Due to their efficient heat and mass transfer potential, impinging jets have received attention in various applications. Heat transfer and flow characteristics of rectangular turbulent impinging jets issued from a 24:1 aspect ratio and 24:1 contraction ratio nozzle were investigated experimentally and numerically. In the heat transfer measurements; a thin stainless-steel foil was utilized to obtain iso-flux boundary conditions on the impingement surface. The target plate was free to translate in the lateral direction and the heat transfer distributions were determined at 0 ≤ x/W ≤ 20 with the micro-thermocouples placed underneath the foil. The measurements were conducted for Re(j) = 8900 − 48600 at nozzle-to-target spacing of 0.5 ≤ H/W ≤ 12.0. Both semi and fully confined jets were investigated. Heat transfer coefficients at Re(j) = 28100, 36800, 45600 and H/W = 4.0 were determined by using adiabatic-wall temperatures and the distributions were compared with those of the wall shear stress. Off-center peaks were observed at high Re(j) and low H/W. Since the wall distributions are susceptible to nozzle-exit conditions, velocity and turbulence profiles at the nozzle-exit were measured for the velocity range of interest. Additionally, near-wall mean velocity and turbulence profiles were determined at Re(j) = 21500 and 36800 at H/W = 4.0 to have a better understanding of the secondary peaks in the wall distributions. Numerical computations were performed by using several different turbulence models (k − ω, k − ε, V 2F and Reynolds stress models). In wall-bounded turbulent flows, near-wall modeling is crucial. Therefore, the turbulence models eliminating wall functions such as the k − ω and V 2F models may be superior for modeling impingement flows. The numerical results showed reasonable agreement with the experimental data for local heat transfer and skin friction coefficient distributions. The occurrence of the secondary peaks was predicted by the k − ω and V 2F models, and for a few cases with the low-Re-k − ε models. Near-wall measurements along with the computed profiles were used to describe the “secondary peak” phenomena. It was shown that the increase in turbulence production in the wall-streamwise direction enhances turbulent momentum and heat transport in the wall-normal direction which lead to secondary peaks in the wall distributions. The possibility of improving surface heat transfer with fully-developed jets was also explored numerically as a case study.

A fundamental study of two turbulent directly opposed impinging jets in a stagnant ambient fluid, unconfined or uninfluenced by walls is presented. By experimental investigation and numerical modeling, the main characteristics of direct impingement of two turbulent axisymmetric round jets under seven different geometrical and flow rate configurations (L*= L/d = { 5, 10, 20 }, where L is nozzle to nozzle separation distance and d is nozzle diameter, and Re = { 1500, 4500, 7500, 11000 }) are discussed. Flow visualization and velocity measurements performed using various laser based techniques have revealed the effects of Reynolds number, Re, and nozzle to nozzle separation, L, on the complex flow structure. Although locally valid, the classical analysis of turbulence is found unable to provide reliable results within the highly unstable and unsteady impingement region. When used to simulate the present flow, the assessment of the performance of three distinct k - epsilon turbulence models showed little disagreement between computed and experimental mean velocities and poor predictions as far as turbulence parameters are concerned.

A fundamental study of two turbulent directly opposed impinging jets in a stagnant ambient fluid, unconfined or uninfluenced by walls is presented. By experimental investigation and numerical modeling, the main characteristics of direct impingement of two turbulent axisymmetric round jets under seven different geometrical and flow rate configurations (L*= L/d = { 5, 10, 20 }, where L is nozzle to nozzle separation distance and d is nozzle diameter, and Re = { 1500, 4500, 7500, 11000 }) are discussed. Flow visualization and velocity measurements performed using various laser based techniques have revealed the effects of Reynolds number, Re, and nozzle to nozzle separation, L, on the complex flow structure. Although locally valid, the classical analysis of turbulence is found unable to provide reliable results within the highly unstable and unsteady impingement region. When used to simulate the present flow, the assessment of the performance of three distinct k - epsilon turbulence models showed little disagreement between computed and experimental mean velocities and poor predictions as far as turbulence parameters are concerned.

