The first experiment completed in the MIR facility was conducted by Dr. Stephan Becker in 2000. This experiment produced new fundamental measurements of the transition process in flat plate boundary layers downstream of a two-dimensional square rib. By use of laser Doppler anemometry (LDA) and the large Matched-Index-of-Refraction (MIR) flow system, data for wall-normal fluctuations and Reynolds stresses were obtained in the near wall region to y + < 0.1 in addition to the usual mean streamwise velocity component and its fluctuations. By varying velocity and rib height, the experiment investigated the following range of conditions: k + = 5.5 to 21, 0.3 < k/δ < 1, 180 < Re k < 740, 6 x 104 < Re x,k ) < 1.5x10 5 , Re θ 660, and -125 < (x-x k )/k < 580. Consequently, results covered boundary layers which retained their laminar characteristics through those where a turbulent boundary layer was established shortly after reattachment beyond the forcing rib. For “large” elements, evolution of turbulent statistics of the viscous layer for a turbulent boundary layer (y + < 30) was rapid even in flows where the mean velocity profile still showed laminar behavior.
Ref: Becker, S., Stoots, C.M., Condie, K.G., Durst, F. and McEligot, D.M., 2002, “LDA-Measurements of Transitional Flows Induced by a Square Rib,” J. Fluids Eng., 124, March 2002, pp. 108-117.
Sample images: Experimental apparatus, model configuration and nomenclature
LDA measurements of the evolution of flow over a two-dimensional square rib
Velocity, velocity fluctuations and Reynolds stresses downstream of a two-dimensional square rib
2. Measurements of Fundamental Fluid Physics of SNF Storage Canisters
The goal of this program was to develop innovative flow visualization methods and reliable predictive techniques for the energy, mass, and momentum transfer in the presence of surface reactions for the passivation treatment operations of spent nuclear fuel (SNF) elements. A generic SNF canister, to store high- and medium-enriched fuels, was developed to serve as a common experimental configuration.
A unique flow visualization technique was developed to model a reaction between the passivating oxygen gas and uranium hydrides found on corroded areas of wet-stored SNF elements. In the experiments, hexanoic acid and sodium metal simulated the oxygen and uranium hydride respectively. The hexanoic acid and sodium metal reacted to form hydrogen, the same byproduct of the passivation reaction. The hydrogen is buoyant compared to the surrounding fluid and therefore rises. To observe the effects of the reaction on the flow field, dye was injected upstream of the sodium. The experiments showed that the rising hydrogen behaved similarly to buoyant plumes, such as those observed in open fires or gases exiting a smokestack. The hydrogen entrained the surrounding fluid, introduced velocity fluctuations and instabilities into the flow, and enhanced fluid mixing. The entrainment caused the dye filaments to stretch and bend, and the buoyant hydrogen created counter-rotating vortices in the flow field that straddled the reaction site. While the hydrogen convected fluid away from the upstream portion of the reaction site and created vortices, these same vortices convected surrounding fluid into the downstream portion of the reaction. High resolution photographs contained in the report document this behavior. It was determined that a local surface reaction can affect the surrounding flow field.
Experiments were also conducted with the INEEL Matched Index of Refraction facility to study the overall flow field of the generic SNF canister. The canister had a perforated basket support plate through which an inlet pipe transported the passivating gases, and experiments were conducted on the perforated support plates with 50, 8, and 4 percent open areas. With the 50 percent open area, the flow was similar to a submerged impinging jet, with formation of a large vortex and entrainment and recirculation through the holes of the plate. With the 4 and 8 percent open area plates, recirculation regions arose downstream of the plate, and two main recirculation zones formed upstream of the plate. Flow did not reenter the inlet plenum through the smaller open area plates as it did with the 50 percent open area plate. The flow with the smaller holes in the support plate was similar to a confined impinging jet.
Velocity and turbulence measurements included the mean distributions of radial and axial components and their root-mean-squared fluctuations. The rms radial velocity fluctuations were considerably larger than the rms axial fluctuations. Large values of the radial fluctuations were observed in the wall jet, and moderate values were observed in its wake. Moderate to high values also appear near the perforated support plate in the region directly affected by the recirculating eddy. Local turbulent intensities were observed to be on the order of 100 percent or greater. The results show that flow in the plenum region of the canister upstream of the perforated support plate is sensitive to the inlet flow considerations. Downstream of the plate, the flow relaminarizes as it flows axially over the SNF elements. It is clear that despite the conceptual simplicity of the canister design, the overall flow pattern within the canister is very complex. An uneven flow distribution near the surface of the elements may lead to uneven passivation treatment of the elements.
