The Johnson-Cook (J-C) plasticity and damage models are widely used for simulating machining processes under large strains, high strain rates, and high temperatures. However, determining the J-C parameters (A, B, n, C, and m) and associated damage parameters remains technically challenging since these parameters can hardly be individually measured. This article presents a multimodal machine learning method to learn the J-C parameters and a simplified damage parameter in response to chip-formation images and cutting force data from orthogonal cutting simulations. The proposed method employs convolutional neural networks to extract spatial features from chip images and utilizes a multilayer perceptron to learn from different data sources. Two numerical studies are conducted to test the method. Study 1 attempts to identify all five J-C parameters simultaneously. Although the non-uniqueness of the J-C model prevents a unique solution, the multimodal approach still provides the most reliable predictions. Study 2 determines the strain-rate hardening and thermal softening effects (C and m), as well as the fracture strain (epsilon(& strns;)(pl )(D)) in the ductile damage criterion. With known A, B, and n from tensile testing, the results demonstrate that C, m, and epsilon(& strns;)(pl )(D)can be uniquely determined. Overall, studies show that multimodal machine learning can be an effective method for inverse analysis in estimating the plasticity and damage model. Moreover, an experimental demonstration was conducted to estimate the parameters from the real-cutting images and force data. The practical aspects and limitations of the approach are also discussed.
Fixed-abrasive lapping has been widely applied in the precision machining of hard and brittle materials. Its well-controlled abrasive motion and dominant two-body removal mechanism enable superior surface quality. However, due to multiple process parameters and the complex removal mechanism, surface quality control still relies heavily on empirical knowledge. To improve the controllability of the lapping process, this study proposes a two-dimensional microscale surface profile simulation model that incorporates pad topography, material removal mechanisms, and microscale contact characteristics. The model innovatively introduces dynamic profile baseline updating, abrasive scratching angle, and elastic recovery. Experiments were conducted on fused quartz under different lapping forces and lapping plate rotational speeds. The close agreement between the simulated and experimental results validates the model's effectiveness in predicting surface profile, surface roughness, and material removal rate. Furthermore, surface roughness analysis was used to optimize both lapping force and plate speed. This work provides new insights into the microscale evolution of surface profiles in fixed-abrasive lapping and offers a theoretical basis for process parameter optimization.
Achieving stable and repeatable keyhole laser welding requires precise control of process parameters, since even small variations can lead to insufficient penetration, overpenetration, or process instabilities, particularly in high-value assemblies where a single defective weld may result in the rejection of the entire component. Numerical simulation plays a central role in process understanding and optimization; however, state-of-the-art multiphysics models are characterized by high computational costs, which prevent their direct use as predictive tools for online process control despite the growing demand for models that can be integrated within real-time monitoring and control frameworks. To overcome these limitations, a physics-informed neural network (PINN) framework is proposed as a near-real-time surrogate model for keyhole laser welding. The approach embeds the transient heat conduction equation, coupled with a double-conical volumetric heat source, directly into the neural network loss function, avoiding the need for large labelled datasets. The model is calibrated through an inverse analysis using a limited set of experiments, establishing empirical correlations between laser power, scanning speed, and heat source geometry. Validation against experimental data and high-fidelity computational fluid dynamics (CFD) simulations shows good agreement, with relative errors typically below 10% for weld depth and width. Once trained, the PINN predicts the thermal field and weld bead geometry within milliseconds, enabling rapid mapping of the process window and supporting laser welding process optimization.
The difference between the position of the load axis and the shear center leads to warping in asymmetric bending of profiles. In a previous study, partial heating during bending was proposed to reduce warping. This work did prove the reduction of warping for an L-shaped cross section using experimental and numerical data and an analytical model for the description of the bending moment and profile warping was created. However, the analytical model was not able to describe the movement of the shear center position and was restricted to the description of L-profiles. The present work generalizes the previous analytical model to describe the movement of the shear center position due to the heating strategy for any cross section. Additionally, descriptions for strain-free fiber, force calculation, and warping have been adjusted to achieve high accuracy when modeling arbitrary profile cross sections. The analysis is carried out on the example of L- and U-shaped profiles consisting of S500MC fine grain steel. Using the adjusted model, the error in the prediction of bending load has been reduced from 30% to 4% and for warping the error changes from 46% to 6%. The reason for the deviation can be attributed to less simplified force and strain-free fiber calculations as well as a simplification approach of bending angle by linearization. A formulation for the shear center movement has been successfully implemented, showing that the shear center moves in the direction of the load axis with the analyzed partial heating strategies.
