UT logoASE logoCAR logo

Author Archive

1-D DSMC simulation of Io’s atmospheric collapse in eclipse

1-D DSMC simulation of Io’s atmospheric collapse in eclipse

Monte Carlo modeling of Io’s [OI] 6300 Å and [SII] 6716 Å auroral emission in eclipse

Modeling of Io’s [OI] and [SII] Auroral Emissions of Io in Eclipse

Far Field Deposition of Scoured Regolith Resulting From Lunar Landings

Morris, A., et al., “Far Field Deposition of Scoured Regolith Resulting From Lunar Landings”, in Proceedings of 27th Intl. Symposium on Rarefied Gas Dynamics. Zaragoza, ES, (2012)

Modeling The Interaction Between a Rocket Plume, Scoured Regolith, and a Plume Deflection Fence

Morris, A., et al., “Modeling the Interaction Between a Rocket Plume, Scoured Regolith, and a Plume Deflection Fence”, in Proc. of 13th Earth and Space ASCE Conf. Pasadena, CA, (2012).

Sensitivity Analysis and Calibration of DSMC Input Parameters

In this work, statistical techniques were employed to study the modeling of a hypersonic shock with the Direct Simulation Monte Carlo (DSMC) method, and to gain insight into how the model interacts with a set of physical parameters. Direct Simulation Monte Carlo (DSMC) is a particle based method which is useful for simulating gas dynamics in rarefied and/or highly non-equilibrium flowfields. A DSMC code was written and optimized for use in this research. The code was developed with shock tube simulations in mind, and it includes a number of improvements which allow for the efficient simulation of 1D, hypersonic shocks. Most importantly, a moving sampling region is used to obtain an accurate steady shock profile from an unsteady, moving shock wave. The code is MPI parallel and an adaptive load balancing scheme ensures that the workload is distributed properly between processors over the course of a simulation. Global, Monte Carlo based sensitivity analyses were performed in order to determine which of the parameters examined in this work most strongly affect the simulation results for two scenarios: a 0D relaxation from an initial high temperature state and a hypersonic shock. The 0D relaxation scenario was included in order to examine whether, with appropriate initial conditions, it can be viewed in some regards as a substitute for the 1D shock in a statistica sensitivity analysis. In both analyses sensitivities were calculated based on both the square of the Pearson correlation coefficient and the mutual information. The quantity of interest (QoI) chosen for these analyses was the NO density profile. This vector QoI was broken into a set of scalar QoIs, each representing the density of NO at a specific point in time (for the relaxation) or a specific streamwise location (for the shock), and sensitivities were calculated for each scalar QoI based on both measures of sensitivity. The sensitivities were then integrated over the set of scalar QoIs to determine an overall sensitivity for each parameter. A weighting function was used in the integration in order to emphasize sensitivities in the region of greatest thermal and chemical non-equilibrium. The six parameters which most strongly affect the NO density profile were found to be the same for both scenarios, which provides justification for the claim that a 0D relaxation can in some situations be used as a substitute model for a hypersonic shock. These six parameters are the pre-exponential constants in the Arrhenius rate equations for the N₂ dissociation reaction N₂ + N [reaction in both directions] 3N, the O₂ dissociation reaction O₂ + O [reaction in both directions] 3O, the NO dissociation reactions NO + N [reaction in both directions] 2N + O and NO + O [reaction in both directions] N + 2O, and the exchange reactions N₂ + O [reaction in both directions] NO + N and NO + O [reaction in both directions] O₂ + N. After identification of the most sensitive parameters, a synthetic data calibration was performed to demonstrate that the statistical inverse problem could be solved for the 0D relaxation scenario. The calibration was performed using the QUESO code, developed at the PECOS center at UT Austin, which employs the Delayed Rejection Adaptive Metropolis (DRAM) algorithm. The six parameters identified by the sensitivity analysis were calibrated successfully with respect to a group of synthetic datasets.