Various numerical models were developed to analyze the complex nature of flow in the SNF canisters. Two- and three-dimensional models were developed to assess different flow features. The general-purpose code provided adequate results for the simulation of the low Re pipe flow. Agreement between the measured and predicted results was adequate, but refinements to the model could lead to improved comparisons. Modeling results that depended on inlet flow conditions to the plenum were found to be very sensitive to the assumptions made to the inlet pipe flow conditions. The selection of inlet boundary conditions is important to accurately predict turbulent flow conditions for these complex geometries. The modeling results are very sensitive to the type of turbulence model used for the simulations. Extremely large variations in the flow field were noted for the different models. The most realistic results were obtained with a low Reynolds number turbulence model with accurate upwinding schemes. Low-order upwinding schemes and poor assumptions for the inlet flow profiles lead to drastic changes in the results.
It is necessary to include the details of the holes in the plate within the numerical model. It may be possible to develop a “porous plate” model for disperse holes, but model limitations may make it difficult to accurately model a wide range of design options. For example, if such a model impeded re-entrant flow into the plenum region, such a limitation may not detect situations where such flow truly exists. If a porous plate model were developed it would have to be developed with care to insure the appropriate physics are properly considered. Numerical convergence is slow for the three-dimensional flow fields modeled. The order of magnitude changes in the velocity field across the computational domain make it difficult to obtain fast convergence. As the next generation of computing platforms is developed, it will be easier to include the necessary details to analyze the details of the flow field. More uniform flow up past the fuel elements will be obtained through the use of a plate that restricts the flow and does not allow it to re-enter the plenum region. Detailed analyses and testing would have to be conducted to truly assess the uniform characteristic of the flow.
Ref: Condie, K.G., McCreery, G.E. and McEligot, 2001, “Measurements of Fundamental Fluid Physics of SNF Storage Canisters,” INEEL/EXT-01-01269, September 2001.
Design of 0.6-scale model for studies of the generic turbulent flow processes of idealized SNF canisters.
Mean flow pattern of a semi-confined impinging jet and its surroundings, Rejet = 2510
3. Physical and Computational Modeling of Airflow Around Buildings
There is a need for information on dispersion and infiltration of chemical and biological agents in complex building environments. A recent collaborative study conducted at the Idaho National Engineering and Environmental Laboratory (INEEL) and Bechtel Corporation Research and Development had the objective of assessing computational fluid dynamics (CFD) models for simulation of flow around complicated buildings through a comparison of experimental and numerical results. The test facility used in the experiments was INEEL’s unique large Matched-Index-of-Refraction (MIR) flow system. The CFD code used for modeling was Fluent, a widely available commercial flow simulation package. For the experiment, a building plan was selected to approximately represent an existing facility. It was found that predicted velocity profiles from above the building and in front of the building were in good agreement with the measurements.
Ref: McEligot, D.M., McCreery, G.E., Pink, R.J, Barringer, C. and Knight, K.J., 2001, “Physical and Computational Modeling for Chemical and Biological Weapons Airflow Applications,” INEEL/CON-02-00860, November 2001.
Experimental model showing LDV measurement location
Building model showing typical recirculation regions
The ultimate goal of the study is the improvement of predictive methods for safety analyses and design of advanced reactors for higher efficiency and enhanced safety and for deployable reactors for electrical power generation, process heat utilization and hydrogen generation. While key applications would be advanced gas-cooled reactors (AGCRs) using the closed Brayton cycle (CBC) for higher efficiency (such as the proposed Gas Turbine - Modular Helium Reactor (GT-MHR) of General Atomics [Neylan and Simon, 1996]), results of the proposed research should also be valuable in reactor systems with supercritical flow or superheated vapors, e.g., steam. Higher efficiency leads to lower cost/kwh and reduces life-cycle impacts of radioactive waste (by reducing waste/kwh). The outcome will also be useful for some space power and propulsion concepts and for some fusion reactor concepts as side benefits, but they are not the thrusts of the investigation. The objective of the project is to provide fundamental thermal fluid physics knowledge and measurements necessary for the development of the improved methods for the applications.
Two experimental models were employed. The first simulated flow in a control rod channel (GA) or a coolant channel, with annular passages interrupted by periodic spacer ribs. Figure B-2 depicts this model schematically. The second experiment addressed a ribbed annulus upstream of an idealized two-stage jet transition from generic coolant channels to a plenum (see Figure B-1b). It is envisioned as a combined experiment focusing on the first stage, a jet issuing from a converging flow into a collecting plenum (or impingement chamber).
