Effect of Neutrally Buoyant Particles on Horizontal Turbulent Flow Through a Tee-Junction

Turbulent particle-laden flows are of high interest due to their presence in many industrial applications. High Reynolds number flows containing solid particles, create complex flows and erosive environments. The effect that the particles have on the turbulence of the surrounding fluid is referred to in the literature as turbulence modulation. This is an area of research in which there is still much to learn to enable a deeper understanding of the physics behind these complex flows. Data that would be of particular usefulness are at higher Reynolds numbers (Re ≥ 100,000), and dense loadings (ΦV ≥ 1%). In this work, turbulent particle-laden flow through a simplified industrial geometry was studied at an upper Reynolds number of 115,000 and particle loadings up to 5% by weight/volume (specific gravity = 1) to address these needs. The flow within a tee junction with the 90-degree branch closed-off downstream was studied. This is analogous to a duct flow but with an exposed region of fluid at the location of the closed-off branch. Super absorbent particles were used as the solid phase, which became index-matched and neutrally buoyant upon saturation with water. Data were acquired using 2-D planar particle image velocimetry (PIV) along the center span of the tunnel. Mean and root-mean-square (rms) velocities were calculated for the fluid phase. Particle loadings studied were 0%, 1%, 3%, and 5 at flow Reynolds numbers of 11,500 and 115,000. Velocity contour plots are presented to provide a macro description of the flow. Three horizontal positions within the shear layer region were selected for profile comparison (x* = −0.45, 0, 0.45). Prior literature suggested that the particles would attenuate the turbulence, however, the result showed no single trend in the current data. The mean velocities were nominally unaffected by loading for a respective Reynolds number case. Turbulence modulation of the flow was found to be sensitive to the Reynolds number, as at x* = −0.45 weakening of the rms was observed in the low Reynolds number case and strengthening in the high Reynolds number case for the same particle loading in the same region of the geometry.

Experimental Study and Modeling of Slug Dissipation in a Horizontal Enlarged Impacting Tee-Junction

Summary A novel experimental and theoretical study on slug dissipation in a horizontal enlarged impacting tee-junction (EIT) is carried out. Both flowing-slug injection and stationary-slug injection into the EIT are studied, and the effects of inlet slug length and liquid-phase fluid properties on the slug dissipation in the EIT are investigated. A total of 161 experimental data are acquired for air-water and air-oil flow. The flowing-slug data (with a horizontal inlet) show that the slug dissipation length increases with increasing mixture velocity, demonstrating a nonlinear trend with a steeper slope at lower mixture velocities. The effect of superficial gas velocity on the slug dissipation length is more pronounced compared with the effect of superficial liquid velocity. For stationary-slug injection into the EIT (with a 5° upward inclined inlet), the injected slug lengths vary between 40d to 100d (d is the inlet diameter). The data reveal that, when increasing the superficial gas velocity or the inlet slug size, the dissipation length in the EIT branches increases. For this case, the ratio of the slug dissipation length to the inlet slug length is higher for air-water compared with air-oil. A slug dissipation model is developed using the slug-tracking approach, which is based on the flow mechanisms of liquid shedding at the back of the slug and liquid drainage and penetration of bubble turning at the front of the slug. These phenomena result in different translational velocities at the back and the front of the slug, which result in the dissipation of the slug body. Evaluation of model predictions against the acquired experimental data shows an average absolute relative error of less than 11%.

Dissipation of Injected Slug in Enlarged Impacting Tee in Single Branch Blocking Configuration

A novel Enlarged Impacting Tee-Junction (EIT), which introduces longer slugs to be dissipated utilizing “Single-Branch-Blocking” is studied experimentally and theoretically under stationary slug-injection conditions to further understand the dissipation mechanism through observation of longer slugs. The EIT test section is designed and constructed, which consists of one inlet pipe connected to a larger, perpendicular pipe allowing flow in both directions. The inlet is 4.6 m of 0.05 m diameter pipe, while the perpendicular “manifold” is 0.074 m in diameter and 5.5 m in length. In order to observe the dissipation of longer slugs, a modification is made to the Normal EIT configuration. The longer slugs in the EIT are generated by blocking one of the EIT branches, allowing flow in only the unblocked branch of the EIT. Thus, the entire injected slug (rather than half in the case of no blocking configuration) dissipates in the branch. For this configuration, stationary slugs are injected into the EIT with lengths of 40d, 50d, 60d, and 70d (with d being the inlet diameter). A total of 64 slug injection tests are conducted utilizing both air-water and air-oil flow. The experimental data show that slug dissipation has a nonlinear increasing relationship with mixture velocity. Furthermore, the data show that higher dissipation length is observed with air-water flow as compared to air-oil flow in the slug injection experiments due to higher shed slug volume of oil. Also, the acquired data are used to validate the EIT slug dissipation model developed by Mohammadikharkeshi (2018). For the Single-Branch-Blocking investigation, comparison between the acquired experimental data and the modified Mohammadikharkeshi (2018) Normal EIT model predictions reveals excellent comparison, with an average discrepancy of 12%.

Experimental and Numerical Investigation of Mixing Phenomena in Double-Tee Junctions

This work investigates mixing phenomena in a pressurized pipe system with two sequential Tee junctions and experiments are conducted for a range of different inlet flow ratios, varying distances between Tee junctions and two pipe branching configurations. Additionally, obtained experimental results are compared with results from previous studies by different authors and are used to validate the numerical model using the open source computational fluid dynamics toolbox OpenFOAM. Two different numerical approaches are used—Passive scalar model and Multiphase model. It is found that both numerical models produce similar results and that they are both greatly dependent on the turbulent Schmidt number. After the calibration procedure, both models provided good results for all investigated flow ratios, double-Tee junction distances, and pipe branching configurations, therefore both numerical models can be applied for a wide range of pipe networks configurations, but passive scalar model is the viable choice due to its much higher computational efficiency. Obtained results also describe the relationship between the double-Tee distances and complete mixing occurrence.