed stripes increase significantly. For cross stripes, the ranges increase by 1.25–3.17 times (3.14–21.75 for friction factors), and for X stripes, 1.29–3.39 times (6.58–43.41 for friction factors), compared with the Dittus–Boelter (Blasius) standards. Notably, the aerothermal performance index for ducts featuring notched cross stripes with 45° ribs and a rib pitch-to-height ratio of 10 reaches 1.47 to 1.11, though this decreases with increasing Reynolds number. Furthermore, the study presents empirical correlations to assess the average Nusselt number and friction factor in a square duct equipped with either cross or X ribbed stripes.

Revisiting hemodynamics and blood oxygenation in a microfluidic microvasculature replica

The complexity of microvascular circulation has led to the development of advanced imaging techniques and biomimetic models. This study developed a multifaceted microfluidic-based microdevice as an in vitro model of microvasculature to replicate important geometric and functional features of in vivo perfusion in mice. The microfluidic device consisted of a microchannel for blood perfusion, mirroring the natural hierarchical branching vascular structures found in mice. Additionally, the device incorporated a steady gradient of oxygen (O2) which diffused through the polydimethylsiloxane (PDMS) layer, allowing for dynamic blood oxygenation. The assembled multi-layered microdevice was accompanied by a dual-modal imaging system that combined laser speckle contrast imaging (LSCI) and intrinsic signal optical imaging (ISOI) to visualize full-field blood flow distributions and blood O2 profiles. By closely reproducing in vivo blood perfusion and oxygenation conditions, this microvasculature model, in conjunction with numerical simulation results, can provide quantitative information on physiologically relevant hemodynamics and key O2 transport parameters that are not directly measurable in traditional animal studies.

An electrochemical-mechanical coupled multi-scale modeling method and full-field stress distribution of lithium-ion battery

The rapid development of lithium-ion batteries has made the market more and more concerned about their lives. A large number of studies have shown that the alternating diffusion-induced stress caused by the cyclic charging and discharging processes can significantly shorten the service life of a battery. In recent years, scientists have tried to reveal the distributions and evolutions of the internal stress in lithium-ion batteries during operation by means of experiments and simulations, so as to propose guidance for their design and application. However, the existing numerical simulation framework is still limited to the particles, electrodes and battery units, rather than a battery cell. It is urgent to develop a method that can obtain the full-field stress distributions in commercial lithium-ion battery cells during working, so as to reveal the influence of stress on their lives, and further improve the efficiency of design and optimization. In this paper, an electrochemical-mechanical coupled multi-scale modeling method for lithium-ion batteries is proposed, which solves the technical problem of cross-scaled modeling and simulation from battery units to battery cells. Taking a pouch lithium-ion cell as an example, the lithium concentration distribution, potential distribution, and current density distribution of the battery unit, as well as the full-field displacement distribution, strain distribution and stress distribution of the battery cell during charging are obtained and analyzed. And the assumption, adaptability and potential applications of the method are discussed. It provides a new technical means and insight into the properties of lithium-ion batteries, and even solid-state batteries in the future.

Implementation and comparison of algebraic and machine learning based tensor interpolation methods applied to fiber orientation tensor fields obtained from CT images

Fiber orientation tensors (FOT) are used as a compact form of representing the mechanically important quantity of fiber orientation in fiber reinforced composites. While they can be obtained via image processing methods from micro computed tomography scans (μCT), the specimen size needs to be sufficiently small for adequate resolution – especially in the case of carbon fibers. In order to avoid massive workload by scans and image evaluation when determining full-field FOT distributions for a plaque or a part, e.g., for comparison with process simulations, the possibilities of a direct interpolation of a few measured FOT at specific support points were opened in this paper. Hence, three different tensor interpolation methods were implemented and compared qualitatively with the help of visualization through tensor glyphs and quantitatively by calculating originally measured tensors at support points and evaluating the deviations. The methods compared in this work include two algebraic approaches, firstly, a Euclidean component averaging and secondly, a decomposition approach based on separate invariant and quaternion weighting, as well as an artificial intelligence (AI)-based method using an artificial neural network (ANN). While the decomposition method showed the best results visually, quantitatively the component averaging method and the neural network behaved better (that is for the type of quantitative error assessment used in this paper) with mean absolute errors of 0.105 and 0.114 when calculating previously measured tensors and comparing the components. With each method providing different advantages, the use for further application as well as necessary improvement is discussed. The authors would like to highlight the novelty of the methods being used with small and CT-based tensor datasets.