Thermoelectric generator (TEG) is a solid state technology based on the Seebeck effect that can generate electrical power from waste heat. For continuous electrical power generation heat exchangers are integrated into the “cold side” and the “hot side” of the TEG such that a temperature difference across the TEG can be established and maintained. This thesis will focus on characterizing two different flat-plate cold side heat exchanger prototypes specifically designed for dissipating the thermal loads from TEG modules. The majority of TEGs modules available have a flat geometry design and a square shape with typical dimension of 40 mm × 40 mm or 56 mm × 56 mm. To maximize the net electrical power generated by the TEGs the cold side heat exchanger is required to have uniform surface temperature distribution, and excellent heat transfer performance with minimum pressure drop. To achieve the previously mentioned requirements, two flat-plate heat exchanger prototypes having two distinct heat transfer techniques were investigated. Each heat exchanger is designed to accommodate an array of 14 TEG modules arranged in two parallel rows with 7 TEGs per row a typical arrangement for large waste energy harvesting applications. The first heat exchanger prototype utilized single-phase forced convection through 140 minichannels (1 mm × 1 mm × 90 mm long) as a heat transfer technique. The second prototype utilized 14 liquid jets, 3 mm in diameter and 40.3 mm apart, impinging on a flat surface located 5 mm above. Each impinging jets was positioned at the centre of the TEG cooling area. An experimental facility was constructed in order to test the minichannels heat exchanger and the impinging jets thermally and hydrodynamically. The heat transfer, pressure drop and temperature distribution results were compared to determine the most appropriate cold side heat exchanger prototype for the TEG POWER system. The TEG POWER system is a waste heat recovery system designed to recoup waste heat from the exhaust gases of commercial pizza ovens. The TEG POWER system is capable of harvesting waste thermal energy produced by an establishment and utilize it for electrical power generation and thermal storage purposes. Heat transfer results indicated that for a given mass flow rate the minichannels heat exchanger has better heat transfer performance compared to the impinging jets heat exchanger. The minichannels heat exchanger design had a thermal conductance of 238 W/C at 0.19 kg/s coolant flow rate compared to 111 W/C for the impinging jets heat exchanger. The total pressure drop and the minor losses for each heat exchanger prototype were measured experimentally. For the minichannels heat exchanger, the total pressure drop is 23.3 kPa at flow rate of 0.235 kg/s. Comparatively, the total pressure drop for the impinging jets heat exchanger was 27.4 kPa at the same flow rate. Fittings losses for the minichannels heat and impinging jets heat exchanger were found to be 50% and 80% respectively. The maximum total measured drop corresponded to pumping power requirements of 5.7 W and 6.8 W for the minichannels and impinging jets heat exchanger respectively. Local and average temperature distributions and their influence on the electrical power generated were studied for both heat exchanger prototypes. It was found that the minichannels heat exchanger offers more uniform surface temperature distribution per row of TEGs compared to the impinging jets heat exchanger. Therefore the minichannels heat exchanger is well suited for cooling two rows of TEGs simultaneously. Based on the thermal and hydrodynamics comparison results the minichannels heat exchanger prototype is recommended for implementation in the TEG POWER system. / Thesis / Master of Applied Science (MASc)

Les instabilités aéroélastiques de bandes d’acier constituent aujourd’hui l’un des problèmes majeurs dans les sections de refroidissement par jets impactants des lignes de recuit continues.En effet, le traitement thermique des nouveaux aciers nécessite de très fortes pentes de température impliquant constamment des augmentations de vitesse de soufflage susceptibles de mettre en jeu des instabilités aéroélastiques. Des flottements ainsi que des divergences de bande ont déjà été constatées et identifiées. Ces deux instabilités impliquent dans la plupart des cas des chocs entre la bande et les buses de soufflage ce qui engendre des défauts de surface sur la bande.Un banc d’essai a été conçu et fabriqué dans le but d’analyser ces instabilités et d’anticiper leur apparition. A partir d’observations, la dynamique structurelle de la bande a été simplifiée à un mode de rotation rigide. Le banc comporte une plaque oscillante en mouvement forcé.Celle-ci est impactée par un dispositif de plusieurs jets axisymétriques turbulents ayant une disposition identique à celle des tours industrielles. Les efforts aérodynamiques stationnaires et instationnaires agissant sur la plaque sont mesurés au moyen de capteurs de pression.L’impact de plusieurs jets en interaction crée de très importants gradients de pression sur la plaque il est donc nécessaire que la grille de prises de pression soit très fine pour que l’estimation des efforts aérodynamiques soit correcte. La plaque est donc instrumentée de 91capteurs de pression sur une surface de 18 cm². Elle peut également être translatée dans les ois directions de l’espace, ce qui permet d’obtenir la distribution des efforts instationnaires ainsi que des coefficients aéroélastiques sur une grande surface de plaque et à différentes distances d’impact.Les mesures de pression stationnaires ont permis d’établir les courbes d’évolution des efforts d’impact des jets sur la plaque en fonction de la distance jet-plaque ainsi que de la géométrie des buses. Les résultats ont permis de déterminer la stabilité statique de la plaque en mouvement de pompage. Les mesures de vitesses