Search

Abstract This paper introduces time-lapse radiography as an in situ technique to image and quantify changes in the internal structure of a porous medium with sub-second temporal resolution. To demonstrate the technique's potential, an experiment was performed using a model system involving flow of a suspension containing ground marble particles through a porous bed of compacted glass beads housed within a pressurized flow rig. During the experiment, particle deposition occurred both within the internal porous structure and on its surface (forming a filter cake). The volume of particles deposited was derived from changes in the grey scale of the radiographs. At the initial stages of the experiment, the volume of particles deposited internally was seen to increase linearly with time. The subsequent growth and compaction of an external filter cake decreased the rate of internal particle deposition. The filter cake's structure was observed to fail owing to increasing stress at higher pressures. The demonstrative experiment illustrates the potential of time-lapse radiography as a new tool to elucidate mechanisms underpinning formation damage, and to optimise drilling fluids and enhanced oil recovery (EOR). A critical assessment of the technique's advantages and limitations to characterise particulate behaviour within porous media is included. Graphical abstract Image 1 Highlights • New method to image and quantify particle intrusion and filter cake evolution. • Formation of a filter cake is preceded by particle deposition in the pores. • Flow through a filter cake can weaken the structure of deposited particles. • Abrupt failure of filter cakes can strengthen the deposited particle agglomerates. [ABSTRACT FROM AUTHOR]

Time-lapse X-ray computed tomography is used to image the evolution of calcite precipitation during flow through microporous quartz over the course of 400 h. The growth rate decreases by more than seven times, which is linked to the clogging of flow paths that restricts flow to some regions of the column. Fewer precipitates are observed as a function of column depth, which is found to be related to a differential nucleation density along the sample. A higher nucleation density closer to the inlet implies more crystal volume increase per unit of time without affecting the rate if normalized to the surface area of crystals. Our overall growth rates measured in porous media are orders of magnitude slower than growth rates derived from traditional precipitation experiments on free surfaces. Based on our time-lapse results we hypothesize a scenario where the evolving distribution of precipitates within a pore structure during precipitation progressively modifies the local transport through the pores. Within less permeable regions the saturation index may be lower than along the main flow paths. Therefore, the reactive crystal surfaces within those regions grow at a slower rate than that expected from the bulk fluid composition. Since the amount of reactive surface area within these less permeable regions increases over time, the overall growth rate decreases without a necessary significant change of the bulk fluid composition along more permeable flow paths. In conclusion, the overall growth rates in an evolving porous media expected from bulk fluid compositions alone can be overestimated due to the development of stagnant sub-regions where the reactive surface area is bath by a solution with lower saturation index. In this context we highlight the value of time-lapse 3D studies for understanding the dynamics of mineral precipitation in porous media. [ABSTRACT FROM AUTHOR]

Minerals, metals, and plastics are indispensable for a modern society. Yet, their limited supply necessitates optimized extraction and recycling processes, which must be meticulously adapted to the material properties. Current imaging approaches perform material analysis on crushed particles imaged with computed tomography (CT) using segmentation and mass characterization. However, their inability to reliably separate touching particles and need to annotate and retrain on new images, leaves untapped potential. By contrast, particle-level characterization unlocks better understanding of particle properties such as mass, appearance and structure. Here, we propose ParticleSeg3D, an instance segmentation method for particle-level characterization with strongly varying properties from CT images. Our approach is based on the powerful nnU-Net, introduces a particle size normalization, employs a border-core representation, and is trained with a diverse dataset. We demonstrate that ParticleSeg3D can be applied out-of-the-box to a large variety of materials without retraining, including materials and properties not present during training. [Display omitted] • A novel method based on nnU-Net for 3D instance segmentation of individual particles. • Generalization to different sizes, shapes, and compositions of various materials. • Generalization to Out-of-Distribution data without requiring retraining. • A new large dataset encompassing diverse particle types and compositions. [ABSTRACT FROM AUTHOR]

3D concrete printing can be used for automatic repair, aiming at increasing efficiency and reducing the use of manpower, as compared to conventional manual repair. Ensuring strong adhesion of the 3D printed materials and characterizing the interface is crucial for the success of this approach. To this end, we first evaluated the adhesion performance of 3D printed concrete based on a low-carbon binder, i.e., calcium sulfoaluminate (CSA) cement. CSA cement pastes with different re-dispersible polymer powder substitution rates were used as coating before 3D printing. We used two approaches, X-ray computed tomography (CT) and scanning electron microscopy (SEM), to characterize the interface region. Results indicate that the yield stress of the 3D printing material and the coating material plays an important role in the adhesion in both the fresh and hardened state. Moreover, the limited adhesion at the overlay interface can be due to the limited contact at the centimeter and micro scale, as reflected by CT and SEM, respectively. Applying a coating can fill the air voids at the interface, while the placement of a coating may not improve the bond strength and can even have a negative impact due to the limited liquid medium at the interface. • Lower yield stress in the 3D printing material can enhance the contact area at the interface. • A coating with a higher yield stress promotes more excellent adhesion in the early stage by preventing slippage. • Applying a coating can enhance adhesion in the hardened state by filling the air voids. • X-ray computed tomography (CT) and scanning electron microscopy (SEM) are adopted to characterize the interface. [ABSTRACT FROM AUTHOR]