Compressive anisotropy effects and crack evolution mechanisms of 3D5d braided composites induced by thermo-oxidative ageing

Ageing cracks and loading directions will affect mechanical behaviors of thermo-oxidative aged 3D braided composites. Here we report ageing crack propagations, compressive responses, in-situ strain evolutions, full-field damage distributions and crack evolution mechanisms of aged three-dimensional five-directional (3D5d) braided composites along out-of-plane (OP) and in-plane (IP) loading based on X-ray micro-computed tomography and digital image correlation technique. Ageing cracks are distributed at both axial ends that corresponding to the yarn arrangement and extend to the interior with the increase in ageing time. Compressive modulus, strength, retention rate and damage mode show ageing weakening behaviors and anisotropy effects. Strain distributions exhibit structure correlation. Ageing cracks affect the surface strain and damage distributions but not internal shear modes during OP loading. Yarn arrangement affects damage evolution and accumulation. Ageing cracks change the damage evolution path in both directions. The synergistic effect of ageing crack with IP loading is greater than that with OP loading. It is expected that this work can provide a clear guideline for the structural optimization and durability design of 3D braided composites.

A hybrid experimental-computational study: Prediction of flow fields and full-field pressure distributions on measured shapes of three-tab asphalt roofing shingles subjected to hurricane velocity winds

A novel hybrid experimental-computational study is performed to predict the flow fields and pressure distributions on the measured three-dimensional shapes of flexible, three-tab asphalt roofing shingles undergoing increasing uplift when exposed to hurricane velocity winds for two hours. To quantify the evolution of shingle shapes, StereoDIC analysis is used to measure the transient, full-field deformed shapes of full-sized, three-tab asphalt shingles that did not show separation or failure when subjected to hurricane velocity winds for two hours. Based on physical observations during wind loading, the authors performed steady state computational fluid dynamics (CFD) simulations to predict the full-field pressure distributions on as-measured, uplifted three-dimensional shingle shapes at selected time instances during wind loading.Simulation predictions clearly show flow recirculation regions on both the front and top of the shingles that remain attached throughout wind loading and control the full-field uplift pressure distribution. For low velocity flow with maximum uplift ≤ 8.4 mm, CFD-predicted pressures are in good agreement with prior measurements. For both low and high-speed flows, the model predictions indicate that high pressures are formed at the leading-edge, upstream of the sealant layer, with maximum pressure occurring near the tab cutouts along the leading-edge of the shingle, providing a physical basis for the observed higher uplift and increased potential for shingle failure in these regions. The combined experimental-computational studies provide a contemporary way to eliminate the difficulties associated with attachment of pressure sensors to flexible materials that can alter shingle response, providing the basis for future design improvements by delineating the physical processes controlling pressure loading and shingle uplift in hurricane velocity winds.

A Hybrid Experimental-Computational Study: Prediction of Flow Fields and Full-Field Pressure Distributions on Measured Shapes of Three-Tab Asphalt Roofing Shingles Subjected to Hurricane Velocity Winds

A Hybrid Experimental-Computational Study: Prediction of Flow Fields and Full-Field Pressure Distributions on Measured Shapes of Three-Tab Asphalt Roofing Shingles Subjected to Hurricane Velocity Winds

A simplified modelling approach for thermal behaviour analysis in hybrid plasma arc-laser additive manufacturing

Hybrid plasma transferred arc (PTA)-laser additive manufacturing (AM) has the potential to build large-scale metal components with high deposition rate and near-net shape. However, the process is complex with many parameters adjustable for process control, which determine the thermal behaviour and thus the final structure and properties of the deposited components. In this study, a three-dimensional steady-state finite element model with two independent circular surface heat sources was developed, validated, and used to analyse the thermal behaviour in hybrid PTA-laser AM of Ti-6Al-4V. Artificial conductivity in three orthogonal directions was applied in the melt pool to compensate for the melt pool convection effect. The predicted melt pool geometry, heat-affected zone and thermal cycles had good agreement with the corresponding experimental data. This model has advantages over the widely used volumetric heat source model, since it is more representative of the energy sources used, giving accurate thermal prediction for a wide range of process parameters. As the heat source parameters in this model are directly linked to the actual arc/laser size, it enables to capture heat source size effect on the hybrid process. In addition, it is easier to calibrate compared to the model with volumetric heat sources due to the fewer empirical parameters involved. It was found that in the investigated ranges of all the parameters, the melt pool geometry is more sensitive to laser power and travel speed compared to arc-laser separation distance and laser beam size. The full-field distributions of the cooling rate and temperature gradient in the hybrid process were obtained and the roles that different process parameters played on them were also studied, which provided useful thermal information for metallurgical analysis.

