In this study, laser metal fusion with feedstock wire is addressed. We investigated how various process parameters affect the fraction of beam energy that is absorbed by the wire and the workpiece and the metal transfer from the feedstock wire to the melt pool. To perform this research, a thermo-fluid dynamic model with tracking of free surface deformation was developed to include the feeding of a solid wire and predict its melting. The fraction of beam energy absorbed by the metal was modeled as a function of local surface curvature and temperature, accounting for multiple Fresnel reflections and absorptions. The model was applied to Titanium alloy (Ti-6Al-4V) with a 1.07 μm laser and a process in conduction mode. Experiments at various wire feeding rates were conducted to evaluate the model’s ability to predict the process and a good agreement was obtained. The different parameters studied were the beam angular position, the wire angular position, the wire feed rate, and the beam-wire offset. The analysis of the simulation results gave a detailed physical understanding of the laser energy use. It highlighted that thermocapillary and Rayleigh-Plateau instabilities can contribute to the transition from continuous to drop metal transfer mode. Damping these instabilities might thus allow using a wider process window.

Additive manufacturing of parts on-site in space requires investigating the feasibility ofadapting to zero-gravity and near-vacuum conditions, a technology applied today on Earthat standard conditions. While a few studies have been conducted for powder bed fusion, afeasibility study remains to be explored for direct energy deposition using a laser beam anda metal wire. This is the purpose of this study, which is conducted using a modelingapproach based on computational fluid dynamics. The simulation model developedincludes melting, re-solidification, vaporization, prediction of beam energy absorptionas a function of the local surface temperature and curvature, ray tracing, tracking of freesurface deformation and metal transfer, and wire-resistive heating. The study is carried outby starting from process parameters suited for stable on-Earth metal deposition. Theseconditions were also studied experimentally to validate the simulation model, leading tosatisfactorily results. A total of three other test cases with ambient pressure lowered downto near-vacuum and/or gravitation down to zero are investigated. It is found that,compared to on-Earth conditions, in-space conditions can induce vaporization of themetal alloy that is large enough to result in a curvature of the melt pool free surface but toosmall to lead to the formation of a keyhole. The in-space conditions can also modify theforce balance at the liquid melt bridge between the wire and the melt pool, leading to smallchanges in the curvature and temperature field at the free surface of the wire tip. Among theobserved consequences are a small increase of the melt pool length and a small elevationof the bead height. More importantly, for process control, changing to in-space conditionsmight also affect the stability of the process, which could be assessed through the width ofthe liquid metal bridge. However, by using appropriate process control to maintain acontinuous liquid metal bridge, it is concluded that direct energy deposition of metal usinga laser and a wire could be used for manufacturing metal parts in-space in a temperedatmosphere.

Laser metal fusion is widely used in production technology to manufacture parts, as in welding, cladding, and additive manufacturing. In this study, conduction mode laser metal fusion is applied without and with metal deposition from a wire feedstock. This manufacturing process encompasses various physical phenomena that are coupled, such as the interaction of anelectro-magnetic wave with the material, phase changes, thermal fluid dynamics, and free surface deformation, which make it complicated to comprehend.

Deeper process knowledge is thus a key to its improvement. Yet, metal is a non-transparent media, which limits experimental observation of this process.

A modelling approach that describes this multi-physics problem paying special attention to convective phenomena was used in this thesis with a two-fold aim:

1) to improve the model reliability,

2) to gain a deeper understandingof the metal fusion and deposition process.

In the first part of this research, metal fusion without wire was addressed. Different beam power density distributions (beam shapes) were investigated. Their effect on the melt pool geometry, which was known from previous experimental studies, could be predicted. Furthermore, as the simulations give access to the melt flow, it could be established that the flow pattern is modified by elongating the beam shape. In addition, a new calculation procedure was introduced to predict the fraction of laser beam energy absorbed by the metal. To validate the model, the predicted melt pool geometry was evaluated through comparison with experimental measurements. The results showed that the proposed absorptivity model that is a function of local surface conditions lead to good agreement with experimental results, with a maximum discrepancy for the melt pool depth of about 10%.

In the second part, the model was applied to study the fusion process with metal transfer from a wire feedstock without and with resistive heating of the filler wire. It was shown that the multipler eflections of beam rays could be ignored at a low laser beam angle whereas with increasing the beam angle the effect became more considerable. It was also found that the laser absorptivity varied up to 50% within the projected laser spot area. The effect of different process parameters such as depositing rate and angle, laser beam angle, position of the wire relative to the beam (offset), and ambient conditions on the metal transfer, thermal flow field, andstability of the process were studied.

The results showed that three different metal transfer modes occurred depending on the offset value. Applying resistive heating on the filler wire decreased the absorptivity. However, this decrease was compensated by the resistive heating, resulting in an increase of the volume of liquid metal. Resistive heating made the melt pool wider due to the augmented role of the thermocapillary force and also the change in flow direction because of the modified position of the melted wire front.

