A Review of X-Ray Flow Visualization With

A Review of X-Ray Flow Visualization With Applications to Multiphase Flows Theodore J. Heindel Bergles Professor of Thermal Science, Department of Mechanical Engineering, Iowa State University, Ames, IA 50011 e-mail: [email protected]

Flow visualization and characterization of multiphase flows have been the quest of many fluid mechanicians. The process is fairly straight forward only when there is good optical access (i.e., the vessel is not opaque or there are appropriate viewing ports) and the flow is transparent, implying a very low volume fraction of the dispersed phase; however, when optical access is not good or the fluid is opaque, alternative methods must be developed. Several different noninvasive visualization tools have been developed to provide high-quality qualitative and quantitative data of various multiphase flow characteristics, and overviews of these methods have appeared in the literature. X-ray imaging is one family of noninvasive measurement techniques used extensively for product testing and evaluation of static objects with complex structures. X-rays can also be used to visualize and characterize multiphase flows. This paper provides a review of the current status of X-ray flow visualization and details various X-ray flow visualization methods that can provide qualitative and quantitative information about the characteristics of complex multiphase flows. [DOI: 10.1115/1.4004367]

1

Introduction

Multiphase flows, composed of gas-liquid, gas-solid, liquid-solid, or gas-liquid-solid mixtures, are commonly found in many process industries such as petroleum-based fuel production, energy generation, commodity and specialty chemical production, mineral processing, textile processing, pulp and paper processing, wastewater treatment, food processing, and biological organism and pharmaceutical production. Dudukovic [1] states that “the heart of chemical transformations in all process and energy industries is multiphase reactor technology, as over 99% of reactor systems require the presence of more than one phase for proper operation.” These flow systems are used to promote solid and/or liquid separation and enhance heat and/or mass transfer. Although the uses of these systems are extensive [2,3], their operation is very complex and an improved understanding of the fundamental hydrodynamic and transport processes is necessary to develop process improvements and optimization, as well as to develop and validate fundamental models of their operation [4,5]. It has been repeatedly stressed that data obtained using a variety of techniques, including noninvasive measurement techniques, are necessary to improve and validate multiphase flow models [2,4,6,7]. With this information, more economical operations can be achieved. The principal difficulty in characterizing and quantifying multiphase flows is the fact that the systems are typically opaque; even an air-water system becomes opaque at fairly low volumetric gas fractions. This necessitates either the use of invasive measurement probes when determining internal flow and transport characteristics or noninvasive (nondestructive) methods. The challenge with invasive probes is that they can alter the internal flow of the multiphase system interfering with realistic process measurements. Noninvasive measurement techniques for multiphase flows are Contributed by the Fluids Engineering Division of ASME for publication in the JOURNAL OF FLUIDS ENGINEERING. Manuscript received January 6, 2011; final manuscript received May 21, 2011; published online July 22, 2011. Assoc. Editor: Malcolm J. Andrews.

Journal of Fluids Engineering

being developed by several research groups in an attempt to provide high-quality quantitative data of various flow characteristics. As a matter of fact, X-ray imaging was one of the first noninvasive measurement techniques used in the 1950s and 1960s to visualize air bubbles in gas-solid fluidized beds [8–10] and gas-liquid reactors [11,12]. As summarized by Dudukovic [13], many multiphase flows are opaque, which limits optical access, as well as accessibility to probing with instrumentation, but experiments are necessary to develop deeper understanding of the flow physics. Dudukovic goes on to state that a multiphase flow experimental system should have: (i) high spatial and temporal resolution for local phase fraction measurements, as well as velocity field measurements of all phases, and (ii) the capability to provide instantaneous and time history snapshots of the flow. Note, however, that a single experimental system that can satisfy all these needs does not currently exist. Multiple noninvasive flow visualization techniques have been reviewed in the literature [13–21]. A variety of tomographic techniques have been developed, where tomography refers to the crosssectional imaging of a system from either transmission or reflection data collected by illuminating the systems from many different directions [22]. As summarized by Marashdeh et al., [23], tomographic systems are generally classified into soft field or hard field measurement systems. In soft field methods like electrical capacitance tomography, a change in the measured property (e.g., capacitance) in one location changes the recorded field throughout the entire domain, resulting in a very complex reconstruction process that could produce multiple solutions. Typically for soft field methods, iterative and optimization techniques are utilized to find the most likely reconstruction. In hard field methods like X-ray tomography, the field lines of the measured property (e.g., X-ray attenuation) remain straight and they are not influenced by property changes outside the line-of-sight. This makes the reconstruction easier, but, because of the detection systems and source strength, data acquisition is typically slow. Of the ionizing radiation (hard field) techniques, X-ray imaging is safest because the sources only emit X-rays when they are powered and their energy can be controlled by varying the input voltage [18,24]. Toye et al. [24] also note that X-rays are preferred over c-rays because the X-rays provide better spatial resolution due to improvements in X-ray detector technology in recent years. X-ray tubes also provide a smaller spot size when compared to c-ray sources of equivalent strength, which also provides improved spatial resolution. Current tomographic techniques for flow visualization and characterization include electrical impedance tomography [25–27], electrical resistance tomography [28,29], electrical capacitance tomography [23,30–38], ultrasonic computed tomography [39–41], c densitometry tomography [13,27,42–49], X-ray computed tomography [50–58], positron emission tomography [59], neutron transmission tomography [60,61] and magnetic resonance imaging [62–64]. Nontomographic noninvasive flow visualization techniques for opaque systems include individual particle tracking methods using radioactive emitting tracer particles [13,65–69], X-ray absorbing particles [70–73], or neutron absorbing particles [74]. Yet others have used radiographic (single or consecutive pictures) [75–77] or fluoroscopic (real-time) [78] imaging to view internal flow characteristics of multiphase and/or opaque flows. Many of the techniques listed above offer trade-offs when implemented [16,18,19,21,23,29,34]. For example, electrical capacitance tomography is very fast but the spatial resolution is generally coarse and the results are sensitive to the reconstruction algorithm [19,21,79,80]. The coarseness and sensitivity are due to the soft fields of the electrical capacitance tomography measurements making the reconstruction process difficult and the resolution, particularly in the center of the object, usually poor [81]. X-ray computed tomography has excellent spatial resolution but poor temporal resolution – it is best for time-averaged phase distributions. For example, Schmit and Eldridge [52] state that the spatial resolution and material compatibility of X-ray computed tomography makes this

C 2011 by ASME Copyright V

JULY 2011, Vol. 133 / 074001-1

Downloaded 01 Aug 2011 to 129.186.252.31. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

technique ideally suited for imaging vapor-liquid fixed-bed contactors. Another advantage of X-ray imaging specific to gas-solid flows, as pointed out by Grassler and Wirth [79], is that X-ray tomography is insensitive to electrostatic charge build-up common in fluidized beds, which may be problematic for electrical capacitance systems. In general, X-ray imaging is one family of noninvasive measurement techniques used extensively for product testing and evaluation of static objects with complex internal structures. As stated above, X-rays can also be used to visualize and characterize multiphase flows. This review provides a summary of the current status of X-ray flow visualization and details various X-ray visualization methods that can provide qualitative and quantitative information about the characteristics of complex multiphase flows. The fundamentals of X-ray radiation will first be described, followed by three imaging categories — radiography, stereography, and computed tomography. New opportunities in X-ray imaging will then be briefly described followed by the conclusions of this review.

2

X-Ray Fundamentals

X-rays are produced by ionizing a target source, such as tungsten, with an electron beam. The electrons are emitted from a cathode and accelerated toward an anode by a high voltage potential between the cathode and anode. When the electrons hit the anode, they are decelerated which is accompanied by the emission of electromagnetic radiation [79,82]. In general, the energy spectrum that is emitted from the X-ray tube is a function of the voltage potential, target material, the angle at which the electron beam hits the target, the angle at which the X-rays are observed, and the material used for the X-ray tube window. X-rays can range in energy from less than 0.1 keV to more than 100 MeV [83]. High energy X-rays are referred to as “hard” X-rays and low energy X-rays are described as “soft” X-rays (this is not to be confused with soft field techniques as described above). Hard X-rays contain the short wavelength X-ray component which can enhance the penetration into thick objects or those with large object densities. The soft X-ray component comprises the longer-wavelength X-rays which enhance the image contrast. Most X-ray sources produce a spectrum of energies and are composed of both hard and soft X-ray components. A polychromatic X-ray beam contains both hard and soft X-rays. The soft (lower energy) X-rays are attenuated more readily than hard (higher energy) X-rays, producing an artifact called beam hardening [84,85]. Beam hardening increases the average energy of the beam as it exits the object because the lower energy spectrum is preferentially absorbed. Beam hardening corrections can be applied to reduce their effects on the resulting X-ray image. X-rays generally penetrate an object of interest in straight lines, and their absorption depends on the material through which they pass. X-rays are scattered, at a very small rate through very small angles, but in most imaging situations, this can be ignored. However, there are applications where scattering is actually utilized. Compton scattering is caused by a photon change in direction and wavelength due to interaction with a loosely bound electron [82]. Compton scattering at larger angles produces a background radiation field that reduces image contrast. If a mono-energetic X-ray beam from a source of intensity I0 traverses through a single phase medium, the X-ray energy will be attenuated following the Beer-Lambert law [86]: l q‘ (1) I ¼ I0 exp q where I is the X-ray energy recorded by a detector, l/q is the mass absorption coefficient for the material, q is the material density, and ‘ is the X-ray path length through the medium. The mass absorption coefficient is a function of the X-ray energy and the atomic number of the absorbing material. The influence of atomic number has been exploited in medical imaging where solutions 074001-2 / Vol. 133, JULY 2011

containing heavy atoms such as iodine, barium, and bromine, are used to enhance X-ray visualization. For a two-phase system composed of phase 1 and 2, Eq. (1) can be rewritten as: I ¼ I0 exp½ðð1 eÞl1 þ el2 Þ‘

(2)

where l1 and l2 are the linear absorption coefficients of phases 1 and 2, respectively, e is the volume fraction of phase 2, and ‘ is the path length of the X-ray beam through the two-phase flow. Mennell et al. [87]. state that in particulate systems, the linear attenuation of X-rays is well defined in terms of medium composition with little moisture-specific dependence. Parameters important to X-ray flow visualization include spatial, temporal, and density resolution [18]. Spatial resolution represents the minimum distance that two high-contrasting points can be separated. Temporal resolution identifies the frequency of successive images at which dynamic objects can be resolved. Density resolution defines the smallest difference in mass attenuation coefficients the system can distinguish. Cartz [83] provides a summary of the factors that affect X-ray image quality and influence resolution. X-rays can be detected through a variety of means. Since X-rays expose photographic film, many of the initial detection techniques were film-based. For newer systems, digital detection through one of two types of electronic detector is preferred [88]. Direct conversion detectors utilize an X-ray photoconductor (e.g., amorphous selenium) to directly convert X-ray photons to an electrical charge. Indirect conversion detectors employ a two-step process where a scintillator converts incident X-rays to visible light, and then the light is converted to an electrical signal through a photodetector. In both cases, the output from these devices is sensed by an electronic readout mechanism with an analog-to-digital converter to produce a digital image. In many X-ray imaging systems, particularly for older types of computed tomography systems, scintillators are used to convert incident X-rays to visible light, where the greater the X-ray energy, the brighter the light. With a linear array of scintillator detectors, the individual signals are electronically captured to quantify the X-ray intensity of a fan beam through a horizontal plane. If the scintillator is a 2D detector and the X-ray source produces a cone beam, a photograph of the screen can be taken to record the X-ray intensity, generating a 2D “shadow” image of a 3D object, which is commonly encountered in simple medical X-rays. Image intensifiers can also be used to capture 2D radiographic images. These imaging devices use electrostatic focusing and acceleration of electrons to produce an output image, and, due to this design, the collected images have a slight bow on the edges. This bow is not part of the object projection and will distort the image unless corrected through calibrations. The accelerated electrons are also susceptible to distortions due to magnetic fields (similar to a CRT screen), which can also be corrected through proper calibrations [71,89].

3

X-Ray Radiography

Radiography is the act of obtaining a shadow image of an object using penetrating radiation such as X-rays or c-rays [83]. As shown in Fig. 1, an X-ray imaging device can record a 2D projection of a 3D object when the object is placed between an X-ray source emitting a cone beam and a 2D detector. Therefore, a radiograph is a photographic record of the X-ray attenuation map of an object. Prior to digital photography, X-ray images were processed in much the same way as traditional photographic film because X-rays affect photographic emulsions in a similar manner as visible light. Digital detectors are now much more common than film and use an electronic capture of the X-ray intensity to produce images that compare very favorably with film. As such, digital radiography has seen a tremendous increase in popularity in the last two decades due largely to improved performance of 2D sensor arrays and enhanced Transactions of the ASME


