Search NASA⌕ Search

SEARCH · Search NASA

Results for “Compressor Modeling Modeling”

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.

At least 289 records · Page 16
All sourcesNASA NTRSDOE OSTI
NASA NTRS1992

Users manual for updated computer code for axial-flow compressor conceptual design

An existing computer code that determines the flow path for an axial-flow compressor either for a given number of stages or for a given overall pressure ratio was modified for use in air-breathing engine conceptual design studies. This code uses a rapid approximate design methodology that is based on isentropic simple radial equilibrium. Calculations are performed at constant-span-fraction locations from tip to hub. Energy addition per stage is controlled by specifying the maximum allowable values for several aerodynamic design parameters. New modeling was introduced to the code to overcome perceived limitations. Specific changes included variable rather than constant tip radius, flow path inclination added to the continuity equation, input of mass flow rate directly rather than indirectly as inlet axial velocity, solution for the exact value of overall pressure ratio rather than for any value that met or exceeded it, and internal computation of efficiency rather than the use of input values. The modified code was shown to be capable of computing efficiencies that are compatible with those of five multistage compressors and one fan that were tested experimentally. This report serves as a users manual for the revised code, Compressor Spanline Analysis (CSPAN). The modeling modifications, including two internal loss correlations, are presented. Program input and output are described. A sample case for a multistage compressor is included.

Glassman, Arthur J.↗
NASA NTRS1978

End-wall boundary layer prediction for axial compressors

An integral boundary layer procedure was developed for the computation of viscous and secondary flows along the annulus walls of an axial compressor. The procedure is an outgrowth and extension of the pitch-averaged methods of Mellor and Horlock. In the present work secondary flow theory is used to develop approximations for the velocity profiles inside a rotating blade row and for the blade force deficit terms in the momentum integral equations. The computer code based on this procedure was iteratively coupled to a quasi-one-dimensional model for the external inviscid flow. Computed results are compared with measurements in a compressor cascade.

Sockol, P. M.↗
NASA NTRS1978

End-wall boundary layer prediction for axial compressors

An integral boundary layer procedure has been developed for the computation of viscous and secondary flows along the annulus walls of an axial compressor. The procedure is an outgrowth and extension of the pitch-averaged methods of Mellor and Horlock. In the present work secondary flow theory is used to develop approximations for the velocity profiles inside a rotating blade row and for the blade force deficit terms in the momentum integral equations. The computer code based on this procedure has been iteratively coupled to a quasi-one-dimensional model for the external inviscid flow. Computed results are compared with measurements in a compressor cascade.

Sockol, P. M.↗
DOE OSTI2025

FedCSpc: A Cross-Silo Federated Learning System With Error-Bounded Lossy Parameter Compression

Cross-Silo federated learning is widely used for scaling deep neural network (DNN) training over data silos from different locations worldwide while guaranteeing data privacy. Communication has been identified as the main bottleneck when training large-scale models due to large-volume model parameters and gradient transmission across public networks with limited bandwidth. Most previous works focus on gradient compression, while limited work tries to compress parameters that can not be ignored and extremely affect communication performance during the training. Here, to bridge this gap, we propose FedCSpc: an efficient cross-silo federated learning system with an XAI-driven adaptive parameter compression strategy for large-scale model training. Our work substantially differs from existing gradient compression techniques due to the distinct data features of gradient and parameter. The key contributions of this paper are fourfold. (1) Our designed FedCSpc proposes to compress the parameter during the training using the state-of-the-art error-bounded lossy compressor – SZ3. (2) We develop an adaptive compression error bound adjustment algorithm to guarantee the model accuracy effectively. (3) We exploit an efficient approach to utilize the idle CPU resources of clients to compress the parameters. (4) We perform a comprehensive evaluation with a wide range of models and benchmarks on a GPU cluster with 65 GPUs. Results show that FedCSpc can achieve the same model accuracy as FedAvg while reducing the data volume of parameters and gradients in communication by up to 7.39× and 288×, respectively. With 32 clients on a 4 Gb size model, FedCSpc significantly outperforms FedAvg in wall-clock time in the emulated WAN environment (at the bandwidth of 1 Gbps or lower without loss of generality).

