The impact dynamics and internal mixing of a droplet onto a liquid-gas interface of lower surface tension was studied both experimentally and numerically, with both the Ohnesorge number (Oh) and the Bond number (Bo) being fixed. Compared to the droplet impact onto a pool of identical liquid, the interfacial Marangoni flow entrains abundant fluid upward and hence induces an additional jet breakup during crater formation (the first breakup), and it facilitates the emergence of the Rayleigh jet breakup (the second breakup) during crater restoration and enhances the vortical mixing beneath the liquid surface. Specifically, with the increase of the impact inertia, the first breakup manifests a nonmonotonic trend of "absence-presence-absence." The former transition of "absencepresence" at a low droplet-based Weber number (We d ) is caused by the shortened path of the Marangoni flow on the faster-growing liquid bridge, and the later transition of "presence-absence" at a high We d is resulted from the reduced displacement velocity of the pool fluid on the expanding crater surface. The second breakup corresponds to the Rayleigh jet breakup without surface tension difference and occurs monotonically beyond a certain We d . Due to the relatively short displacement time of the Marangoni flow on the crater surface compared to the time for crater formation, the critical condition for the emergence of the second jet breakup could be described by the critical reservoir-fluid-based We number (We r ). The critical We r contains two parts: the Bo-dependent critical We r0 without surface tension difference, and the increased viscous dissipation from the wrap-up motion of the Marangoni flow. Furthermore, capillary waves are also induced by the Marangoni flow during crater restoration, and the accompanied vorticity generation causes the mixing pattern to exhibit multiple vortex rings and even a clawlike structure, which is substantially enhanced compared to the vortical mixing without surface tension difference.

Hypergolic ignition of a bipropellant is an intrinsically nonpremixed physicochemical process that involves both chemistry and fluidic mixing. While the ignition delay time (IDT) of various hypergolic propellants has been extensively measured by using the prevalent droplet test, the effect of fluidic mixing on the hypergolic ignition process has rarely been studied. Compared with the well-understood droplet mixing within the same liquid, a prominent feature of bipropellant droplet mixing is the substantial surface tension difference, which induces a Marangoni effect upon droplet coalescence; however, it has not been addressed by previous hypergolic ignition studies and still remains inadequately understood. In this work, we numerically study the internal mixing of colliding droplets of different surface tensions with implications for hypergolic propellant ignition. The results show that the Marangoni effect substantially enhances droplet mixing compared with the situation without surface tension difference, and indicate the Marangoni effect could be a pivotal physical mechanism in hypergolic ignition. In particular, we identify an interesting phenomenon in that the Marangoni effect yields a nonmonotonic variation of internal mixing with increasing impact inertia, and this provides a likely interpretation of the nonmonotonic variation of the IDT experimentally observed by Zhang et al. [Combust. Flame 173, 276-287 (2016)]. Although binary droplet collision is employed as the specific object of the study, the present results could also provide insight into the hypergolic ignition of a bipropellant using various mixing methodologies, such as impinging jet and droplet-pool impact.

No abstract

The Marangoni-driven film spreading in the fully miscible droplet-reservoir system was experimentally studied. Three stages of film spreading were identified over a relatively long time, with the power-law exponent n ( R~ tn, R is film radius, t is time) nonmonotonically transitioning from 3/4, to 1/4, and back to 3/4. The variation of the Marangoni stress was found to be responsible for the three-stage power-law relationships, with the tangential stretching effect of the expanding liquid-air interface governing Stages I and III, and the normal diffusion effect dominating Stage II. This work presents a unified interpretation on the inconsistent power-law relationships reported from previous studies on Marangoni-driven film spreading for miscible fluids.