|
Rocket
Combustor Experiments and Analyses
A number of projects are underway at Purdue University to provide basic data and physical insight on rocket combustion processes that determine performance and life. Experiments focus on sub- and real-scale combustors at realistic pressure conditions. Processes of interest include steady state and dynamic combustion of main chamber injectors for the oxidizer-rich staged combustion cycle, liquid film cooling, autoignition of hydrocarbon fuels, ignition of non-toxic hypergolic propellants, and the mixing and subsequent thermal choking of a fuel-rich rocket exhaust/ducted air flow. Results from these and other studies are presented.
Thermal
and Fluid Analyses for Gas Turbine Cooled Vane and Blade
In a high temperature gas turbine, turbine vane and blade cooling designs require key technologies. The metal temperature of turbine cooled vanes and blades must be predicted in the design stage as accurately as possible to reduce the period of engine development. Because life prediction of turbine cooled vanes and blades is strongly dependent on the metal temperature, the problem of metal temperature prediction is essentially equal to obtaining the convective heat transfer boundary conditions on external and internal surfaces of the cooled vane and blade. Kawasaki Heavy Industries has focused attention on the accuracy of commercial computational fluid dynamic (CFD) software based on advanced computing in order to improve the accuracy of three-dimensional metal temperature predictions.
This paper presents an accurate analysis
of heat and fluid distributions for turbine cooled vanes
and blades using the commercial CFD software FLUENT.
Investigations consisted of five tasks. Task 1 investigated
static pressure (Mach number) and heat transfer coefficient
for the vane external surface (NASA C3X). Task 2 investigated
film cooling effectiveness on the surface of a flat
plate, a semi-circular cylinder incorporated with flat
plate, and a two-dimensional vane. Task 3 investigated
mass flow rate for internal flow in the turbine cooled
blade for an internal cooling structure consisting of
typical serpentine passages with turbulence promoters
(ribs) and pin fins arranged in the trailing edge region.
Furthermore, heat transfer coefficient for a rectangular
channel with consecutive turbulence promoters was also
investigated. Our calculations compared adequately well
with published experimental data as well as our own
fundamental test data. Additionally, the suitable selection
of turbulence model and near wall treatment, mesh sizes
(y+), and run time were also investigated. Task 4 involved
on a coupled analysis of the external gas-pass flows
of the first turbine vane and blade to investigate the
accuracy of static pressure (Mach number), mass flow
rate, and other features. Finally, Task 5 involved analyses
of heat transfer of the turbine cooled blade by thermal
conjugation of the internal and external fields of a
first-stage turbine blade consisting of convection heat
transfer and thermal conduction. Comparison of CFD results
with actual engine test data clearly show that the analytical
method, based on the commercial CFD software FLUENT,
is useful for prediction of blade temperature and may
be applied routinely in the design stage of turbine
cooled vanes and blades.
Thermo-Mechanical
Modeling and Analysis for Turbopump Assemblies
Transient and quasi-steady thermal effects have a strong influence on the design of stationary and rotating components in rocket turbopumps. In traditional design systems, there are many obstacles that the engineer must overcome in the process of performing an accurate analysis. One of the primary hurdles is the needed synthesis of information (geometry, internal and external flow conditions, etc.) that is required to completely analyze the 2D and 3D effects of the various thermal conditions at the component and assembly level. More efficiently and effectively analyzing turbopump system thermal and transient response is one of the most challenging aspects of turbopump design and integration into the propulsion system.
Integrating cost- and time-effective
analysis of turbopump steady state and transient thermo-mechanical
effects into the preliminary and final design process
is the overall goal of an ongoing Phase II SBIR project
administered by NASA MSFC. Automating the thermo-mechanical
analysis process is the proposed innovation. Integrating
the new methodology within an existing multi-disciplinary
turbopump design system (owned by NASA and select turbopump
manufacturers) is the primary commercialization path.
The SBIR subtopic calls for advancements in the area
of integrated multi-disciplinary design and analysis
systems for important vehicle subsystems such as turbopumps.
The integrated thermo-mechanical analysis capability
is required by turbopump designers to meet requirements
such as better performance and reliability, lower product
cost, reduced size and weight, and improvements in product
development cycle time. This project also has tremendous
dual-use potential for designers of industrial turbomachinery
products.
Cryogenic turbopump design is influenced
by complex interaction of design goals and constraints
that vary from application to application. A successful
design must effectively balance the customer targets
for performance, stress, vibration, weight, cost and
life. By combining an integrated pump design system
with robust optimization technology, this project lays
the foundation for significant gains in advanced turbopump
designs. The deterministic analytical tools include
hydrodynamic evaluation using streamline curvature and
CFD techniques, finite element structural evaluation
with shell and solid models, and a 1-D internal flow
network.
