This paper proposes geometry parameterization of a complete single centrifugal compressor stage and applies CFD-driven optimization using artificial neural networks and a Kriging surrogate model. Kinematic velocity triangle analysis is used to arrive at an initial design, which is then improved by using automated optimization algorithms and fundamental flow physics using CFD simulation. By allowing the design to evolve, guided by CFD, new and untested optimum designs are possible. This work deals specifically with design and optimization of the flow path. Design for structural aspects such as vibration, rotor dynamics and other mechanical aspects is outside the scope of this work. The CAD parameterization enables robust specification of the flow path geometry while maintaining a sufficiently small set of parameters for practical design space exploration. The parameterization includes an impeller with optional splitter blades, a vaneless diffuser and volute. Steady-state, RANS-based CFD is employed in the analysis of both the rotationally-periodic components and the volute geometry. Direct optimization and response surface optimization are demonstrated for rotationally periodic components to maximize design-point efficiency of the flow path using a multi-objective genetic algorithm (MOGA). Improvements in total-total isentropic efficiency of between 4 and 5 percent are achieved. Optimization of the flow path of the volute is likewise demonstrated. In the case of the volute, a Kriging response-surface model is used and a 1.4 percentage point improvement is shown. Further research in the utilization various implementations of Artificial Intelligence (AI) machine learning techniques in conjunction with parameterized turbomachinery flow paths to enable enhanced designs to be generated effectively is proposed.

A turbulent transition model has been applied to fluid flow problems that can be laminar, turbulent, transitional, or any combination. The model is based on a single additional transport equation for turbulence intermittency. While the original model was developed for external flows, a slight modification in model constants has enabled it to be used for internal flows. It has been successfully applied to such flows for Reynolds numbers that ranged from 100 to 100,000 in circular tubes, parallel plate channels, and circular tubes with an abrupt change in diameters. The model is shown to predict fully developed friction factors for the entire range of Reynolds numbers as well as velocity profiles for both laminar and turbulent regimes.

An experimentally validated computational model is used to determine quantitative information on convective heat transfer coefficients on all of the external surfaces of a generic residence. The motivating application for this investigation is the need for overall convective heat transfer coefficients to enable accurate design and analysis for devices such as roof top solar panels. The numerical solutions provide detailed distributions of the local heat transfer coefficient on all of the surfaces of the residence that are otherwise difficult to obtain by experimental measurements alone. The unsteady fluid flow behavior around residential buildings and its affect on heat transfer were found to be significant at higher wind speeds.

The canonical problem of flow over a square cylinder has been studied extensively in the scientific literature. Nevertheless, there are some critical issues which are not fully understood. Here, an extensive review of the literature is presented and brought together in a single repository. Next, remaining questions are identified such as: Which CFD models are most able to calculate fluid drag and heat transfer between the fluid and the cylinder? What are the mesh requirements for hydrodynamic and thermal analysis? How important is the blockage effect for cylinders that are placed in confined spaces (such as wind tunnels)? Do upstream effects significantly alter the results (such as upstream flow development, velocity profile, and turbulence intensity)? What aspect ratio is sufficient for a three-dimensional prism to approximate a two-dimensional square cylinder? Finally, how do three-dimensional flow patterns differ from those in two-dimensions? This manuscript attempts to answer these questions and provide practical recommendations to academic and industrial scientists. One key result from this work is the development of new correlating equations for both the drag coefficient and the Nusselt number for a wide range of Reynolds numbers and thermal conditions. The results presented here agree very well with accepted correlations from the literature, however these new correlations cover a much wider range of Reynolds numbers than previously published correlating equations.

A broad ranging study of the fluid flow in packed beds of uniform diameter spheres has been implemented by the method of numerical simulation. The parametric extent of the present study encompassed both laminar and turbulent flows, random and regular packings of the spheres, bed containments in both rectangular ducts and circular pipes, and beds of either bounded or unbounded lateral extent. The regular packings included bodycentered cubic, face-centered cubic, and simple cubic. Comparisons were made between the simulation-based results and relevant experimental data using variables that emerged from the experimental correlations. In particular, the Reynolds number range of the Ergun flow regime for packed beds of unbounded extent was identified. In addition, for random packed beds of bounded extent, very good agreement between the predicted and experimentally determined friction factors was found to exist. The level of agreement was sufficiently good to validate the simulation model and its implementation. The pressure drop-fluid flow results are presented in terms of friction factor-Reynolds number relationships. These dimensionless quantities were defined in terms of physical variables commonly used throughout the packed-bed literature.