SZ3↗
NASA NTRS2014

Numerical Simulation of Boundary Layer Ingesting (BLI) Inlet-Fan Interaction

Future civil transport designs may incorporate engine inlets integrated into the body of the aircraft to take advantage of efficiency increases due to weight and drag reduction. Additional increases in engine efficiency are predicted if the inlet ingests the lower momentum boundary layer flow. Previous studies have shown, however, that efficiency benefits of Boundary Layer Ingesting (BLI) ingestion are very sensitive to the magnitude of fan and duct losses, and blade structural response to the non-uniform flow field that results from a BLI inlet has not been studied in-depth. This paper presents an effort to extend the modeling capabilities of an existing rotating turbomachinery unsteady analysis code to include the ability to solve the external and internal flow fields of a BLI inlet. The TURBO code has been a successful tool in evaluating fan response to flow distortions for traditional engine/inlet integrations, such as the development of rotating stall and inlet distortion through compressor stages. This paper describes the first phase of an effort to extend the TURBO model to calculate the external and inlet flowfield upstream of fan so that accurate pressure distortions that result from BLI configurations can be computed and used to analyze fan aerodynamics and structural response. To validate the TURBO program modifications for the BLI flowfield, experimental test data obtained by NASA for a flushmounted S-duct with large amounts of boundary layer ingestion was modeled. Results for the flow upstream and in the inlet are presented and compared to experimental data for several high Reynolds number flows to validate the modifications to the solver. Quantitative data is presented that indicates good predictive capability of the model in the upstream flow. A representative fan is attached to the inlet and results are presented for the coupled inlet/fan model. The impact on the total pressure distortion at the AIP after the fan is attached is examined.

Inlet Fan Interactions↗
NASA NTRS1957

Performance Comparison at Mach Numbers 1.8 and 2.0 of Full Scale and Quarter Scale Translating-Spike Inlets

The performance of a full-scale translating-spike inlet was obtained at Mach numbers of 1.8 and 2.0 and at angles of attach from 0 deg to 6 deg. Comparisons were made between the full-scale production inlet configuration and a geometrically similar quarter-scale model. The inlet pressure-recovery, cowl pressure-distribution, and compressor-face distortion characteristics of the full-scale inlet agreed fairly well with the quarter-scale results. In addition, the results indicated that bleeding around the periphery ahead of the compressor-face station improved pressure recovery and compressor-face distortion, especially at angle of attack.

Anderson, B. H.↗
NASA NTRS1998

Minnowbrook II 1997 Workshop on Boundary Layer Transition in Turbomachines

The volume contains materials presented at the Minnowbrook II - 1997 Workshop on Boundary Layer Transition in Turbomachines, held at Syracuse University Minnowbrook Conference Center, New York, on September 7-10, 1997. The workshop followed the informal format at the 1993 Minnowbrook I workshop, focusing on improving the understanding of late stage (final breakdown) boundary layer transition, with the engineering application of improving design codes for turbomachinery in mind. Among the physical mechanisms discussed were hydrodynamic instabilities, laminar to turbulent transition, bypass transition, turbulent spots, wake interaction with boundary layers, calmed regions, and separation, all in the context of flow in turbomachinery, particularly in compressors and high and low pressure turbines. Results from experiments, DNS, computation, modeling and theoretical analysis were presented. Abstracts and copies of viewgraphs, a specifically commissioned summation paper prepared after the workshop, and a transcript of the extensive working group reports and discussions are included in this volume. They provide recommendations for future research and clearly highlight the need for continued vigorous research in the technologically important area of transition in turbomachines.

