Predicted Temperature Profiles Research Articles

Artificial Intelligence Applied to Microwave Heating Systems: Prediction of Temperature Profile through Convolutional Neural Networks

Microwave heating, which is caused by the interaction of electromagnetic radiation and materials, has become an important component in industrial operations across numerous industries. Despite their importance, conventional numerical simulations of microwave heating are computationally intensive. Concurrently, advances in artificial intelligence (AI), particularly machine learning algorithms, have transformed data processing by increasing accuracy while decreasing computational time. This study tackles the difficulty of efficient and accurate modelling in microwave heating by combining convolutional neural networks (CNNs) with traditional simulation techniques. The major goal of this research is to use CNNs to forecast temperature profiles in a variety of industrial materials, including susceptors, semi-transparent, and microwave-transparent materials, under varying power settings and heating periods. This unique strategy greatly reduces prediction times, with up to 60-fold speed increases over standard methods. Our research is based on examining the electromagnetic and thermal responses of these materials under microwave heating. This study’s findings emphasise the need for extensive datasets and show the transformational potential of CNNs in optimising material processing. It uses artificial intelligence to pave the way for more effective and exact simulations, supporting breakthroughs in industrial microwave heating applications.

Optimum Learning Model for Temperature Profile Prediction in Additive Manufacturing Process

Optimum Learning Model for Temperature Profile Prediction in Additive Manufacturing Process

Modelling the Response of Timber Beams Under Fire

AbstractA fundamental requirement for analysing timber structures under fire is to consider the degradation of material properties with temperature. Therefore, the objective of this study is to propose a model that accounts for the variation of the thermo-physical properties, the development of char, and its evolution with temperature. This model integrates a sequential coupling of heat transfer analysis with structural response. The degradation of the material properties is accounted for through the regulatory approach recommended in Eurocode 5. The stress analysis employs an elasto-plastic model with nonlinear isotropic hardening. Implementation of the model is achieved within the Abaqus suite of finite element software using external subroutines. The model's predictions align well with experimental data, accurately reproducing both thermal and structural responses. Specifically, the model accurately predicts temperature profiles, displacements, and the depth of the charred layer, which initiates above 300 °C. Additionally, for rectangular sections, it was observed that exposure of all faces to fire results in a non-rectangular residual section. Furthermore, employing the temperature-dependent thermal property curves suggested by EC5 yields satisfactory results when predicting the fire resistance of softwood timber structures.

Developing reference-based correlations for temperature distribution in VVER reactor using gene expression programming and single-heated channel approach

Temperature distribution in different channels of a light-water nuclear reactor is essential for Multiphysics coupling approaches in core calculation. Machine learning algorithms trained by well-structured datasets can be applied for the prediction of temperature profiles in different components of the core including coolant, clad and fuel. In this study, gene expression programming (GEP) as an evolutionary machine learning technique is used for developing proper nondimensionalized correlations to estimate the temperature profiles in the VVER-1000 reactor. The single-heated channel approach based on solving mass, energy, and momentum equations is used as a mathematical model for generating datasets to train the GEP. To determine the most influential parameters on the temperature profiles a two-step sensitivity analysis including Pearson correlation coefficient and one-at-a-time sensitivity measure is performed. According to Wilks’ statistical technique, 59 samples are generated to uniformly distribute the uncertainties in the input parameters. By considering 80 % of the acquired data as a training dataset, three reference-based correlations for coolant, clad and center of fuel are developed. Mean Square Error (MSE) and coefficient of determination are used to measure the error of each approach. The high predictability (i.e., R2 > 0.95) presented in the temperature profiles indicates that the GEP could correlate input parameters successfully. The calculations for relative error show that at some axial directions, the maximum relative error can get to 5 %, while the average relative error is 0.21, 0.23, and 1.71 % respectively for coolant, clad, and fuel temperature profiles.

AFSD-Nets: A Physics-Informed Machine Learning Model for Predicting the Temperature Evolution During Additive Friction Stir Deposition

Abstract This study models the temperature evolution during additive friction stir deposition (AFSD) using machine learning. AFSD is a solid-state additive manufacturing technology that deposits metal using plastic flow without melting. However, the ability to predict its performance using the underlying physics is in the early stage. A physics-informed machine learning approach, AFSD-Nets, is presented here to predict temperature profiles based on the combined effects of heat generation and heat transfer. The proposed AFSD-Nets includes a set of customized neural network approximators, which are used to model the coupled temperature evolution for the tool and build during multi-layer material deposition. Experiments are designed and performed using 7075 aluminum feedstock deposited on a substrate of the same material for 30 layers. A comparison of predictions and measurements shows that the proposed AFSD-Nets approach can accurately describe and predict the temperature evolution during the AFSD process.

