SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
This report describes the work being undertaken to study the scalar mixing in curved shear layers. First, the motivation for this work and its objectives are described. Second, a description of the experimental rig that has been built is given. Third, some preliminary results (flow visualizations) are discussed, and finally, future steps that will be taken to complete the study are outlined.
The first goal of this research is to understand the dynamics of three dimensional vortical structures, especially the interaction of the ribs and two dimensional rollers. Because of vortex stretching effects, this requires accurate and well resolved simulations. Another goal is to distinguish temporally- and spatially-growing shear layers in terms of the dynamics and observed vortical structures. This is important since the former is much easier to compute, but the error incurred in using temporal results for certain applications is unknown. The third goal is to investigate the effects of asymmetry on passive scalar mixing and fast-chemistry product formation. In this report a brief summary of a new numerical method is presented, followed by results for mixing layers and wake flows.
Direct numerical simulations of the time evolution of homogeneous stably stratified turbulent shear flows have been performed for several Richardson numbers Ri and Reynolds numbers R(sub lambda) in earlier works. The results show excellent agreement with length scale models developed from laboratory experiments to characterize oceanic turbulence. When the Richardson number Ri is less than the stationary value Ri(sub s), the turbulence intensity grows at all scales, and the growth rate appears to be a function of Ri. The size of the vertical density inversions also increases. On the other hand, when Ri is greater than or equal to Ri(sub s) the largest turbulent eddies become vertically constrained by buoyancy when the Ellison (turbulence) scale L(sub E) and the Ozmidov (buoyancy) scale L(sub O) are equal. At this point, the mixing efficiency is maximal and corresponds to a flux Richardson number R(sub f) = 0.20. The vertical mass flux becomes counter-gradient when epsilon = 19(nu)N(exp 2) and vertical density overturns are suppressed in less than half a Brunt-Vaisala period. The results of the simulations were also recast in terms of the Hydrodynamic Phase Diagram introduced in fossil turbulence models. The so-called point of fossilization occurs when epsilon = 4DCN(exp 2); Gibson proposed 13DCN(exp 2). This value is in agreement with indirect laboratory observations and field observations. Finally, the validity of the steady-state models to estimate vertical eddy diffusivities in the oceanic thermocline is discussed.
There is significant industry demand for a method of analyzing the shear buckling of composite plates with cutouts. This method should be able to easily accommodate frequent changes in model design; inflexibility is the major drawback to current finite element methods. The approach taken here is broken into two problems, prebuckling and buckling. To solve for the prebuckling stresses, complex variable equations are used in conjunction with boundary collocation. The least squares approach is utilized to improve the accuracy of the results. The buckling problem is solved using the Ritz method. A product of the Ritz method is a complicated integral equation which is solved using numerical integration. To date, the aforementioned method of determining the prebuckling stresses was verified against infinite plate theory and finite elements.
The shear-lag model used previously by Nairn (1988) to derive a fracture mechanics analysis of longitudinal splitting in double-edge notched unidirectional composites was used to investigate the effect of fibers bridging across the longitudinal split in these composites. Using the new analysis, the longitudinal splitting fracture toughness, G(Lc), was determined for Hercules AS4/3501-6 graphite/epoxy laminates and for K-polymer based laminates containing Magnamite IM-6 as the graphite reinforcing fiber. Results show that the inclusion of the fiber bridging in the fracture analysis significantly affects the reported fracture toughness.
Recent experimental studies suggest that the hairpin vortex plays an important (and perhaps dominant) role in the dynamics of turbulent flows near walls. In this study, a numerical procedure is developed to allow the accurate computation of the trajectory of a 3D vortex having a small core radius. For hairpin vortices which are convected in a shear flow above a wall, the calculated results show that a 2D vortex containing a small 3D disturbance distorts into a complex shape with subsidiary hairpin vortices forming outboard of the original hairpin vortex. As the vortex moves above the wall, it induces unsteady motion in the viscous flow near the wall: numerical solutions suggest that the boundary-layer flow near the wall will ultimately erupt in response to the motion of the hairpin vortex and in the process a secondary hairpin vortex will be created. The computer results agree with recent experimental investigations.
An evaluation is made of the influence of interfacial microstructure and shear strength on the mechanical properties of a 30 vol pct uniaxially-aligned SiC fiber-reinforced reaction-bonded Si3N4-matrix composite whose interface microstructure was varied through control of fabrication conditions and by heat-treatment in an oxidizing environment. The carbon-rich coating of the as-produced SiC fibers was stable in composites fabricated at 1200 C in an N or N + 4-percent H mixture for 40 hrs. This coating was degraded in composites fabricated at 1350 C in N + 4 percent H for 40 and 72 hrs, as well as after heat-treatment in an oxidizing environment at 600 C for 100 hrs even after fabrication at 1200 C in N. This degradation occurred via carbon removal.
The physical process by which sheared Alfven waves can accelerate electrons to a Lorentz factor of 10,000 to 100,000 within 5 km of the stellar surface is applied to a study of gamma-ray bursts, taking both resonant and nonresonant scattering into account. Several very encouraging features of the model are discussed. Although the field is oscillatory, virtually all the charges are ejected from the system, resulting in very little backheating of the stellar surface. The particle number density is accounted for naturally in terms of BA0 and m, which in principle are known from the physical manifestation of the agent causing the crustal disturbance. The resulting gamma-ray spectrum compares very favorably with the observation. The model restricts the geometry of the emission region, in the sense that only the Compton upscattering of soft photons from a warm polar cap can produce the correct spectral shape.