Fig. 1

Schematic of Radiographic Imaging

computer power capable of acquiring, processing, and displaying the large data sets. Major advantages of digital radiography over more traditional film radiography are the speed with which images can be acquired and the flexibility in manipulating and storing the images. Traditional film radiography can be likened to a shadow picture of an object, with lighter images associated with object areas of greatest density. The image quality depends on the contrast developed in the radiograph, which is a function of the X-ray attenuation characteristics of the materials in the system of interest. Contrast can be reduced by several mechanisms; some are the result of X-ray absorption, which is highly dependent on X-ray energy level, while others are the result of X-ray scattering. X-ray scatter, resulting in secondary X-rays, cause a general fogging of the image, which reduces the image sharpness, clarity, contrast, and resolution. Scattering can be reduced by passing the X-ray beam through a metal filter prior to reaching the object of interest to reduce the amount of soft X-rays reaching the object. An antiscatter grid can also be placed directly in front of the detector to minimize the effects of scatter. A blurring of the object edges can also be caused by a penumbra effect if the X-ray source (spot) size is too large or the object is too close to the source [83]. If the object in the imaging region is dynamic, such as a fluid flow process, the radiographic image will be blurry unless the detector has a fast enough response, and the imaging device connected to the detector has a fast enough shutter speed to freeze the desired motion, where “fast enough” depends on the motion of interest. With a fast enough system and successive imaging, a digital 2D radiographic movie of a 3D dynamic process can be created [89–91]. Radiographic imaging of dynamic processes has also been referred to as fluoroscopic imaging, where the original definition of fluoroscopy was to observe an X-ray image on a fluorescent screen in real time [83]. The term “fluoroscopic imaging” has also been used when X-ray imaging is completed in real time using image intensifiers [92]. 3.1 Examples of X-Ray Radiography for Flow Visualization. Fluidized beds are found in many industrial applications, including chemical and fuel production, power generation, mineral and powder processing, and pharmaceutical production. Their operation is based on a gas or liquid moving through a granular bed at a sufficient velocity to suspend the particles, and these beds are useful because they have good mixing, low pressure drop, and high heat and mass transfer rates. Gassolid fluidized beds have been studied extensively, but they are difficult to visualize because the systems are opaque due to the dense particle-laden flow, limiting the choice of experimental techniques. X-ray imaging provides a good option for flow visualization. In 1955, Grohse [8] was one of the first researchers to use X-rays to visualize and characterize gas-solid fluidized beds. He used a commercially available X-ray unit to qualitatively assess density variations in silicon powder fluidized beds as gas flow rate Journal of Fluids Engineering

increased. Three different aeration methods were used and he was able to visualize local bed density variations through X-ray imaging. Radiographic images, coupled to cine´ photography, were used by Rowe and Yacono [93] in 1976 to characterize bubble size, shape, and rise velocity in a fluidized bed with a rectangular cross-section. They used a 22.9 cm diameter image intensifier coupled to a 0–50 frame/sec cine´ camera. The image intensifier was slightly smaller than the extent of the fluidized bed. Typically, 200 frames were used to measure their desired characteristics for each test condition. One drawback at the time of this study was that much of the image analysis was done by hand. Rowe et al. [94] used the same equipment and a calibration wedge to quantify bed voidage (density) variations along the X-ray beam path from the radiographic images. Yates et al. [95] used a similar system to compare the bubble dynamics in a fluidized bed filled with two different particle types. They maintained the gas flow rate at the minimum fluidization velocity and then injected a second gas stream through a single nozzle to produce a stream of bubbles within the fluidized bed. Using a 25 frame/sec video recorder coupled to an image intensifier, they were able to record bubble coalescence of the rising bubble stream. They also concluded that bubbles within a fluidized bed have no defined boundary between the bubble shell and surrounding emulsion phase. The use of a single projection X-ray fan beam to quantify solids loading in a dilute suspension was described by Mennell et al. [87]. They described using soft X-rays to measure solids concentration and the challenges in interpreting the signal, such as the nonlinear nature of the attenuation and beam hardening effects. They concluded that to minimize uncertainties in the solid fraction measurements, (i) the fan beam detecting elements should be as close as possible to each other or arranged in an offset array to minimize dead space effects, (ii) the lowest possible X-ray voltage should be used or the object should be as far as possible from the source to minimize scattering effects, (iii) fast sampling detectors should be used to minimize dynamic bias effects, (iv) a point source of X-rays should be used and the source should be as far as possible from the object of interest to minimize X-ray field divergence, and (v) beam hardening effects must be considered to account for the nonlinear nature of X-ray attenuation. Hulme and Kantzas [78] used an image intensifier to characterize bubbles in a 10 cm diameter cold flow fluidized bed. X-ray images were acquired at 30 frames/s over a period of approximately 2 mins for several different flow conditions. They developed image processing tools to identify bubbles in individual frames and then tracked them from frame to frame. They were able to estimate mean bubble diameter, average vertical bubble velocity, and bubble number density at various locations within their fluidized bed from the X-ray images. He et al. [96] extended this work to record the bubble characteristics in a pseudo-2D fluidized bed with inner width of 22.5 cm and thickness of 5 cm. The effective image diameter was about 17 cm, which was less than the column width. To overcome this limitation, six locations were visualized for each test condition. Using an image intensifier and digital camera with a temporal resolution of 40 ms, Jenneson and Gundogdu [97] visualized nanoparticle agglomeration and breakup in a small fluidized bed filled with 50 nm zinc oxide particles. Different gas flow rates revealed different flow conditions, including the observation of channeling. In addition to fluidized beds, X-ray radiographic imaging has been used to visualize other multiphase flow systems. Gupta et al. [92] used X-rays to visualize the fluid flows in packed beds. They used barium chloride dissolved in water to enhance the contrast of water flowing through a packed bed of polyethylene beads in which air was flowing. In the resulting images, only the flowing water was observed. They also completed a few imaging trials, in which mercury was the liquid. Single radiographs showed liquid percolating through the porous media with the flowing gas JULY 2011, Vol. 133 / 074001-3


modifying the flow direction. Basavaraj et al. [77] extended this study and obtained radiographic images to quantify liquid holdup in a 2D packed bed of expanded polystyrene beads in which an aqueous barium chloride solution trickled through it. They were able to quantify liquid holdup and hysteresis through image analysis of the radiographs. Roels and Carmeliet [98] used radiographic projections to investigate porous media; in this case the determination of 2D moisture profiles in porous materials. The moisture profile was determined as a function of time, showing the water absorption characteristics in various 2D material sections. The authors concluded that 2D X-ray radiographic techniques provide “good resolution in space, time, and moisture content” and the method is extremely useful to study transient moisture flow in various materials. Trickle bed reactors are gas-liquid-solid reactors in which a fluid flows over a packed bed of stationary particles and gas may flow in the same or opposite direction of the liquid. Van der Merwe et al. [99] used X-ray radiography to visualize the trickle flow liquid saturation and distribution as a function of time, prewetting procedure, and liquid and gas velocities. In this study, the trickle bed was composed of a 40 mm ID column filled with 2.5 mm diameter alumina particles. The radiographic images produced a 2D projection image of the 3D flow through the bed of stationary particles. Although the 3D nature of the fluid flow was lost, the 2D projections provided new insights into trickle bed hydrodynamics. Sun et al. [76] used X-ray radiography to visualize the 2D liquid metal front in lost foam casting. They compared maps of the liquid metal front as a function of time with local density variations in the solidified casting, as well as to any material defects. They concluded that faster liquid metal filling produced more material defects. Caulk [100] claims the X-ray images of Sun et al. [76], especially when viewed in full motion video, provide the best available examples of mold filling defects in the lost foam casting process. Single projection images using cone beam X-ray systems and 2D imaging devices were used by Uchida and Okamoto [101] to develop an X-ray imaging method to track powder flows in a screw feeder. They injected small amounts of tracer powder with high X-ray absorption characteristics (i.e., tungsten powder) and then tracked the “tracer crowd” as it moved downstream with the bulk material. They also described the required image processing steps to interpret the X-ray images. They operated two different X-ray systems during different tests, one producing a vertical projection and one producing a horizontal projection, to provide qualitative information of powder movement in the screw feeder. This imaging system was used by Uchida and Okamoto [102] to measure the diffusion coefficient in powder-filled screw feeders, which is a measure of powder mixing effectiveness. 3.2 Flash X-Ray Radiography. Similar to high-speed stopmotion photography, flash X-ray radiography (FXR) is an X-ray radiographic imaging process where an intense burst of radiation is produced for a fraction of a second to record high-speed events that are obscured by dust, smoke, or light. FXR can also capture images of inclusions or voids inside opaque objects that are part of these dynamic events. Areas where FXR are commonly found include high-velocity ballistics, explosives, weapons development (e.g., imaging projectile penetration), nondestructive testing, and medical and biological studies [103–105]. As cited by Charbonnier [106], FXR has been used since the late 1950s to study dynamic events such as metals casting and flow through rocket motors. It has also been extensively used to visualize projectiles penetrating solid objects. The early systems used film radiography and elaborate setups to capture the dynamic event of interest on a series of films, which were typically mounted on large rotating drums. Triantafillopoulos and Farrington [107] state there are several challenges associated with FXR imaging. First, a relatively high 074001-4 / Vol. 133, JULY 2011

voltage is required to obtain the required X-ray intensity during the short duration pulse. Second, a fast detector system is needed since the intensity per pulse is limited. Third, relatively large X-ray spot sizes are needed because it is not possible to provide adequate cooling for small spot sizes during the high intensity, short duration pulse; this will adversely affect the spatial resolution and image sharpness. Finally, the system parameters cannot be adjusted during the short duration pulse; this implies optimal imaging parameters require significant trial-and-error and experience. Relative to fluid mechanics, FXR was first used in 1965 to visualize vapor development from boiling in metal tubes under high pressures [11]. A specially designed thin-walled titanium test section was utilized, and FXR provided a visual record of the flow conditions within the opaque structure. This study was expanded to compare high-speed flash photography and FXR of air-water flows at high liquid flow rates [12]. The FXR system provided a 100 keV energy pulse over a duration of approximately 107 seconds, providing stop-motion X-ray images of their flows of interest. Good qualitative agreement was obtained between both image-recording techniques. Additionally, the images taken with FXR were superior to those of conventional flash photography at very high liquid flow rates because the X-rays penetrated the regions made opaque by the multiphase flow. Rowe and Partridge [10] also used FXR in 1965 to visualize gas bubbles in a gas-solid system. They investigated the effect of fluidized particle diameter on bubble size and velocity in fluidized beds composed of glass beads or natural silver sand. They used various cylindrical beds and provided fluidization air such that no bubbles would form. They then injected supplemental air to follow individual bubbles as they rose through the bed using an image intensifier and camera setup. Romero and Smith [9] used two flash X-ray radiographic systems and image sequencing to estimate bubble rise velocity in a fluidized bed. Smith et al. [108] utilized this X-ray imaging system to determine bubble size and rise velocity in gas-liquid-solid slurry bubble columns. Several researchers have used FXR to image flow systems that are commonly found in the pulp and paper industry. For example, Farrington [109] used FXR to image 25 lm diameter tungsten tracer fibers suspended in 2% by weight hardwood fibers. The short pulse duration of FXR was needed to provide stop-motion imaging of the suspended tungsten fibers in suspensions flowing at speeds of up to 10 m/s. Farrington was able to demonstrate that FXR is a potentially powerful technique for investigating high speed multiphase flows in general and concentrated fiber flows in particular. He extended this technique to investigate fiber sheet formation and quality [110] and black liquor spray formation [111]. Triantafillopoulos and Farrington [107] further extended this technique to visualize flow phenomena in opaque coating flows. Zavaglia and Lindsay [112] used FXR and an X-ray tracer fluid (a silver nitrate solution) to visualize fluid motion during impulse drying. Finally, in a series of studies by Heindel and co-workers [75,113–119], FXR was used with quasi-2D flow systems to characterize and quantify gas flows in various fiber suspensions. An example of these results is shown in Fig. 2 where FXR was used to freeze the bubble dynamics in an air-water system (Fig. 2(a)) and then the results were compared to air-water-Rayon fiber systems with 1% by mass fiber (Figs. 2(b)–2(f)). In all cases of this study, a 20 cm 2 cm bubble column was used with a superficial gas velocity of 0.8 cm/s [119]. Image analysis was then completed to determine effective bubble size distributions. Note the bubbles in Fig. 2 appear as dark objects because the images are from the actual 25.2 cm 20 cm negatives (i.e., the color is inverted). 3.3 Other X-Ray Radiographic Imaging Methods. A variation of X-ray radiography, called digital speckle radiography (DSR), was used by Grantham and Forsberg [120] to characterize granular flows in opaque hoppers. In this method, a thin vertical layer of X-ray absorbing granular material (tungsten) was sandwiched between aluminum powder of the same size distribution. Transactions of the ASME


Fig. 2 Single FXR images of air bubbles in various 1% by mass Rayon fiber suspensions contained in a 20 cm 2 cm bubble column with a superficial gas velocity of 0.8 cm/s [119]. The dark amoeba-like structures in each image represent air bubbles and the bubbles appear as dark objects because the images are from the actual 25.2 cm 20 cm negatives – the color is inverted.

Continuous X-rays were transmitted through the thin layer and recorded with an image intensifier and digital camera. Using methods developed for speckle metrology, cross-correlations of successive images were applied to quantify particle displacement from image to image. With this method, localized particle movement was determined from which strain rates were calculated. They observed clear differences in particle motion when two different hopper outlet sizes were imaged. Most X-ray flow visualization studies use a polychromatic X-ray source in which the X-ray energy spectrum is a broad function; in these systems, X-ray attenuation is a function of X-ray wavelength and leads to beam hardening effects. The advanced photon source (APS) at Argonne National Lab (Argonne, IL) produces a monochromatic highly focused X-ray beam and was used to characterize a spray’s projected density map along the beam path [121]. By scanning the spray region, a two-dimensional projection map of the spray’s mass distribution was developed. The penetrating effects of the X-rays allowed dense sprays to be quantified, which is not possible using traditional laser-based techniques. The APS at Argonne was also used by Coutier-Delgosha et al. [122] to develop an ultra-fast X-ray imaging procedure to simultaneously acquire vapor volume fraction, vapor velocity field, and liquid velocity field in a high speed cavitating flow. To track fastmoving X-ray-absorbing small particles, the X-ray source must have a very high intensity like that produced at the APS. The high intensity X-ray beam was directed through the imaging region and then recorded at 10,000 frames/sec with a high speed APX-RS Photron camera. Silver coated hollow glass spheres (6–25 lm in diameter) were used as seed particles. Vapor volume and velocity were determined using phase contrast enhancement to track the Journal of Fluids Engineering

interface because of the large difference in density between the vapor and liquid. Liquid velocity was determined by identifying tracer particles due to absorption differences and phase contrast enhancement. The authors were able to show representative velocity fields for the cavitating flow. Synchrotron X-ray imaging was also used by Kim et al. [123] to visualize water diffusion in a simulated fuel cell. They produced both 2D radiographs and 3D tomographs (see Sec. 5). Although the 3D tomographs produced more quantitative results of time-averaged data, the 2D radiographs provided qualitative results of transient behavior.