Experimental and numerical analysis of temperature distributions in SA 387 pressure vessel steel during submerged arc welding

Abstract The present study aims to explore experimental investigations and numerical simulations for temperature distributions at heat-affected zones within SA 387-Gr.11-Cl.2 steel during the submerged arc welding (SAW) process. Experimental endeavors entailed welding steel plates under controlled conditions, precisely measuring temperatures at key locations by thermocouples. A special program based on 3D Goldak’s double ellipsoidal model was developed in ANSYS Parametric Design Language for moving heat source calculations in the finite-element analysis (FEA). For welding an 8 mm thick plate with one pass, the suitable parameters were found to be 600 A current, 31 V voltage, and 10 mm·s−1 welding speed. The experimental cooling periods were found to be slower than predicted by FEA. When temperature distributions were compared between experimental and FEA results, an average variation of 1.88% at peak temperatures and 11.8% at completion time was observed. The results showed the temperature distribution at various time steps, illustrating the transient nature of the welding process. The results highlight the capacity of the FEA model to predict temperature profiles during SAW accurately, presenting a potent tool for optimizing welding parameters without extensive trial and error.

Statistical analysis of modelling approaches for CFD simulations of high-pressure natural gas releases

The use of computational fluid dynamics (CFD) in process safety to estimate the risk of a given incidental scenario has become ever more present in common industry practice. The simulation of high-pressure, compressible natural gas jets is often performed by modelling its source with a simpler notional diameter approach, such that the highly computationally expensive nearfield zone need not to be simulated; this is particularly determining when simulating a gas release in complex scenario like liquid natural gas (LNG) regasification plants. In this study, we analysed the structure of compressible and incompressible jets, using Birch 1984 (B84) and Birch 1987 (B87) models. In this work, a study on the positioning of the notional diameter with respect to the real orifice of the released gas is performed, along with a statistical analysis to assess the limits of the simpler model approaches.It was found that no spacing is needed between the virtual and real sources, as the potential core generated by the simpler model is as large as the fully simulated nearfield zone by the compressible model. Additionally, an end-of-transition zone position correlation is reported. The incompressible models can be used instead of the fully compressible model for a wide range of release conditions, with both models providing accurate predictions of axisymmetrical mole fraction, temperature, and velocity profiles between 2.5 and 130 bar of storage pressure at a 1-inch orifice diameter. However, as the diameter increases, B84 is not a viable model for a “full bore” (10-inch diameter size) release at 65 bar. While B84 is reliable, B87 is the superior model for its ability to account for the compressible effects of the expansion. Therefore, B87 should be used when simulating cases where temperature is of particular interest to the user.

TemproNet: A transformer-based deep learning model for seawater temperature prediction

Accurate prediction of seawater temperature is crucial for meteorological model understanding and climate change assessment. This study proposes TempreNet, a deep learning model based on a transformer and convolutional neural network, to accurately predict subsurface seawater temperature using satellite observations in the South China Sea. TemproNet uses multivariate sea surface observations such as sea level anomaly (SLA), sea surface temperature (SST), and sea surface wind (SSW) as model inputs, which employs a hierarchical transformer encoder to extract the multi-scale feature, uses a lightweight convolutional decoder to predict seawater temperature. We train and validate the model using the CMEMS temperature dataset and compare its accuracy with Attention-Unet, LightGMB, and ANN. Experimental results show that TemproNet has significantly outperformed other models with RMSE and R2 of 0.52 °C and 0.83 in a 32-layer temperature profile prediction task over 200 m in the South China Sea. In addition, we fully demonstrate the error of our model in space, in time, and at different depths, showing the efficiency and stability of our model. The input sensitivity analysis showed that SST contributed more to predicting shallow water temperature, while SLA significantly impacted the prediction of mid-deep water temperature. The results of this study provide an innovative and reliable solution for seawater temperature prediction and have important implications for meteorological model understanding and climate change assessment.

