Carderock Division Title with American Flag Backdrop - Select to go to home page
 ABOUT US  OUR CAPABILITIES  DEPARTMENTS  PRODUCTS AND SERVICES  WORKING WITH US  NEWS  EMPLOYMENT

Multi-Phase Computational Fluid Dynamics

Table of Contents

Introduction

As the US Navy enters the 21st century, its ships are coming under increasingly stringent local and global environmental regulations which place limitations on overboard discharges and may, in fact, limit their operability. This has spurred the US Navy to evaluate their ships' present environmental performance and then, if necessary, to develop environmentally-friendly shipboard technologies. While some of the necessary technologies are commercially available, in many cases the US Navy must develop their own, custom, technology. In some cases, where the technology involves large systems with complex fluid dynamics, multi-phase computational fluid dynamics (MP-CFD) can be brought to bear to provide the physical insights that will enable the development of improved technologies as well as provide a test-bed for design optimization.

The Multi-Phase Computational Fluid Dynamics (MP-CFD) group has been using MP-CFD analyses to help with the analysis and design of compensated fuel/ballast tanks, and biological reactors. We have also been active in the optimization of oily- and solid-waste incinerators. CFD and chaos theory are being utilized to reduce pollutants from diesel engine combustion. Another project involves the optimization of stack gas flows. These CFD problems involve two-fluid dynamics, multi-phase flows, heat transfer and chemical reactions. While well-versed in the use of computational tools, the group has a high regard for the practical aspects of hardware design ensuring that advanced technology is possible to implement. Environmental compliance requires the measurement and prediction of minute quantities of pollutants; we are therefore actively engaged in determining the uncertainty of our simulations so as to assess the confidence of our predictions.

Two-Fluid Dynamics

Ships' Compensated Ballast

The MP-CFD group has been assisting the Environmental Quality Branch (Code 633) in the analysis and design of ships' compensated fuel/ballast tanks. Within these tanks complex interactions between diesel fuel and water occur during refueling. The MP-CFD group is using two-fluid CFD tools to predict regions of high fuel entrainment potential and the amount of water remaining in the tanks at the end of refueling. Fuel entrainment and particle tracking algorithms are being developed to predict fuel/water mixing and the subsequent advection of fuel droplets in the water phase. Experiments are providing the physics for model development and the data from code validation. The following two figures show how CFD can be used to assess the flows in the existing tanks, give guidance for redesign and then quantify the relative effectiveness of the modifications.



Baseline tank In this figure the fuel (red) is advecting (left to right) through a manhole into a water-filled (blue) compartment. Buoyancy forces cause the flow to impact the ceiling of the downstream compartment, potentially causing fuel/water mixing.

Modified tank This figure shows the flow through the same tank which has undergone minor structural modifications. The momentum of the buoyant flow has been reduced to a fraction of its unmodified-tank value - the potential for fuel/water mixing has been significantly reduced. In addition, such modifications significantly improve the efficiency of refueling.

Double-hull Tanker Groundings

We assisted the Structures Directorate in investigating the tradeoffs between single- and double-hull tankers. We performed two-fluid simulations of accidental grounding and hull-puncture scenarios for two double-hull tanker configurations. (Paper and animations)

Stratified Flows

The MP-CFD group has investigated fundamental physical processes involved in vortices rising through a stratified fluid towards a free surface. The fluid was stratified in temperature and salinity. Our results showed persistence of stratification at the free surface long after the physical deformation of the free surface, caused by vortex motion, had died out. A paper summarizes our work in this area.

Selected Two-Fluid Dynamics References

Incineration

A system for thermal treatment of shipboard solid waste is under development under a U.S. Navy Advanced Technology Demonstration (ATD) project. Tests have been conducted at the Naval Research Laboratory to test the most innovative element, a plasma-fired eductor (PFE) that serves as the primary reactor for organic waste (primarily paper and cardboard) destruction. The PFE is designed to assure extremely rapid gasification of the organic feed by exposure to extremely hot air from a plasma torch. Destruction efficiency has been determined for several PFE configurations to identify the important design considerations and constraints.   A secondary combustion chamber (SCC) will be designed to oxidize the syngas generated from the PFE. The ATD PAWDS will be built and operated to demonstrate full-scale waste destruction in a land-based reactor. Computational Flow Dynamics (CFD) modeling has been employed to develop a bridge from the laboratory system to the ATD design.