4

X-Ray Stereography

Stereographic measurement methods use information from two 2D projections to calculate the 3D location of features in an object [124]. This can be accomplished by analyzing two images of an object which are taken at different positions either due to a rotation or translation of the sample. Alternatively, with two identical source/detector pairs like those found in [89], and using appropriate software controls for the two CCD cameras, two image projections can be acquired simultaneously. For example, Fig. 3 shows a schematic representation of this process where one source/detector pair provides a radiographic projection of the x-z coordinate information as a function of time, while the other source/detector pair provides a radiographic projection of the y-z coordinate information as a function of time. As long as the point of interest is identifiable in both images, its corresponding 3D coordinate as a function of time can be determined. Hence, with proper software, stereographic imaging can produce a 3D map of a feature of interest. With successive continuous JULY 2011, Vol. 133 / 074001-5


Fig. 3 Schematic representation of X-ray stereographic imaging

images, the 3D location of a flow feature, such as a particle, bubble, or phase front, can be determined as a function of time. Characteristics such as object rise and settling velocity or breakup and coalescence rate, could then be determined from these data. Thus, X-ray stereography provides a means for 3D flow visualization of dynamic characteristics in opaque systems at a temporal resolution comparable to either the frame rate of the digital cameras or the response time of the detector systems, whichever is slower. Although not required, accuracy of the coordinate determination can be increased through the use of markers at known locations in the object of interest. For example, Doering [124] used markers and was able to measure points in several small static samples with an accuracy of 0.1 mm. Jensen and Gray [125] also used markers to visualize molten aluminum flow in lost foam castings. Lee et al. [126–128] utilized stereographic imaging technology to develop stereoscopic tracking velocimetry (STV) for systems with optical access (like 3D PIV) to simultaneously track several particles dispersed in a carrier fluid. STV consists of two imaging devices, which are oriented at some separation angle relative to each other, that simultaneously acquire time-sequenced images of the tracking particles from their respective location, and then particle position as a function of time is identified in each frame so 3D motion can be determined. The challenge with STV is in the efficient identification and tracking of multiple particles. Lee and Cha [126] utilized artificial neural networks to improve their STV tracking. An extension of STV to X-ray imaging is called X-ray particle tracking velocimetry (XPTV) [129]; this method tracks several X-ray absorbing objects (particles) simultaneously as a function of time. As stated by Seeger et al. [129], the advantages of XPTV are many and include: (i) can track velocities in opaque flows or in flows with high gas fractions, (ii) can measure many points simultaneously, (iii) can record 3D velocity components, and (iv) is noninvasive. The disadvantages include (i) low image frequency, (ii) large tracer particles are typically required, and (iii) the method utilizes X-rays. A significant challenge with XPTV is identifying tracer particles with the desired fluid flow characteristics (e.g., neutrally buoyant) but yet differentially attenuate X-rays, which is based primarily on density differences. Lee et al. [127] mentioned the STV algorithms of Lee and Cha [126] could be extended to X-ray imaging and XPTV advancement, and provided directions for development. This method is similar to that used by Schober and Kantzas [130] who tracked radioisotopetagged polyethylene particles in a 10 cm diameter fluidized bed. They tracked the specially-tagged particles with 2D c-ray cameras. XPTV is safer than this method because the particles to be tracked are not radioactive. However, the effective density and X-ray attenuation requirements of XPTV tracer particles may limit the type of material and the tracer particle effective diameter that is ultimately used for XPTV. By tracking neutrally buoyant X-ray absorbing particles, Seeger and co-workers [70,71,131] were able to record the three074001-6 / Vol. 133, JULY 2011

dimensional liquid velocity field in a slurry bubble column. Seeger et al. [71] used X-ray stereographic imaging to track multiple solid particles in a gas-liquid-solid three-phase system. The solid particles were polymethylmethacrylate cubes with dimensions of 2 mm on a side. Similar cubic X-ray tracer particles were fabricated from polyurethane foam with tin alloy cylindrical inserts. In this case, the tin alloy absorbed the X-ray energy and the foam provided buoyancy to produce manufactured particles with an effective density and dimensions similar to the bulk solid material. Sixty-five manufactured particles were added to 100 g of particles and then mixed in 1 L of glycerin. They used highly viscous glycerin as the liquid to provide slow particle velocities to make particle tracking easier (their system was limited to 25 images/s). Air was then injected at a superficial gas velocity of 1.6 mm/s through 91 hypodermic needles. Four hundred and sixty image pairs were acquired over a 18.4 s period to determine 3D solid particle velocity fields. Assuming the distribution of the Xray absorbing particles was similar to the solid phase, the local solid holdup could also be estimated. Seeger et al. [71] concluded this method can simultaneously provide 3D velocity fields and solid holdup but, as designed, was limited in the results that could be obtained. They further concluded that improvements could be made with smaller X-ray absorbing seeding particles and faster camera and image intensifier pairs. Kertzscher et al. [131] improved this system to track intruder particles in a gas-liquid system. Improvements included better mapping between the two image intensifiers in terms of an isocenter correction and matching of whole particle trajectories. Owens et al. [132] also manufactured neutrally buoyant, X-ray absorbing particles to track fluid flow with a focus on packed columns. In this case, they used a 2.88 mm diameter high-density polyurethane foam cylinder 2.88 mm in height with a 0.81 mm diameter lead solder core as tracking particles. They then used a single X-ray source-detector pair to track injected particles as they traveled through a packed bed. Images were acquired at 30 frames/s. Their biggest challenge was identifying the particles in the packed bed because the particles were moving too fast relative to the frame rate of the recording device. However, several trajectories were recorded. Shimada et al. [133] followed the techniques of Seeger and coworkers [70,71,131] to develop an XPTV technique for the visualization and characterization of highly viscous (500 Pa-s) opaque slurry flows. The application Shimada et al. [133] was addressing was the casting of solid propellants. By using 1 mm diameter lead balls as tracers, they were able to develop velocity maps in the highly viscous slurry flows. Lee and Kim [72] utilized a coherent X-ray source from the synchrotron radiation source at the Pohang Accelerator Laboratory (PAL; Pohang, Korea) to develop X-ray particle image velocimetry (PIV) for opaque fluids that exploit phase contrast imaging. They focused on refraction-based edge enhancement and identification because the monochromaticity of the X-ray beam is not required. With this technique, they were able to track 3 lm alumina microspheres in a 750 lm diameter opaque Teflon tube and reproduce the expected velocity profile. Kim and Lee [134] extended this technique to produce a highly coherent X-ray beam; the resulting beam induced a classic Fresnel edge diffraction in the radiographs which was used to identify the edge of particles in the opaque fluid. As stated by the authors, the diffraction-based edge detection techniques are influenced by several parameters including the X-ray beam monochromaticity, the source size, and the detector resolution. They used this diffraction-based edge detection method to determine red blood cell location as a function of time and the resulting velocity profile in a 490 lm wide by 1390 lm deep microchannel through which a sample of blood was pumped. X-ray phase-contrast imaging was also used by Im et al. [135] for particle tracking in polydispersed particle-laden flows in optically opaque systems. The authors used a monochromatic X-ray beam from the APS at Argonne National Laboratory to probe a glycerin flow in a 860 lm ID Teflon tube seeded with 10 lm diameter silverTransactions of the ASME


coated hollow glass spheres. They were able to measure the velocity profile in the opaque tube using this technique. Kakimoto et al. [136] used a single X-ray source and image intensifier to visualize tracer particle movement in molten silicon. Their challenge was in creating a tracer particle that was thermally and chemically stable and allowed the observation of convection patterns in the molten silicon. They determined that a tungsten particle encased in silica (SiO2) and then coated with carbon was the best tracer particle. The effective tracer particle density was close to that of molten silicon and the carbon coating provided good wettability in the molten silicon. A similar process was used by Munakata and Tanasawa [137] to visualize silicon melt convection under radio frequency heating. In this case, stereography was used to identify the 3D structure of the melt convection currents by tracking several tracer particles. The authors used 0.5 mm diameter zirconia particles covered with a 0.25 mm layer of quartz glass, and then coated with carbon to avoid wettability problems of the quartz and molten silicon. Tracking intruder particles in a fluidized bed using X-ray stereographic imaging has been completed by Drake et al. [73,138]. They showed that they can successfully track large intruder particles (9 mm in diameter) in a fluidized bed composed of 500–600 lm glass beads. Drake and Heindel [73] summarized system improvements to enhance the tracking abilities, including enhanced image analysis techniques, better intruder particle manufacturing capabilities, and faster camera systems. Intruder particle manufacturing capabilities, for particles specific to X-ray flow visualization, have been recently addressed by Drake et al. [139]. They describe a manufacturing process for XPTV tracer particles that satisfy specific particle characteristics including high X-ray attenuation, uniform shape, specified effective density, and desired diameter. Finally, X-ray stereography imaging was also used by Morgan and Heindel [140,141] to visualize the Brazil nut effect, where a dense object migrates to the top of a bed of granular media when exposed to vibration. They used a normalized cross correlation method of template matching to track the location of a large intruder particle in a vibration-induced fluidized bed. By tracking the large intruder particle, the X-ray visualization revealed strong convection currents within the vibrating bed over a range of operating conditions. Figure 4 shows a sample sequence of images of this effect where a large intruder particle is shown in a bed of almonds [141]. In Fig. 4, the left and right images are perpendicular X-ray radiographic projections taken at the same instant in time, and each successive frame is separated by about 55 ms. In this application, X-ray imaging is one of the few experimental techniques that can be used to gather quantitative information of this effect.

5

X-Ray Computed Tomography

As stated by Wellington and Vinegar [142], X-ray computed tomography (CT) imaging was first developed by G. N. Hounsfield in 1972, for which he shared a Nobel prize in medicine for this contribution. It is based on the mathematical process of reconstructing a function based on multiple projections, which was originally proposed by Radon in 1917 [22]. The reconstruction produces a two-dimensional cross-sectional image of an object showing internal details. Tomography literally means “the picture of a slice” [83]. The process of CT imaging includes an X-ray source illuminating the object of interest and projecting the transmitted X-ray intensity onto an imaging device (Fig. 5). Projections from several hundred orientations are collected into so-called sinograms and reconstructed with standard algorithms, such as filtered back projection [85,143], generating an image of the object cross-section. If the X-ray source is a fan beam, a single slice is reconstructed. The object or source/detector pair are then moved a small amount perpendicular to the fan beam to generate another slice. By acquiring enough slices, a 3D reconstruction can be generated by stacking the individual slices. Alternatively, as shown in Fig. 5, an Journal of Fluids Engineering

Fig. 4 Five sequential frames of an intruder particle in a granular bed of almonds vibrated at 10 Hz [141]. The left and right images are perpendicular X-ray radiographic projections taken at the same instant in time, and each successive frame is separated by about 55 ms.

X-ray cone beam can produce several slices in a single scan, producing a 3D reconstructed image [89]. To reconstruct a reliable image, it is essential that each projection be of the same object. This is not possible in a dynamic setting like a multiphase flow, so each projection is acquired over a time scale that averages any signal fluctuation. Hence, it takes time to acquire several hundred projections, and X-ray CTs produce time-averaged details of the object’s internal features, similar to a medical CAT scan. Using a 2D detector array, multiple slices can be collected during a single exposure; each slice is from one horizontal row of pixels. Two methods can be applied in reconstructing these data. The first is to reconstruct each row of pixels as a slice and stack the slices into a 3D volume, as shown in Fig. 5. The second method is to use a cone beam reconstruction algorithm to limit the distortions introduced in rows farthest away from the central fan beam of the CT system [22]. JULY 2011, Vol. 133 / 074001-7


Fig. 5 Schematic representation of X-ray computed tomography (CT) imaging

There are many mathematical algorithms that can be used in CT reconstruction [22]. One of the most common methods is filtered back projection, which was originally developed for stationary objects. Ikeda et al. [144] have shown that when applied to reconstructing dynamic flows, the result is a time-average of the varying quantities as long as the image acquisition time is much larger than the characteristic time scale of the variations. Behling and Mewes [145] provide a good description of the reconstruction process. First, the measurement plane is divided into a rectangular grid of small imaging pixels. The reconstruction process calculates the amount of attenuation that took place in each pixel. The resulting bitmapped image of gray scale values represents the local time-averaged attenuation coefficient at each pixel. Since the attenuation coefficients are material properties that can be related to local density variations, the gray scale maps can be interpreted as material distributions, and with proper calibrations, can quantify time-averaged phase distributions. Wu et al. [146] describes the filtered back projection reconstruction technique that is used in most medical imaging CT scanners as well as the algebraic reconstruction techniques (ART), which are iterative. Reconstruction using ART methods is generally an error minimization process and [146] used a genetic algorithm for reconstruction optimization. During the filtered back projection reconstruction process, the intensity data in the sinogram files are converted to CT numbers or CT values that have a range based on the computer system, with 12-bit most common (i.e., 4096 possible values) [84]. The CT values correspond to the gray scale value in the resulting reconstructed image. An important feature of the CT reconstruction is that the CT values usually map linearly to the effective attenuation coefficient of the material in each voxel (3D pixel) [84]. Hence, the variation in X-ray local attenuation is closely related to variations in local density [147]. The information is then used to determine local timeaveraged phase fractions in a two-phase system. Using Eq. (1), the local time-averaged phase fraction is determined by first determining the intensity values for each phase, I1 and I2. The local time-averaged void fraction is then related to the intensity recorded during the two-phase operation (I) and the pure systems (I1 and I2): e¼

lnðIÞ lnðI1 Þ lnðI2 Þ lnðI1 Þ

(3)

where e is the local time-averaged void fraction of phase 2. Equation (3) assumes the incident radiation, I0, is constant [148]. 074001-8 / Vol. 133, JULY 2011

As summarized by Kak and Slaney [22], commercial (medical) CT systems typically calculate the linear attenuation coefficient, and the numbers that are recorded by the system are integers that range from 1000 to 3000. This range is an international standard scale called the Hounsfield Scale, with a Hounsfield unit defined as H¼

ðl lwater Þ 1000 lwater

(4)

where lwater is the attenuation coefficient of water. The value of H ¼ 0 corresponds to water and H ¼ 1000 is assumed to correspond to air (assuming lair ¼ 0). Within each pixel from a reconstructed fan beam, or voxel from a reconstructed cone beam, the associated CT number is proportional to the effective attenuation coefficient. Using Eq. (4) and redefining the reference linear attenuation coefficient (lwater), Ikeda et al. [144] define the CT value of a reconstructed element (i, j), CTij, as CTij ¼ alij þ b

(5)

where a and b are constants. In a two phase system, the attenuation coefficient of element (i, j), lij, can be expressed in terms of void fraction of phase 2: m (6) lij ¼ 1 eij q1 lm 1 þ eij q2 l2 where q1 and q2 are the phase 1 and 2 densities, respectively, and m lm 1 and l2 are their respective mass attenuation coefficients. Ikeda et al. [144] go on to show that eij ¼