Prediction of pressure and temperature profiles for two-phase flow in multilayered gas wells by DTS data

Downhole temperature and pressure data are crucial information to help us understand the fluid flow dynamics in the reservoir. With the advancement of downhole monitoring technology, high-precision and high-resolution monitoring instruments facilitate qualitative and quantitative utilization of transient downhole temperature and pressure data in monitoring applications. The distributed temperature sensors (DTSs) overcome the limitations of traditional monitoring techniques and achieve real-time temperature monitoring in complex downhole conditions. However, the physical phenomena and response mechanisms hidden in the temperature data obtained by (DTSs) in transient, multilayered combined production and two-phase gas wells have not been thoroughly studied yet.Gas-water two-phase fluid flow in the reservoir is the prime consideration of this study. In this paper, based on the principles of mass, energy conservation, Darcy's law, and the second laws of thermodynamics, a forward model is established to predict pressure and temperature distribution in a multilayered vertical gas well system by coupling the flow model and thermal model. The model considers the changes of fluid high-pressure properties and microheat effects such as the Joule-Thomson effect, viscous friction to reach the desired solution. Subsequently, the model is solved numerically and used for sensitivity analysis of parameters influencing temperature and pressure distribution, such as gas production rates, permeability, Joule-Thomson coefficient. At last, the model is applied to a multilayer combined production well. The results of the example show the pressure, temperature and production profile can be predicted precisely.

Online thermal field prediction for metal additive manufacturing of thin walls

Various data-driven modeling methods have been developed to predict the thermal field in metal additive manufacturing (AM). The generalization capability of these models has been shown with simulation, but rarely tested with online physical printing. Instead, this paper aims to study a practical issue in metal AM, i.e., how to predict the thermal field of yet-to-print parts online when only a few sensors are available. This work proposes an online thermal field prediction method using mapping and reconstruction, which could be integrated into a metal AM process for online performance control. Based on the similarity of temperature curves (curve segments of a temperature profile of one point), the thermal field mapping applies an artificial neural network to estimate the temperature curves of points on the yet-to-print layer from measured temperatures of certain points on the previously printed layer. With measured/predicted temperature profiles of several points on the same layer, the thermal field reconstruction proposes a reduced order model (ROM) to construct the temperature profiles of all points on the same layer, which could be used to build the temperature field of the entire layer. The training of ROM is performed with an extreme learning machine (ELM) for computational efficiency. Fifteen wire arc AM experiments and nine simulations are designed for thin walls with a fixed length and unidirectional printing of each layer. The test results indicate that the proposed prediction method could construct the thermal field of a yet-to-print layer within 0.1 s on a low-cost desktop computer (Intel Core i7-3770 CPU @ 3.40GHz processor, 24.0 GB RAM). Meanwhile, the method has acceptable generalization capability in most cases from lower layers to higher layers in the same simulation, as well as from one simulation to a new simulation on different AM process parameters. More importantly, after fine-tuning the proposed method with limited experimental data, the relative errors of all predicted temperature profiles on a new experiment are sufficiently small, which demonstrates the applicability and generalization of the proposed thermal field prediction method in online applications for metal AM.

Modeling and in vivo experimental validation of 1,064 nm laser interstitial thermal therapy on brain tissue.

Laser interstitial thermal therapy (LITT) at 1064 nm is widely used to treat epilepsy and brain tumors; however, no numerical model exists that can predict the ablation region with careful in vivo validation. In this study, we proposed a model with a system of finite element methods simulating heat transfer inside the brain tissue, radiative transfer from the applicator into the brain tissue, and a model for tissue damage. To speed up the computation for practical applications, we also validated P1-approximation as an efficient and fast method for calculating radiative transfer by comparing it with Monte Carlo simulation. Finally, we validated the proposed numerical model in vivo on six healthy canines and eight human patients with epilepsy and found strong agreement between the predicted temperature profile and ablation area and the magnetic resonance imaging-measured results. Our results demonstrate the feasibility and reliability of the model in predicting the ablation area of 1,064 nm LITT, which is important for presurgical planning when using LITT.

Drying capacity of a continuous vibrated fluid bed dryer – Statistical and mechanistic model development

The drying capacity of a continuous vibrated fluid bed dryer was studied using a DoE by varying microcrystalline cellulose content in the formulation, water amount in the twin-screw granulation, inlet air temperature, air flow rate and the acceleration of the horizontal fluid-bed. Temperature and humidity profiles were measured along the dryer using wireless sensors. For the parameter space explored in this study, acceleration was the most influential process parameter of the dryer regarding the resulting granule moisture content. An empirical model was developed that allowed for fast and accurate moisture content prediction that could be incorporated into an enhanced control strategy. In addition, a mechanistic model was formulated that allow for prediction of temperature and moisture profiles, and most importantly the moisture content of the granules inside the dryer. The mechanistic model can be integrated to other unit operation models to provide overall understanding of an integrated continuous process line. The mechanistic model also makes it possible to define the equipment design requirements (e.g., length of the dryer) to meet the specific needs in terms of drying capacity, temperature and moisture profile.