The increasing demand for lightweight materials in the automotive industry has driven significant advancements in joining technologies for dissimilar materials, particularly steel-aluminum (Fe-Al) assemblies. This study reviews the mechanical behavior of Fe-Al spot joints under tensile-shear loading, examining recent joining techniques including resistance spot welding (RSW), rivet welding (RW), friction stir spot welding (FSSW), ultrasonic spot welding (USW), riveting, clinching, and flow drill screwing (FDS). The analysis focuses on key parameters influencing shear performance-such as aluminum sheet thickness, joining area size, and failure mechanisms-while also evaluating the impact of fasteners versus direct welding, intermetallic compound (IMC) formation. This work could provide a foundation for creating preliminary design abacuses and predictive models to joint behavior in lightweight auto-body assembly.
The grinding principle of a spherical worm wheel is used to determine the coordinate system for the tooth surface of an equivalent helical gear that envelops the worm wheel. The equation for the tooth surface is reverse-engineered to construct a model. The dressing principle for the grinding wheel based on the virtual center distance method is used to examine the motion structure of the gear grinding machine and dressing motion characteristics of the grinding wheel. A coordinate system for the diamond roller dressing of the spherical worm grinding wheel is established. The four-axis linkage of the machine tool enables the grinding wheel to undergo deflection, thereby completing the dressing cycle. The grinding process is simulated, and theoretical predictions are experimentally validated. The results confirm the accuracy of the internal helical gear with the worm wheel, and this demonstrates the rationality and effectiveness of the proposed dressing approach. Overall, this study presents significant advancements in the field of gear manufacturing and provides valuable insights for future developments in grinding technology.
Cryogenic cutting has been considered as an optional beneficial method to enhance machined surface integrity and machinability when cutting titanium alloys and nickel-based superalloys. Its advantages and disadvantages on the machinability and surface integrity of copper alloy remain controversial. Most existing research focuses on summarizing phenomenological experimental observations without mentioning the underlying mechanisms. To provide insights for understanding machinability and surface integrity evolution with cryogenic cooling from a material constitutive behavior perspective, this research systematically studies the effect of cryogenic cooling on multiple evaluative aspects of machinability (cutting force, chip formation) and machined surface integrity (geometrical, physical, and microstructural properties) of copper alloy with varied cutting parameters. Cutting tests are performed under different cutting conditions (both orthogonal and oblique cutting settings, varied cutting depth, feed, and linear speed), and Gleeble compression tests are performed from cryogenic temperature to dry cutting temperature. It is discovered that the cryogenic cooling method has its advantages in surface residual stress, especially at low-feed and high-speed cutting conditions, but it raises cutting force, deteriorates surface roughness, and surface material side flow, especially at large feed and low cutting speed conditions. Such transitions are induced by a large high strain hardening rate variation of copper between cryogenic temperature and dry cutting temperature, which is further attributed to recrystallization suppression and deformation twin activation of low temperature and summarized as low-temperature-induced enhanced toughness. The output helps understand the cooling effect on ductile metals and for deciding the cryogenic cooling strategy in industrial applications.
The rigidity of the additively manufactured objects can be tailored by manipulating the infill lattice type and density. In this research, an island-type novel infill structure termed the Hexagonal-Zigzag pattern is introduced, and its mechanical performance is investigated. In this pattern, the zigzag raster reflects the repeating hexagonal-shaped cell constituting the parallel-oriented islands, and a 90 deg rotation of the pattern in each layer distributes the island span along both transverse and longitudinal directions of the printing contour. A mathematical model is established to illustrate the effect of the infill parameters on hexagon unit cell size and relative infill density. The compression test is executed on some rectangular test samples to characterize the nature of this pattern and explore its performance over other existing infill patterns: Honeycomb, Zigzag, and Triangle. The test result reveals that the compressive strength and elastic modulus of the proposed infill pattern are comparable with those of the Honeycomb infill pattern. Furthermore, the highest plateau stress and absorbed energy density demonstrate the performance of the proposed infill pattern. In addition, the experiment is extended to investigate the mechanical behavior of the proposed infill pattern with various hexagon unit cell sizes. The experimental results reveal the related pros and cons, such as increasing compressive strength and energy density with the enlargement of unit cell size, exhibiting its efficiency, whereas the decreasing elastic modulus and early densification represent the weakness.