Intrinsic stress determination based on the combination of photoelasticity and digital gradient sensing

A compact stress measurement system for transparent objects was proposed based on the combination of photoelascity and DGS methods. The interferometry fringe and speckle image of the deformed specimen were captured from perpendicular and slant directions, respectively. The fundamentals of the stress calculation and imaging model of oblique camera viewing were analyzed mathematically. After phase-shifting computation, image correlation, and geometric transformation, full-field distributions of the major and minor intrinsic principal stresses were obtained simultaneously. A validation experiment was conducted on a circular disk under diametral compression by the proposed method. Comparison between the experimental resulsts and theoretical data calculated according to the theory of elasticity verified the feasibility and accuracy. Moreover, the combined method was also successfully applied for the intrinsic principal stresses determination on transparent wrinkling thin film with defects subjected to shearing forces.

Dynamic full-field refractive index distribution measurements using total internal reflection terahertz digital holography

Massive usage scenarios prompt the prosperity of terahertz refractive index (THz RI) measurement methods. However, they are very difficult in measuring the full-field dynamical RI distributions of either solid samples without a priori thickness or liquid samples. In this study, we propose total internal reflection THz digital holography and apply it for measuring RI distributions for both solid and liquid samples dynamically. An RI measurement model is established based on an attenuated total reflection prism with a pitching angle. The pitching angle and the field of view can be numerically calculated from the spectrogram of the off-axis Fresnel hologram, which solves the adjustment of the visually opaque prism irradiated by the invisible THz beam. Full-field RI distributions of the droplets of solid-state soy wax and distilled water are obtained and compared with THz time-domain spectroscopy. The evaporation of an ethanol solution droplet is recorded, and the variation of the RI distribution at the sample–prism interface is quantitatively visualized with a temporal resolution of 10 Hz. The proposed method greatly expands the sample range for THz RI measurements and provides unprecedented insight into investigating spontaneous and dynamic THz phenomena.

Transport of intensity equation-based terahertz lensless full-field phase imaging.

Terahertz (THz) phase imaging is widely spreading in various scenarios, among which full-field phase distributions are commonly retrieved by digital holography or ptychography. In this Letter, the transport of the intensity equation reconstruction method is applied into the THz band. An algorithm named the lensless US-transport of intensity equation (TIE) is proposed to accommodate to an in-line configuration. The object phase is retrieved by primarily conducting iterations between the axial intensity derivative and the phase distribution at the recording plane and subsequent backward diffraction propagation. This method is applicable to both isolated and extended weakly absorbing samples with higher reconstruction quality and remarkably less time cost than holographic phase retrieval algorithms. It can also be attempted in other non-interferometric geometries or using low-cost partially coherent THz sources, which significantly broaden the application scope of THz phase imaging.

Optical method for identification and quantification of full-field stress distributions and evolution in assembled lining structures based on additively printed models and phase-shifting methods

The accurate knowledge and characterization of the full-field stress distributions and evolution in assembled lining structures underlie the design, construction, and maintenance of tunnel linings. However, it is challenging for traditional field monitoring and laboratory experiments to identify directly and quantify accurately such distributions and evolution. Numerical simulations, as alternative solutions, are limited, as their results are significantly dependent on the accuracy of mechanical parameters, constitutive laws, failure criteria, mesh optimization, and boundary conditions, which are difficult to determine experimentally. In this study, a plane-stress-model-based photoelastic testing method was developed for the direct observation and quantification of the full-field stress distribution and evolution in an assembled lining structure based on additive manufacturing or three-dimensional printing technology and phase shifting methods. Three-dimensional printing techniques and stress- sensitive photopolymers were employed to fabricate transparent assembled models based on the information of real lining structures. The model split-merge technique, phase-shifting method, and unwrapping algorithm were combined to determine the full-field distribution of the principal stress difference and shear stress within the assembled segments. The proposed method was verified based on a comparison of the calculated and measured fringe orders at different loading stages. The experimental results indicated that the largest stress concentration was around the joint that connected the adjacent Segment L1 and basic Segment B1, which formed the potential critical failure zone of the assembled structure. An approximate X-shaped shear band was also identified. Comparisons of the results verified that the boundary and contact conditions, which are difficult to simulate, can lead to significantly different simulation results with respect to the distribution and evolution of the principal stress difference and shear stress. This study shows that the proposed method can effectively and quantitatively characterize the full-field stress distributions and evolution within complex assembled lining structures.