Development of temperature differential maps for rigid pavements in tropical climates for quick estimation of thermal stresses

ABSTRACT Accurate estimation of the thermal stresses in rigid pavements is essential for the assessment of its service life. Thermal stress (an outcome of temperature differential) within the slab is significantly dependent on local weather conditions. This article presents a methodology for prediction of temperature profiles in rigid pavement. In this work, partial differential equations governing heat transfer through concrete pavement are numerically solved. The boundary conditions of the heat transfer model were formulated based on the real-time weather data. The temperature profiles predicted by solving the model were used to calculate the temperature difference between the top and bottom of the rigid pavement slab. The proposed methodology was implemented using 30 years of gridded weather data for whole of India. From the predicted temperature profiles, the annual M P T D and M N T D of the pavement slab were calculated. The mean values for each grid point observed over 30 years were used to develop temperature differential maps. The predicted values of maximum positive and negative temperature differentials had significantly higher values compared to the current codal recommendations of India. Developed maps can be used for estimating the design temperature differentials in places having similar climatic conditions presented in this study. Abbreviations: DLC: dry lean concrete; GSB: granular subbase; IRC: Indian Road Congress; MMMT: mean monthly maximum temperature; MMRH: mean monthly relative humidity; MTD: maximum temperature differential; MPTD: maximum positive temperature differential; MNTD: maximum negative temperature differential; PDE: partial differential equation; PQC: pavement quality concrete; SG: subgrade; TD: temperature differential

An improved theory of thermoelasticity for ultrafast heating of materials using short and ultrashort laser pulses

This paper introduces an improved thermoelastic theory designed specifically to address the challenges posed by fast and ultrafast heating of materials. The proposed theory incorporates a time delay in the temperature gradient and introduces the rate of temperature gradient as an additional term in the heat equation. This term accounts for the phase lag and the precedence of the temperature gradient and heat flux vector. By leveraging the sign of the temperature gradient rate, the theory automatically determines the precedence of either the temperature gradient or the heat flux vector. To verify the effectiveness of the proposed theory, experiments reported in the literature were simulated using several theories, namely the classical, Lord-Schulman (LS), Green-Lindsay (GL), dual-phase-lag (DPL), and the presented theory. The simulations considered the effects of relaxation times and pulse duration on the results. The findings reveal that for pulse durations shorter than 50 ps, the second-order time derivative of the temperature gradient becomes a crucial term in the energy equation for accurately predicting temperature profiles. Additionally, as the pulse duration decreases, the sensitivity to relaxation times increases. Regarding the surface temperature corresponding to the material ablation threshold, it is observed that for an 80 fs laser pulse, the ablation temperature falls between the boiling and critical temperatures, significantly higher than the boiling temperature associated with the ablation by a 270 ps pulse duration. Lastly, the simulation results exhibit a high level of agreement with the experimental data, particularly in terms of displacements, further validating the accuracy of the presented simulations.

Enhancing solar chimney performance in urban tunnels: Investigating the impact factors through experimental and theoretical model analysis

Efficient and sustainable ventilation in urban tunnels is crucial for combating air pollution and safeguarding human health. This study investigates the design factors impacting solar chimney performance in urban tunnels to optimize ventilation efficiency. Experimental trials analyzed the effects of blockage ratio, chimney height, and solar radiation on temperature distribution and ventilation rate. The results demonstrate that increased chimney height and solar radiation positively influence airflow velocity at the chimney outlet, enhancing ventilation. The temperature rise near absorber is higher than that closed to glazing wall. Temperature distribution within the chimney follows a distinctive horizontal two-piecewise semi-parabolic decay pattern, enabling accurate prediction of temperature profiles along the cavity depth. Novel analytical models predict temperature distribution, airflow velocity, and ventilation rate within the solar chimney system, aiding precise design and optimization. Remarkably, the blockage ratio has limited impact on ventilation rate, allowing for disregarding vehicle blockage effects in solar chimney design for urban tunnels. Matching chimney width to tunnel width and ensuring a relatively high chimney height are emphasized for optimal functionality. The study holds substantial implications for ventilation system design in urban environments, promoting healthier and more sustainable cities.