des jets libres ont été effectuées paranémométrie à fil chaud et ont permis de déterminer leurs propriétés statistiques.Les mesures de coefficients aéroélastiques sur la plaque en rotation ont été effectuées surune seule géométrie de soufflage, pour différentes vitesses réduites. Les résultats mettent en évidence l’importance des effets de bords sur la stabilité de plaque. Des méthodes de post traitements ont proposées afin d’extrapoler les résultats à différentes largeurs de bande. Ils sont confrontés aux travaux de Regardin et al. (réf. [1]) et mettent en évidence des désaccords avec le cas réel. Des suggestions sont apportées afin d’améliorer la représentativité du banc vis-à-vis des bandes industrielles. / Aeroelastic instabilities of steel strips impinged by arrays of cooling gas jets have becomeone of the main issues in cooling sections of continuous annealing lines. Indeed, the new steeltreatments require very high temperature variation rates which involve increases in jetvelocities that are likely to onset some aeroelastic instabilities. Strip flutter and divergencehave already been observed and identified. These two aeroelastic instabilities imply a strongrisk of contact with the blowing boxes, which can seriously blemish the strip.An experimental test rig was designed and built in order to analyze and predict of theseinstabilities. From observations, the strip’s structural dynamics was simulated by a rigidrotation mode. The rig included a forced oscillating plate which is impinged by an array ofaxisymmetric jets having the exact industrial geometry. The plate was instrumented withpressure sensors to measure the steady and unsteady surface pressures. The impingement ofinteracting jets creates very large pressure gradients on the plate and therefore a tight mesh ofpressure taps (91 over an 18cm² jet impingement surface) was necessary to allow a goodestimation of the aerodynamic loads The plate could also be moved in the three coordinatedirections as to obtain surface mappings of the unsteady jet forces and aeroelastic coefficientscan be obtained over a wide area and different jet-to-plate distances.The variation of the impinging aerodynamic forces was established as a function of the jetto-plate distance for different nozzle geometries. These results were used to determine the jetstatic stability in plunging motion. Velocity and turbulence measurements in free jets werecarried out using hot wire anemometry in order to determine their statistical properties.Aeroelastic coefficient measurements were carried out on the oscillating plate with onlyone nozzle geometry and different reduced velocities. Results show that the plate’s stability ismainly dependent on the boundary effects. Post processing methods are suggested in order toapply the results to larger plates. Results are compared to the data of Regardin et al. (réf. [1])and emphasize some discrepancies with respect to the real case. Finally some improvementsto the test-rig are suggested for it to be more representative of the industrial situation.

L’objectif de ces travaux est d’étudier expérimentalement la dispersion d’hydrogène pour évaluer l’impact des fuites chroniques ou accidentelles qui peuvent intervenir en milieu libre ou encombré, à faible ou à forte pression, sur un véhicule fonctionnant avec une pile à combustible. Les fuites étudiées sont assimilées à des jets verticaux, turbulents, axisymétriques, à densité variable, et issus d’orifices cylindriques de 1 à 3 mm de diamètre. Un banc expérimental a été conçu pour étudier ces fuites : l’hydrogène a été remplacé par de l’hélium pour des raisons de sécurité. Il résiste à une pression de 200 bars et permet de positionner un obstacle dans le jet. La technique BOS (Background Oriented Schlieren) a été adaptée aux jets millimétriques et à la présence d’un obstacle. Un soin particulier à été apporté à la mise en place de cette technique. Les résultats obtenus sont en accord avec ceux de la littérature quand il en existe. De nouvelles lois de similitude sont proposées, plus représentatives de la physique des jets : libres subsoniques, libres sousdétendus, et subsoniques en présence d’un obstacle (sphère). À partir de l’analyse de la structure compressible des jets sous-détendus, de nouvelles lois sont aussi proposées pour estimer la position et le diamètre du disque de Mach puis la longueur du cône potentiel. Enfin, deux lois d’estimation du volume et de la masse inflammables de jets libres d’hydrogène sont proposées : elles sont fonction du débit massique de la fuite. / The aim of this work is to experimentally investigate the hydrogen dispersion to evaluate the impact of chronic or accidental leaks that may occur in a free or in a congested environment, from a low or a high pressure tank, on a fuel cell vehicle. The leaks are assimilated to vertical turbulent and axisymmetric jets with variable density. They are issued from cylindrical orifices from 1 to 3 mm diameter. An experimental set-up was designed to investigate the leaks : hydrogen has been replaced by helium for safety reasons. It supports a 200 bar pressure and allows to position an obstacle in the jet flow. The BOS (Background Oriented Schlieren) technique has been adapted to millimeter jets and to the presence of an obstacle. Particular attention has been given to the development of this technique. The results show good agrement with the available litterature data.New similarity laws are proposed, more representative of the flows of : subsonic free jets, under-expanded free jets, and subsonic impinging (a sphere) jets. From the structure analysis of compressible under-expanded free jets, new relations are proposed to estimate the Mach disk position and diameter as well as the potential core length. Finally, two new laws are found from the analysis of flammable volume and flammable mass : they are based on the leak mass flow rate.