LaGraff John E.↗
NASA NTRS1998

Compressibility and Turbulence Effects Due to Airfoil Clocking in Axial Compressors

Axial compressors have inherently unsteady flow fields because of relative motion between rotor and stator airfoils. This relative motion leads to viscous and inviscid (potential) interactions between blade rows. As the number of stages increases in a turbomachine, the buildup of convected wakes can lead to progressively more complex wake/wake and wake/airfoil interactions. Variations in the relative circumferential positions of stators or rotors can change these interactions, leading to different unsteady forcing functions on airfoils and different compressor efficiencies. In addition, as the Mach number increases the interaction between blade rows is intensified due to potential effects. The current study uses an unsteady, two-dimensional Navier-Stokes approach to investigate the unsteady aerodynamics of stator clocking in a 1-1/2 stage compressor, typical of high-pressure compressors used in advanced commercial jet engines. The effects of turbulence are modeled with both algebraic and two-equation models. Results include surface pressures, efficiencies, boundary layer quantities and turbulence quantities. In addition, the growth of turbulence and the effects of compressibility on airfoil are discussed.

Dorney, Daniel J.↗
DOE OSTI2025

A Survey on Error-Bounded Lossy Compression for Scientific Datasets

Error-bounded lossy compression has been effective in significantly reducing the data storage/transfer burden while preserving the reconstructed data fidelity very well. Many error-bounded lossy compressors have been developed for a wide range of parallel and distributed use cases for years. They are designed with distinct compression models and principles, such that each of them features particular pros and cons. In this article, we provide a comprehensive survey of emerging error-bounded lossy compression techniques. The key contribution is fourfold. (1) We summarize a novel taxonomy of lossy compression into six classic models. (2) We provide a comprehensive survey of 10 commonly used compression components/modules. (3) We summarized pros and cons of 47 state-of-the-art lossy compressors and present how state-of-the-art compressors are designed based on different compression techniques. (4) We discuss how customized compressors are designed for specific scientific applications and use-cases. We believe this survey is useful to multiple communities including scientific applications, high-performance computing, lossy compression, and big data.

Error-Bounded Lossy Compression↗
DOE OSTI2024

Advanced Compressors for CO2-Based Power Cycles and Energy

Pumped thermal energy storage (PTES) is a cost-effective method for grid-scale, long-duration electrical energy storage (LDES). PTES uses a heat pump cycle to transfer thermal energy from a low temperature reservoir (LTR) to a high temperature reservoir (HTR), later using a heat engine cycle to reverse the process. A key component of the PTES system, the heat pump compressor, represents a significant driver to the cost, performance and operating characteristics of the PTES system. For grid-scale charging (>100 MW), traditional compressor scaling charts indicate that the operating conditions needed would be best served by a multistage axial compressor. While frame gas turbine compressors at these power ratings exist and operate at higher pressure ratios than needed for the CO2 PTES system, the inlet pressure and fluid density of this application exceed experience values. A conceptual design of a large-scale CO2 axial compressor was completed, including mean-line estimates of the compressor performance at full power conditions. The results of the conceptual design were used to refine the PTES cycle design, and updated operating conditions provided for further aero design optimization. A roughly one third scale of the first three stages of the 100 MW compressor was designed, built, and tested at relevant PTES charging conditions. Results of this test program are compared against high-resolution computational fluid dynamics models.

Compressor Heat↗
NASA NTRS1992

Stability and performance of notch filter control for unbalance response

Many current applications of magnetic bearings for rotating machinery employ notch filters in the feedback control loop to reduce the synchronous forces transmitted through the bearings. The capabilities and limitations of notch filter control are investigated. First, a rigid rotor is examined with some classical root locus techniques. Notch filter control is shown to result in conditional stability whenever complete synchronous attenuation is required. Next, a nondimensional parametric symmetric flexible three mass rotor model is constructed. An examination of this model for several test cases illustrates the limited attenuation possible with notch filters at and near the system critical speeds when the bearing damping is low. The notch filter's alteration of the feedback loop is shown to cause stability problems which limits performance. Poor transient response may also result. A high speed compressor is then examined as a candidate for notch filter control. A collocated 22 mass station model with lead-lag control is used. The analysis confirms the reduction in stability robustness that can occur with notch filter control. It is concluded that other methods of synchronous vibration control yield greater performance without compromising stability.