CTij CTij;1 CTij;2 CTij;1

(7)

where eij is the time-averaged phase fraction of phase 2 in element (i, j), CTij is the time-averaged CT value at location (i, j) of the flowing system, and CTij,1 and CTij,2 are the time-averaged CT values of 100% phase 1 and 100% phase 2 at location (i,j), respectively, which are acquired through calibration. By comparing Eq. (3) to Eq. (7), the CT value is proportional to ln(I) which, through Eq. (1), is proportional to the linear attenuation coefficient. Equation (7) can be used to determine the local time-averaged phase fraction in a gas-liquid system, where phase 1 and 2 correspond to liquid and gas, respectively. However, Eq. (7) is not Transactions of the ASME


directly applicable to a gas-solid system unless a system of 100% solid can be produced for the CTij,1 calibration. This is generally not possible for most granular systems commonly found in fluidized beds because the solid is in powder form with air filling the voids between the particles. In this case, a CT calibration of a homogeneous static granular bed is used in the calibration [56,149]: eij;b ¼

CTij;b CTij;s CTij;g CTij;s

(8)

where eij,b is the local void fraction of the static (bulk) granular system, CTij,b is the local CT value of the bulk system, and CTij,s and CTij,g are the local CT values of the solid and gas, respectively. The CT value of the 100% solid (CTij,s) is not known but can be calculated by measuring the void fraction of the bulk system, eb, q eb ¼ 1 b (9) qp where qb is the bulk density and qp is the particle density provided by the manufacturer. The bulk density is measured for the granular system by recording the bed mass and volume (i.e., qp ¼ mbed/Vbed, where mbed is the bed mass and Vbed is the bed volume). Assuming the bulk granular system is homogenous, eb ¼ eij,b, Eqs. (7)–(9) can be combined to yield: CTij CTij;b þ CTij;g CTij eb eij ¼ (10) CTij;g CTij;b Therefore, finding the local time-averaged gas holdup from a single dynamic gas-solid experiment requires three different X-ray CT files: (1) the X-ray CT of the dynamic bed (flow, CTij), (2) the X-ray CT of a static bed of the same material (bulk, CTij,b), and (3) the X-ray CT of an empty bed (gas, CTij,g). Equation (7) can be used for gas-liquid systems and Eq. (10) for gas-solid systems to produce a 3D matrix of local time-averaged gas holdup (void fraction) values. Software has been developed by many groups to visualize these 3D gas holdup maps at any location within the imaging volume. Additionally, the programs can generate “slices” of the 3D image to produce 2D images of any plane within the imaging region. For example, Fig. 6 show a representation of a 3D volume where y-slices pass through the center of the volume in the x-z plane, x-slices pass through the center of the volume in the y-z plane, and z-slices show x-y planes at any height within the imaging volume. Software can also apply a color map to the slice images to enhance visualization. The colorizing scales can be independently adjusted for each material to improve image contrast. A sample CT image is shown in Fig. 7 to highlight these capabilities. This figure shows the local time-averaged gas holdup in a 10.2 cm fluidized bed filled with 500–600 lm glass beads and operated at a superficial gas velocity of Ug ¼ 3Umf, where Umf is the minimum fluidization velocity. As shown in Fig. 7, local gas holdup values can be extracted from the qualitative gas holdup maps to produce quantitative graphs of the local time-averaged gas holdup along individual rays through the system of interest. When reconstructing X-ray CT images, artifacts within the image can cause large sources of error. Artifacts come from three primary sources: beam hardening, ring artifacts, and abrupt changes

Fig. 6

Sample X-ray CT imaging planes


in density [84]. Beam hardening is the most common artifact found in X-ray CT reconstructions and causes the edges of the object to appear brighter than the center, even when the material is homogeneous [84]. It can be reduced by filtering low energy “soft” X-rays with metal filters. It can also be corrected through a wedge calibration [84]. Ring artifacts are caused by nonuniform response of adjacent detector elements to changes in X-ray energy. These artifacts can be minimized through proper calibration [84,89]. Finally, abrupt changes in system density result in sharp changes in intensity signals that leads to streaks in the reconstructed image due to mathematical relationships in the reconstruction algorithms. The best way to minimize streaking in the reconstructed image is through system design to minimize sharp density differences in the imaging region (e.g., replace metallic bolts in an acrylic containment system with nylon bolts). For example, Heindel et al. [150] filled air voids in a control valve housing with wax and modified the handle and flange to minimize reconstruction artifacts when imaging cavitation from a butterfly valve. Many studies involved with CT imaging also use phantoms to quantify spatial or density resolution. Phantoms are defined as static test objects containing features of known size, spacing, and contrast that are used to quantify spatial and density resolution [151], and are useful to characterize a CT system. 5.1 Examples of X-Ray CT Imaging of Fluid Flows. A medical computer-assisted tomography (CAT) scanner was used by Lutran et al. [152] in the early 1990s to visualize liquid flow through a small packed bed of glass beads. In this case, the 30.5 cm high packed bed was located on a horizontal conveying table that translated the bed through the CT scanning region in 3 mm steps. Vertical CT images were acquired in 3 mm increments through the 6 cm square bed. The water was doped with barium bromide to enhance the difference in X-ray absorption between the glass beads and water. The authors showed many qualitative images in different vertical planes to assess differences between liquid flow rate, particle size, and liquid surface tension, as well as the influence of system prewetting, flow history, and water inlet configuration. An X-ray CT system was developed by Toye et al. [24] specifically for X-ray CT flow visualization of liquid holdup and maldistribution in packed beds. The X-ray system produced a fan beam that was captured with a linear array of 1024 photodiodes, each 1.66 mm wide. The source-detector pair rotated around a 0.6 m ID fixed packed bed. This facility was used by Toye and co-workers [51,153] to characterize the fluid flow patterns and liquid holdup in packed beds. A fixed X-ray CT system with a 450 keV fan beam, a linear detector array, and a 15.2 cm diameter object turntable was used by Schmit et al. [154] to study liquid flow patterns and liquid holdup in trickle bed reactors. This study was extended by Schmit et al. [52,53] to examine the hydrodynamics in a 15.2 cm diameter trickle bed reactor containing structured packing material. Many of the images in Ref. [53] suffered from ring artifacts, which were attributed to the dynamics of the flow and liquid droplets moving into and out of the image during the data acquisition cycle. Trickle bed hydrodynamics were also the focus of van der Merwe et al. [155] who used a cone beam X-ray CT system. In this system, a 40 mm ID trickle bed reactor was placed on a turntable to allow radiographs to be acquired through a full 360 rotation. The rotational acceleration was set very low to minimize centrifugal forces on the fluid flow. After CT reconstruction, extensive image processing procedures were developed to allow for the identification of the solid packing, liquid regions, and gas regions. This allowed the calculation of gas-liquid, gas-solid, and liquid-solid interfacial areas. A medical CAT scanner was modified by Kantzas [50] to allow for both horizontal or vertical flow systems. They showed that their system could be used to measure local holdup in a 10 cm ID fluidized bed and a 4.5 cm ID trickle bed reactor. Kantzas and JULY 2011, Vol. 133 / 074001-9


Fig. 7 Sample X-ray CT imaging results of the time-averaged gas holdup in a 10.2 cm diameter fluidized bed filled with 500600 lm glass beads and operated at 3Umf with side air injection rates (Qside) of 0, 0.1, and 0.2Qmf, where Qmf is the minimum fluidization flow rate. Local values can be extracted along any ray at a given horizontal location.

co-workers extended this study to quantify local and planar-average void fractions in various polyolefin resin fluidized beds [156] and void fraction and channeling in polyethylene resin filled fluidized beds [157]. The same CT system was used by Wu et al. [158] to study the hydrodynamics in three different diameter fluidized beds. X-ray CT imaging has also been used to study fluid flow in porous media, with applications to air sparging for in situ groundwater cleanup when contaminated with hydrocarbons [159], two-fluid coreflood experiments for enhanced oil recovery [142], and fluid flow through soil samples [160,161]. For example, Chen et al. [159] used a modified medical CT scanner to visualize air sparged through water-saturated packed sands. The reconstructed images provided high-resolution visualizations of the air distribution patterns and showed that operating conditions had a significant impact on local air flow patterns. Du et al. [147] utilized CT imaging to study CO2 and N2 foam flow in a sandstone sample initially saturated with surfactant solution; the application was to improve the understanding of enhanced oil recovery when CO2 is injected into sandstone. The CT images were helpful in understanding the CO2 foam formation and how it propagated in the axial direction. A medical X-ray CT system was used by Morooka et al. [162] and Mitsutake et al. [163] to quantify void fraction within rod bundles that are common in nuclear reactors. Their focus was to improve the understanding of steam-water two-phase flow in rod bundles [162], and then develop a void fraction model and compare the predictions to experimental measurements [163]. Boden et al. [164] utilized an unbaffled stirred tank reactor (STR) with a gas-inducing impeller to test a method of single projection X-ray computed tomography. This method is based on the assumption of rotationally symmetric phase distributions, which produces a rather fast CT image relative to typical X-ray CT data acquisition cycles. They also implemented beam hardening and radiation scattering corrections in their reconstruction algorithms. They validated this technique using a rotating phantom object. 074001-10 / Vol. 133, JULY 2011

With this method, they were able to calculate time-averaged void fraction with a systematic error below 3.5% of the measured value. Note, however, that the assumption of a rotationally symmetric phase distribution limits how this technique can be utilized. Ford et al. [165] used X-ray CT imaging to show that the assumption of axisymmetry was not valid in baffled stirred tank reactors, which are common reactors found in the process industries. They showed that the gas holdup distribution varied depending on the flow regime. Within a given flow regime, the general gas holdup patterns were similar, even when the STR was operated at different conditions within the respective flow regime. Grassler and Wirth [79] used a fan beam X-ray CT system and a linear detector array to quantify solids holdup in a circulating fluidized bed. Their focus was on technique development, so limited solid holdup data were provided. Heindel and co-workers have used the X-ray CT system described in Ref. [89] to produce detailed local time-averaged phase distribution maps within bubble columns [55] and stirredtank reactors [165]. They have extended this work to fluidized beds, including a 9.5 cm ID [166–168], a 10.2 cm ID [56,169], and a 15.2 cm ID [169,170] cold-flow fluidized bed. The fluidized bed data have been used to validate computational fluid dynamics (CFD) models of this complex gas-solid flow [168,171–174]. CT imaging was also completed of an impinging water jet to measure the hydraulic jump at any angular location [175], as well as the cavitation field from a butterfly control valve [150]. Table-top X-ray CT systems have been developed in recent years to analyze small material samples, typically on the order of 1 cm in size. These systems produce microtomography images where the voxel size can be as small as 1 lm or less. Micro-CT systems are ideal for studying flow through porous media where there may be a need to image individual pores. Moreno-Atanasio et al. [176] have recently reviewed X-ray microtomography and how detailed microCT data can be coupled to computer simulations of the fluid flow in porous and granular material. One drawback of these table-top CT systems is that the flow system is typically mounted on a turntable Transactions of the ASME


Fig. 8 Fast X-ray CT system of Hori et al. [180]. The diameter of the measured object is 300 mm.

because the X-ray source and detector are fixed in the micro-CT device. 5.2 Fast X-Ray CT Systems. One of the limitations of X-ray CT imaging of fluid flows is the poor temporal resolution (e.g., it may take several minutes to acquire the needed data for an X-ray CT reconstruction) and most data are local time-averaged values. Many research groups are developing X-ray CT techniques to improve the temporal resolution. Prasser et al. [54] have developed a fast X-ray CT system that provides a spatial resolution of 2.6 mm with a scanning rate of 263 complete CTs per second. They visualized gas-liquid flow in a 50 mm vertical pipe and compared their measurements to those obtained with a wire mesh sensor. The spatial and temporal resolution of the wire mesh sensor were better than the X-ray CT system, but the wire mesh was completely intrusive. Although individual small bubbles were not captured in the CT images, the void fraction measurements between the two methods were still in good agreement. As cited by Harvel et al. [177], Hori et al. [178] used 122 cadmium tungstenate (CdWO4) detectors and 18 X-ray sources mounted in a circular array to develop a high-speed X-ray CT system to measure void fractions in gas-liquid vertical flows. The detector plane was slightly above the X-ray source plane and the X-ray fan beam was directed at a slight incline. This allowed the X-ray beam to pass underneath the detector before the flow region but be recorded by the detector after the flow regions. The system was considered high speed because it could acquire a void fraction distribution slice image every 4 ms. Harvel et al. [60,177] used this system to determine average and instantaneous void fractions in a vertical annulus. Hori et al. [179,180] further enhanced the CT imaging system to produce CT slice images in a scanning time of under 0.5 ms. To do this, they distributed 60 stationary X-ray tubes and 584 Journal of Fluids Engineering

stationary X-ray detectors in separate concentric circles (Fig. 8), and then activated each source independently in quick succession to prevent scattering from adjacent sources. As shown in Fig. 8, the X-ray sources are located in a plane slightly above the detector system and then directed at a slight decline toward the detectors. A fast X-ray CT system was also developed by Kai et al. [181] to visualize gas bubbles in a fluidized bed. The fast CT system used in this study consisted of 18 electron guns arranged in a semicircular arc and housed in a single vacuum chamber. Each electron gun was focused on a tungsten target to produce a 24 3 mm thick Xray fan beam. Each fan beam was recorded by a linear array of 32 detectors. Because some detectors recorded the signal from multiple X-ray beams, a total of 122 detectors were used in the system. A switching mechanism was used to sequence the electron guns, producing 18 individual X-rays over a 4 ms period resulting in 250 CT slices per second. This system allowed for 3D reconstruction of large bubbles in a 4.6 cm ID fluidized bed. A significant drawback of the imaging system used by Kai et al. [181] was that only multiphase flows contained in vessels with an OD smaller than 5 cm can be imaged; this limitation is a result of their physical configuration. A redesigned system could accommodate larger diameter flow systems. A fast X-ray CT system was theoretically described by Wu et al. [146]. Their system acquired limited projection data and then used a genetic algorithm to optimize the reconstruction. In general, they showed that the reconstructed CT simulations produce better results when there are more radiographic projections, and there were tradeoffs between temporal and spatial resolution and system cost. Bieberle et al. [182–185] have developed electron beam X-ray tomography that can provide cross-sectional images of multiphase flows with a spatial resolution of about 1 mm and a frame rate approaching 10 kHz. This method directs a high frequency electron beam to a tungsten target to produce a fast moving X-ray fan beam, JULY 2011, Vol. 133 / 074001-11