Regarding the great importance of fast cooling in steel industry for the production processes, a deep understanding of heat transfer and fluid dynamics must be held. A steel plate is heated up until a maximum temperature of 700 ⁰ C to be the n cooled down seconds later by a configuration of multi ple impinging water jets. Different flow rates are used simultaneously by different adjacent jets to perform quenching over the sample, so different hardness is obtained in the material over a small area . Temperature drop in time i s measured and monitored by embedded thermocouples and LabVIEW program. To achieve greater understanding of the quenching performance with different flow rates , several parameters are selected to be varied in order to achieve the best working conditions. Jet diameter takes values between 4 and 10 mm, initial temperature of quenching varies from 400 to 700 ⁰ C , subcooling temperature is tested for 65 and 75 K, and jet velocity varies between 1.9 and 3.9 m/s. The result of total number of 9 expe riments shown that v ariation of jet diam eter does not influence substantially on the cooling rate if flow rate is kept constant. High initial quenching temperature (600 - 700 ⁰ C ) led to slightly higher cooling rate in the stagnation region of water jets. The peak value of heat transfer rate in the upwash flow zone was more highlighted for initial quenching temperature 600 ⁰ C and below it. High er values of subcooling and jet velocity produce better cooling rates. The result shown higher jet velocity at one column of water jets changes position of upw ash flow slightly toward the adjacent column of jets with lower jet velocity. In general, the result shown that all the studied parameters did not have negative effect on obtaining various cooling rates over the steel plate.

Cette thèse est consacrée à l'étude des propriétés aéroacoustiques de jets supersoniques impactant une paroi par simulation des grandes échelles. Ces simulations sont réalisées à partir des équations de Navier-Stokes 3-D instationnaires compressibles exprimées pour des coordonnées cartésiennes ou cylindriques. Afin de résoudre ces équations, des schémas numériques de différenciation spatiale et d'intégration temporelle peu dispersifs et peu dissipatifs sont utilisés. Les écoulements étudiés étant supersoniques, une procédure de capture de choc est également implémentée afin de supprimer les oscillations de Gibbs de part et d'autre des chocs.Dans un premier temps, un jet rond libre et quatre jets ronds impactant une paroi avec un angle de 90 degrés sont simulés sur des maillages cylindriques. Ces jets sont supersoniques, sous-détendus, et sont caractérisés par un nombre de Reynolds calculé à partir du diamètre du jet de Re=60.000, et par un nombre de Mach parfaitement détendu de Mj=1.56. Les résultats du jet libre sont tout d'abord présentés. Ils sont comparés aux résultats de plusieurs études expérimentales et de modèles afin de valider l'approche numérique utilisée. Notamment, les différentes composantes acoustiques spécifiques aux jets sous-détendus comme le bruit de choc large-bande et le bruit de screech sont observées et analysées. Les résultats obtenus pour les quatre jets impactant une paroi sont ensuite examinés. Dans ce cas, la présence d'une boucle de rétroaction aéroacoustique entre les lèvres de la buse et la paroi est montrée. Pour finir, le comportement aérodynamique et aéroacoustique des jets est étudié, et comparé à différentes études numériques et expérimentales de la littérature. Quatre jets plans supersoniques idéalement détendus impactant une paroi avec un angle de 90 degrés sont ensuite calculés. Ils ont un nombre de Reynolds évalué à partir de la hauteur de la buse de Re=50.000 et un nombre de Mach de Mj=1.28. Une boucle de rétroaction aéroacoustique entre la buse et la paroi est de nouveau mise en évidence. Une combinaison de modèles associant un modèle d'onde stationnaire aérodynamique-acoustique et un modèle de stabilité de jet plan 2-D avec des couches de mélange infiniment minces est alors proposée. Ce modèle permet de déterminer à la fois les fréquences les plus probables de la boucle de rétroaction