Knospe, C. R.↗
NASA NTRS2002

Computational Modeling of Vortex Generators for Turbomachinery

In this work computational models were developed and used to investigate applications of vortex generators (VGs) to turbomachinery. The work was aimed at increasing the efficiency of compressor components designed for the NASA Ultra Efficient Engine Technology (UEET) program. Initial calculations were used to investigate the physical behavior of VGs. A parametric study of the effects of VG height was done using 3-D calculations of isolated VGs. A body force model was developed to simulate the effects of VGs without requiring complicated grids. The model was calibrated using 2-D calculations of the VG vanes and was validated using the 3-D results. Then three applications of VGs to a compressor rotor and stator were investigated: 1) The results of the 3-D calculations were used to simulate the use of small casing VGs used to generate rotor preswirl or counterswirl. Computed performance maps were used to evaluate the effects of VGs. 2) The body force model was used to simulate large part-span splitters on the casing ahead of the stator. Computed loss buckets showed the effects of the VGs. 3) The body force model was also used to investigate the use of tiny VGs on the stator suction surface for controlling secondary flows. Near-surface particle traces and exit loss profiles were used to evaluate the effects of the VGs.

Chima, R. V.↗
NASA NTRS2002

Overview of High-Fidelity Modeling Activities in the Numerical Propulsion System Simulations (NPSS) Project

A high-fidelity simulation of a commercial turbofan engine has been created as part of the Numerical Propulsion System Simulation Project. The high-fidelity computer simulation utilizes computer models that were developed at NASA Glenn Research Center in cooperation with turbofan engine manufacturers. The average-passage (APNASA) Navier-Stokes based viscous flow computer code is used to simulate the 3D flow in the compressors and turbines of the advanced commercial turbofan engine. The 3D National Combustion Code (NCC) is used to simulate the flow and chemistry in the advanced aircraft combustor. The APNASA turbomachinery code and the NCC combustor code exchange boundary conditions at the interface planes at the combustor inlet and exit. This computer simulation technique can evaluate engine performance at steady operating conditions. The 3D flow models provide detailed knowledge of the airflow within the fan and compressor, the high and low pressure turbines, and the flow and chemistry within the combustor. The models simulate the performance of the engine at operating conditions that include sea level takeoff and the altitude cruise condition.

Veres, Joseph P.↗
NASA NTRS2009

SWIFT Code Assessment for Two Similar Transonic Compressors

One goal of the NASA Fundamental Aeronautics Program is the assessment of computational fluid dynamic (CFD) codes used for the design and analysis of many aerospace systems. This paper describes the assessment of the SWIFT turbomachinery analysis code for two similar transonic compressors, NASA rotor 37 and stage 35. The two rotors have identical blade profiles on the front, transonic half of the blade but rotor 37 has more camber aft of the shock. Thus the two rotors have the same shock structure and choking flow but rotor 37 produces a higher pressure ratio. The two compressors and experimental data are described here briefly. Rotor 37 was also used for test cases organized by ASME, IGTI, and AGARD in 1994-1998. Most of the participating codes over predicted pressure and temperature ratios, and failed to predict certain features of the downstream flowfield. Since then the AUSM+ upwind scheme and the k- turbulence model have been added to SWIFT. In this work the new capabilities were assessed for the two compressors. Comparisons were made with overall performance maps and spanwise profiles of several aerodynamic parameters. The results for rotor 37 were in much better agreement with the experimental data than the original blind test case results although there were still some discrepancies. The results for stage 35 were in very good agreement with the data. The results for rotor 37 were very sensitive to turbulence model parameters but the results for stage 35 were not. Comparison of the rotor solutions showed that the main difference between the two rotors was not blade camber as expected, but shock/boundary layer interaction on the casing.