Purpose The purpose of this study is to quantify the relationship between the fluid flow and pressure drop for perforated plates. The homogenization of non-uniform fluid flows is often accomplished by passing the fluid through perforated plates. The underlying principle for the accomplishment of flow homogenization is a tradeoff of pressure drop for flow uniformity. Design/methodology/approach The investigation, implemented by numerical simulation, is based on turbulent flow in pipes and across perforated plates. The approach is as follows: (a) to devise a model to determine pressure drop's fluid flow information from a single-aperture, (b) to obtain this information for apertures of different shapes, (c) to determine this type of information for perforated plates situated in a circular pipe, (d) to compare the entire perforated-plate pressure drop with that for a single-aperture modular and (e) to analyze two identical perforated plates in series. Findings The pressure drop results for the single-aperture modular model agreed very well with those for a whole perforated plate in a round pipe, therefore negating the need to simulate the more complex situation. In addition to the parametric study with aperture shape and Reynolds number, porosities (20-60 per cent) and plate thicknesses were also varied. The results obtained here compared favorably with experimental data. Originality/value This work demonstrates an efficient method for analyzing and obtaining useful pressure drop information for perforated plates. For the first time, the porous media approach for modeling perforated plates is compared directly to complete, full-scale perforated plate applications and identical plates in series.

Dual scales of permeability are encountered in many situations, and one such instance is in water treatment where packed beds of permeable spheres are essential. For design purposes, it is necessary to know the relationship between the rate of fluid flow and the corresponding pressure drop in the packed beds. By the use of numerical simulation, a comprehensive investigation of the pressure drop in different geometric bed configurations and different degrees of sphere permeability was performed. Three ordered packings were considered (simple cubic, body-centered cubic, and face-centered cubic) and were studied over a range of Reynolds numbers between 0.01 and 150. A convenient dimensionless parameter zeta, for sphere permeability, was varied from 0.0001 to 0.01, where larger values of zeta represent higher permeability. In addition, results for nonpermeable spheres were obtained for comparison. Multiple pressure drop correlations for the different operating parameters are presented along with displays of flow visualization to enhance the understanding of the fluid mechanics occurring in the beds. The present numerical results are compared with well-established existing experimental-based correlations for nonpermeable spheres, and excellent agreement was found for the applicable cases.

Purpose This paper aims to investigate and understand the fluid mechanics of piezometer rings, a device frequently encountered in engineering practice. Design/methodology/approach The investigation, implemented by numerical simulation, is based on turbulent flow in a pipe with a 90-degree bend. The pipe Reynolds numbers ranged from approximately 50,000 to 200,000. Two rings, with different dimensions, were investigated. Each ring consisted of four radially deployed straight segments of tubing which connect the pipe to a surrounding circular ring. The interconnections between the pipe and the ring were situated at 90-degree intervals around the circumference of the pipe. Findings The focus was directed to optimal circumferential locations of the radial connections, the optimal circumferential locations for accurate pressure measurements and the pressure drop penalty incurred by the use of a piezometer ring. For both of the investigated piezometer ring configurations, it was found that measurement locations situated just beyond the points of intermediate circumferential pressure variations were suitable for determining accurate values. The pressure drop was seen to increase because of the presence of the ring. For the smaller ring configuration, the increase in relative pressure drop was on the order 15 per cent, whereas the larger ring configuration lead to a 10 per cent increase. Originality/value This is the first attempt known to the authors to investigate and understand the fluid mechanics of piezometer rings.