Laser-directed energy deposition (L-DED) offers unique advantages for fabricating large-scale metallic components and repairing high-value parts. However, recurring interlayer porosity, particularly while depositing targeted geometry and dimensions, remains a major limitation affecting structural integrity. In this study, systematic deposition strategies were developed to mitigate interlayer porosity by controlling track overlap and optimizing energy apportionment, the two aspects that have not been reported together in previous L-DED studies. Experimental analysis showed that increasing the percentage overlap from 30% to 40% significantly reduced porosity, whereas defining the overlap based on the full width at half maximum (FWHM) provided a more geometry-representative approach. A 30% FWHM overlap was found to be most effective in disrupting periodic porosity recurrence. Additionally, introducing skewed track alignment minimized valley-to-valley overlap across layers, further reducing defect formation. Complementary to geometric strategies, interlayer laser polishing with circular and line beams facilitated pore closure while refining the interlayer microstructure. A key novelty of this work lies in coupling overlap optimization with energy apportionment between powder and substrate, achieved by adjusting the stand-off distance (SoD), which is quantified by a unique experimental approach. This enhanced molten pool flow and ensured improved remelting of the previously deposited layer, which, when combined with a 30% FWHM overlap, effectively eliminated visible interlayer porosity, validated by micro-computed tomography analysis. The integrated approach of optimized overlap, energy apportionment, and interlayer polishing enabled defect-free fabrication of straight walls as well as complex turbine blade profiles, while simultaneously enhancing strength and ductility.
Additive/subtractive hybrid manufacturing (ASHM), which enables in situ machining to alleviate tool interference issues and improve surface quality during additive manufacturing, holds significant potential for fabricating complex, high-performance components. However, due to the differing characteristics of machined and additively built surfaces, variations in powder spreading, melt-pool flow, and solidification behavior make the alternating interface a weak point in interfacial bonding. This study investigates the influence of substrate surface condition and powder layer thickness on melt-pool behavior during laser powder bed fusion of GH3536 powder. The thermal-flow behavior under varying substrate surface conditions was simulated using a computational fluid dynamics-based model and validated by single-track laser scanning experiments. Peak temperature and melt-pool lifetime were used to assess thermal behavior. The results showed that rough substrates promoted continuous and stable melt tracks, while smooth substrates were prone to defects such as necking and balling. For smooth substrates, increasing the powder layer thickness to 80 & micro;m significantly improved thermal behavior, with the peak temperature and melt-pool lifetime increasing by 15.65% and 43.89%, respectively, compared to the 40 & micro;m layer used for rough substrates. To enhance interfacial bonding and microstructural uniformity at the interface, a variable powder layer thickness strategy was proposed. This study provides practical guidance for determining layer thickness in ASHM to improve interfacial bonding.
Monitoring melt pool behavior in laser powder bed fusion additive manufacturing is essential for ensuring process stability and detecting anomalies such as spatter, plume generation, and irregular melt pool shapes, all of which influence part integrity. However, conventional image-based deep learning approaches for this task, while accurate, are computationally intensive and difficult to deploy in real-time production environments. To address this challenge, this article presents a lightweight, feature-driven deep learning framework for multi-label defect classification. We develop a model that leverages a compact set of statistical, morphological, and texture features extracted from melt pool images, enabling concurrent classification of multiple defect types with minimal computational overhead. The dataset used in this study includes melt pool images from the Additive Manufacturing Metrology Testbed (AMMT) at the National Institute of Standards and Technology (NIST), providing both in situ monitoring data and ex situ characterization via high resolution X-ray computed tomography (XCT). Experimental benchmarking against a standard image-based model confirms the efficiency of our approach: it achieves F1 scores exceeding 98% across all categories while reducing model complexity by 16-fold. Furthermore, compared to conventional image-based pipelines, the proposed framework achieves a 2.3 & times; speedup. Crucially, we validate these in situ classifications against ex situ XCT data, demonstrating that specific multi-label defect combinations correspond to measurable grayscale shifts as a proxy for internal porosity. This work thus offers a scalable, physically validated pathway for real-time quality management in additive manufacturing.
