This paper provides a description of a recent comprehensive Analysis of Alternatives (AoA) for icing research. A team of NASA icing subject matter experts (SMEs) was assembled from NASA’s Glenn and Langley Research Centers, which began by gathering inputs from icing stakeholders in industry, Government, and academia. Those inputs, which were grouped into a number of themes, were used by the SMEs to identify needed icing technical elements. Each of these high-level technical elements included the icing physics, modeling capabilities, and experimental capabilities needed to address a particular element. A multivariable analysis technique called TOPSIS (Technique for Order Preference by Similarity to Ideal Solutions) was then applied to prioritize the technical elements for several potential future realities. A large number of icing research needs were identified and analyzed to determine the highest priorities for NASA key focus areas and those that will endure into the future. The primary outcome of the AoA is the identification of legacy and newly identified capability areas in priority order to guide future investment.

This paper evaluates simulation predictions against experimental test data of icing clouds that were produced during 2018 ice crystal icing physics tests conducted at the NASA Propulsion Systems Laboratory icing wind tunnel. Aero-thermal and cloud parameters are set and known upstream at the tunnel inlet and spray system, but change as the cloud and air thermodynamically interact as the flowing masses reach the tunnel test section. Utilizing the ANSYS Fluent Discrete Phase Model function, 3D computational fluid dynamics (CFD) simulations were performed, capturing the thermodynamic interactions between the test parameters, and providing predictions of the aero thermal and cloud conditions at the tunnel test section. Simulation predictions were compared with test data that were measured at the tunnel exit plane. Evaluations focused on the cloud concentration (total water content), humidity content, and air temperature. CFD simulation predictions showed areas of agreement and disagreement. Simulations showed that cloud concentration profiles at the test section are strongly related to the initial spray nozzle pattern used at the tunnel inlet. Experimental data suggest that greater dispersion of the cloud occurred as the simulated cloud predicted areas of high and low cloud concentration compared to test data profiles. Simulations, however, captured the magnitude and location of the change in humidity content and the change in air temperature due to the presence of the cloud reasonably well, when compared to test data. This result would suggest that while the ANSYS simulation did not fully predict the spreading of the cloud as measured during experiment, it did capture evaporation and the molecular movements of air and vapor relatively well.

In ice-crystal icing, ice erosion can occur simultaneously with ice accretion. This paper presents a proof-of-concept experimental study that attempts to isolate and quantify ice erosion. The experiments took place at the NASA Icing Research Tunnel and Propulsion Systems Laboratory in 2017 and 2018. The test procedure involved first generating an ice accretion on various NACA 0012 airfoil models using a supercooled liquid water or mixed-phase icing cloud and, subsequently, exposing that ice accretion to an ice crystal cloud. The resulting ice mass decay rates were analyzed in terms of an erosion term, defined as the ratio of the eroded mass flux to impinging mass flux. The erosion term generally increased with airspeed with a value of 1.8% at air velocities of 270 knots and increasing to just over 6% at 360 knots. A preliminary correlation for ice erosion is developed for use in ice accretion models subject to the conditions tested. The study revealed several challenges, both experimental and analytical, including spray bar nozzle freeze-out as well as measurement of the cloud melt ratio to ensure freeze-out.

This paper provides a characterization of the temperature and humidity variations within icing cloud flows generated during the second Fundamental Ice Crystal Icing Physics test conducted in NASA’s Propulsion Systems Laboratory, in June of 2018. The tests were conducted at nominal airflow velocities of 85, 135, and 185 m/s, a static pressure of 44.8 kPa, and a total temperature of 7.2 ⁰C. A custom rearward facing probe and commercial total air temperature probe were used to obtain temperature and humidity profiles of the icing cloud flow leaving the exit duct plane. Additional background humidity measurements were also obtained from a commercial isokinetic probe that measured total water content. The probes captured the changes in temperature and humidity that took place at the test section after the cloud was activated. As the main altitude airflow interacts with the icing cloud, their thermodynamic properties undergo substantial changes. Due to the particle nature of the icing cloud, the droplets and ice crystals tend to redistribute within the cloud as it propagates with the main flow. Data will be presented that reflects how these factors affected the thermodynamic property distribution within the icing cloud. A discussion is given of the thermal balances in the icing cloud as a basis for interpreting the measured results.