Ref: McEligot, D.M., Condie, K.G., Foust, T.D., Jackson, J.D., Kunugi, T., McCreery, G.E., Pink, R.J., Pletcher, R.H., Satake, S.I., Shenoy, A., Stacey, D.E., Vukoslavcevic, P. and Wallace, J.M., 2002, Fundamental Thermal Fluid Physics of High Temperature Flows in Advanced Reactor Systems,” INEEL-EXT-2002-1613, December 2002.
The goal of this collaboration is the improvement of predictive methods for Generation IV reactor systems and associated Advanced Fuel Cycle Initiative (AFCI) and Nuclear Hydrogen Initiative (NHI) activities. The INEEL fabricated a large-scale model for simulating flow in supercritical water reactor (SCWR) coolant passages and employed laser Doppler velocimetry to collect velocity measurements in a closely-packed array of fuel rods separated by periodic grid spacers as used in several SCWR concepts.
Ref: McEligot, D.M., Condie, K.G., McCreery, G.E., Hochreiter, L.E., Jackson, J.D., Pletcher, R.H., Wallace, J.M., Yoo, J.Y., Ro, S.T., Lee, J.WS. and Park, S.O., 2003, “Advanced Computational Thermal Fluid Physics (CTFP) and its Assessment for Light Water Reactors and Supercritical Reactors,” INEEL-EXT03-01215 Rev 5, December 2003.
The impact of turbine blade surface roughness on aerodynamic performance and heat loads is well known. Over time, as the turbine blades are exposed to huge heat loads, the external surfaces of the blades are degraded and become rough. Also, for film-cooled blades, surface degradation can have a significant impact on film-cooling effectiveness. Surface roughness on turbine airfoils can begin, and increase dramatically, from surface oxidation, surface corrosion, deposition of unburned fuel or foreign materials, spallation of thermal barrier coatings (TBC), erosion and/or pitting during routine engine operation. Surface degradation results in an increase in heat loads and friction losses. Many studies have been conducted on the effects of surface degradation/roughness on engine performance but most investigations have modeled the rough surfaces with uniform or two-dimensional roughness patterns. There is a clear requirement to investigate and measure the effects of realistic surface roughness on the flow through a turbine stage. The objective of the present investigation is to conduct measurements that will reveal the influence of realistic surface roughness on the near-wall behavior of the boundary layer. Measurements have been conducted at the Matched-Index-of-Refraction (MIR) Facility oil tunnel at the Idaho National Engineering and Environmental Laboratory (INEEL) with a two-component, TSI fiberoptic-based laser Doppler velocimeter (LDV) operated in the forward-scatter mode. A zero-incidence flat plate model (flow parallel to the plate) of a turbine blade has been developed that produces a transitional boundary layer, elevated freestream turbulence and an accelerating freestream in order to simulate conditions on the suction side of a high-pressure turbine blade. Extensive boundary layer measurements have been completed over a smooth plate version of the model and a version with an embedded strip of realistic rough surface. The realistic rough surface was developed by scaling actual turbine blade surface data that was provided by the U. S. Air Force Research Laboratory. The results indicate that bypass transition occurred very early in the flow over the model and that the boundary layer remained unstable throughout the entire length of the test plate. Results from the rough patch study indicate the boundary layer thickness and momentum thickness Reynolds numbers increased over the rough patch and the shape factor increased over the rough patch but then decreased downstream of the patch. It was also found that flow downstream of the patch experienced a gradual re-transition to laminar-like behavior but in less time and distance than in the smooth plate case. Additionally, the rough patch caused a significant increase in streamwise turbulence intensity and streamwise normal turbulence intensity downstream of the patch and the rough surface caused the skin friction coefficient to vary dramatically over the rough patch where the skin friction coefficient was nearly double the smooth plate value. Finally, the rough patch caused the Reynolds stresses to increase in the region close to the plate surface. The LDV system used with the MIR oil tunnel arrangement has permitted mean and fluctuating velocity measurements very close to the wall (down to y + < 1). These detailed results should be valuable to assess and guide development of computational fluid dynamics models that could reliably predict flows over a high-pressure turbine with realistically roughened blades or vanes. Therefore, these results may be of benefit to the gas turbine industry by improving design and reducing manufacturing and maintenance costs.
Ref: McIlroy, H. M. Jr., 2004, “The Boundary Layer Over Turbine Blade Models with Realistic Rough Surfaces,” PhD Dissertation, University of Idaho, December 2004.