Purpose The purpose of this study is to examine the physical processes experienced by a particle-laden gas due to various types of collisions, different heat transfer modalities and jet axis switching. Here, attention is focused on a particle-laden gas subjected to jet axis switching while experiencing fluid flow and heat transfer. Design/methodology/approach The methodology used to model and solve these complex problems is numerical simulation treated here as a two-phase turbulent flow in which the gas and the particles keep their separate identities. For the turbulent flow model, validation was achieved by comparisons with appropriate experimental data. The considered interactions between the fluid and the particles include one-way fluid–particle interactions, two-way fluid–particle interactions and particle–particle interactions. Findings For the fluid flow portion of the work, emphasis was placed on the particle collection efficiency and on independent variables that affect this quantity and the trajectories of the fluid and of the particles as they traverse the space between the jet orifice and the impingement plate. The extent of the effect depended on four factors: particle size, particle density, number of particles and the velocity of the fluid flow. The major effect on the heat transferred to the impingement plate occurred when direct heat transfer between the impinging particles and the plate was taken into account. Originality/value This paper deals with issues never before dealt with in the published literature: the effect of jet axis switching on the fluid mechanics of gas-particle flows without heat transfer and the effect of jet axis switching and the presence of particles on jet impingement heat transfer. The overall focus of the work is on the impact of jet axis switching on particle-laden fluid flow and heat transfer.

The method of numerical simulation has been used here to establish the heat transfer characteristics of a fluid jet impinging on a target surface. In particular, it was demonstrated that a fluid mechanic phenomenon designated as jet axis switching has a tremendous effect on both the magnitude and surface distribution of the impingement heat transfer coefficient. The neglect of this phenomenon, which has been common in the literature on jet impingement heat transfer, gives rise to significant inaccuracies in the predicted values of the heat transfer coefficient. As an essential prelude to the heat transfer analysis, the fluid mechanics of jet impingement were set forth in detail in order to document the jet-axis switching phenomenon. For this purpose, color contour and velocity vector diagrams are displayed to show the change of shape experienced by the jet as it passes through an unconfined space. The local heat transfer coefficient at all points of the impingement plate was determined. The highest values of the Nusselt number do not occur when the plate is nearest to the origin of the jet. Off-axis peaks of the local Nusselt number were found to exist at locations between z/b = 0 and 3 for respective impingement plate positions Xmax/b = 30 and 10. The numerical predictions compared favorably with experimental results from the literature.

This investigation presents numerical results for both laminar and turbulent flow and heat transfer for an in-line tube bank ranging in size from 1 to 20 tube rows over a Reynolds number range of 100-1000. Both the longitudinal and transverse pitches were fixed at 1.5D. It was demonstrated that the most useful heat transfer results were expressible as a total-tube-bank-averaged Nusselt number value, which is in contradiction to other investigations. For sufficiently lengthy tube banks, the existence of a fully developed regime characterized by an axially unchanging array-average Nusselt number was identified. It was found that the highest array-average heat transfer coefficients occurred in the initial portion of the tube bank, also in contradiction with information conveyed in some of the literature. A special case in which only a single tube in the array was thermally active was investigated in deference to experiments conducted under that condition. The present results obtained by numerical simulation compared favorably to existing experimental data. (C) 2018 Elsevier Ltd. All rights reserved.

Heat transfer coefficients for turbulent pipe flow are typically envisioned as axially varying from very high values at the pipe inlet to a subsequent monotonic decrease to a constant fully developed value. This distribution, although well enshrined in the literature, may not be universally true. Here, by the use of high accuracy numerical simulation, it was shown that the initially decreasing values of the coefficient may attain a local minimum before subsequently increasing to a fully developed value. This local minimum may be characterized as an undershoot. It was found that whenever a turbulent flow laminarizes when it enters a round pipe, the undershoot phenomenon occurs. The occurrence of laminarization depends on the geometry of the pipe inlet, on fluid-flow conditions in the upstream space from which fluid is drawn into the pipe inlet, on the magnitude of the turbulence intensity, and on the Reynolds number. However, the presence of the undershoot does not affect the fully developed values of the heat transfer coefficient. It was also found that the Fanning friction factor may also experience an undershoot in its axial variation. The magnitude of the heat transfer undershoot is generally greater than that of the Fanning friction factor undershoot.