Identifying the coefficients of thermal expansion of an electronic device composed of various materials using the virtual fields method

A method for inverse determination of the coefficients of thermal expansion of an electronic device that is composed of multiple materials, i.e. dissimilar materials, is proposed. The virtual fields method is applied for obtaining the coefficients of thermal expansion. Appropriate virtual displacement fields have been automatically obtained instead of using any arbitrary virtual displacement. This leads to an accurate and effective determination of the coefficient of thermal expansion. Full-field displacement distributions measured by digital image correlation are used as input for the inverse analysis. The validity of the proposed method has been demonstrated by measuring samples made of various materials. Furthermore, the proposed method is applied to samples simulating electronic devices with and without constraints on the boundary. Results show that the coefficients of thermal expansion of an object composed of dissimilar materials are obtained properly. The proposed method is expected to be used for determining the coefficients of thermal expansion of actual electronic devices.

The surface effects on solid and hollow nanowires under diametral loading

The size effects of energetic surfaces on the elastic behavior of solid and hollow nanowires are investigated under diametral loading for the first time. In order to model the diametral compression test on nanowires, the structural version of the locally-exact theory is developed for solid and hollow nanowires (LETNW) by considering energetic surfaces attached at the boundaries of bulk materials and modeled using the Gurtin-Murdoch model. The developed analytical solutions employ Fourier series representations of displacement fields in the cylindrical coordinate system that exactly satisfy the governing differential equations, facilitating rapid convergence of full-field stress distributions. The solution's accuracy is validated against the in-house finite-element (FE) program which is currently the computational standard, where excellent agreement is obtained. The surface effects on the local stresses are then illustrated by varying the surface parameters, loading angle, dimension of nanowires, as well as residual surface stress. The efficiency of the LETNW is attributed to its analytical nature and thus consumes much less execution time relative to the FE implementation. What's more, instabilities associated with the solution of stress fields are observed with surface properties in certain crystallographic orientations for both LETNW and FE approaches. Contrary to the LETNW technique, the bifurcation phenomenon is shown to depend on the finite-element mesh refinement. The locally-exact theory provides an alternative and independent way of identifying surface bifurcations in solid and hollow nanowires and hence is a good tool in assessing the finite-element method as well as other approaches’ predictive capability for this class of materials.

Experimental heat transfer and flow simulations of rectangular channel with twisted-tape pin-fin array

In this study, an innovative twisted-tape pin-fin array was proposed for passive heat transfer enhancement (HTE) to improve the aerothermal performance of channel flows. The full-field Nusselt number distributions, Fanning friction coefficients and thermal performance factors of an improved channel were measured at Reynolds numbers of 5000, 7500, 10,000, 12,500, and 15,000. The turbulent channel flows were numerically explored using the ANSYS-Fluent code to perform correlative analysis between the aerothermal measurements and numerical flow results. A set of aerothermal data and numerical flow results was selected to illustrate the properties of HTE and the responsive flow mechanisms. Because of the vortical flows induced by each twisted tape in the array, the average measured Nusselt numbers and Fanning friction coefficients of the improved channel were 5.26–4.73 and 39.4–38.5 times higher than those of a plain tube, respectively; thus, the thermal performance factor (TPF) of the improved channel was 1.55–1.39 when 5000 <Re< 15,000. Compared with the pin-fin channels with different geometries reported in the literature, the twisted-tape pin-fin channel designed in this study provided a higher degree of HTE and thus a higher TPF at the same pressure drop penalty and pumping power. On the basis of the experimental results, two empirical correlations that were used to evaluate the average endwall Nusselt number and Fanning friction coefficient of the designed pin-fin channel were proposed for relevant applications.

Thermal performance of channel flow with detached and attached pin-fins of hybrid shapes under inlet flow pulsation

A newly devised hybrid pin-fin array composed by the attached rhombuses and the hexagon pin-fins was proposed for heat transfer enhancements. The flow pulsation at channel entrance was stimulated by venting the airflow periodically at the frequencies of 1/4, 1/6 and 1/12 Hz with the time-mean Reynolds number, Re, controlled at 1500, 2000, 5000, 10,000, 15,000 and 20,000 to give rise the Strouhal number, Sr, in the range of 0.00028–0.017. At each Re and Sr tested, the open period of vented flow was adjusted to 1/4, 1/2 or 3/4 pulsation cycle to alter the pulsating waveform. A selective set of full-field Nusselt number distributions and Fanning friction factors illustrated the general and individual hydrothermal impacts in association with Re, Sr and the frequency and waveform of the pulsating flow. As Sr increased continuously, the area-averaged time-mean Nusselt numbers were initially reduced but recovered and then raised above than the steady-flow references. The Fanning friction factors of pulsating flows were initially reduced and then increased from the steady-flow baselines as Sr increased. With the significant heat transfer elevations by the present pin-fin array from the smooth tubular flow references, the thermal performance factors at Re≥5000 were above unity. Two sets of empirical correlations were devised for evaluating the time-mean area-averaged Nusselt numbers and Fanning friction factors of the present pin-fin channel with steady and pulsating flows.