The present research focuses on the hot deformation behavior, microstructure, texture evolution, and fracture mechanisms of wire arc additively manufactured Inconel 625 alloy. The deposited material exhibits columnar, cellular, and dendritic grain structures, with grain widths ranging between 10 mu m and 40 mu m and lengths extending up to 150 mu m. The deformation characteristics were studied across a temperature range of 700-900 degrees C. At medium temperatures, distinct serrations were observed, transitioning from B type to C type as the temperature increased. At elevated temperatures, B-type serrations reappeared, attributed to interactions between C14-Ni2Nb Laves phases and mobile dislocations, as revealed by transmission electron microscopy (TEM) analysis. Conversely, C-type serrations were associated with the nucleation and growth of deformation twins. Grain size and boundaries under varying deformation temperatures were examined using electron backscattered diffraction (EBSD) and TEM, which revealed the occurrence of dynamic recrystallization (DRX). It was observed that the preferred orientation for DRX nucleation in the Inconel 625 alloy is along the < 001 > direction. At 900 degrees C, recrystallized grains were prominent, with EBSD results confirming both discontinuous dynamic recrystallization (DDRX) and continuous dynamic recrystallization (CDRX), where CDRX acted as a secondary nucleation mechanism. Below 900 degrees C, cracks primarily nucleate due to stress concentrations near Nb-rich phases. At 900 degrees C, crack initiation was influenced by slip band impingement at grain boundaries combined with stress concentrations around Nb-rich phases. Furthermore, void formation at grain boundary triple junctions caused by grain boundary sliding contributed to ductile fracture. The high-temperature deformation behavior of wire arc additive manufacturing-deposited Inconel 625 is strongly influenced by the interplay of DDRX and CDRX mechanisms.
Metallic biomedical implants require surfaces with improved wettability, decontamination, and biocompatibility to ensure successful osseointegration and long-term durability. Conventional surface functionalization methods are often hindered by high costs, long process times, and complex setups. The atmospheric pressure dielectric barrier discharge (DBD) plasma jet offers a promising alternative, providing a low-power, cost-effective, and low-temperature approach with rapid processing times. This study investigates the application of a DBD plasma jet for surface modification of additively manufactured Ti-6Al-4V ELI (Ti Gr23) to enhance wettability and promote bone cell mineralization, which is crucial for better osseointegration. Improved surface wettability of the plasma-treated Ti Gr23 samples is demonstrated by a significant reduction in the water contact angle (WCA) following only 2 s of treatment. X-ray photoelectron spectroscopy (XPS) analysis reveals that plasma interaction induces the formation of oxide metallic bonds, such as TiO2 and Ti2O3, while reducing carbon contamination, leading to increased surface energy. A parametric study was conducted to evaluate the effects of treatment time, gas flowrate, and input power on the treated area size, WCA reduction, and process cost. The optimal conditions of 8 s of treatment time, 2 slm gas flowrate, and 7.4 W power achieved the largest treated area (70 mm(2)), maximum WCA reduction (58 deg), and minimal processing cost ($26/m(2)). Cytocompatibility tests confirmed that plasma treatment had no cytotoxic effects on Ti Gr23 implants. Additionally, plasma treatment of Ti Gr23 surfaces resulted in a 45% increase in bone cell mineralization, highlighting its dual role as a surface modification strategy for improving cell adhesion and as an effective sterilization approach against hospital-acquired pathogens such as MRSA for smooth and rough surfaces. Overall, plasma treatment enables simultaneous enhancement of wettability, cell adhesion, and decontamination on rough bioimplant surfaces, otherwise difficult to achieve with conventional methods like ethanol rinsing.
With recent advances in computer numerical control (CNC) systems, modern machine tools have become increasingly intelligent, capable of automatically compensating for process errors and abnormalities. A well-known source of errors in modern milling processes is associated with tool eccentricity and cutter runout that produce rough surface finish and lead to accelerated tool wear. This article presents a novel strategy where the CNC machine tool senses the eccentricity/runout related errors on-the-fly and compensates for them using its own feed drive system. A general formulation is developed, which reveals the force/vibration frequency spectrum of the milling process suffering from tool eccentricity/radial runout. The tool eccentricity is then compensated by commanding the machine tool feed drives with microcircular trajectory at the spindle frequency, which then cancels the circular (eccentric) motion of the tool center point. The commanded circular trajectory parameters, i.e., the amplitude and the phase, are adjusted automatically by iteratively learning the dynamic response of the feed drive system and using the tool eccentricity-induced process response based on the data collected either via an accelerometer or a force sensor. The overall learning (adaptation) process is formulated as a convex optimization problem, and various simulation studies are provided to demonstrate the optimality and the convergence of the approach. The effectiveness of the proposed strategy is validated through various simulation studies and actual milling experiments.
