High-pressure gas atomization (HPGA) is a widely used method for producing metal powders using high-velocity gas jets, offering high efficiency for large-scale production. Achieving small and spherical powders is critical for this process, which requires a comprehensive understanding of the primary breakup of liquid metals. However, the highly turbulent nature of gas jets complicates the breakup process, making it difficult to control. Here, we explore the influence of vortical structures on the primary breakup during atomization using large-eddy simulations for an annular-slit, close-coupled gas atomizer with molten aluminum and nitrogen gas. We extract individual droplets from the instantaneous flow field and classify them as fibers, ligaments, or spheroids based on their sphericity and aspect ratio. In the near field (z/D < 4), smaller and more spherical droplets are produced compared to the far field (z/D > 4). To analyze the effects of turbulence on the droplet breakup process, we track individual droplets to investigate how strong adjacent vortical structures influence droplet breakup, focusing on the near field. Approximately 70% of the droplets that evolve into spheroids detach far from the nozzle inlet (r/D > 1.5) and experience frequent breakups, averaging more than four times during their lifetime. The droplets undergoing breakup interact with strong vortical structures over 10 times more frequently than those that remain intact. Conditionally averaged flow fields further show that the droplets continuously interact with strong vortical structures before the breakup, generating opposing rotational forces. After the breakup, the maximum magnitude of the surface normal vorticity, which represents the rotational force acting on the droplet interface, decreases by nearly 35%. A comparison of the Weber number (We) for droplets interacting with strong and weak vortical structures indicates that droplets overlapping with strong vortical structures maintain higher We values (35 < We < 80). This range corresponds to the multimode breakup, ultimately leading to droplet breakup. Our findings provide valuable insights into improving nozzle designs from the perspective of recirculation zones and vortical structures, contributing to the production of high-quality spherical powders in HPGA.

Backflow (BF) events, distinguished by negative wall-shear stress (τx), are rare phenomena occurring in the near-wall region of fully developed wall turbulence. Although these events manifest as small-scale patches of viscous scales, they originate from collisions between large-scale structures (LSSs). Hence, we explore the formation of BF, focusing particularly on interactions with the surrounding LSSs to elucidate the associated inner–outer interactions. We perform direct numerical simulations of turbulent channel flows at Reτ = 180 and 550, including a narrow box simulation at Reτ = 550 to restrict the LSSs. We observe the presence of wide BFs, which are absent at the lower Reynolds number and in the narrow box simulation. These wide BFs have widths significantly larger than the mean size of typical BF regions. Temporal tracking of the BFs with surrounding LSSs and vortical structures reveals that wide BFs result from symmetric collisions between streamwise-aligned high- and low-speed LSSs, whereas narrow BFs stem from asymmetric collisions. In the symmetric collisions, the upstream high-speed structure overrides the downstream low-speed structure, forming a wide shear layer and a significant velocity jump at the interface. This induces a strong prograde vortex near the wall, which elongates laterally and descends owing to the downwash motion of the high-speed structure, ultimately inducing wide BF regions. Conversely, the narrow BF regions develop from the asymmetric collisions occurring at the sides of the spanwise-aligned LSSs, forming narrow, laterally tilted shear layers. The large-scale collisions also induce extreme positive-τx events, particularly noticeable over broad streamwise extents during symmetric collisions. These insights into BF dynamics can inform the development of novel drag reduction strategies by manipulating LSS collisions.

We examined the capability of an unsupervised deep learning network to capture the spatial organizations of large-scale structures in a cross-stream plane of a fully developed turbulent channel flow at Reτ≈180. For this purpose, a generative adversarial network (GAN) is trained using the instantaneous flow fields in the cross-stream plane obtained by a direct numerical simulation (DNS) to generate similar flow fields. Then, these flow fields are examined by focusing on the turbulent statistics and the spatial organizations of coherent structures. We extracted the intense regions of the streamwise velocity fluctuations (u) and the vortical structures in the cross-stream plane. Comparing the DNS and GAN flow fields, it is revealed that the network not only presents the one-point and two-point statistics quite accurately but also successfully predicts the structural characteristics hidden in the training dataset. We further explored the meandering motions of large-scale u structures by measuring their waviness in the cross-stream plane. It is shown that as the size of the u structures increases, they exhibit more aggressive waviness behavior which in turn increases the average number of vortical structures surrounding the low-momentum structures. The success of GAN in this study suggests its potential to predict similar information at a high Reynolds number and, thus, be utilized as an inflow turbulence generator to provide instantaneous boundary conditions for more complicated problems, such as turbulent boundary layers. This has the potential to greatly reduce the computational costs of DNS related to a required large computational domain at high Reynolds numbers.

Townsend의 attached-eddy hypothesis(AEH)는 로그 영역 통계량을 높이 y에 따라 성장하는 자기 유사성 구조들을 통해서 설명한다. 로그 영역 Reynolds 응력 생산과 모멘텀 이송에 관여하는 구조를 active, 아닌 구조를 inactive로 제안했다. 원론의 분리 기준이 모호하여 개별 자기 유사 구조의 에너지 기여를 분석할 필요가 있다. 본 연구는 파이프 난류 유동 직접수치모사 데이터에 spectral linear stochastic estimation 기반의 개별 구조 에너지 분리를 수행하였다. 개별 구조의 premultiplied 2차원 스펙트럼의 유동방향 및 횡방향 파장 간의 선형 비례 관계, 1차원 스펙트럼의 구조 높이 스케일링을 통해 자기 유사성을 확인하였다. 관측 지점 약 2.5배의 높이의 구조가 에너지에 가장 강하게 기여해 active 구조의 특징을 보였지만 더 높은 관측 지점에서는 해당 구조의 에너지 기여가 약해지기 때문에 inactive 특성을 보여 AEH의 예측과 같이 같은 구조가 위치에 따라 다르게 분류될 수 있음을 보였다.