In the discussed turbopump optimization
case, the optimization plan utilizes a hierarchical
set of optimization loops to characterize the effects
of the independent design parameters on dependent hydrodynamic
and structural results. The inner optimization loops,
which are run the most often, are formed with the fastest-running
engineering models. Higher fidelity, more time-consuming
computations are saved for the less frequently traversed
outer loops. This analytical pyramid is geared towards
time-effective optimization in the multidisciplinary
design space.
By including thermo-mechanical modeling
of the entire turbopump assembly, this project will
greatly extend the capabilities for preliminary turbopump
design and optimization. Designers have an ongoing need
for evaluation of transient and steady-state fits and
clearances for critical parts, thermal stress, low cycle
fatigue, and overall life and reliability. The value
of the early consideration of these critical design
issues cannot be overstated.
CFD-based
Design of Turbopump Inlet Duct for Reduced Dynamic Loads
Fluid flow, moving from a straight duct into a bend, experiences a centrifugal force that causes an adjustment in the pressure distribution. The adjustment process gives rise to a secondary circulation at the bend exit that displaces the velocity maximum toward the outer wall. Consequently the absolute flow angle is no longer axially symmetric. If rotating turbomachinery is located near the bend exit, the blades will experience unsteady loads as a consequence of the variable relative flow angle. Therefore, it is of interest to know how to minimize the contribution of the (bending) duct to unsteady blade loads. In the present study the diameter, bend radius, and cross-sectional shape have been varied systematically using an automated tool that generates structured grids. A series of time-independent, three-dimensional computational fluid dynamic (CFD) simulations has been performed using Loci-Chem at turbulent Reynolds numbers Re through a 90 degree bend. Chem is a density-based, finite-volume, parallel Navier-Stokes solver using the Loci framework. Low resolution grids, easily capable of resolving the secondary circulation, have been used to investigate a large parametric space. The number of grid points varied from ~82,000 to 237,000. Each case typically required no more than 4 hours to converge using 4-10 CPUs, achieving 6 or more orders of magnitude reduction in residuals. Predictions for wall static pressure, mean velocity profile, and secondary circulation, all compare favorably with laboratory experiments for the case of water flow (Re ~43000) through a duct with 90 deg bend and constant circular cross section. The relative flow angle experienced by a turbopump, located immediately downstream of the elbow, may vary by 2 deg or more for a hypothetical wheel speed of ~20000 rpm and duct diameter of ~8 inches. Such variations may have serious consequences for turbopump lifetime and performance. The presentation will summarize results from all of the simulations.
Experimental
and Modeling Studies of Liquid Hydrocarbon Rocket Injectors:
Status and Future Research Directions
Recent renewed interest in rocket engines using liquid hydrocarbon/liquid oxygen propellants has resulted in an increased focus on understanding and modeling of suitable injectors. Previous work in this area has been undertaken by the research group at Penn State for RP-1/LOX and ethanol/LOX propellant systems. This work forms a basis from which to assess our current understanding and plan future work related to liquid hydrocarbon injectors. The paper will focus on previous work conducted at Penn State as well as some recently planned studies to highlight the current understanding of liquid hydrocarbon rocket injectors. Comparisons with modeling done for these studies, where it exists, will be reviewed. Futures needs in terms of coordination experimental and modeling studies will be described.
Summary
of Recent Inducer Testing at MSFC and Future Plans
Marshall Space Flight Center (MSFC) has continued to improve its water test capabilities while remaining committed to providing test support to both internal and external customers. Inducer test capability in particular has grown beyond basic steady-state performance testing to include high-fidelity, high-density unsteady instrumentation and direct measurement of hydrodynamic-induced forces and moments. These advances have been driven by customer demands for more detailed information to supplement and validate the ever-improving computational analyses and by the evolving understanding of the interrelationship of local and system dynamics, cavitation, and fluid unsteadiness. In the last two years MSFC has successfully measured the hydrodynamic-induced forces and moments on two axial pump designs in the presence of cavitation and subjected miniature pressure transducers to a cavitating environment to evaluate durability. Project-sponsored activities will continue through early 2004 with three water tests planned on two different designs. Measurements will include inducer steady and cavitation performance and blade loads through miniature pressure transducers installed on the blade surfaces. Inducer backflow characteristics and the corresponding dynamic environment will also be measured in order to assess potential interactions with upstream components. Long-term plans include reviving the water test loop developed at the California Institute of Technology for measuring pump dynamic transfer functions.
|