The interaction of a circular synthetic jet with a laminar cross-flow boundary layer was investigated experimentally in the Matched-Index-of-Refraction flow facility at Idaho National Laboratory. Two orifice orientations were investigated, straight and inclined. For each orifice, phase-averaged and time-averaged PIV measurements were made at Lο/Dο = 1.0 and 2.0 with ReUο = 250 and r = 1.12. Refractive index matching between the working fluid and the model material permitted experimental measurements of the flow field inside the actuator orifice and cavity simultaneously. At Lο/Dο = 1.0, the vortex ring formed at the orifice during the expulsion portion of the actuator cycle blocks the boundary layer causing the flow to divert over and around the ring. This vortex ring does not escape the near-vicinity of the orifice and is subsequently re-ingested. At the same stroke, inclining the orifice axis 30ο downstream leads to a jet comprised of a train of vortex rings that penetrates the cross-flow. At Lο/Dο = 2.0, both the straight and inclined orifices create large discrete vortex rings that penetrate deep into the cross-flow, and consequently do not affect the boundary layer much beyond the near-field of the orifice.
Ref: Shuster, J.M., Pink, R.J., McEligot, D.M. and Smith, D.R., 2005, “Interaction of a Circular Synthetic Jet with a Cross-Flow Boundary Layer,” 35th AIAA paper 2005-4749, Fluid Dynamics Conference and Exhibit, 6-9 June 2005, Toronto, CA.
Mean-velocity-field and turbulence data are presented that measure turbulent flow phenomena in an approximately 1:7 scale model of a region of the lower plenum of a typical prismatic gas-cooled reactor (GCR) similar to a General Atomics Gas-Turbine-Modular Helium Reactor (GTMHR) design. The data were obtained in the Matched-Index-of-Refraction (MIR) facility at Idaho National Laboratory (INL) and are offered for assessing computational fluid dynamics (CFD) software. This experiment has been selected as the first Standard Problem endorsed by the Generation IV International Forum.
Results concentrate on the region of the lower plenum near its far reflector wall (away from the outlet duct). The flow in the lower plenum consists of multiple jets injected into a confined cross flow - with obstructions. The model consists of a row of full circular posts along its centerline with half-posts on the two parallel walls to approximate geometry scaled to that expected from the staggered parallel rows of posts in the reactor design. The model is fabricated from clear, fused quartz to match the refractive-index of the working fluid so that optical techniques may be employed for the measurements. The benefit of the MIR technique is that it permits optical measurements to determine flow characteristics in complex passages in and around objects to be obtained without locating intrusive transducers that will disturb the flow field and without distortion of the optical paths. An advantage of the INL system is its large size, leading to improved spatial and temporal resolution compared to similar facilities at smaller scales. A three-dimensional (3-D) Particle Image Velocimetry (PIV) system was used to collect the data. Inlet jet Reynolds numbers (based on the jet diameter and the time-mean bulk velocity) are approximately 4,300 and 12,400. Uncertainty analyses and a discussion of the standard problem are included.
The measurements reveal developing, non-uniform, turbulent flow in the inlet jets and complicated flow patterns in the model lower plenum. Data include three-dimensional vector plots, data displays along the coordinate planes (slices) and presentations that describe the component flows at specific regions in the model. Information on inlet conditions is also presented.
Ref: McIlroy, H. M. Jr., McEligot, D. M., and Pink, R. J., “Measurement of Flow Phenomena in a Lower Plenum Model of a Prismatic Gas-Cooled Reactor,” J. of Eng. for Gas Turbines & Power, 132, Feb. 2010, pp. 022901-1 – 022901-7.
The concept of experimental model-ability for CFD validation is demonstrated by using experimental and numerical models of swirling jets as an example. The experiment is designed to contain relevant physics (e.g. swirl) but without regard for generality or direct scaling to realistic geometries. Also, the design allows for direct measurement of the inflow. The swirling jets experience vortex breakdown, have non-symmetric inlet profiles, and have Reynolds numbers of 3650, 2560, and 550 with Rossby numbers of 0:63, 0:59, and 0:73 respectively. It is demonstrated that the numerical simulations will provide more accurate results when the inlet conditions of these nonsymmetric swirling jets are fully described. The geometry of the experimental model and several typical turbulence closure models are discussed to support the idea of experimental model-ability. Effects of the geometrical design on the numerical simulation are also analyzed.
Ref: Wilson, B.M., Smith, B.L., Spall, R. and McIlroy, H.M. Jr., 2009, “A Non-Symmetrical Swirling Jet as an Example of a highly Model-able Assessment Experiment,” ICONE17-75362, Proceedings of ICONE17 2009, 17th International Conference on Nuclear Engineering Brussles, Belgium, July 16-19, 2009
The centerline velocities of the experimental, asymmetric and symmetric numerical models for Re = 3650.