Eductor

Eductor static temp The PFE is a two-dimensional axi-symmetric steady state model. This figure shows the static temperature profile of the PFE. Cellulose enters at a rate of 0.03 kg/sec. Air enters at 50% stoichiometry. The hot plasma torch gas enters along the centerline at 5000 degrees K, while particulate and carrier air enter from an annulus about the torch gas. Gasification begins as the particles rapidly absorb energy from the high enthalpy torch gas. A narrow region along the wall remains relatively cool.

Vortex Incinerator

Incinerator animation Animation showing isotherms.

Incinerator mass fraction This 2 dimensional slice shows the results of a steady state 3 dimensional solution using 30 GPH DFM as fuel with 20% excess air slide shows the static temperature of a cross section of the vortex incinerator, where the fuel burner enters the main incineration chamber (tangentially, off center). An optimal fuel burner will combust all fuel, leaving only hot gases (products of combustion) to enter the main incineration chamber.

Incinerator static temperature This 2-dimensional slice shows the results of a steady state 3 dimensional solution using 30 GPH DFM as fuel with 20% excess air slide shows the static temperature of a cross section of the vortex incinerator, where the fuel burner enters the main incineration chamber (tangentially, off center). The products of combustion enter the main incineration chamber at very near the adiabatic flame temperature. These slices give us some insight into the nature and strength of our right hand vortex. Since the DFM burner assembly has done itıs job, we see the highest temperature in the main incineration chamber in an area surrounding the waste stream on one side.

Incinerator velocity contour This 2-dimensional slice shows the results of a steady state 3 dimensional solution using 30 GPH DFM as fuel with 20% excess air slide shows the velocity contour of a cross section of the vortex incinerator, where the fuel burner enters the main incineration chamber (tangentially, off center). These slices give us some insight into the nature and strength of our right hand vortex. The high velocity gases produced as a result of the DFM combustion process leaving the burner assy into the main incineration chamber.

Incinerator particle tracks This shows the results of a steady state 3-dimensional solution using 30 GPH DFM as fuel with 20% excess air slide shows the velocity contour of a cross section of the vortex incinerator, where the fuel burner enters the main incineration chamber (tangentially, off center). This figure shows the waste stream mass in kg. The waste stream enters at 30 GPH, and is composed of 98 percent water, and 2 percent organic matter and incombustible ash. This shows 1500 individual particle tracks, the majority of which are consumed in the first 2/3rds of the main chamber.

Diesel Engine Combustion

Diesel Engine Emissions Reduction / Diesel Engine Efficiency

Effects of Thermal Radiation

The MP-CFD group has been conducting applied research in the area of reducing diesel exhaust emissions and improving the efficiency of diesel engines. The primary exhaust emissions of concern are soot and NOx. We have investigated the role of thermal radiation in the formation of NO using the KIVA3-V diesel combustion code. The following three figures show a 60 degree symmetric section of a diesel engine cylinder volume between the piston face and the cylinder head. The crank shaft angle is about 1 degree below top-dead-center. In each figure, the piston face is the bottom surface, and the cylinder head is the top surface.

Temperature plot This figure shows the temperature during combustion in the diesel cylinder. Note the near adiabatic flame temperature in the bowl just above the piston face. This is where the fuel becomes almost completely atomized after being injected from the cylinder head down towards the piston face.

Radiation plot This figure shows the radiative heat transfer rate resulting from gas phase radiation of the combustion products CO2 and H2O, and infrared soot radiation. Note the area of peak radiative heat transfer is consistent with the highest combustion temperatures.

NOx plot This figure shows the mass fraction of NOx being formed during the diesel combustion process. The highest NOx levels are seen where the temperatures and radiative heat transfer rates are highest.

Chaos in Diesel Combustion

The MP-CFD group is also exploring the influence of chaotic structures in the diesel combustion process, and how they might be used to control combustion to minimize exhaust emissions while maximizing real cycle efficiency. We are conducting this work in collaboration with the Engine Research Center at the University of Wisconsin. We are using a parallel version of KIVA-3, developed by Osman Yasar at SUNY Brockport, to calculate thousands of diesel combustion cycles in the search for the period multiplying phenomena that lead to chaos. Chaos might then be used to control such parameters as fuel injection frequency and dwell angle, where research has shown that multiple fuel injection pulses can reduce both NOx and soot formation. The following four figures show two different forms of fuel injection and how they affect the heat release during the combustion cycle.