aéroacoustique et leurs natures plane ou sinueuse.Enfin, les simulations de deux jets plans supersoniques impactant une paroi avec des angles de 60 et 75 degrés sont réalisées grâce à l'utilisation de deux maillages cartésiens, par une méthode de recouvrement de maillages. Les modifications des propriétés de la boucle de rétroaction aéroacoustique lorsque l'angle d'impact dévie de 90 degrés sont ainsi étudiées. / In this PhD work, supersonic impinging jets are simulated using large-eddy simulation in order to investigate their aerodynamic and acoustic fields. In practice, the unsteady compressible Navier-Stokes equations are solved on Cartesian or cylindrical meshes. Low-dissipation and low-dispersion numerical methods are used for spatial differentiation and time integration. As the jets are supersonic, a shock-capturing filtering is also applied in order to avoid Gibbs oscillations near shocks.A free round jet and four round jets impinging normally on a flat plate are first simulated on cylindrical meshes. They are underexpanded, and have a Reynolds number based on the nozzle diameter of Re=60.000 and a fully expanded Mach number of Mj=1.56. The results for the free jet are first presented. They are compared with experimental results and predictions given by models in order to validate the numerical setup. Acoustic components specific to underexpanded jets such as broadband shock-associated noise and screech noise are obtained. The results for the four impinging jets are then examined. An aeroacoustic feedback mechanism establishing between the nozzle lips and the flat plate is found to generate tones. Finally, the flow and acoustic properties of the jets are studied and compared with numerical and experimental data.Four ideally expanded jets impinging normally on a flat plate are then simulated. They have a Reynolds number based on the nozzle height of Re=50.000 and a Mach number of Mj=1.28. An aeroacoustic feedback mechanism is again observed between the nozzle lips and the flat plate. A combination of models based on an aeroacoustic feedback model and a vortex sheet model of the jet is then proposed. The model appears able to predict the most likely tone frequencies of the feedback mechanism, and the symmetric or antisymmetric nature of the corresponding jet oscillation.Finally, two ideally expanded jets impinging on a flat plate with angles between the jet direction and the plate of 60 and 75 degrees are simulated using two Cartesian meshes. The effects of the angle of impact on the properties of the aeroacoustic feedback mechanism are finally studied.

Numerical solutions based on explicit filtered LES for computing turbulent flow field, of free round jets and impinging round jet on cuboidal cavities, are presented and discussed in this dissertation work. One-parameter fourth-order explicit filter is implemented to account for sub-grid scale effects. Compact difference schemes proposed by Hixon & Turkel involving only bidiagonal matrices is used to evaluate spatial derivatives. Compact schemes with overall fourth order accuracy and eight order accuracy are used in simulating free and impinging jets respectively. Simulations of free round jets are used for validating LES approach. 6 simulations of free round jet, in three levels of computational grids at three different Reynolds number, are performed to understand the effects of Reynolds number and turbulent length scales. Energy in the smaller length scales are found to be higher for higher Reynolds number. Potential core collapse is found to occur at shorter distance for high Reynolds number jets. Accurate computation of smaller length scales of turbulence is found to be essential for high Reynolds number flows. LES of subsonic impinging jets are performed on cuboidal cavities to understand the physical phenomenon. High intensity, low frequency sounds are captured, in the presence of cavity, as reported by other research works. Lip-thickness is found to have an effect on the intensity of sound produced. Matching of Jet shear layer roll up frequency with cavity’s natural frequency to produce resonance phenomenon is attempted and observations are presented.