Chima, Rodrick V.↗
NASA NTRS1990

The numerical simulation of multistage turbomachinery flows

The need to account for momentum and energy transport by the unsteady deterministic flow field in modeling the time-averaged flow state within a blade row passage embedded in a multistage compressor is assessed. It was found that, within the endwall regions, large-scale three-dimensional unsteady structures existed which caused significant transport of momentum and energy across the time-averaged stream surface of a stator flow field. These experiments confirmed that the tranport process is dominated by turbulent diffusion in the midspan region. A model was then proposed for simulating this transport process, and a limited study was undertaken to assess its validity.

Adamczyk, J. J.↗
NASA NTRS1981

Supersonic Stall Flutter of High Speed Fans

An analytical model is developed for predicting the onset of supersonic stall bending flutter in axial flow compressors. The analysis is based on a modified two dimensional, compressible, unsteady actuator disk theory. It is applied to a rotor blade row by considering a cascade of airfoils whose geometry and dynamic response coincide with those of a rotor blade element at 85 percent of the span height (measured from the hub). The rotor blades are assumed to be unshrouded (i.e., free standing) and to vibrate in their first flexural mode. The effects of shock waves and flow separation are included in the model through quasi-steady, empirical, rotor total-pressure-loss and deviation-angle correlations. The actuator disk model predicts the unsteady aerodynamic force acting on the cascade blading as a function of the steady flow field entering the cascade and the geometry and dynamic response of the cascade. Calculations show that the present model predicts the existence of a bending flutter mode at supersonic inlet Mach numbers. This flutter mode is suppressed by increasing the reduced frequency of the system or by reducing the steady state aerodynamic loading on the cascade. The validity of the model for predicting flutter is demonstrated by correlating the measured flutter boundary of a high speed fan stage with its predicted boundary. This correlation uses a level of damping for the blade row (i.e., the log decrement of the rotor system) that is estimated from the experimental flutter data. The predicted flutter boundary is shown to be in good agreement with the measured boundary.

Adamczyk, J. J.↗
NASA NTRS1983

Design of shock-free compressor cascades including viscous boundary layer effects

A computer code that generates shock-free transonic compressor cascade shapes while taking into account viscosity effects is developed. The mathematical model for the inviscid flow field is the full potential equation. The Kutta-Joukowski condition is satisfied by varying the free stream angle at downstream infinity. A boundary fitted computational grid of C-type is generated using a sequence of conformal mapping and nonorthogonal coordinate stretching and shearing transformations. The inviscid calculation is performed sequentially on up to four consecutively refined grids thereby accelerating the convergence of the solution process. The full potential equation is solved using a finite area technique and rotated, type-dependent finite differencing. Artificial viscosity of the first order is added in a fully conservative form. Shock-free cascade airfoil shapes are obtained using the fictitious gas concept of Sobieczky and the method of characteristics in the rheograph plane. Viscosity effects are incorporated via a boundary layer displacement thickness. The integral boundary layer code is based on Rotta's turbulence model and assumes transition region of zero length.

Dulikravich, G. S.↗
NASA NTRS2006

High Fidelity System Simulation of Multiple Components in Support of the UEET Program

The High Fidelity System Simulation effort has addressed various important objectives to enable additional capability within the NPSS framework. The scope emphasized High Pressure Turbine and High Pressure Compressor components. Initial effort was directed at developing and validating intermediate fidelity NPSS model using PD geometry and extended to high-fidelity NPSS model by overlaying detailed geometry to validate CFD against rig data. Both "feedforward" and feedback" approaches of analysis zooming was employed to enable system simulation capability in NPSS. These approaches have certain benefits and applicability in terms of specific applications "feedback" zooming allows the flow-up of information from high-fidelity analysis to be used to update the NPSS model results by forcing the NPSS solver to converge to high-fidelity analysis predictions. This apporach is effective in improving the accuracy of the NPSS model; however, it can only be used in circumstances where there is a clear physics-based strategy to flow up the high-fidelity analysis results to update the NPSS system model. "Feed-forward" zooming approach is more broadly useful in terms of enabling detailed analysis at early stages of design for a specified set of critical operating points and using these analysis results to drive design decisions early in the development process.

Plybon, Ronald C.↗