This paper presents water content and particle measurements from the second Fundamental Ice Crystal Icing Physics test conducted in the NASA Propulsion Systems Laboratory. The water content measurements came from the Science Engineering Associates Multi-Element Probe and the Isokinetic Probe whereas the Artium Technologies, Inc. Phase Doppler Interferometer and High Speed Imaging instruments measured the particle size distribution. Ice particle accretion due to the ingestion of ice-crystals is attributed to numerous jet engine powerloss events. The NASA PSL is an altitude jet-engine test facility which has recently added a capability to inject ice particles into the flow. This work focuses on characterizing the icing cloud by traversing the instruments across the test section at four different conditions where velocity and water content were varied. For this test, the nozzle patterns were selected to make a uniform cloud in the center 6-inch diameter of the test section. Radial trends for the ice and water content measurements are presented and the melt ratio is calculated at the measurement locations. Particle size distributions are also presented. The measured water content, melt ratio, and particle size decreased moving out radially. Measured water content trends show uniformity to within 19.2% for the center 6 inches of the cloud, as intended. In addition, the total water content increased at centerline with velocity, whereas the melt ratio decreased.

Numerical simulations of fluid flow and collection efficiency for a Science Engineering Associates (SEA) Robust probe are presented. The cases cover a range of freestream velocity from 10-135 m/s, and particle sizes from 1-200 µm. This combination produces results for Stokes numbers in the range of 0.07-1092. A blended modified Stokes number (defined in the paper) is introduced which covers a range of 0.07-146. Of particular interest is the collection efficiency of the sensing element (a forward-facing half-pipe), Etot, in the low Stokes number regime. It was observed that the collection efficiency of the sensing area approached one for large Stokes numbers (>100) as expected. As blended modified Stokes number approached one, the Etot dropped to approximately 0.5. By blended modified Stokes numbers of 0.3, the Etot has dropped to essentially zero. Plotting collection efficiency versus blended modified Stokes number, the different freestream velocities collapsed quite well on to a single curve. The paper presents a discussion on the flowfield and local collection efficiency, beta, in an effort to explain the predictions. Of particular interest is what is essentially a shadow zone in plain sight. As the Stokes number is reduced, the local beta along the centerline of the half-pipe goes to zero. This occurs at Stokes numbers where the overall collection efficiency, Etot, is still appreciable. An empirical relation was developed to allow for corrected collection efficiency values at low Stokes numbers. The correlation remains within 2% for blended modified Stokes numbers above 1, and within 6% for all cases.

This paper presents the evaluation of a thermodynamic ice crystal icing model, previously presented to describe the possible mechanisms of icing within the core of a turbofan jet engine. It has been proposed that there are two types of distinct ice accretions based on a surface energy balance: freeze-dominated icing and melt-dominated icing. In the former, ice accretion occurs where a freeze fraction (0 to 1) of melted ice crystals freezes on a surface, along with the existing ice of the impinging water and ice mass. This freeze-dominated icing is characterized by having strong adhesion to the surface. In the latter, icing occurs from accumulated unmelted ice on a surface, where a melt fraction (0 to 1) dictates the amount of unmelted impinged ice. This melt-dominated icing is characterized by weakly bonded surface adhesion. The experimentally observed ice growth rates suggest that only a small fraction of the impinging ice remains on the surface, implying a mass loss mechanism such as splash, runback, bounce, or erosion. This mass loss parameter must be determined in conjunction with the fraction of freezing liquid water or fraction of melting ice on an icing surface. This loss parameter, however, along with the freeze and melt fraction, are the only experimental parameters that are currently not measured directly. Using reported icing growth rates from published ice crystal icing experiments, a methodology is proposed to determine these unknown parameters. This work takes reported ice accretion data from tests conducted by the National Aeronautics and Space Administration (NASA) in 2016 and tests NASA collaborated on with the National Research Council (NRC) of Canada in 2012 that examined the fundamental physics of ice crystal icing. Those research efforts sought to generate icing conditions representative of those that occur inside a jet engine when ingesting ice crystals. This paper presents the fundamental equations of the thermodynamic model, the methodology used to determine the aforementioned unknown icing parameters, and results from model evaluation using experimental data. In addition, this paper builds on the previously proposed model by adding a transient conduction term to explain ice growth behavior at the onset of experimental tests that was observed to be different from steady-state ice growth that occurred later in the test run. With the addition of this energy term, this becomes a quasi-steady model. A key finding from this work suggests that mass loss fractions can exceed 0.90 for steady ice growth periods. In addition, due to conductive heat fluxes when using a warmer-than-freezing airfoil, lower mass loss fraction values were calculated during the initial transient period.