This study presents a numerical analysis of the effects of a rigid flat wall with oscillating motion on the pressure wave propagation during a single spherical cavitation bubble collapse at different initial bubble positions. Different nondimensional distances S = 0.8, 0.9, 1.0, 1.1, 1.2 and 1.3 were considered to investigate the effects of initial in-phase and out-of-phase oscillations of the flat wall. Numerical simulations of cavitation bubble collapse near an oscillating wall were conducted using a compressible two-phase flow model. This model incorporated the Volume of Fluid (VOF) interface-sharpening technique on a general curvilinear moving grid. The numerical results were consistent with published experimental data. The simulation examined the impact of oscillating walls on bubble behavior and the resulting pressure peaks observed on the wall surface. The numerical results demonstrate the significant impact of wall oscillation conditions on bubble collapse and migration behavior, and consequently, the generation of pressure waves with significantly different propagation and pressure peaks induced by shock impact on the rigid wall. Different behaviors were observed in the trendlines of the pressure peaks and maximum jet velocity under in-phase and out-of-phase oscillating walls, with distinct values. At S ≥ 1.0, a higher-pressure peak on the wall was observed in the case of the out-of-phase oscillating condition, whereas a higher-pressure peak was found in the case of the in-phase condition at S < 1.0. The highest-pressure peak was found at S = 0.8 in trend lines of in-phase and S = 1.1 in trend lines of out-of-phase oscillation effects.

High-pressure gas atomization (HPGA) is a widely used method for producing metal powders using high-velocity gas jets, offering high efficiency for large-scale production. Achieving small and spherical powders is critical for this process, which requires a comprehensive understanding of the primary breakup of liquid metals. However, the highly turbulent nature of gas jets complicates the breakup process, making it difficult to control. Here, we explore the influence of vortical structures on the primary breakup during atomization using large-eddy simulations for an annular-slit, close-coupled gas atomizer with molten aluminum and nitrogen gas. We extract individual droplets from the instantaneous flow field and classify them as fibers, ligaments, or spheroids based on their sphericity and aspect ratio. In the near field (z/D &amp;lt; 4), smaller and more spherical droplets are produced compared to the far field (z/D &amp;gt; 4). To analyze the effects of turbulence on the droplet breakup process, we track individual droplets to investigate how strong adjacent vortical structures influence droplet breakup, focusing on the near field. Approximately 70% of the droplets that evolve into spheroids detach far from the nozzle inlet (r/D &amp;gt; 1.5) and experience frequent breakups, averaging more than four times during their lifetime. The droplets undergoing breakup interact with strong vortical structures over 10 times more frequently than those that remain intact. Conditionally averaged flow fields further show that the droplets continuously interact with strong vortical structures before the breakup, generating opposing rotational forces. After the breakup, the maximum magnitude of the surface normal vorticity, which represents the rotational force acting on the droplet interface, decreases by nearly 35%. A comparison of the Weber number (We) for droplets interacting with strong and weak vortical structures indicates that droplets overlapping with strong vortical structures maintain higher We values (35 &amp;lt; We &amp;lt; 80). This range corresponds to the multimode breakup, ultimately leading to droplet breakup. Our findings provide valuable insights into improving nozzle designs from the perspective of recirculation zones and vortical structures, contributing to the production of high-quality spherical powders in HPGA.

Backflow (BF) events, distinguished by negative wall-shear stress (τx), are rare phenomena occurring in the near-wall region of fully developed wall turbulence. Although these events manifest as small-scale patches of viscous scales, they originate from collisions between large-scale structures (LSSs). Hence, we explore the formation of BF, focusing particularly on interactions with the surrounding LSSs to elucidate the associated inner–outer interactions. We perform direct numerical simulations of turbulent channel flows at Reτ = 180 and 550, including a narrow box simulation at Reτ = 550 to restrict the LSSs. We observe the presence of wide BFs, which are absent at the lower Reynolds number and in the narrow box simulation. These wide BFs have widths significantly larger than the mean size of typical BF regions. Temporal tracking of the BFs with surrounding LSSs and vortical structures reveals that wide BFs result from symmetric collisions between streamwise-aligned high- and low-speed LSSs, whereas narrow BFs stem from asymmetric collisions. In the symmetric collisions, the upstream high-speed structure overrides the downstream low-speed structure, forming a wide shear layer and a significant velocity jump at the interface. This induces a strong prograde vortex near the wall, which elongates laterally and descends owing to the downwash motion of the high-speed structure, ultimately inducing wide BF regions. Conversely, the narrow BF regions develop from the asymmetric collisions occurring at the sides of the spanwise-aligned LSSs, forming narrow, laterally tilted shear layers. The large-scale collisions also induce extreme positive-τx events, particularly noticeable over broad streamwise extents during symmetric collisions. These insights into BF dynamics can inform the development of novel drag reduction strategies by manipulating LSS collisions.