Fig. 9 Fast X-ray CT system of Mudde [81]. The fluidized bed had a 23 cm ID with a 5 mm wall thickness.

resulting in multiple angular linear projections through the object. Although multiple angular views are gathered, they do not encompass 360 which provides incomplete data sets that result in image artifacts, such as shape distortions and streaks; these artifacts can be reduced using iterative reconstruction methods. A binary algebraic reconstruction technique (ART) with an additional smoothing operation derived from level-set theory was implemented by Bieberle and Hampel [182] for use in this system. The equipment needed for this imaging method was housed in an evacuated commercial electron beam welded box, which limited the scale of the multiphase flow that could be imaged (i.e., 50 mm). This research group has used this system to visualize gas flows in gas-liquid systems [185] and gas-solid systems [184]. Although this method was extremely fast, only a single plane was imaged at one time and limited angle CT data were obtained which relied on algebraic reconstruction techniques. Bieberle and Hampel [182] also state that this X-ray CT method was rather expensive to implement, but the data it can produce may justify the expense for certain problems of interest. Mudde [57] has also developed a fast CT system consisting of three X-ray sources arranged symmetrically around the flow of interest. Each source produced a fan beam that was directed toward 30 scintillation detectors. This allowed the entire flow region to be imaged simultaneously from three different directions. The resulting data were acquired at a rate of about 100 frames/s and then reconstructed to form a cross-sectional image of the phase distribution. Mudde [58,81] extended this system to include a second plane of 30 scintillation detectors associated with each source (Fig. 9). In this configuration, CTs were used to estimate the bubble rise velocity in a 23 cm ID fluidized bed. Although the frame rate for this system is fast for X-ray tomographic imaging, image threshholding techniques were needed to remove noise and resulted in visualizing bubbles that were at least 3 cm in diameter (13% of the column diameter). Smaller bubbles were not resolved due to signal noise. Nevertheless, this X-ray CT can record bubble properties, like volume and velocity, with reasonable accuracy. 5.3 Other X-Ray CT Flow Visualization Developments. It is fairly straight forward to determine time-averaged phase fractions in two component systems, like gas-liquid or gas-solid two-phase flows. Incorporating a third phase requires an additional measurement. Behling and Mewes [145] developed a dual-energy X-ray CT technique to measure time-averaged local void fraction in a gas-liquid-solid three-phase flow. They generated an X-ray fan beam at two sufficiently different X-ray energies (70 Kv and 200 kV) and obtained calibration intensities for pure systems of air, water, and PVC particles at both energies. They then operated a three-phase slurry column by bubbling gas through the PVC-water mixture. 074001-12 / Vol. 133, JULY 2011

X-ray intensities were then determined for the operating systems at each X-ray energy. These measurements were sufficient to determine time-averaged local phase fractions of each phase. Hu et al. [186], developed a new CT reconstruction technique that allowed CT reconstructions from only two independent projections. They incorporated “smoothing equations” to compensate for the lack of projection data found in typical CT imaging. They further utilized a rotating filter wheel composed of lead, copper, and air filter sections to take advantage of the X-ray beam hardening effects. This allowed them to discriminate between “hardened” X-rays that passed through the copper filter and “softened” X-rays that passed through the air region. The lead section, which made up half the filter wheel, blocked the X-ray beam completely. By operating the wheels 180 degrees out of phase, the two X-ray sources could be operated simultaneously while eliminating any scattering effects. With this system, Hu et al. [186] were able to image an airwater-oil three-phase system. Note that sodium iodide was added to the aqueous phase to enhance the signal difference between the oil and water. The authors concluded that the main application of this system would likely be for stratified-type or slug-type flow regimes where large interfaces separate the flow materials. X-ray tomographic microscopy was used by Waelchi et al. [187] to produce 3D images of stationary gas-liquid flows (i.e., annular flow) in a 250 lm wide by 100 lm deep microchannel. They used a synchrotron X-ray source and edge-enhancement radiographic projections to determine the 3D images because absorption-based methods do not provide a sufficient difference in absorption rates between the two fluids for the given path lengths. The edge-enhancement technique uses Fresnel diffraction in this case [86]. By selecting the appropriate energy level and sampleto-detector distance, the edge-enhanced projections produced a spatial resolution of approximately 4 lm. A new CT technique for axisymmetric multiphase flows was proposed by Wu et al. [188]. They used a flash X-ray system to produce single projection data with a pulse width of 100 ns to freeze the motion, and they acquired several images in sequence. They then assumed axisymmetry and used a Tikhonov Regularization method to reconstruct the time-averaged gas holdup in a 100 mm diameter airlift reactor. The authors completed a detailed analysis to show that their method was adequate, subject to the assumption of axisymmetric flow conditions. Finally, Charvet et al. [189] recently described an X-ray holotomographic technique to quantify the 3D liquid distribution in a fibrous filter. They utilized the European Synchrotron Radiation Facility (ESRF) to provide a parallel monochromatic X-ray beam. By rotating the filter sample, they produced 1500 independent image projections and then reconstructed them using filtered back projection to produce a voxel size of 0.7 lm. A single scan Transactions of the ASME


produced a traditional CT image providing a 3D mapping of the linear attenuation coefficient. For X-ray holotomography, multiple (two to four) CT scans were completed with different distances between the sample and detector for each scan. By combining the CT images obtained with different object-detector separation distances, they produced a 3D mapping of the electron density. This technique allowed Charvet et al. [189] to differentiate between fiber, liquid, and gas in the 3D image and to quantify the various volume fractions within their fibrous filter.

6 Current X-Ray Flow Visualization Challenges and Opportunities Typical noninvasive measurement techniques can either capture time-averaged 3D or time-varying 2D phase characteristics; however, most industrial multiphase applications involve locally unsteady or chaotic 3D flows. Hence, additional noninvasive measurement techniques must be developed to completely characterize and quantify various multiphase processes. Fast X-ray CT imaging will be further developed, although there are trade-offs between number of projections that influence spatial resolution and the speed of projection acquisition. Assumptions in the reconstruction symmetry will also affect resolution and accuracy of the reconstructed object. A key issue in quantifying the voids in a gas-liquid or gas-solid system is determining the boundary of the interface. In the case of a large void with good contrast, simple image manipulation methods are sufficient. However, as the signal to noise ratio in the image is reduced, interface detection becomes a challenge because of quantum mottle effects in the image processing. Further developments in image processing methods are to be expected. Medical imaging has used tracers that preferentially absorb X-rays to visualize dynamic biological processes, whereby the tracer is either ingested by or injected into a patient. These techniques could be extended to fluid flow studies where one fluid is tagged with an X-ray absorbing salt like potassium iodide. Preliminary trials have been completed in our laboratory to show it is possible to visualize water with dissolved potassium iodide mixing with plain water, but further development is needed before this technique produces useful data. XPTV will become a powerful tool to measure local velocities in a multiphase system, but additional development in tracer particle manufacturing is needed [139]. Tracers must be developed for specific system conditions and be small enough to follow the fluid flow. Additionally, multiple high speed detector/camera systems are required for this technique, which will increase the cost of developing XPTV. Finally, data storage, manipulation, and processing could become an issue with X-ray flow visualization. Depending on the imaging method, systems can easily produce gigabytes of data when studying a “simple” multiphase flow. Terabytes of data could be produced when imaging 3D time-resolved systems over long time periods. The process of analyzing and visualizing such large data sets will remain a challenge.

7

Conclusions

X-rays have been used for over 50 years to visualize the fluid flow in multiphase systems. Initially, simple radiographic imaging was done using traditional X-ray film processing. With the introduction of digital X-ray detection and CT imaging, stereoscopic and computed tomographic imaging have been used to characterize and quantify multiphase flow characteristics. X-ray visualization techniques have provided useful tools for multiphase flow characterization. Many research groups now have access to the required X-ray hardware and further technique development is expected.

References [1] Dudukovic, M. P., 2007, “Relevance Of Multiphase Reaction Engineering To Modern Technological Challenges,” Ind. Eng. Chem. Res., 46(25), pp. 8674– 8686.


[2] Dudukovic, M. P., Larachi, F., and Mills, P. L., 1999, “Multiphase Reactors Revisited,” Chem. Eng. Sci., 54, pp. 1975–1995. [3] Reese, J., Silva, E. M., Tang, S.-T., and Fan, L.-S., 1999, “Industrial Applications of Three-Phase Fluidization Systems,” Fluidization Solids Handling and Processing - Industrial Applications, W.-C. Yang, ed. Noyes Publications, Westwood, NJ, pp. 582–682. [4] Sundaresan, S., 2000, “Modeling the Hydrodynamics of Multiphase Flow Reactors: Current Status and Challenges,” AIChE J., 46(6), pp. 1102–1105. [5] Syamlal, M., ed., 2006, Report on Workshop on Multiphase Flow Research, Morgantown, WV, June 6-7, 2006. Report No. DOE/NETL-2007/1259. [6] Joshi, J. B. and Ranade, V. V., 2003, “Computational Fluid Dynamics for Designing Process Equipment: Expectations, Current Status, and Path Forward,” Ind. Eng. Chem. Res., 42, pp. 1115–1128. [7] Mudde, R. F., 2005, “Gravity-Driven Bubbly Flows,” Annu. Rev. Fluid Mech., 37, pp. 393–423. [8] Grohse, E. W., 1955, “Analysis of Gas-Fluidized Solid Systems by X-Ray Absorption,” AIChE J., 1(3), pp. 358–365. [9] Romero, J. B. and Smith, D. W., 1965, “Flash X-Ray Analysis of Fluidized Beds,” AIChE J., 11(4), pp. 595–600. [10] Rowe, P. N. and Partridge, B. A., 1965, “An X-Ray Study of Bubbles in Fluidised Beds,” Trans. Inst. Chem. Eng., 43, pp. T157–T175. [11] Bennett, A. W., Hewitt, G. F., Kearsey, H. A., Keeys, R. K. F., and Lacey, P. M. C., 1965, “Flow Visualization Studies of Boiling at High Pressure,” Proc. Inst. Mech. Eng., 180, pp. 1–11. [12] Hewitt, G. F. and Roberts, D. N., 1969, “Studies of Two-Phase Flow Patterns by Simultaneous X-Ray and Flash Photography,” United Kingdom Atomic Energy Authority (UKAEA), Harwell, Berkshire, Technical Report No. [13] Dudukovic, M. P., 2000, “Opaque Multiphase Reactors: Experimentation, Modeling and Troubleshooting,” Oil Gas Sci. Technol., 55(2), pp. 135–158. [14] Yates, J. G. and Simons, J. R., 1994, “Experimental Methods in Fluidization Research,” Int. J. Multiphase Flow, 20, pp. 297–330. [15] Williams, R. A. and Beck, M. S., eds., 1995, Process Tomography: Principles, Techniques and Applications, Butterworth-Heinemann Ltd., Oxford. [16] Beck, M. S. and Williams, R. A., 1996, “Process Tomography: A European Innovation and Its Applications,” Meas. Sci. Technol., 7, pp. 215–224. [17] Chaouki, J., Larachi, F., and Dudukovic, M. P., eds., Non-Invasive Monitoring of Multiphase Flows, Elsevier, New York. [18] Chaouki, J., Larachi, F., and Dudukovic, M. P., 1997, “Noninvasive Tomography and Velocimetric Monitoring of Multiphase Flows,” Ind. Eng. Chem. Res., 36, pp. 4476–4503. [19] Boyer, C., Duquenne, A. M., and Wild, G., 2002, “Measuring Techniques in Gas-Liquid and Gas-Liquid-Solid Reactors,” Chem. Eng. Sci., 57, pp. 3185– 3215. [20] Yates, J. G., Cheesman, D. J., Lettieri, P., and Newton, D., 2002, “X-Ray Analysis of Fluidized Beds and Other Multiphase Systems,” Kona, 20, pp. 133–143. [21] Van Ommen, J. R. and Mudde, R. F., 2008, “Measuring the Gas-Solids Distribution in Fluidized Beds – A Review,” International Journal of Chemical Reactor Engineering, 6, pp. R3. [22] Kak, A. C. and Slaney, M., 1998, Principles of Computerized Tomographic Imaging, Society of Industrial and Applied Mathematics, New York. [23] Marashdeh, Q., Fan, L. S., Du, B., and Warsito, W., 2008, “Electrical Capacitance Tomography – A Perspective,” Ind. Eng. Chem. Res., 47(10), pp. 3708–3719. [24] Toye, D., Marchot, P., Crine, M., and L’Homme, G., 1996, “Modelling of Multiphase Flow in Packed Beds by Computer-Assisted X-Ray Tomography,” Meas. Sci. Technol., 7(3), pp. 436–443. [25] Ceccio, S. L. and George, D. L., 1996, “A Review of Electrical Impedance Techniques for the Measurement of Multiphase Flows,” J. Fluids Eng., 118, pp. 391–399. [26] George, D. L., Torczynski, J. R., Shollenberger, K. A., O’Hern, T. J., and Ceccio, S. L., 2000, “Validation of Electrical-Impedance Tomography for Measurements of Material Distribution in Two-Phase Flows,” Int. J. Multiphase Flow, 26, pp. 549–581. [27] Tortora, P. R., Ceccio, S. L., O’Hern, T. J. Trujillo, S. M. and Torczynski, J. R., 2006, “Quantitative Measurement of Solids Distribution in Gas-Solid Riser Flows Using Electrical Impedance Tomography and Gamma Densitometry Tomography,” Int. J. Multiphase Flow, 32(8), pp. 972–995. [28] Fransolet, E., Crine, M., Marchot, P., and Toye, D., 2005, “Analysis of Gas Holdup in Bubble Columns with Non-Newtonian Fluid Using Electrical Resistance Tomography and Dynamic Gas Disengagement Technique,” Chem. Eng. Sci., 60, pp. 6118–6123. [29] Toye, D., Fransolet, E., Simon, D., Crine, M., L’Homme, G., and Marchot, P., 2005, “Possibilities and Limits of Application of Electrical Resistance Tomography in Hydrodynamics of Bubble Columns,” The Canadian Journal of Chemical Engineering, 83, pp. 4–10. [30] Makkawi, Y. T. and Wright, P. C., 2002, “Fluidization Regimes in a Conventional Fluidized Bed Characterized by Means of Electrical Capacitance Tomography,” Chem. Eng. Sci., 57, pp. 2411–2437. [31] Pugsley, T., Tanfara, H., Malcus, S., Cui, H., Chaouki, J. and Winters, C., 2003, “Verification of Fluidized Bed Electrical Capacitance Tomography Measurements with a Fiber Optic Probe,” Chem. Eng. Sci., 58, pp. 3923–3934. [32] Warsito, W. and Fan, L.-S., 2003, “3D-ECT Velocimetry for Flow Structure Quantification of Gas-Liquid-Solid Fluidized Beds,” The Canadian Journal of Chemical Engineering, 81(3-4), pp. 875–884. [33] Warsito, W. and Fan, L.-S., 2003, “ECT Imaging of Three-Phase Fluidized Bed Based on Three-Phase Capacitance Model,” Chem. Eng. Sci., 58, pp. 823–832.