The inviscid evolution of a two-dimensional shear layer is simulated numerically by a scheme based on a kinematic decomposition of the unsteady flow. Lagrangian and Weber transformations of the incompressible Euler equations result in a Clebsch representation that separates the flowfield into rotational and irrotational components. These transformations produce the initial construction of the flowfield and define its subsequent evolution.
A two-dimensional disturbance evolving from a strictly linear, finite-growth-rate instability wave, with nonlinear effects first becoming important in the critical layer is considered. The analysis is carried out for a general weakly nonparallel mean flow using matched asymptotic expansions. The flow in the critical layer is governed by a nonlinear vorticity equation which includes a spatial-evolution term. As in Goldstein and Hultgren (1988), the critical layer ages into a quasi-equilibrium one and the initial exponential growth of the instability wave is converted into a weak algebraic growth during the roll-up process. This leads to a next stage of evolution where the instability-wave growth is simultaneously affected by mean-flow divergence and nonlinear critical-layer effects and is eventually converted to decay. Expansions for the various streamwise regions of the flow are combined into a single composite formula accounting for both shear-layer spreading and nonlinear critical-layer effects and good agreement with the experimental results of Thomas and Chu (1989) and Freymuth (1966) is demonstrated.
The equilibrium structure of homogeneous turbulent shear flow is investigated from a theoretical standpoint. Existing turbulence models, in apparent agreement with physical and numerical experiments, predict an unbounded exponential time growth of the turbulent kinetic energy and dissipation rate; only the anisotropy tensor and turbulent time scale reach a structural equilibrium. It is shown that if a residual vortex stretching term is maintained in the dissipation rate transport equation, then there can exist equilibrium solutions, with bounded energy states, where the turbulence production is balanced by its dissipation. Illustrative calculations are presented for a k-epsilon model modified to account for net vortex stretching.
Gravitational microlensing by a single star plus external shear is considered. It is shown that for a general cusp the magnification probability distribution follows pc(A)dA of about (A exp -7/2)dA for sufficiently large magnifications. An adaptive grid technique is developed to calculate the magnification probability distributions. The results could be useful for cases of microlensing where the surface-mass density is low.
The problem considered is the development of the necessary plate stiffnesses for use in a general purpose program for buckling and vibration of composite plate assemblies. The required stiffnesses are for the assumption of sinusoidal response along the plate length with transverse shear included. The method is based on the exact solution of the plate differential equations for a composite laminate having fully populated A, B, and D matrices which leads to a differential equation of tenth order.
3D, compressible, unsteady, Reynolds-averaged Navier-Stokes equations are presently solved by a finite-volume and alternating-direction-implicit method in order to simulate supersonic and hypersonic turbulent shear flows. The effect of turbulence is incorporated via a modified Baldwin-Lomax eddy viscosity model which reflects the influence of high-speed compressibility, multiple walls, near-wall vortices, and turbulent memory effects, as well as local equilibrium effects. Attention is given to the simulation of the flow around the nozzle-afterbody of a generic, scramjet-propelled hypersonic vehicle; computed pressure distributions are consonant with experimental surface and off-surface flow surveys.
A finite element computational procedure has been developed to find linear and nonlinear (von Karman) first-ply failure loads of composite laminates subjected to in-plane and transverse loads. The FEM is based on first-order shear deformation theory and several phenomenological failure criteria. It is found that the failure loads predicted by various failure criteria differ from one another by a maximum of 35 percent in the case of linear loads and 50 percent in the case of nonlinear loads; the failure locations differ from each other in a random way. It is demonstrated that the difference between the linear and nonlinear failure loads is large in the case of transversely loaded simply supported laminates and thin plates. The difference is found to be quite small in the case of in-plane (tensile) loading and thick laminates.
The time-dependent Navier-Stokes equations are numerically integrated for two-dimensional incompressible viscous flow in a shear-driven square cavity. Using a time-splitting method and finite differences on a staggered mesh, the momentum and pressure equations are directly solved by a tensor product method where one finite difference direction is diagonalized by eigenvalue decomposition. The effects of increasing Reynolds number are studied and the developing boundary layer is captured by using a finely clustered mesh. At Re = 30,000 the flow is in a continuously developing unsteady regime. Power spectrum plots indicate that the unsteady flow oscillates with one fundamental frequency and exhibits some characteristics of transition between laminar and turbulent states.
Direct numerical simulations of compressible homogeneous turbulent shear flow are used to provide insight into compressibility effects on turbulence. The simulations show a reduction in the growth rate of the turbulence compared to the incompressible case. Examination of the turbulent kinetic energy budget shows that the reduced growth rate is due to an increase in the dissipation rate due to the divergence of the velocity and to the pressure-dilatation correlation which acts to transfer energy between internal energy and kinetic energy. The structure of the turbulence is also examined. Visualizations of the flowfields reveal the presence of eddy shocklets. These shock structures are important contributors to the increased dissipation of compressible turbulence. A mechanism for the generation of the shocks is suggested.
The paper substantiates the notion that selective amplification and direct resonance, based on linear theory, does not provide a selection mechanism for the well-defined streak spacing of about 100 wall units observed in wall-bounded turbulent shear flows. For the direct resonance theory, it is shown that the streaks are created by the nonlinear self-interaction of the vertical velocity rather than that of the directly forced vertical vorticity. It is proposed that the selection mechanism must be inherently nonlinear and correspond to a self-sustaining process. For the case of plane Poiseuille flow the 100-wall-unit criterion corresponds to a critical Reynolds number of 1250, based on the centerline velocity and the channel half-width, which is close to the usually quoted value of about 1000. In plane Couette flow, it corresponds to a critical Reynolds number of 625, based on the half-velocity difference and the half-width.