Chapter 33 Skin Burns John P. Abraham, Corresponding Author John P. Abraham University of St. Thomas, School of Engineering, St. Paul, MN, USA Corresponding author: JPABRAHAM@stthomas.eduSearch for more papers by this authorBrian D. Plourde, Brian D. Plourde University of St. Thomas, School of Engineering, St. Paul, MN, USASearch for more papers by this authorLauren J. Vallez, Lauren J. Vallez University of St. Thomas, School of Engineering, St. Paul, MN, USASearch for more papers by this authorBrittany B. Nelson-Cheeseman, Brittany B. Nelson-Cheeseman University of St. Thomas, School of Engineering, St. Paul, MN, USASearch for more papers by this authorJohn R. Stark, John R. Stark Department of Mechanical Engineering, University of Kansas, Lawrence, KS, USASearch for more papers by this authorEphraim M. Sparrow, Ephraim M. Sparrow Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN, USASearch for more papers by this authorJohn M. Gorman, John M. Gorman Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN, USASearch for more papers by this author John P. Abraham, Corresponding Author John P. Abraham University of St. Thomas, School of Engineering, St. Paul, MN, USA Corresponding author: JPABRAHAM@stthomas.eduSearch for more papers by this authorBrian D. Plourde, Brian D. Plourde University of St. Thomas, School of Engineering, St. Paul, MN, USASearch for more papers by this authorLauren J. Vallez, Lauren J. Vallez University of St. Thomas, School of Engineering, St. Paul, MN, USASearch for more papers by this authorBrittany B. Nelson-Cheeseman, Brittany B. Nelson-Cheeseman University of St. Thomas, School of Engineering, St. Paul, MN, USASearch for more papers by this authorJohn R. Stark, John R. Stark Department of Mechanical Engineering, University of Kansas, Lawrence, KS, USASearch for more papers by this authorEphraim M. Sparrow, Ephraim M. Sparrow Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN, USASearch for more papers by this authorJohn M. Gorman, John M. Gorman Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN, USASearch for more papers by this author Book Editor(s):Devashish Shrivastava, Devashish Shrivastava US Food and Drug Administration, Silver Spring, NY, 10903 USASearch for more papers by this author First published: 27 April 2018 https://doi.org/10.1002/9781119127420.ch33Citations: 2 AboutPDFPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShareShare a linkShare onFacebookTwitterLinked InRedditWechat Summary Skin burns are very common injuries that affect people of all ages in all parts of the world. Burns are particularly harmful because damage can be life threatening or, in other cases, can require long-term care accompanied by great physical and emotional pain and economic costs. Full-thickness burns extend through the full depth of the dermal layer into the subcutaneous tissue and require extensive medical treatment. The procedure to calculate the extent of burn injuries requires two steps. First, temperatures within the tissue must be determined. Next, translating these temperatures to an injury criterion must be achieved. The aforementioned calculations were performed using a high-fidelity modeling program; however, the gold standard is, and should be, experimental/clinical evidence. Scald burns or other skin burns are serious concerns for human health and safety. It is possible to reduce the frequency and impact of skin burns by controlled exposure temperatures. Citing Literature Theory and Applications of Heat Transfer in Humans, Volume 1 RelatedInformation

The performance of the perforated plates in fluid-flow applications is evaluated by measuring the pressure drop of the working fluid. The purpose of this investigation is to determine how different parameters affect the capability of the perforated plates and modify the design by using a design of experiment analysis, namely Taguchi method for optimization. The flow characteristics, which were obtained by the CFD software package ANSYS-CFX, were used for this analysis. The design parameters which affect the pressure loss are Reynolds number (A), porosity (B), non-dimensional thickness of the plate (C), and hole pattern (D). The level of importance of the design parameters are determined by use of analysis of variance method. According to the analysis, the optimum values are obtained for the case A8B2C2D1 (Re = 15000, porosity = 50.3, t / D = 1, and staggered hole). The most effective design parameter on the results is found as porosity (92%), while the least effective is the hole pattern (0.2%). A special dividend of this work was to demonstrate the capabilities of the Taguchi method as a powerful means of increasing the effectiveness of numerical simulation.

The pressure loss caused by a single isolated bend, which receives a fully developed flow and allows the flow to return to a fully developed state in its downstream tangent, is well documented and understood. However, when two or more bends are separated by a length insufficient to restore fully developed flow, the truncation of the pressure loss occurring downstream of the first bend and the distorted flow received by the second bend prevent the extrapolated use of single-bend data for the two-bend case. Using numerical simulation, a comprehensive investigation of the pressure losses caused by two-bend combinations was performed. Two in-plane bend combinations, having a "U" and "S" shape, and an out-of-plane bend combination were examined. For each bend combination, simulations were performed for many separating lengths ranging from zero to 100 pipe diameters for Reynolds numbers of 2. 10(5), 5.10(4), and 1.10(4). For the "U" combination, the two-bend pressure loss decreased monotonically as the separating distance decreased. For the "S" and out-of-plane configurations, the pressure loss was at a minimum when the separating length was between three and five pipe diameters. This is consistent with the existing results from the highest quality experiments on the topic. In addition to results, the present work presents a fundamentals-based comprehensive explanation of the observed pressure loss trends using dimensionless quantities to measure the distortion of the flow at cross sections within the bend combination.