JULY 2011, Vol. 133 / 074001-13


[34] Makkawi, Y. T. and Wright, P. C., 2004, “Electrical Capacitance Tomography for Conventional Fluidized Bed Measurements – Remarks on the Measuring Technique,” Powder Technol., 148(2), pp. 142–157. [35] Gamio, J. C., Castro, J., Rivera, L., Alamilla, J., Garcia-Nocetti, F., and Aguilar, L., 2005, “Visualization of Gas-Oil Two-Phase Flows in Pressurized Pipes Using Electrical Capacitance Tomography,” Flow Meas. Instrum., 16(2), pp. 129–134. [36] Ismail, I., Gamio, J. C., Bukhari, S. F. A., and Yang, W. Q., 2005, “Tomography for Multi-Phase Flow Measurement in the Oil Industry,” Flow Meas. Instrum., 16(2-3), pp. 145–155. [37] Warsito, W. and Fan, L.-S., 2005, “Dynamics of Spiral Bubble Plume Motion in the Entrance Region of Bubble Columns and Three Phase Fluidized Beds Using 3D ECT,” Chem. Eng. Sci., 60, pp. 6073–6084. [38] Du, B., Warsito, W., and Fan, L. S., 2006, “Imaging the Choking Transition in Gas-Solid Risers Using Electrical Capacitance Tomography,” Ind. Eng. Chem. Res., 45(15), pp. 5384–5395. [39] Utomo, M. B., Warsito, W., Sakai, T., and Uchida, S., 2001, “Analysis of Distributions of Gas and TiO2 Particles in Slurry Bubble Column Using Ultrasonic Computed Tomography,” Chem. Eng. Sci., 56, pp. 6073–6079. [40] Vatanakul, M., Zheng, Y., and Couturier, M., 2004, “Application of Ultrasonic Technique in Multiphase Flows,” Ind. Eng. Chem. Res., 43, pp. 5681–5691. [41] Zheng, Y. and Zhang, Q., 2004, “Simultaneous Measurements of Gas and Solid Holdups in Multiphase Systems Using Ultrasonic Technique,” Chem. Eng. Sci., 59, pp. 3505–3514. [42] Kumar, S. B., Moslemian, D., and Dudukovic, M. P., 1995, “A c-Ray Tomographic Scanner for Imaging Voidage Distribution in Two-Phase Flow Systems,” Flow Meas. Instrum., 6(1), pp. 61–73. [43] Kumar, S. B. and Dudukovic, M. P., 1996, “Computer Assisted Gamma and X-Ray Tomography: Applications to Multiphase Flow Systems,” Non-Invasive Monitoring of Multiphase Flows, J. Chaouki, F. Larachi and M. P. Dudukovic, ed., Elsevier, New York, pp. 47–103. [44] George, D. L., Torczynski, J. R., O’Hern, T. J., Shollenberger, K. A., Tortora, P. R., and Ceccio, S. L., 2001, “Quantitative Electrical-Impedance Tomography in an Electrically Conducting Bubble-Column Vessel,” ASME 2001 Fluids Engineering Division Summer Meeting, New Orleans, LA, May 29-June 1, 2001. [45] Yin, F., Afacan, A., Nandakumar, K., and Chuang, K. T., 2002, “Liquid Holdup Distribution in Packed Columns: Gamma Ray Tomography and CFD Simulation,” Chem. Eng. Process., 41, pp. 473–483. [46] Jin, H., Yang, S., He, G., Guo, Z., and Tong, Z., 2005, “An Experimental Study of Holdups in Large-Scale P-Xylene Oxidation Reactors Using the c-Ray Attenuation Approach,” Chem. Eng. Sci., 60, pp. 5955–5961. [47] Mudde, R. F., Bruneau, P. R. P., and Van Der Hagen, T. H. J. J., 2005, “TimeResolved c-Densitometry Imaging within Fluidized Beds,” Ind. Eng. Chem. Res. 44, pp. 6181–6187. [48] Rados, N., Shaikh, A., and Al-Dahhan, M., 2005, “Phase Distribution in a High Pressure Slurry Bubble Column Via a Single Source Computed Tomography,” The Canadian Journal of Chemical Engineering, 83, pp. 104– 112. [49] Shaikh, A. and Al-Dahhan, M., 2005, “Characterization of the Hydrodynamic Flow Regime in Bubble Columns Via Computed Tomography,” Flow Meas. Instrum., 16(2), pp. 91–98. [50] Kantzas, A., 1994, “Computation of Holdups in Fluidized and Trickle Beds by Computer-Assisted Tomography,” AIChE J., 40(7), pp. 1254–1261. [51] Marchot, P., Toye, D., Pelsser, A.-M., Crine, M., L’Homme, G. L., and Olujic, Z., 2001, “Liquid Distribution Images on Structured Packing by X-Ray Computed Tomography,” AIChE J., 47(6), pp. 1471–1476. [52] Schmit, C. E. and Eldridge, R. B., 2004, “Investigation of X-Ray Imaging of Vapor-Liquid Contactors: 1 – Studies Involving Stationary Objects and a Simple Flow System,” Chem. Eng. Sci., 59, pp. 1255–1266. [53] Schmit, C. E., Perkins, J., and Eldridge, R. B., 2004, “Investigation of X-Ray Imaging of Vapor-Liquid Contactors: 2 – Experiments and Simulations of Flows in an Air-Water Contactor,” Chem. Eng. Sci., 59, pp. 1267–1283. [54] Prasser, H. M., Misawa, M., and Tiseanu, I., 2005, “Comparison between Wire-Mesh Sensor and Ultra-Fast X-Ray Tomograph for an Air-Water Flow in a Vertical Pipe,” Flow Meas. Instrum., 16(2), pp. 73–83. [55] Hubers, J. L., Striegel, A. C., Heindel, T. J., Gray, J. N., and Jensen, T. C., 2005, “X-Ray Computed Tomography in Large Bubble Columns,” Chem. Eng. Sci., 60(22), pp. 6124–6133. [56] Franka, N. P. and Heindel, T. J., 2009, “Local Time-Averaged Gas Holdup in a Fluidized Bed with Side Air Injection Using X-Ray Computed Tomography,” Powder Technol., 193, pp. 69–78. [57] Mudde, R. F., 2010, “Time-Resolved X-Ray Tomography of a Fluidized Bed,” Powder Technol., 199(1), pp. 55–59. [58] Mudde, R. F., 2010, “Double X-Ray Tomography of a Bubbling Fluidized Bed,” Ind. Eng. Chem. Res., 49(11), pp. 5061–5065. [59] Dechsiri, C., Ghione, A., Van De Weil, F., Dehling, H. G., Paans, A. M. J., and Hoffmann, A. C., 2005, “Positron Emission Tomography Applied to Fluidization Engineering,” The Canadian Journal of Chemical Engineering, 83, pp. 88–96. [60] Harvel, G. D., Hori, K., Kawanishi, K., and Chang, J. S., 1999, “Cross-Sectional Void Fraction Distribution Measurements in a Vertical Annulus TwoPhase Flow by High Speed X-Ray Computed Tomography and Real-Time Neutron Radiography Techniques,” Flow Meas. Instrum., 10(4), pp. 259–266. [61] Prasser, H. M., 2008, “Novel Experimental Measuring Techniques Required to Provide Data for CFD Validation,” Nucl. Eng. Des., 238(3), pp. 744–770. [62] Rees, A. C., Davidson, J. F., Dennis, J. S., Fennell, P. S., Gladden, L. F., Hayhurst, A. N., Mantle, M. D., Muller, C. R., and Sederman, A. J., 2006, “The

074001-14 / Vol. 133, JULY 2011

[63]

[64] [65] [66]

[67] [68]

[69]

[70]

[71]

[72]

[73]

[74]

[75] [76] [77]

[78]

[79]

[80] [81] [82] [83] [84]

[85] [86] [87]

[88]

[89] [90]

[91]

[92]

Nature of the Flow Just above the Perforated Plate Distributor of a Gas-Fluidised Bed, as Imaged Using Magnetic Resonance,” Chem. Eng. Sci., 61(18), pp. 6002–6015. Muller, C. R., Holland, D. J., Sederman, A. J., Mantle, M. D., Gladden, L. F., and Davidson, J. F., 2008, “Magnetic Resonance Imaging of Fluidized Beds,” Powder Technol., 183(1), pp. 53–62. Powell, R. L., 2008, “Experimental Techniques for Multiphase Flows,” Phys. Fluids, 20(4), pp. 040605–040622. Devanathan, N., Moslemian, D., and Dudukovic, M. P., 1990, “Flow Mapping in Bubble Columns Using CARPT,” Chem. Eng. Sci., 45(8), pp. 2285–2291. Sun, J. G. and Chen, M. M., 1999, “Radioactive Tracer Techniques,” Instrumentation for Fluid-Particle Flow, S. L. Soo, ed., Noyes Publications, Park Ridge, NJ, pp. 354–401. Rados, N., Shaikh, A., and Al-Dahhan, M. H., 2005, “Solids Flow Mapping in a High Pressure Slurry Bubble Column,” Chem. Eng. Sci., 60, pp. 6067–6072. Chen, P., Gupta, P., Dudukovic, M. P., and Toseland, B. A., 2006, “Hydrodynamics of Slurry Bubble Column During Dimethyl Ether (DME) Synthesis: Gas-Liquid Recirculation Model and Radioactive Tracer Studies,” Chem. Eng. Sci., 61(19), pp. 6553–6570. Fraguio, M. S., Cassanello, M. C., Larachi, F., and Chaouki, J., 2006, “Flow Regime Transition Pointers in Three-Phase Fluidized Beds Inferred from a Solid Tracer Trajectory,” Chem. Eng. Process., 45(5), pp. 350–358. Seeger, A., Affeld, K., Goubergrits, L., Kertzscher, U., and Wellnhofer, E., 2001, “X-Ray-Based Assessment of the Three-Dimensional Velocity of the Liquid Phase in a Bubble Column,” Exp. Fluids, 31, pp. 193–201. Seeger, A., Kertzscher, U., Affeld, K., and Wellnhofer, E., 2003, “Measurement of the Local Velocity of the Solid Phase and the Local Solid Hold-up in a Three-Phase Flow by X-Ray Based Particle Tracking Velocimetry (XPTV),” Chem. Eng. Sci., 58, pp. 1721–1729. Lee, S.-J. and Kim, G.-B., 2003, “X-Ray Particle Image Velocimetry for Measuring Quantitative Flow Information inside Opaque Objects,” J. Appl. Physics, 94(5), pp. 3620–3623. Drake, J. B., Tang, L., and Heindel, T. J., 2009, “X-Ray Particle Tracking Velocimetry in Fluidized Beds,” Proceedings of the 2009 ASME Fluids Engineering Division Summer Meeting, Vail, CO, August 2–9, Paper No. FEDSM2009-78150. Furui, S., Umekawa, H., Tsuzuki, M., Okura, M., Ozawa, M., and Takenaka, N., 2005, “Visualization of Fluidized-Bed Heat Exchanger in Upward/Downword Flow Condition by Neutron Radiography,” Nucl. Instrum. Methods Phys. Res. A, 542, pp. 161–167. Heindel, T. J., 2002, “Bubble Size in a Cocurrent Fiber Slurry,” Ind. Eng. Chem. Res., 41(3), pp. 632–641. Sun, W. L., Littleton, H. E., and Bates, C. E., 2002, “Real-Time X-Ray Investigations on Lost Foam Mold Filling,” AFS Transactions, 110, pp. 1347–1356. Basavaraj, M. G., Gupta, G. S., Naveen, K., Rudolph, V., and Bali, R., 2005, “Local Liquid Holdups and Hysteresis in a 2-D Packed Bed Using X-Ray Radiography,” AIChE J., 51(8), pp. 2178–2189. Hulme, I. and Kantzas, A., 2004, “Determination of Bubble Diameter and Axial Velocity for a Polyethylene Fluidized Bed Using X-Ray Fluoroscopy,” Powder Technol., 147, pp. 20–33. Grassler, T. and Wirth, K.-E., 2000, “X-Ray Computer Tomography - Potential and Limitation for the Measurement of Local Solids Distribution in Circulating Fluidized Beds,” Chem. Eng. J., 77, pp. 65–72. Du, B., Warsito, W., and Fan, L.-S., 2005, “ECT Studies of Gas-Solid Fluidized Beds of Different Diameters,” Ind. Eng. Chem. Res., 44, pp. 5020–5030. Mudde, R. F., 2011, “Bubbles in a Fluidized Bed: A Fast X-Ray Scanner,” AIChE J., in press. Selman, J., 1994, The Fundamentals of X-Ray and Radium Physics, Charles C. Thompson, Publisher, Springfield, IL. Cartz, L., 1995, Nondestructive Testing, ASM International, Materials Park, OH. Ketcham, R. A. and Carlson, W. D., 2001, “Acquisition, Optimization and Interpretation of X-Ray Computed Tomography Imagery: Applications to the Geosciences,” Comput. Geosci., 27, pp. 381–400. Hsieh, J., 2003, Computed Tomography: Principles, Design, Artifacts, and Recent Advances, SPIE, Bellingham, WA. Pedrotti, F. L., Pedrotti, L. S., and Pedrotti, L. M., 2007, Introduction to Optics, Pearson Prentice Hall, Upper Saddle River, NJ. Mennell, J., Byrne, B., and Yan, Y., 2000, “Appraisal of Radiometric Techniques to Determine Absolute Solids Fraction in Pneumatic Suspensions of Particulate Solids,” Flow Meas. Instrum., 11(3), pp. 213–221. Chotas, H. G., Dobbins, J. T., and Ravin, C. E., 1999, “Principles of Digital Radiography with Large-Area, Electronically Readable Detectors: A Review of the Basics,” Radiology, 210(3), pp. 595–599. Heindel, T. J., Gray, J. N., and Jensen, T. C., 2008, “An X-Ray System for Visualizing Fluid Flows,” Flow Meas. Instrum., 19(2), pp. 67–78. Heindel, T. J., Hubers, J. L., Jensen, T. C., Gray, J. N., and Striegel, A. C., 2005, “Using X-Rays for Multiphase Flow Visualization,” Proceedings of the 2005 ASME Fluids Engineering Division Summer Meeting and Exhibition, Houston, TX, June 19–23, Paper No. FEDS2005-77359. Heindel, T. J., Jensen, T. C., and Gray, J. N., 2007, “Visualizing Fluid Flows with X-Rays,” Proceedings of FEDSM2007: 5th Joint ASME/JSME Fluids Engineering Conference, San Diego, CA, July 30 – August 2, Paper No. FEDSM2007-37023. Gupta, G. S., Litster, J. D., Rudolph, V., and White, E. T., 1997, “Flow Visualization Study in Porous Media Using X-Rays,” Steel Res., 68(10), pp. 434–440.