This study investigates the additive manufacturing (AM) processing, microstructural evolution, and resulting mechanical and thermal properties of multimaterial components combining 17-4PH stainless steel and pure copper (Cu) fabricated via laser powder directed energy deposition (LP-DED). Conventional tooling steels exhibit limited thermal conductivity, significantly constraining production throughput in high-volume processes. Incorporating Cu, with its superior thermal conductivity, could significantly enhance tool performance, though Cu and steel present metallurgical incompatibilities when processed via AM. A systematic investigation was conducted across compositions ranging from 0 to 100 wt% Cu, revealing critical thresholds influencing solidification behavior, defect formation, microstructure, hardness, and thermal transport. Optical microscopy, electron backscatter diffraction (EBSD), hardness testing, and thermal conductivity measurements provided comprehensive process-structure-property correlations. Severe hot cracking occurred at low-Cu contents (6-25 wt%), aligning generally well with crack susceptibility modeling, with an unexpected discrepancy at 25 wt%. Porosity remained low (>= 99% dense) throughout the compositional spectrum. EBSD analysis revealed a transformation from columnar martensitic structures at low-Cu contents to equiaxed FCC Cu-dominated structures at higher Cu concentrations, highlighting the complex microstructural transitions driven by Cu-induced changes in solidification and phase stability. Hardness decreased from 330 HV (pure 17-4PH) to 62 HV (pure Cu), consistent with microstructural changes. Concurrently, thermal conductivity improved substantially from 13.5 W/m K to 367.9 W/m K, emphasizing Cu's dominant role in thermal transport. The findings highlight the feasibility of leveraging compositional gradients between 17-4PH and Cu to achieve tailored tooling with optimized thermal and mechanical performance.
Cyberattacks have been rising steadily since the 1990s, and today the manufacturing and industrial sectors have become prime targets. U.S. manufacturing is seen as a lucrative target because of its rich space for exploitation, the fear of production halts, and the lack of a reliable, self-sufficient supply chain that can support operations during crises. With the growing use of interconnected technologies, entry points for attackers are more numerous than ever. Traditional methods such as signature-based or static defenses have proven ineffective, while artificial intelligence (AI)-driven approaches have shown promise but often lack consistency, performing well in some areas while failing in others. This study addresses that challenge by examining existing AI models used for cyber-threat detection, evaluating their advantages and limitations, and proposing a more reliable alternative. This article proposes a lightweight and easily deployable deep hybrid learning (DHL) model trained and tested on the TON_IoT dataset. The model was compared against ten of the most widely used machine learning (ML) and deep learning (DL) models in cybersecurity and achieved superior performance with 98.13% accuracy, 98.82% precision, 98.24% recall, and 98.53% F1. This study provides practical recommendations to strengthen industrial systems and protect U.S. manufacturing enterprises from the growing wave of cyber threats.
High-accuracy modeling of machine tool dynamics is essential for advanced process planning and monitoring. However, modeling high-speed multi-axis machines is challenging due to the inherent coupled and nonlinear multibody dynamics and structural flexibility. This complex modeling task is addressed by a new approach in which the control dynamics and the open-loop plant dynamics are characterized by a multiple-input and multiple-output (MIMO) linear time-invariant (LTI) system coupled with a generalized disturbance, which is able to capture the open-loop coupled nonlinear dynamics. As a case study, different machine tool topologies of a flexible linear drive coupled with a rotary drive are systematically analyzed using the proposed modeling approach. The identification procedure for the proposed method requires capturing the internal structural vibration between the drives. This article also presents a method to reconstruct the internal structural vibration using data from the embedded encoders as well as a low-cost microelectromechanical systems (MEMS) inertial measurement unit (IMU) mounted on the machine table. This modeling-building approach is nonintrusive and practical for industrial implementation. The experimental validation shows a 2-6% error in predicting the tracking error and motor force/torque. Especially, the vibratory inter-axis coupling effect and posture-dependency are accurately predicted.