Fuel Injection This figure shows fuel being injected in a single pulse starting at a crank shaft angle of —10.5 deg (0 deg is top-dead-center).

Fuel Injection This figure shows the heat release as a function of crank shaft angle for the single pulse fuel injection. NOx and soot production are strong functions of the peaks in heat release. Suitable modifications to the heat release map caused by multiple pulse fuel injection can reduce the NOx and soot produced.

Fuel Injection This figure shows fuel injected over the same dwell angle as the single pulse case, but in five, equally spaced pulses. The amount of fuel injected is the same as in the single pulse case.

Heat Release This figure shows the heat release for the fuel injected in multiple pulses. More fuel spray fronts provide better diffusion flame burn characteristics in the cylinder, which favor soot oxidation, and reduce NOx production.

Post-Detonation Combustion in Closed Compartments

The MP-CFD group has done numerical simulations of post-detonation combustion in closed compartments to determine the effect on quasi-steady pressure in the compartments of late burning of non-ideal explosives. This work has been summarized in papers 1 through 5 in the bibliography.

Selected Bibliography

Biological Reactors

Biological waste reactor tanks use biological agents to decompose shipboard liquid and solid waste. The biological agents require aeration to function, so a well-mixed air distribution in the tank is needed. Unlike land-based systems, these shipboard-based reactors are volume- and weight-limited. Thus optimization of the many design parameters (flow rates, geometry, aeration...) is critical for meeting strict overboard discharge requirements. The MP-CFD group has performed numerical simulations of aeration in these tanks. The following figures show aeration patterns, as they evolve in time, in a cylindrical test tank filled with water. The air is injected through two rectangular holes at the bottom of the tank.

Reactor at 0.5 sec This figure shows the volume fraction of the air after 0.5 second has elapsed. The air bubbles have formed a plume which is rising through the water in the tank.

Reactor at 100 sec This figure shows the volume fraction of the air after 100 seconds have elapsed. The air bubble plume has now expanded to fill a greater volume in the tank.

Velocity vectors This figure shows the velocity in the water induced by the rising air bubbles. Note the patterns of strong recirculation. Our Australian colleagues have recently shown similar behavior in a glass of Guinness Stout.

Reactor at 550 sec This slide shows the distribution of air bubbles in the tank after about 550 seconds have elapsed. The air bubbles have now occupied a significant portion of the tank.

Stack-Gas Flows

The investigation of stack gas flows is of practical interest for a variety of commercial applications. Analyses are currently being performed on a gas ejector to determine the governing fluid dynamic and geometric parameters, as well as to determine some optimization of these parameters which will result in the greatest efficiency.

Velocity magnitudes This figure shows the predicted vertical velocity contours for the case of a passive ejector.

Velocity vectors This figure details the entrainment of the surrounding air into the ejector. (It also demonstrates an inefficiency found in straight-walled passive ejectors, in that a reversed flow develops at the outlet.)

Computational Assets

The MP-CFD group has the following dedicated computational assets in place for fast and efficient solutions to three dimensional problems involving complex physics:

  1. 25 node Beowulf cluster — this cluster runs under Linux 6.1, and has HPF, F90, F77, C++, and C compilers, along with MPI and PVM libraries for parallel processing. Each node in the cluster has a Pentium III processor running at 600 Mhz and 256 Mbytes of memory.
  2. 2-processor Compaq Alpha server
  3. R12000 SGI O2 workstation with 1 Gb of memory.
  4. 4-processor R10000 SGI Origin 200 server with 2 Gb of memory.
  5. 2-processor R10000 SGI Octane workstation with 500 Mb of memory.
  6. 2 Pentium III processor Dell NT Precision 410 workstation with 500 Mb of memory.

Links


Carderock Division, Naval Surface Warfare Center, 9500 MacArthur Blvd., West Bethesda, MD 20817-5700


This is an official U.S. Navy web site.

The Carderock Division is a field activity of the Naval Sea Systems Command

and a vital division within the Naval Surface Warfare Center.

See our Privacy Policy, Disclaimers and Accessibility Statement.

Contact us through our Feedback Form.


Navy Information: Navy Jobs | Freedom of Information Act (FOIA)