Transactions of the ASME


[93] Rowe, P. N. and Yacono, C. X. R., 1976, “The Bubbling Behaviour of Fine Powders When Fluidised,” Chem. Eng. Sci., 31(12), pp. 1179–1192. [94] Rowe, P. N., Santoro, L., and Yates, J. G., 1978, “The Division of Gas between Bubble and Interstitial Phases in Fluidised Beds of Fine Powders,” Chem. Eng. Sci., 33(1), pp. 133–140. [95] Yates, J. G., Cheesman, D. J., and Sergeev, Y. A., 1994, “Experimental Observations of Voidage Distribution around Bubbles in a Fluidized Bed,” Chem, Eng, Sci, 49(12), pp. 1885–1895. [96] He, Z., Wu, B., Chandrasekaran, B., Bellehumeur, C., and Kantzas, A., 2007, “X-Ray Fluoroscopy Measurements and CFD Simulation of Hydrodynamics in a Two Dimensional Gas-Solids Fluidized Bed,” 2007 ECI Conference on the 12th International Conference on Fluidization - New Horizons in Fluidization Engineering, Vancouver, Canada, pp. 481–488. [97] Jenneson, P. M. and Gundogdu, O., 2006, “In Situ X-Ray Imaging of Nanoparticle Agglomeration in Fluidized Beds,” Appl. Phys. Lett., 88, p. 034103. [98] Roels, S. and Carmeliet, J., 2006, “Analysis of Moisture Flow in Porous Materials Using Microfocus X-Ray Radiography,” Int. J. Heat Mass Transfer, 49(25), pp. 4762–4772. [99] Van Der Merwe, W., Nicol, W., and De Beer, F., 2007, “Trickle Flow Distribution and Stability by X-Ray Radiography,” Chem. Eng. J., 132(1), pp. 47–59. [100] Caulk, D. A., 2006, “A Foam Engulfment Model for Lost Foam Casting of Aluminum,” Int. J. Heat Mass Transfer, 49(21), pp. 3831–3845. [101] Uchida, K. and Okamoto, K., 2006, “Measurement of Powder Flow in a Screw Feeder by X-Ray Penetration Image Analysis,” Meas. Sci. Technol., 17(2), pp. 419–426. [102] Uchida, K. and Okamoto, K., 2008, “Measurement Technique on the Diffusion Coefficient of Powder Flow in a Screw Feeder by X-Ray Visualization,” Powder Technol., 187(2), pp. 138–145. [103] Jamet, F. and Thomer, G., 1976, Flash Radiography, Elsevier, New York. [104] Grady, D. E. and Kipp, M. E., 1994, “Experimental and Computational Simulation of the High Velocity Impact of Copper Spheres on Steel Plates,” Int. J. Impact Eng., 15(5), pp. 645–660. [105] Primex Physics International, 1996, Flash X-Ray Seminar Notes, San Leandro, CA. [106] Charbonnier, F., 1985, “Flash Radiography,” Nondestructive Testing Handbook. Vol. 3 - Radiography and Radiation Testing, L. E. Bryant and P. McIntire, eds., American Society for Metals: Metals Park, OH, pp. 491–531. [107] Triantafillopoulos, N. G. and Farrington, T. E., Jr., 1988, “Flash X-Ray Radiography Techniques for Visualizing Coating Flows,” 1988 TAPPI Coating Conference, New Orleans, LA, May 8–12, pp. 47–52. [108] Smith, G. B., Gamblin, B. R., and Newton, D., 1995, “X-Ray Imaging of Slurry Bubble Column Reactors: The Effect of System Pressure and Scale,” Trans. Inst. Chem. Eng., Part A 73(A6), pp. 632–636. [109] Farrington, T. E., Jr., 1986, “A More Fundamental Approach to the Problem of High Consistency Forming,” 1986 TAPPI Engineering Conference, Atlanta, GA, September 22–25, pp. 709–717. [110] Farrington, T. E., Jr., 1988, “Soft X-Ray Imaging Can Be Used to Assess Sheet Formation and Quality,” Tappi J., 71(5), pp. 140–144. [111] Farrington, T. E., Jr., 1988, “Flash X-Ray Imaging of Kraft Black Liquor Sprays,” Tappi J., 71(2), pp. 89–92. [112] Zavaglia, J. C. and Lindsay, J. D., 1989, “Flash X-Ray Visualization of Multiphase Flow During Impulse Drying,” Tappi J., 72(9), pp. 79–85. [113] Garner, A. E. and Heindel, T. J., 2000, “The Effect of Fiber Type on Bubble Size,” J. Pulp Pap. Sci., 26(7), pp. 266–269. [114] Heindel, T. J. and Monefeldt, J. L., 1997, “Flash X-Ray Radiography for Visualizing Gas Flows in Opaque Liquid/Fiber Suspensions,” 6th International Symposium on Gas-Liquid Two-Phase Flows, Vancouver, BC, June 22–26. [115] Heindel, T. J. and Monefeldt, J. L., 1998, “Observations of the Bubble Dynamics in a Pulp Suspension Using Flash X-Ray Radiography,” Tappi J., 81(11), pp. 149–158. [116] Heindel, T. J., 1999, “Bubble Size Measurements in a Quiescent Fiber Suspension,” J. Pulp Pap. Sci., 25(3), pp. 104–110. [117] Heindel, T. J. and Garner, A. E., 1999, “The Effect of Fiber Consistency on Bubble Size,” Nord. Pulp Pap. Res. J., 14(2), pp. 171–178. [118] Heindel, T. J., 2000, “Gas Flow Regime Changes in a Bubble Column Filled with a Fiber Suspension,” The Canadian Journal of Chemical Engineering, 78(5), pp. 1017–1022. [119] Heindel, T. J. and Omberg, S. J., 2001, “Gas Bubble Size in a Synthetic Fiber Slurry,” in 4th International Conference on Multiphase Flow, E. E. Michaelides, ed., ICMF-2001, New Orleans, LA. [120] Grantham, S. G. and Forsberg, F., 2004, “Measurement of Granular Flow in a Silo Using Digital Speckle Radiography,” Powder Technol., 146(1), pp. 56–65. [121] Kastengren, A. L., Powell, C. F., Riedel, T., Cheong, S.-K., Im, K.-S., Liu, X., Wanng, Y. J., and Wang, J., 2008, “Nozzle Geometry and Injection Duration Effects on Diesel Sprays Measured by X-Ray Radiography,” J. Fluids Eng., 130, pp. 041301-1 – 041301-12. [122] Coutier-Delgosha, O., Vabre, A., Hocevar, M., Delion, R., Dazin, A., Lazaro, D., Gmar, M., Fezzaa, K., and Lee, W.-K., 2009, “Local Measurements in Cavitating Flow by Ultra Fast X-Ray Imaging,” Proceedings of the 2009 ASME Fluids Engineering Division Summer Meeting, Vail, CO, August 2–6, Paper No. FEDS2009-78407. [123] Kim, J., Je, J., Kaviany, M., Son, S. Y., and Kim, M. H., 2010, “Visualization of Water Distribution in Gas Diffusion Layer of Fuel Cell Using 2-D and 3-D X-Ray Radiography,” Proceedings of the 7th International Conference on Multiphase Flow, Tampa, FL, May 30 – June 4. [124] Doering, E. R., 1992, “Three-Dimensional Flaw Reconstruction Using a RealTime X-Ray Imaging System,” Ph.D. thesis, Iowa State University, Ames, IA.


[125] Jensen, T. C. and Gray, J. N., 2006, “Model-Assisted Reconstruction of Volumetric Data,” U.S. Patent No. 7,092,484, [126] Lee, D. J. and Cha, S. S., 2003, “Three-Dimensional Measurement of Fluid Flow in Stereoscopic Tracking Velocimetry,” Proceedings of SPIE, Anaheim, CA, pp. 73–85. [127] Lee, D. J., Cha, S. S., and Ramachandran, N., 2004, “Three-Dimensional High-Resolution Optical/X-Ray Stereoscopic Tracking Velocimetry,” 2004 ASME International Mechanical Engineering Congress and Exposition, Anaheim, CA, Paper No. IMECE2004-62450. [128] Lee, D. J., Cha, S. S., and Ramachandran, N., 2006, “Application of Stereoscopic Tracking Velocimetry for Experimental and Numerical Investigation of Directional Solidification,” Exp. Therm. Fluid Sci., 30(3), pp. 203. [129] Seeger, A., Affeld, K., Kertzscher, U., Goubergrits, L., and Wellnhofer, E., 2001, “Assessment of Flow Structures in Bubble Columns by X-Ray Based Particle Tracking Velocimetry,” 4th International Symposium on Particle Image Velocimetry, Gottingen, Germany. [130] Schober, L. and Kantzas, A., 2004, “Analysis and Interpretation of Multiple Radioactive Particle Tracking Data on Laboratory Scale Air/Polyethylene Fluidized Beds,” The Canadian Journal of Chemical Engineering, 82, pp. 220– 226. [131] Kertzscher, U., Seeger, A., Affeld, K., Goubergrits, L., and Wellnhofer, E., 2004, “X-Ray Based Particle Tracking Velocimetry – A Measurement Technique for Multi-Phase Flows and Flows without Optical Access,” Flow Meas. Instrum., 15(4), pp. 199–206. [132] Owens, S. A., Kossmann, A., Farone, J., Eldridge, R. B., and Ketcham, R. A., 2009, “Flow Field Visualization in Structured Packing Using Real Time X-Ray Radiography,” Ind. Eng. Chem. Res., 48(7), pp. 3606–3618. [133] Shimada, T., Habu, H., Seike, Y., Ooya, S., Miyachi, H., and Ishikawa, M., 2007, “X-Ray Visualization Measurement of Slurry Flow in Solid Propellant Casting,” Flow Meas. Instrum., 18(5), pp. 235–240. [134] Kim, G.-B. and Lee, S.-J., 2004, “Velocity Field Measurements of Blood Flows Using a Synchrotron X-Ray Imaging Technique,” 2004 Proceedings of SPIE International Symposium on Biomedical Optics, San Jose, CA, January 24–29. [135] Im, K.-S., Fezzaa, K., Wang, Y. J., Liu, X., Wang, J., and Lai, M. C., 2007, “Particle Tracking Velocimetry Using Fast X-Ray Phase-Contrast Imaging,” Appl. Phys. Lett., 90(9), pp. 091919-1. [136] Kakimoto, K., Eguchi, M., Watanabe, H., and Hibiya, T., 1988, “Direct Observation by X-Ray Radiography of Convection of Molten Silicon in the Czochralski Growth Method,” J. Cryst. Growth, 88(3), pp. 365–370. [137] Munakata, T. and Tanasawa, I., 1999, “Study on Silicon Melt Convection During the Rf-Fz Crystal Growth Process: I. Experimental Flow Visualization,” J. Cryst. Growth, 206(1-2), pp. 23–26. [138] Drake, J. B., Franka, N. P., and Heindel, T. J., 2008, “X-Ray Particle Tracking Velocimetry for Applications in Fluidized Beds,” Proceedings of the 2008 ASME International Mechanical Engineering Congress and Exposition, Boston, MA, October 31 – November 6, Paper No. IMECE2008-66224. [139] Drake, J. B., Kenney, A. L., and Heindel, T. J., 2011, “Developing Tracer Particles for X-Ray Particle Tracking Velocimetry,” Proceedings of the 2011 ASME-JSME-KSME Joint Fluids Engineering Conference, AJK2011, Hamamatsu, Japan, July 24–29, Paper No. AJK2011-11009. [140] Morgan, T. B. and Heindel, T. J., 2007, “Observations of Dense Particle Motion in a Vibration-Excited Granular Bed,” Proceedings of IMECE2007: 2007 ASME International Mechanical Engineering Congress and Exposition, Seattle, WA, November 11–15, Paper No. IMECE2007-42306. [141] Morgan, T. B. and Heindel, T. J., 2010, “X-Ray Particle Tracking of Dense Particle Motion in a Vibration-Excited Granular Bed,” Proceedings of the 2010 ASME International Mechanical Engineering Congress and Exposition, Vancouver, British Columbia, Canada, November 12–18, Paper No. IMECE2010-39106. [142] Wellington, S. L. and Vinegar, H. J., 1987, “X-Ray Computerized Tomography,” J. Pet. Technol., 39(8), pp. 885–898. [143] Barrett, H. H. and Swindell, W., 1981, Radiological Imaging: The Theory of Image Formation, Detection, and Processing, Academic, New York. [144] Ikeda, T., Kotani, K., Maeda, Y., and Kohno, H., 1983, “Preliminary Study of Application of X-Ray CT Scanner to Measurement of Void Fractions in Steady State Two-Phase Flows,” J. Nucl. Science Technol., 20(1), pp. 1–12. [145] Behling, M. and Mewes, D., 2006, “Dual-Energy X-Ray Tomographic Measurement of Local Phase Fractions in 3-Phase Bubble Columns,” 2006 ASME Joint U.S. - European Fluids Engineering Summer Meeting, Miami, Fl, July 17–20, Paper No. FEDSM2006-98142. [146] Wu, C., Cheng, Y., Ding, Y., Wei, F., and Jin, Y., 2007, “A Novel X-Ray Computed Tomography Method for Fast Measurement of Multiphase Flow,” Chem. Eng. Sci., 62(16), pp. 4325–4335. [147] Du, D.-X., Beni, A. N., Farajzadeh, R., and Zitha, P. L. J., 2008, “Effect of Water Solubility on Carbon Dioxide Foam Flow in Porous Media: An X-Ray Computed Tomography Study,” Ind. Eng. Chem. Res., 47(16), pp. 6298– 6306. [148] Hewitt, G. F., 1982, “Measurement Techniques: Overall Measurements - Measurement of Void Fraction,” Handbook of Multiphase Systems, G. Hetsroni, ed., Hemisphere Publishing Corp., New York, Chap. 10.2.1.2. [149] Drake, J. B. and Heindel, T. J., 2011, “The Repeatability and Uniformity in 3D Fluidized Beds,” Powder Technol. (submitted). [150] Heindel, T. J., Jensen, T. C., Drake, J. B., McCormick, N., and Riveland, M. L., 2010, “3D X-Ray CT Imaging of Cavitation from a Butterfly Valve,” Proceedings of the 7th International Conference on Multiphase Flow, Tampa, FL, May 30 – June 4, Paper No. 1746.