Simulating Comet Impacts on the Moon

  • DSMC mass density contours for vapor plumes generated by two different oblique impacts. It is observed that there is less downrange focusing of vapor in the 60 degree impact, as well as lower velocities in the downrange and vertical directions. The red semi-circle indicates the hemispherical SOVA-DSMC interface (the SOVA hydrocode solves for the region within this boundary).
  • Mass density and radial velocity contours over the SOVA interface (a 20 km hemisphere centred at the point of impact) for two different oblique impacts. In the velocity contours, only material with velocity below lunar escape velocity (2380 m/s) is shown. It can be seen that for the 60 degree (more vertical) impact, a larger fraction of the material crossing the interface is below escape velocity- thus more likely to be retained in a transient atmosphere and ultimately, deposited in cold traps.
  • Mass density and radial velocity contours over the SOVA interface (a 20 km hemisphere centred at the point of impact) for two different oblique impacts. In the velocity contours, only material with velocity below lunar escape velocity (2380 m/s) is shown. It can be seen that for the 60 degree (more vertical) impact, a larger fraction of the material crossing the interface is below escape velocity- thus more likely to be retained in a transient atmosphere and ultimately, deposited in cold traps.
  • Mass density and radial velocity contours over the SOVA interface (a 20 km hemisphere centred at the point of impact) for two different oblique impacts. In the velocity contours, only material with velocity below lunar escape velocity (2380 m/s) is shown. It can be seen that for the 60 degree (more vertical) impact, a larger fraction of the material crossing the interface is below escape velocity- thus more likely to be retained in a transient atmosphere and ultimately, deposited in cold traps.
  • Mass density and radial velocity contours over the SOVA interface (a 20 km hemisphere centred at the point of impact) for two different oblique impacts. In the velocity contours, only material with velocity below lunar escape velocity (2380 m/s) is shown. It can be seen that for the 60 degree (more vertical) impact, a larger fraction of the material crossing the interface is below escape velocity- thus more likely to be retained in a transient atmosphere and ultimately, deposited in cold traps.
  • DSMC mass density contours for vapor plumes generated by two different oblique impacts. It is observed that there is less downrange focusing of vapor in the 60 degree impact, as well as lower velocities in the downrange and vertical directions. The red semi-circle indicates the hemispherical SOVA-DSMC interface (the SOVA hydrocode solves for the region within this boundary).

Over the last four billion years, a large amount of cometary material is estimated to have impacted the Moon. Water ice is thought to be the major constituent of comet nuclei, and analysis of hydrogen isotopes present in lunar minerals suggests the possibility of a cometary source for lunar water. We simulate comet impacts on the Moon, with a view to studying the nature of deposition of cometary water in the Moon’s permanently shadowed craters (cold traps), where temperatures are low enough to trap water over geological time scales.

On impact, a comet vaporizes. The dense regions closest to the point of impact are simulated by our collaborators at the Planetary Science Institute in Arizona, using the SOVA hydrocode. We then use a Direct Simulation Monte Carlo (DSMC) code designed to handle rarefied planetary scale flows to track the evolution of the water vapor plume, and the eventual deposition of water molecules in cold traps. We are currently carrying out a parametric study of the influence of various parameters (comet density, impact angle, velocity, location etc.) on the final deposition pattern. Our aim is to investigate whether cometary delivery can account for current observations of hydrogen on the Moon, and the influence of parameters such as impact angle, velocity and location on the extent and nature of final retention of water.

        Other publications and presentations

Simulations of Turbulent Spots and Wedges over Textured Surfaces

  • Figure 1 - Laminar to turbulent transition. Source - Viscous Fluid Flow by F. White.
  • Turbulent spot moving over straight riblets  (left) and turned riblets (right). Turbulent spots at three time snapshots are superimposed on one picture.  The turned riblets inspired asymmetric development of turbulent spots.  This is evidence that realistic and relatively small surface textures can modify turbulent spot behavior on a macroscopic level.
  • Turbulent wedge visualized by iso-surface of streamwise velocity colored by distance from the bottom wall. Low speed streaks (streaks that juts away from the wall) can be observed. Low speed streaks is a probable link in the cycle of turbulence self-regeneration. If low-speed streaks can be controlled, then maybe we can control spreading of turbulence.
  • Turbulent spot visualized by iso-surface of swirling strength colored by free stream velocity. Swirling strength Flow is from bottom left to top right. Notice the overall arrowhead shape as well as the intricate forest of hairpin (hooks and legs) coherent structure. Even though turbulence is chaotic, there is still order within it.

Turbulent spots are arrowhead shaped pockets of turbulent that form in the late stages of laminar to turbulent transition process (red circle in the schematic below). These spots increase in size as they travel downstream and form fully turbulent flow as they merge together. My research is looking at the formation and growth mechanisms of turbulent spot as well as interactions of millimeter scale surface textures with spots. If laminar to turbulent transition can be delayed using surface textures, then drag could be reduced.

The simulations are done using a channel flow spectral DNS code modified with immersed boundary to allow for boundary layer simulations. Rex ranges from about 524,000 to 675,000.