Copper's high electrical and thermal conductivity makes it appealing for use in various industrial products. However, challenges related to its weight and performance in extreme environments limit copper's usage in aerospace applications. Previous research has demonstrated that incorporating multilayer graphene (MLG) on a copper substrate via chemical vapor deposition (CVD) improves the performance of this conductor in high-temperature applications without incurring a weight penalty. Incorporating a high percentage of large-area graphene, needed for better performance, is difficult while fabricating wires. This work shows a method for making high-quality graphene-copper composite wires from 25 & micro;m and 50 & micro;m copper foils, consolidated into a wire via repeated annealing and roller drawing reductions. A copper foil wire without graphene is compared to the composite to highlight graphene's benefits. This research correlates the composite's resulting material properties to the microstructure and creation process. The final results suggest that graphene content aids in consolidation, removing one of the primary defects in manufacturing wires from foils. Reducing porosity through improved consolidation prevents early fracture under tensile loading. In addition, the specific conductivity at room temperature for bilayer graphene (BLG), few-layer graphene, and MLG samples was comparable to that of bare copper wire. Graphene content also improves the resulting high-temperature electrical properties by protecting the wire from further oxidation. Based on the data presented in this article, recommendations are provided for further reducing void defects and enhancing the quality and performance of copper-graphene composite wires.
Additive friction stir deposition (AFSD) is a solid-state additive manufacturing process with significant potential for titanium alloys, yet its applicability to large-scale Ti-6Al-4V builds has remained largely unexplored. This work presents an assessment of residual stresses along with the characterization of microstructure and mechanical properties in a 200 & times; 35 & times; 67 mm3 AFSD Ti-6Al-4V block. The comprehensive characterization includes scanning electron microscopy (SEM), electron backscatter diffraction (EBSD), hardness mapping, tensile testing, and full-field contour method residual stress analysis. It demonstrates that AFSD produces a uniform basket-weave alpha + beta microstructure from the substrate/first layer interface to the top and along the edges of the block. The deposited block without postheat treatment exhibits consistent hardness and reproducible tensile behavior, matching or exceeding wrought standards. Most critically, this is the first report of residual stress in AFSD Ti-6Al-4V, revealing exceptionally low magnitudes with longitudinal stresses limited to +100-150 MPa in the core and -100 to -150 MPa at the top surface, while build direction stresses remain negligible (-40 to +40 MPa), corresponding to similar to 13% of yield strength compared to fusion-based additive manufacturing processes, where residual stresses often reach 30-50% of yield. AFSD uniquely achieves large-scale, defect-free, and stress-minimized deposits without any postprocessing. These results establish AFSD as a robust and industrially viable route for both near-net-shape fabrication and repair of aerospace-grade titanium structures.
Ultrasonic-assisted resistance spot welding (URW) has emerged as a superior technique compared to conventional resistance spot welding (RSW) for multiple thin aluminum foils-to-tab welding applied in the manufacturing of pouch cell batteries, producing larger and higher quality welds. However, in situ experimental observation of the nugget formation during URW is challenging due to the enclosed weld region and the transient nature of the process. In this study, a numerical modeling framework is implemented, leveraging a baseline finite element model (FEM) of RSW to systematically evaluate individual and coupled impacts of various ultrasonic effects on the thermal, mechanical, electrical, and flow fields of the weld stack. A coupled FEM-computational fluid dynamics model and a cavitation energy coupled FEM have been utilized for the first time to study the melt flow under ultrasonic pressure variation and the effects of cavitation energy on temperature distribution, respectively. To experimentally verify the occurrence of acoustic cavitation during URW, in situ acoustic signals have been monitored with a microphone. Coupling all ultrasonic effects, including reduced contact resistance, acoustic softening, and increased electrical conductivity, along with cavitation energy, underpredicts the nugget size, contrasting experimental observations. The reduction in contact resistance proves to be a dominating factor that results in a smaller-sized URW nugget in joining thin foils to the tab. These findings highlight the need to modify the contact resistance model and to incorporate additional ultrasonic mechanisms to enable predictive modeling of weld nugget evolution for URW.