An Effective Temperature Profile Prediction in Additive Manufacturing Process Using Fractional GNU Global Herding Optimization-based Deep Learning Technique

An Effective Temperature Profile Prediction in Additive Manufacturing Process Using Fractional GNU Global Herding Optimization-based Deep Learning Technique

Study on the Influence Law of Temperature Profile of Water Injection Well

Due to the lack of knowledge on the influence law of the temperature profile of layered water injection wells, it is still highly challenging to quantitatively diagnose the water injection profile of layered water injection wells using distributed optical fiber temperature sensing (DTS). In this paper, a temperature profile prediction model for layered water injection wells has been developed by considering the micro-thermal effect and non-isothermal reservoir seepage. The influence of various single-factor changes on the temperature profile of layered water injection wells is simulated and analyzed. Orthogonal experiment analysis results demonstrate that the sensitivity of different factors on wellbore temperature from strong to weak is the injection temperature of the water, injection time, water injection rate, wellbore diameter, formation thermal conductivity, wellbore trajectory, and the permeability of injection formations (Tinj＞t＞Qinj＞D＞Kt＞θ＞k). The injection temperature of water, injection time, and water injection rate are the dominant factors affecting the temperature profile of water injection wells. The results of this paper provide a theoretical foundation for the accurate evaluation of the water injection profile and water injection scheme optimization for the layered water injection wells.

Two-temperature model for ultrafast melting of Au-based bimetallic films interacting with single-pulse femtosecond laser: Theoretical study of damage threshold

Ultrashort-pulse laser excitation and associated thermal damage of single-layer metallic films have been extensively studied both theoretically as well as experimentally. However, study of ultrafast heat transfer mechanism in bimetallic films, a topic of immense current interest for designing next-generation optical components with better thermal resilience, is limited to a low fluence regime. Moreover, available results for damage threshold (DT) of bimetallic films have several ambiguities. The current work is an attempt towards development of comprehensive theory for the DT of Au-based bimetallic films, analyzing its dependence on associated physical parameters and bridging the gap in existing literature. This has been achieved by developing fully implicit two-temperature model-based Python code. Excellent agreement of the predicted temperature profiles with recently reported [E. L. Gurevich, Y. Levy, S. V. Gurevich, and N. M. Bulgakova, Phys. Rev. B 95, 054305 (2017)] femtosecond laser irradiated Au film validates the code. Subsequently, the code is utilized to investigate the influence of different metallic substrates, $M$ ($M$ = Ni, Cr, Cu) on ultrafast melting damage of $\mathrm{Au}\text{/}M$ films excited by single-pulse femtosecond laser. Simulations carried out for increasing Au layer thickness revealed a nonmonotonic variation of DT for all substrates. This study also brought out an important observation that for every substrate metal, there exists an optimum Au thickness for maximizing the incipient and the complete melting threshold. Maximum achievable enhancement for $\mathrm{Au}\text{/}M$ films over pure Au are found to be 30%, 22%, and 14% with Cu, Ni, and Cr substrates, respectively. The observed distinction in maximum DT among three different $\mathrm{Au}\text{/}M$ targets is explained by developing appropriate thermophysical models. Finally, an empirical function is proposed here for expressing nonmonotonic variation of two-layer DTs. Very good agreement between the substrate-specific peak positions predicted by the proposed function and effective electron diffusion length arising from thermophysical analysis makes the modeling robust and applicable for a wide range of laser wavelengths as well as arbitrary combinations of bimetallic films.

FEM modeling for predicting temperature profile of heat affected zone in single spark EDM process

FEM modeling for predicting temperature profile of heat affected zone in single spark EDM process

Modelling of a continuous kneader reactor for the polymerization of partially neutralized acrylic acid

AbstractA mathematical model and analysis of the continuous polymerization of partially neutralized acrylic acid (AA) in a continuous kneader reactor is presented here as an initial attempt to simulate the synthesis of a superabsorbent polymer. A detailed kinetic model has been used to describe the copolymerization of AA and sodium acrylate (NaA) in aqueous medium. This model is used to describe batch and continuous operations. The polymerization is initiated by a mixture of potassium persulphate (K2S2O8, KPS) and hydrogen peroxide (H2O2) as oxidizing agent and ascorbic acid (AsA) as reducing agent. A novel set of kinetic parameters has been estimated by fitting experimental data from different literature sources. The operation of a continuous kneader reactor modelled as a plug‐flow reactor with axial dispersion is theoretically investigated to predict temperature profile, total and individual monomer conversion, consumption of KPS, H2O2, and AsA, and polymer average molecular weights. The simulation results show the presence of a hot spot close to the reactor entrance that could be potentially severe during startup and could have a detrimental impact on polymer quality. This model is a first step in the direction of achieving optimal operating protocols and exploring improved polymerization reactor designs.