JULY 2011, Vol. 133 / 074001-15


[151] ASTM, 1997, “Standard Guide for Computed Tomography (CT) Imaging,” Technical Report, American Society for Testing and Materials. [152] Lutran, P. G., Ng, K. M., and Delikat, E. P., 1991, “Liquid Distribution in Trickle-Beds. An Experimental Study Using Computer-Assisted Tomography,” Ind. Eng. Chem. Res., 30(6), pp. 1270–1280. [153] Toye, D., Marchot, P., Crine, M., Pelsser, A. M., and L’Homme, G., 1998, “Local Measurements of Void Fraction and Liquid Holdup in Packed Columns Using X-Ray Computed Tomography,” Chem. Eng. Process., 37(6), pp. 511–520. [154] Schmit, C. E., Cartmel, D. B., and Eldridge, R. B., 2001, “The Experimental Application of X-Ray Tomography to a Vapor-Liquid Contactor,” Chem. Eng. Sci., 56, pp. 3431–3441. [155] Van Der Merwe, W., Nicol, W., and De Beer, F., 2007, “Three-Dimensional Analysis of Trickle Flow Hydrodynamics: Computed Tomography Image Acquisition and Processing,” Chem. Eng. Sci., 62(24), pp. 7233–7244. [156] Kantzas, A. and Kalogerakis, N., 1996, “Monitoring the Fluidization Characteristics of Polyolefin Resins Using X-Ray Computer Assisted Tomography Scanning,” Chem. Eng. Sci., 51(10), pp. 1979–1990. [157] Kantzas, A., Wright, I., and Kalogerakis, N., 1997, “Quantification of Channelling in Polyethylene Resin Fluid Beds Using X-Ray Computer Assisted Tomography (CAT),” Chem. Eng. Sci., 52(13), pp. 2023–2035. [158] Wu, B., Yu, G., Bellehumeur, C., and Kantzas, A., 2007, “Dynamic Flow Behavior Measurements in Gas-Solid Fluidized Beds Using Different Non-Intrusive Techniques and Polyethylene Powder,” Flow Meas. Instrum., 18(5), pp. 197–203. [159] Chen, M.-R., Hinkley, R. E., and Killough, J. E., 1996, “Computed Tomography Imaging of Air Sparging in Porous Media,” Water Resour. Res., 32(10), pp. 3013–3024. [160] Anderson, S. H., Peyton, R. L., Wigger, J. W., and Gantzer, C. J., 1992, “Influence of Aggregate Size on Solute Transport as Measured Using Computed Tomography,” Geoderma, 53(3), pp. 387–398. [161] Heijs, A. W. J., De Lange, J., Schoute, J. F. T., and Bouma, J., 1995, “Computed Tomography as a Tool for Non-Destructive Analysis of Flow Patterns in Macroporous Clay Soils,” Geoderma, 64, pp. 183–196. [162] Morooka, S., Ishizuka, T., Iizuka, M., and Yoshimura, K., 1989, “Experimental Study on Void Fraction in a Simulated BWR Fuel Assembly (Evaluation of Cross-Sectional Averaged Void Fraction),” Nucl. Eng. Des., 114(1), pp. 91–98. [163] Mitsutake, T., Morooka, S., Suzuki, K., Tsunoyama, S., and Yoshimura, K., 1990, “Void Fraction Estimation within Rod Bundles Based on Three-Fluid Model and Comparison with X-Ray CT Void Data,” Nucl. Eng. Des., 120(2), pp. 203–212. [164] Boden, S., Bieberle, M., and Hampel, U., 2008, “Quantitative Measurement of Gas Hold-up Distribution in a Stirred Chemical Reactor Using X-Ray ConeBeam Computed Tomography,” Chem. Eng. J., 139(2), pp. 351–362. [165] Ford, J. J., Heindel, T. J., Jensen, T. C., and Drake, J. B., 2008, “X-Ray Computed Tomography of a Gas-Sparged Stirred-Tank Reactor,” Chem. Eng. Sci., 63, pp. 2075–2085. [166] Deza, M., Battaglia, F., and Heindel, T. J., 2007, “Computational Modeling of Biomass in a Fluidized Bed Gasifier,” Proceedings of IMECE2007: 2007 ASME International Mechanical Engineering Congress and Exposition, Seattle, WA, November 11–15, Paper No. IMECE2007-43097. [167] Franka, N. P., Heindel, T. J., and Battaglia, F., 2007, “Visualizing Cold-Flow Fluidized Beds with X-Rays,” Proceedings of IMECE2007: 2007 ASME International Mechanical Engineering Congress and Exposition, Seattle, WA, November 11–15, Paper No. IMECE2007-43073. [168] Deza, M., Franka, N. P., Battaglia, F., and Heindel, T. J., 2009, “CFD Modeling and X-Ray Imaging of Biomass in a Fluidized Bed,” J. Fluids Eng., 131(11), p. 111303. [169] Franka, N. P., Drake, J. B., and Heindel, T. J., 2008, “Minimum Fluidization Velocity and Gas Holdup in Fluidized Beds with Side Port Air Injection,” Proceedings of FEDSM2008: 2008 ASME Fluids Engineering Division Summer Meeting, Jacksonville, FL, August 10–14, Paper No. FEDSM2008-55100. [170] Min, J., Drake, J. B., Fox, R. O., and Heindel, T. J., 2008, “CFD Modeling of Cold-Flow Fluidized Beds and Validation with X-Ray Computed Tomography,” 2008 AIChE Annual Meeting, Philadelphia, PA, November 16–21.

074001-16 / Vol. 133, JULY 2011

[171] Deza, M., Battaglia, F., and Heindel, T. J., 2008, “A Validation Study for the Hydrodynamics of Biomass in a Fluidized Bed,” Proceedings of FEDSM2008: 2008 ASME Fluids Engineering Division Summer Conference, Jacksonville, FL, August 10–14, Paper No. FEDSM2008-55158. [172] Deza, M., Battaglia, F., and Heindel, T. J., 2009, “Modeling a Biomass Fluidizing Bed with Side Port Air Injection,” Proceedings of the 2009 ASME Fluids Engineering Division Summer Meeting, Vail, CO, August 2–6, Paper No. FEDSM2009-78372. [173] Gavi, E., Heindel, T. J., and Fox, R. O., 2010, “Modeling Fluidization in Biomass Gasification Processes,” Proceedings of the 7th International Conference on Multiphase Flow, Tampa, FL, May 30 – June 4, Paper No. 1956. [174] Min, J., Drake, J. B., Heindel, T. J., and Fox, R. O., 2010, “Experimental Validation of CFD Simulations of a Lab-Scale Fluidized-Bed Reactor with and without Side-Gas Injection,” AIChE J., 56(6), pp. 1434–1446. [175] DiVall, M. and Heindel, T. J., 2009, “X-Ray Flow Visualization of a Circular Hydraulic Jump,” Proceedings of the 2009 ASME Fluids Engineering Division Summer Meeting, Vail, CO, August 2–6, Paper No. FEDSM2009-78035. [176] Moreno-Atanasio, R., Williams, R. A., and Jia, X., 2010, “Combining X-Ray Microtomography with Computer Simulation for Analysis of Granular and Porous Materials,” Particuology, 8(2), pp. 81–99. [177] Harvel, G. D., Hori, K., Kawanishi, K., and Chang, J. S., 1996, “Real-Time Cross-Sectional Averaged Void Fraction Measurements in Vertical Annulus Gas-Liquid Two-Phase Flow by Neutron Radiography and X-Ray Tomography Techniques,” Nucl. Instrum. Methods Phys. Res. A, 371(3), pp. 544–552. [178] Hori, K., Kawanishi, K., Hamamura, H., Ochi, M., and Akai, M., 1994, “A High Speed X-Ray Computed Tomography Scanner for Multipurpose Flow Visualization and Measurement,” Proceedings of the Fourth International Topical Meeting on Nuclear Thermal-Hydraulics, Operations, and Safety, Taipei, Taiwan, April 6–8, pp. 43-D-1–43-D-6. [179] Hori, K., Fujimoto, T., and Kawanishi, K., 1998, “Development of Ultra-Fast X-Ray Computed Tomography Scanner System,” IEEE Trans. Nucl. Sci., 45, pp. 2089–2094. [180] Hori, K., Fujimoto, T., Kawanishi, K., and Nishikawa, H., 2000, “Development of an Ultrafast X-Ray Computed Tomography Scanner System: Application for Measurement of Instantaneous Void Distribution of Gas–Liquid Two-Phase Flow,” Heat TransferAsian Res., 29(3), pp. 155–165. [181] Kai, T., Misawa, M., Takahashi, T., Tiseanu, I., and Ichikawa, N., 2005, “Observation of 3-D Structures of Bubbles in a Fluidized Catalyst Bed,” The Canadian Journal of Chemical Engineering, 83, pp. 113–118. [182] Bieberle, M. and Hampel, U., 2006, “Evaluation of a Limited Angle Scanned Electron Beam X-Ray CT Approach for Two-Phase Pipe Flows,” Meas. Sci. Technol., 17, pp. 2057–2065. [183] Bieberle, M., Fischer, F., Schleicher, E., Franke, M., Menz, H. J., Mayer, H. G., Laurien, E., and Hampel, U., 2010, “Ultrafast Multiphase Flow Imaging by Electron Beam X-Ray Computed Tomography,” Proceedings of the 7th International Conference on Multiphase Flow, Tampa, FL, May 30 – June 4. [184] Bieberle, M., Fischer, F., Schleicher, E., Menz, H. J., Mayer, H. G., and Hampel, U., 2010, “Ultrafast Cross-Sectional Imaging of Gas-Particle Flow in a Fluidized Bed,” AIChE J., 56(8), pp. 2221–2225. [185] Bieberle, M., Schleicher, E., Fischer, F., Koch, D., Menz, H. J., Mayer, H. G., and Hampel, U., 2010, “Dual-Plane Ultrafast Limited-Angle Electron Beam X-Ray Tomography,” Flow Meas. Instrum., 21(3), pp. 233–239. [186] Hu, B., Stewart, C., Hale, C., Lawrence, C., Hall, A., Zwiens, H., and Hewitt, G., 2005, “Development of an X-Ray Computed Tomography (CT) System with Sparse Sources: Application to Three-Phase Pipe Flow Visualization,” Exp. Fluids, 39(4), pp. 667–678. [187] Waelchli, S., Rudolph Von Rohr, P., and Stampanoni, M., 2005, “Multiphase Flow Visualization in Microchannels Using X-Ray Tomographic Microscopy (XTM),” J. Flow Visualization Image Process., 12, pp. 1–13. [188] Wu, C., Cheng, Y., Liu, M., and Jin, Y., 2008, “Measurement of Axisymmetric Two-Phase Flows by an Improved X-Ray-Computed Tomography Technique,” Ind. Eng. Chem. Res., 47(6), pp. 2063–2074. [189] Charvet, A., Rolland Du Roscoat, S., Peralba, M., Bloch, J. F., and Gonthier, Y., 2011, “Contribution of Synchrotron X-Ray Holotomography to the Understanding of Liquid Distribution in a Medium During Liquid Aerosol Filtration,” Chem, Eng, Sci, 66(4), pp. 624–631.

Transactions of the ASME


A Review of X-Ray Flow Visualization With

A Review of X-Ray Flow Visualization With

Suggest Documents

Flow visualization and flow cytometry with

Flow Visualization in Stirred Vessels: A Review of ... - CiteSeerX

Synchrotron Xray diffraction experiments with a

Uandbook of Flow Visualization

Visualization of Vortex Flow Patterns with a Uniformly Perforated ...

TEACHING FLUID MECHANICS WITH FLOW VISUALIZATION VIDEOS

Flow Visualization on Naca2421 Airfoil with and

Flow visualization in parallel-plate ducts with

Assessment of a Novel Flow Visualization

Visualization of Superfluid Helium Flow

Turbulent Laser - Flow Visualization

Turbulent Laser - Flow Visualization

Fluid Flow Visualization - CiteSeerX

Buoyancy - Flow Visualization

Scanning xray microscope with 75nm resolution

In situ microfocused Xray beam characterization with a lensless

Schlieren visualization of a stratified flow around a ... - Springer Link

Xray Fluorescence Tomography of Aged

Karman Vortex Street - Flow Visualization

Context-Controlled Flow Visualization in

Karman Vortex Street - Flow Visualization

DC Load Flow Visualization

XRAY T3R Parts List

Redox flow batteries: a review