A new model for fluid flow of a liquid through a hollow fiber filtration cluster has been set forth and numerically implemented. The key feature of the approach is modeling of the intrinsic permeability of the individual fibers. This fiber permeability, in addition to the geometrically based permeability of the cluster, gives rise to a situation characterized by two scales of permeability. The fluid mechanic problem is governed by the Navier-Stokes equations supplemented by pressure loss terms for the fiber wall given by the Darcy porous medium model. A number of independent parameters were investigated: the Reynolds number, the magnitude of the permeability, the geometric arrangement of the fibers, the center-to-center separation distances of the fibers, the outer to inner diameter ratio of each fiber, and the boundary condition at the inner surface of a fiber. It was found that for reasonable values of the prescribed intrinsic permeability, significant reductions in the pressure drop occur. Also considerably affected are the patterns of fluid flow within the fiber cluster. It was found that the fiber permeabilities have a major impact on the pressure drop results, suggesting predictions based on the impermeable fibers are, for the most part, in error.

A numerical-based model was developed and implemented to determine the spatial and temporal temperature distributions within skin tissue resulting from thermal contact with a heated and high thermal conductivity metallic medium. In the presence of wet tissue, boiling is likely to occur, thereby affecting the probability of inducing burns. This investigation deals with how contact between a hot, highly conductive metallic material and skin gives rise to burns. In particular, the study focuses on the likelihood that metals typically used in cooking or industrial applications may cause burns. Insofar as the surfaces under consideration are above the boiling temperature of water, a mathematical model including phase change was developed. That model allowed different thermophysical properties to be respectively employed for dry and wet tissues. Multiple processes and their governing parameters were investigated to assess their impact on burn severity, including the temperature of the metal, the duration of contact, the contact resistance between the surface and the skin, the temperature range over which phase change occurred, and the cooling environment after the exposure. It was discovered that the most important parameters are the surface temperature and exposure duration. The other conditions/parameters had lesser impacts on the results.

Abstract This paper describes a three-part numerical investigation of fluid flow and heat transfer in a never-previously-studied pipe bend situation. The investigation deals with downstream fluid-flow and heat transfer processes which are affected by upstream flow disturbances. The studied physical situation is a 90° pipe bend fitted with a wall-adjacent obstruction that partially blocks the inlet cross section. The first phase of the work consisted of validating numerical simulation results with experimental data. In the second phase, the impact of the inlet flow distribution on the pressure drop is determined. Heat transfer downstream of the bend exit comprises the third section of the paper. The heat transfer results are reported in terms of the circumferentially averaged Nusselt numbers which are displayed as a function of axial position for Reynolds numbers between 100 and 10,000. It was found that the disturbances caused by the blockage significantly enhance the Nusselt number values.

The existence and characteristics of an overshoot phenomenon in the axial velocity distribution that occurs at the centerline of a turbulent pipe flow is investigated and documented by means of numerical simulation. A complementary phenomenon is also encountered in which the axial variation of the wall shear stress experiences an undershoot. These occurrences are not restricted to the case of a uniform velocity profile at the pipe inlet. The magnitude of the inlet turbulence intensity was found to play a major role in the downstream development of the flow. In particular, the magnitude of the overshoot showed a dependence on the value of the inlet turbulence intensity; the higher the intensity value, the lower the magnitude of the overshoot. Evidence was presented that enabled the attribution of the velocity peak and the wall shear undershoot to an initial tendency for the flow to laminarize. In particular, the presence of the velocity peak was related to different patterns of radial flow. When a peak was present, there was a radial inflow of fluid toward the centerline of the pipe followed downstream by a radial outflow from the centerline to the wall. The suppression of the velocity peak was accomplished by a very high value of the turbulence intensity at the inlet which neutralized the tendency towards laminarization. Other evidence of the laminarization tendency was obtained by examining the magnitude of the turbulence viscosity.