 

Videos:

Simulating the Atmosphere of Jupiter’s Moon Io

Io has one of the most dynamic atmospheres in the solar system due in part to an orbital resonance with Europa and Ganymede that causes intense tidal heating and volcanism. The volcanism serves to create a myriad of volcanic plumes across Io’s surface that sustain temporally varying local atmospheres. The plumes primarily eject sulfur dioxide (SO2) that condenses on Io’s surface during the relatively cold night. During the day, insolation warms the surface to temperatures where a global partially collisional atmosphere can be sustained by sublimation from SO2 surface frosts. Both the volcanic and sublimation atmospheres serve as the source for the Jovian plasma torus which flows past Io at ~57 km/s. The high energy ions and electrons in the Jovian plasma torus interact with Io’s atmosphere causing atmospheric heating, chemical reactions, as well as altering the circumplanetary winds. Energetic ions which impact the surface can sputter material and create a partially collisional atmosphere. Simulations suggest that energetic ions from the Jovian plasma cannot penetrate to the surface when the atmospheric column density is greater than 1015 cm−2. These three mechanisms for atmospheric support (volcanic, sublimation, and sputtering) all play a role in supporting Io’s atmosphere but their relative contributions remain unclear.

In the present work, the Direct Simulation Monte Carlo (DSMC) method is used to simulate the interaction of Io’s atmosphere with the Jovian plasma torus and the results are compared to observations. These comparisons help constrain the relative contributions of atmospheric support as well as highlight the most important physics in Io’s atmosphere.  These rarefied gas dynamics simulations improve upon earlier models by using a three-dimensional domain encompassing the entire planet computed in parallel. The effects of plasma heating, planetary rotation, inhomogeneous surface frost, molecular residence time of SO2 on the exposed non-frost surface, and surface temperature distribution are investigated.

 

Modeling Volcanic Plumes on Jupiter’s Moon Io

  • Ground-level number density contours for gas emerging from a half-annular vent.  Streamlines converge on the focal point along the symmetry plane, and a jet of gas shoots out to the right.
  • Side-view of a plume emerging from a half-annular vent seen along the symmetry plane.  A large region of high gas density forms above the focal point where the gas shocks and turns.
  • A giant plume (red) erupts from Io's equator over the course of an Io day.  As surface frost warms, a sublimation atmosphere (blue) is produced.  Plume material becomes suspended in the sublimation atmosphere and spreads over a huge area.  Eventually the surface cools and the atmosphere and plume material sticks to the surface.
  • A comparison of Pele's deposition ring with and without plasma bombardment.  Plasma causes the canopy to inflate and the ring to become thicker, more uniform, and more diffuse.  Adding plasma produces better agreement with the observed ring (Galileo spacecraft observation on right).
  • The simulated deposition ring of the Pele plume, with a Galileo image of Pele on Io's surface inset.  The simulation captures the ovoid shape and the angle of the major axis, as well as the sharper north end and north/south gas jets in the interior.  Fans of low density to the east and west compare well with the black fans seen in observations, and in simulations of Pele which include dust the particles are seen to fall in these areas.
  • A giant plume at 30 degrees north latitude in Io's sublimation atmosphere at noon.  The plume suppresses net sublimation where its canopy intersects the atmosphere, and plume material spreads out over the top of the atmosphere in a huge area around the plume source.
  • A simulated Pele plume.  Gas and dust rises from the vent region (the blue box in the middle).  The gas falls back on itself under gravity, and when falling gas meets rising gas an umbrella-shaped canopy shock is formed, and the canopy gas falls to the ground in a large ring.  Dust decouples from the gas before the canopy, is spread out by the canopy shock, and falls to the ground much closer to the plume source.
  • An SO2 plume at Io's north pole being bombarded by Jupiter's plasma torus.  Plasma inflates and heats plume canopies asymmetrically.
  • Galileo image of the lava lake at the center of the Pele plume, which I take as the source of the plume for my simulations.

Io is the most volcanically active body in the solar system, and its volcanic plumes rise hundreds of kilometers above the surface.  They rise far above the atmosphere, and I model this plume expansion into a near-vacuum with Direct Simulation Monte Carlo.  I simulate Pele, one of the largest plumes, in 3D using observations of the caldera to guide my choice of source geometry.  My goal is to explain the physics behind the deposition pattern and plume structure seen in observations. I also simulate plumes alongside other features of Io’s environment, like its sublimation atmosphere and Jupiter’s plasma torus, to understand how plumes fit into the big picture.