Applying the model at near-vacuum and no gravity conditions, it was obtained that directed energy deposition of metal with laser and wire could be used for manufacturing metal parts in space. However, the process window could need some adjustment as in-space conditions result in some narrowing of the liquid bridge between wire and workpiece compared to on-Earth.

A computational fluid dynamics approach is used to analyse the influence of beam shaping in fusion welding on melt thermal flow. Three beam shapes are studied at several welding travel speeds: a reference Gaussian profile and its elliptic elongations along and transverse to the welding travel direction. It is found that these beam shapes change not only the intensity and direction of the melt thermocapillary flow but also the flow pattern. For instance, and contrary to the other profiles, the beam shape elongated along the welding travel direction generates melt front vortices that assist metal pre-heating. It can result in deeper penetration, larger melt volume, and lower amount of thermal energy diffused into the heat affected zone. The simple elongation of a beam profile has thus a non-linear effect on the melt flow and in turn on the seam geometry as well as the temperature gradients in the heat affected zone.

Laser is an efficient and widely used heat source in metal processing suchas welding and additive manufacturing. It has some great advantages compared to the other conventional heat sources like electron beam and arc namely: ability of handling complicated joint geometries and producing large components. Laser beam welding encompasses many complex physical phenomena such asheat transfer, metal melting, flow and solidification, free surface deformation, evaporation and possibly vaporization. The aim of this research work istwo-fold: gain deeper process understanding and improve the model reliability. Deeper process understanding is sought on the effect of beam shaping on themelt pool. To achieve improved model reliability, a good support consists in using qualitative experimental data representing the process. Thus, 3D validation of the melt pool geometry is performed while it was usually 2D inprevious research works. Furthermore, a new calculation procedure for laser absorption is introduced. To conduct this research work, a Computational Fluid Dynamics approach is used. A solver, capable of tracking the deformation of the melt free surface, is developed in OpenFOAM. Concerning beam shaping, it is found that not only the melt pool size as previously known but also the melt flow pattern is modified through elongating the beam shape.This last result could not be revealed by former studies as the non-transparent media hinders optical observation. New in-process quantitative measurements performed by a project partner are used to test the model. Weaknesses of the former absorptivity models are highlighted, as well as the limitations of the proposed model. Finally, the results show that the proposed absorptivity model function of local surface conditions leads to much better agreement with experimental results compared to the former constant absorptivity model. The maximum discrepancy compared to the experimental measurement, which is observed for the melt pool depth, can indeed be reduced to about 10%.

Laser är en effektiv och allmänt använd värmekälla vid svetsning och additiv tillverkning. Den har några viktiga fördelar jämfört med andra konventionella värmekällor såsom elektronstråle och elektrisk ljusbåge, nämligen: den kan ofta användas till komplicerade svetsgeometrier, och den kan producera stora komponenter. Lasersvetsning involverar olika sammansatta fysikaliska fenomen såsom värmeöverföring, metallsmältning, flöde, stelning, ytdeformation, avdunstning och i vissa fall förångning. Syftet med mitt forskningsarbete är tvåfaldigt: att få en djupare processförståelse och att förbättra modellens tillförlitlighet. Fördjupad processförståelse eftersträvades för att förstå hur formen på laserstrålen påverkar svetssmältan. För att uppnå förbättrad modellsäkerhet behövs experimentella data av hög kvalitet som representerar processen. Således utfördes 3D-validering av smältgeometrin medan det vanligtvis var 2D i tidigare forskningsarbeten. Dessutom har en ny modell för laserabsorption föreslagits. I forskningen har numerisk strömningssimulering (Computational Fluid Dynamics) använts för att simulera processen och en numerisk lösare, som kan spåra deformationen av den rörliga smälta ytan, är utveckladi programvaran OpenFOAM. Beträffande laserstrålens utbredning visar resultaten att svetssmältans storlek och även svetssmältansflöde modifieras genom att laserstråleformen förlängs. Medan den förra är känd från tidigare experimentella studier upptäcktes den senare inte före denna studie eftersomdet icke-transparenta mediet hindrar optisk observation. Nya (in-process) kvantitativa mätningar utförda av en projektpartner har använts för att testa modellerna. Svagheter i den tidigare absorptionsmodellen framhävdes, liksom begränsningarna i den föreslagna modellen. Slutligen visade resultaten att den föreslagna modellen där laserabsorptionen är en funktion av lokala ytförhållanden ledde till en bättre overensstämmelse med mätningar jämfört med den tidigare modellen med konstant laserabsorbtion. Den maximala avvikelsen jämfört med experimentell mätning, som observerades med avseende på smältbassängsdjupet, kunde reduceras till cirka 10%.