Heat transfer brought about by impinging jets is common in several industrialapplications such as the cooling of heated objects. The purpose of this investigationis to create a numerical model of an impacting jet, then to use it in parametricresearch and in the cooling of an electronic device. In this work, COMSOL Multiphysics software has been used to carry out a numericalanalysis of a cooling jet impacting a vertical surface. Water and air are the two fluidscompared to each other, and three different materials were selected for modelling theheated object. The model created can be used in many areas. For example, the caseof the air jet hitting an aluminum object is commonly used in the cooling of electronicdevices while the water jet hitting a brick is used in building and construction. Inother words, the highlight of this project is to create a useful tool that can be easilymodified in order to investigate a specific area. To set an example, the cooling of aninsulated gate bipolar transistor (IGBT) has been investigated in this work. Equations related to the conservation of mass and momentum, coupled to the energyconservation equation, have been solved. Several assumptions were carefully andreasonably considered to simplify the simulation and ensure the accuracy andreliability of the results. The effect of the type of the heated material and the distancebetween the tube nozzle and the heated object have been investigated in the study.The results show that the jet is effective in cooling heated objects and can be used inparametric research. In fact, surface temperature and jet velocity have been displayedand discussed and it has been found that the most optimal combination for efficientheat transfer is a shorter distance between the tube nozzle and the heated object, amaterial with higher thermal conductivity, and water as the jet fluid. The impingingjet device has been tested in the cooling of Insulated Gate Bipolar Transistor (IGBT)and the results showed that the temperature of the IGBT drops by 14.57%. Despite the accuracy of the results, it is important to recognize the limits of thenumerical model such as the discretization of the physical domain, the resolution ofthe mesh, the assumptions, and the simplifications. In addition, this investigation waslimited to three variable factors. Therefore, further studies are recommended tofurther optimize the cooling effect of the jet, such as the study of nozzle shape, jetvelocity, jet mass flow rate and a 3D-simulation. / Värmeöverföring som orsakas av påverkande strålar är vanligt förekommandeinom flera industriella tillämpningar, såsom kylning av uppvärmda objekt. Syftetmed denna undersökning är att skapa en numerisk modell av en påverkande stråleoch sedan använda den i parametrisk forskning och för att kyla en elektroniskenhet. I detta arbete har COMSOL Multiphysics-programvaran använts för att genomföraen numerisk analys av en kylstråle som påverkar en vertikal yta. Vatten och luft ärde två vätskor som jämförs med varandra, och tre olika material valdes för attmodellera det uppvärmda objektet. Den skapade modellen kan användas inommånga områden. Till exempel används fallet med luftstrålen som träffar ettaluminiumobjekt vanligtvis vid kylning av elektroniska enheter, medanvattenstrålen som träffar en tegelsten används inom bygg- ochkonstruktionsbranschen. Med andra ord är höjdpunkten i detta projekt att skapa ettanvändbart verktyg som lätt kan modifieras för att undersöka ett specifikt område.Som ett exempel har kylningen av en isolerad gatebipolär transistor (IGBT) undersökts i detta arbete.  Ekvationer relaterade till mass- och rörelsebevarande, kopplade tillenergibevarelsesekvationen, har lösts. Flera antaganden beaktades noggrant ochrimligtvis för att förenkla simuleringen och säkerställa noggrannheten ochtillförlitligheten hos resultaten. Effekten av uppvärmt materials typ och avståndetmellan tubmunstycket och det uppvärmda objektet har undersökts i studien.Resultaten visar att strålen är effektiv för att kyla uppvärmda objekt och kananvändas inom parametrisk forskning. Faktum är att yttemperaturen och strålenshastighet har visats och diskuterats, och det har konstaterats att den mest optimalakombinationen för effektiv värmeöverföring är ett kortare avstånd mellantubmunstycket och det uppvärmda objektet, ett material med högre termiskledningsförmåga och vatten som strålens fluid. Den påverkande strålenheten hartestats för att kyla en isolerad gatebipolär transistor (IGBT) och resultaten visadeatt temperaturen på IGBT sjunker med 14,57%. Trots resultatens noggrannhet är det viktigt att erkänna begränsningarna hos dennumeriska modellen, såsom diskretiseringen av det fysiska området, upplösningenpå nätet, antaganden och förenklingar. Dessutom var denna undersökningbegränsad till tre variabla faktorer. Därför rekommenderas ytterligare studier för attytterligare optimera strålens kylningseffekt, såsom studiet av munstyckets form,strålens hastighet, strålens massflödesintensitet och en 3D-simulering.