 

Modeling the Plumes on Saturn’s Moon Enceladus

  • The DSMC output is fed into the free-molecular model and the flow is propagated further into the far field. Eight point sources are placed on the planet surface according to Spitale and Porco. Free-molecular particles are launched from these sources with velocities obtained from the DSMC simulation. The far-field is less diffused and streakier. The individual sources are more distinguishable in this case.
  • This is the near-field gas translational temperature of the flow out of the vent. The translational temperature is defined based on the mean random kinetic (thermal) energy of the molecules. This temperature is equal to the thermodynamic temperature when the flow is in equilibrium. The translational temperature drops as the gas expands into vacuum. The flow is in equilibrium in the core of the plume where the density is high and collisions are frequent but is in non-equilibrium at the plume edges where density is low and collisions are few. This is expected since collisions are responsible for equilibrium.
  • The DSMC output is fed into the free-molecular model and the flow is propagated further into the far field. Eight point sources are placed on the planet surface according to Spitale and Porco. Free-molecular particles are launched from these sources with velocities obtained from the DSMC simulation. The gas plume is diffused enough that the individual sources cannot be distinguished.
  • This is a comparison between the simulated gas density data and the in-situ gas density data obtained by Cassini during one of the flybys.
  • 10-nm dust grain trajectories, plotted with the gas streamlines. They are affected by the gas flow the most, as can be seen from the spreading of the beam of grains towards the top. The grains are ice (density is 920 kg/m3) launched at the same speed as the gas at the vent. They are launched with only an upwards velocity. As the grains are moving up, they are affected by the gas flow and may be deflected.
  • This is the velocity distribution of the gas particles that escape the top of the inner DSMC domain as the gas flow becomes free-molecular. The tangential velocity is the velocity component tangential to the planet surface and the normal velocity is the velocity component normal to the planet surface.
  • Images taken by the Cassini spacecraft of the south polar plume of Enceladus. Note that the visible plume is the ice grains
  • Velocity distribution of gas molecules in a two-phase flow at an altitude of 10 km. Both gas and dust grains exit the vent at the same speed, with the mass flow rate of the grains 10x that of the gas.
  • This is the near-field gas number density contour of the flow out of the vent. The flow is axisymmetric, thus our simulation domain is a one-degree wedge instead of a 360-degree cylindrical domain. The gas is water vapor and it expands out of a 3-m vent at Mach-5. The gas number density drops as the flow expands into vacuum. Expansion waves emanate from the vent edges and turn the gas flow near the vent edges. Gas number density drops across these waves, creating a free-molecular (collisionless) region at the plume edges.
  • The DSMC output is fed into the free-molecular model and the flow is propagated further into the far field. Eight point sources are placed on the planet surface according to Spitale and Porco. Free-molecular particles are launched from these sources with velocities obtained from the DSMC simulation. Since the 10-nm grains are affected a lot by the gas flow, the far-field is pretty diffused too, similar to the gas plume.
  • Images taken by the Cassini spacecraft of the south polar plume of Enceladus. Note that the visible plume is the ice grains
  • 1-micron grains are barely, if at all, affected by the gas flow. The larger the grains are, the more mass they have. As a result, the larger heavier grains are not affected by the gas flow as much as the smaller lighter ones. The grains are ice (density is 920 kg/m3) launched at the same speed as the gas at the vent. They are launched with only an upwards velocity. As the grains are moving up, they are affected by the gas flow and may be deflected.
  • The second figure is the 100-nm grain trajectories. They are still affected by the gas flow, but not as much as the 10-nm grains. The grains are ice (density is 920 kg/m3) launched at the same speed as the gas at the vent. They are launched with only an upwards velocity. As the grains are moving up, they are affected by the gas flow and may be deflected.
  • Images taken by the Cassini spacecraft of the south polar plume of Enceladus. Note that the visible plume is the ice grains
  • Images taken by the Cassini spacecraft of the south polar plume of Enceladus. Note that the visible plume is the ice grains

The Cassini spacecraft first detected a plume near the warm south pole of the Saturnian moon Enceladus in 2005. The discovery of the plume not only helped to explain some phenomena that have been puzzling scientists for a long time but also brought about the exciting possibility of finding liquid water on Enceladus, making it a possibly favorable environment for life. Therefore, more flybys have been made over the moon and have yielded spectacular images, details of the plume structure and composition, as well as the possible locations of the plume sources. Observations found that the plume is composed of gas (mostly water vapor) with tiny entrained ice particles. Based on the images and data from Cassini, we construct a hybrid model of the plume. Our model divides the plume into two regimes: the collisional flow in the near-source region and the collisionless flow in the far-field region. The direct simulation Monte Carlo (DSMC) method is used to simulate the collisional gas flow in the near-source region as the gas has only begun to expand and is therefore, still relatively dense and warm. Once the flow becomes collisionless further out, the DSMC output is fed into a computationally less expensive free-molecular model to propagate the flow further into the far field. The simulation results are directly compared to the in-situ measurements made by Cassini. Our objective is to attempt to deduce the nature of the plume sources and hopefully, answer the question of whether there is liquid water on Enceladus.