Identifying two-phase flow patterns is fundamental to successfully design and subsequently optimize highprecision heat transfer equipment, given that the heat transfer efficiency and pressure gradients occurring in such thermo-hydraulic systems are dependent on the flow structure of the working fluid. This paper shows that with visualization data and artificial neural networks, the flow pattern images of condensation of R-134a refrigerant in inclined smooth tubes can be classified with more than 98% accuracy. The study considers 10 classes of flow pattern images acquired from previous experimental works for a wide range of flow conditions and the full range of tube inclination angles. Although not the focus of this paper, the use of a Principal Component Analysis allowed feature dimensionality reduction, dataset visualization, and decreased associated computational cost when used together with multilayer perceptron neural networks. In addition, the superior two-dimensional spatial learning capability of convolutional neural networks allowed improved image classification and generalization performance. In both cases, the classification was performed sufficiently fast to enable real-time implementation in two-phase flow systems.

In this paper, an adaptive neuro-fuzzy inference system (ANFIS) with fuzzy C-means clustering (FCM) structure identification is proposed to model condensation heat transfer and flow regimes in flattened smooth tubes with different aspect ratios. The FCM-ANFIS model was trained by using experimental data points for six effective chosen parameters of saturation temperature, heat flux, mass flux, aspect ratio and hydraulic diameter of the flattened tube, and vapor quality. Three flow regimes of annular flow, stratified, and intermittent flow were linked to the effective parameters based on the experimental data. Two models were proposed to predict the condensation heat transfer coefficient and the flow regime of R134a and R410a in flattened smooth tubes. Three statistical criteria were used to ascertain the accurateness of the models compared to the experimental results. It is found that while among benchmarked cases, the proposed model for the condensation heat transfer coefficient performs well, the best result with the lowest error (MAE = 0.029, RMSE = 0.036 and MRE = 2.83%) is when T-sat = 45 degrees C, q ‘’ = 10 kW/m(2), G = 100 kg/m(2). s, beta = 6 and D-h = 2.3 mm. On the other hand, in the worst-performing case when the errors are MAE = 0.239, RMSE = 0.239 and MRE = 13.36%, the predicted results are still in the uncertainty range of the experimental result when T-sat = 45 degrees C, q ‘’ = 5 kW/m(2), G = 200 kg/m(2). s, beta = 6 and D-h = 2.3 mm.

Fluid dynamics models for laser material processing with metal fusion in conduction mode generally assume a constant absorptivity. This parameter is known to govern the process. However, it used to be pre-set by extrapolating absorptance measurements made at different conditions or adjusted to reproduce experimental bead shapes. In this study a new approach isd eveloped. It consists in predicting the absorptance as a function of local surface conditions, including the surface temperature. The proposed absorptance model is applied to the metal alloyTi-6Al-4V. It is found that the absorptance of this alloy changes with surface temperature over awide range of beam incidence angles. Thermo-fluid simulations with tracking of the free-surface deformation are performed for conduction mode beam welding test cases with a Yb fibre laser and different travel speeds. It is found that the absorptivity coefficient commonly used for this process clearly underestimates the absorptance and the melt pool geometry predicted for the process conditions of this study. The computational results are also compared against experimental results and good quantitative agreement of the melt pool depth, width, length, free surface contour geometry, and the curvature of the end depression left afterre-solidification at the laser switch-off location is obtained. The results show that the absorptance field predicted depends on the melt pool development stage, on the spatial location within the beam spot, and on the process conditions.

The effect of different shapes of laser beam power density distribution was investigated numerically with respect to the thermo-hydrodynamics of the melt pool during welding. The process addressed is conduction mode bead on plate welding of the Titanium alloy Ti-6Al-4V. A new solver based on the volume of fluid method to track the deformation of the melt free surface was developed in the OpenFOAM software. Experiments were conducted for the purpose of validating the model. In addition to the traditional cross-cut images of the weld bead, top view images of the melt pool were analysed to perform the validation along the 3-space dimensions. A good agreement between numerical predictions and experimental measurements was obtained, thus promising aconfident utilization of this simulation model when investigating the influence of beam shapes on the resulting weld seam. The effectof three different beam shapes on the melt pool velocity flow, temperature fields, and melt geometry were studied. It was found that the melt pool size was largest for an elliptical power density distribution with the major axis along the welding direction. The results also showed that the laser beam with Gaussian power density distribution resulted in the deepest penetration.

In this study, natural convection of non-Newtonian power-law fluids around an array of elliptic cylinders has been investigated numerically. The governing equations have been solved using an in-house computational fluid dynamics code based on the well-known finite volume method. It is assumed that the flow and temperature fields are laminar, steady, and two-dimensional. Furthermore, due to the low-temperature difference between the tube walls and the surrounding fluid, the changes in the physical properties of the fluids are neglected. The numerical results are validated against the available experimental and numerical results. The results show that by increasing the non-Newtonian fluid power-law index, the ratio of average Nusselt number of the ith cylinder to the average Nusselt number of a single cylinder under identical thermal conditions decreases. Moreover, it is found that the increase in the ratio of the distance between elliptic centers and the elliptic vertical diameter increases the ratio of the average Nusselt number of ith cylinder to the average Nusselt number for a single cylinder. Finally, a mathematical expression is given for the array averaged Nusselt number.