This paper describes an ice-crystal icing experiment conducted at the NASA Propulsion System Laboratory during June 2018. This test produced ice shape data on an airfoil for different test conditions similar to those inside the compressor region of a turbo-fan jet engine. Mixed-phase icing conditions were generated by partially freezing out a water spray using the relative humidity of flow as the primary parameter to control freeze-out. The paper presents the ice shape data and associated conditions which include pressure, velocity, temperature, humidity, total water content, melt ratio, and particle size distribution. The test featured a new instrument traversing system which allowed surveys of the flow and cloud. The purpose of this work was to provide experimental ice shape data and associated conditions to help develop and validate ice-crystal icing accretion models. The results support previous experimental observations of a minimum melt-ratio threshold for accretion to occur as well as the existence of a plateau region where the icing severity is high for a range of melt ratios. However, a maximum limit for melt ratio, which is suggested in the ice crystal icing literature, was not observed perhaps complicated by the potential for some supercooling of the water at these conditions.

This paper reports on temperature and humidity measurements from a series of ice-crystal icing tunnel experiments conducted in June 2018 at the Propulsion Systems Laboratory at the NASA Glenn Research Center. The tests were fundamental in nature and were aimed at investigating the icing processes on a two-dimensional NACA0012 airfoil subjected to artificially generated icing clouds. Prior to the tests on the airfoil, a suite of instruments, including total temperature and humidity probes, were used to characterize the thermodynamic flow and icing cloud conditions of the facility. Two different total temperature probes were used in these tests which included a custom designed rearward facing probe and a commercial self-heating total temperature probe. The rearward facing probe, the main total temperature probe, was designed to reduce and mitigate the contaminating effects of icing and ingestion of ice crystals and water droplets at the probe's inlet. The probe also serves as an air-sample inlet for a light absorption based humidity measurement. The paper includes a section which discusses total temperature and humidity measurement considerations, and another section which provides an analysis of the main probe's performance characteristics. A computational fluid dynamic model of the flow around the probe was also conducted to gain insight into the trajectory of the flow entering the probe inlet. The experiments included a series of tests in which the relative humidity of the facility flow was swept through with increasingly larger values. The data showed that the rearward facing probe can reasonably capture the flow's total temperature and humidity under mild to moderate icing conditions but produces anomalous results under more intense icing conditions. The experimental data was also compared to an in-house developed thermodynamic model which takes into account the interaction of the main flow with the icing cloud. Comparison to the thermodynamic model showed that the rearward facing probe measured the predicted trends.

Engine Icing Performance loss: rollback, surge, flameout, and even internal engine damagePartial melting and refreeze of ice inside engine core (Mason et al., 2006). Ingestion of ice crystals and aggregates, mixed-phase droplets, or supercooled liquid dropletsNeed to better understand the conditions and properties that lead to engine icing.Simulation and analysis (physical and computational, and modeling)Test facilities (PSL, NRC, ...). Thermal and computational models and analysisProbesMultiple probes (aerothermal probes and ice cloud characterization probes and techniques). Total temperatureTraditional total temperature probes (vented forward facing)Heated total temperature probes (Goodyear). Rearward facing (developmental). Total temperature relevance. Thermal interaction between the icing cloud and air flow impinging particles contribute to kinetic heating effect (Gent et al., 2000). Measurement considerations Temperature sensor accuracy. Incomplete recovery of total temperature. Thermal surfaces (sources and sinks). Flow effects (viscous losses). Debris contamination, including icing and ice ingestion.

This paper presents an analysis of local total temperature and humidity experimental measurement taken in atmospheric ice cloud flows. The measurements were obtained in a series of tests in NASA's Propulsion Systems Laboratory. The probe used in the tests is referred to as the Rearward Facing Probe which was designed to mitigate the contamination effects of ice accretion and ingestion into the probe. The data provided important insights in the interaction of the ice cloud and the atmospheric flow. For the majority of the test runs, small temperature drops in the range of 0.6 to 2.8 C and up to 1.5 g/kg of water vapor rise were found as a result of the interaction. Under certain very low temperature or high TWC conditions, the interaction with the cloud produced a warming of the airflow. A thermal model based on evaporative and convective heat transfer mechanisms between the spray droplets and the airflow showed good agreement with the experimental data. Detailed analyses of the response of the probe under various flow, thermodynamic, and cloud conditions, are provided in the paper.

The particle size characterization portion of the 2017 Propulsion Systems Laboratory Cloud Calibration is described. The work focuses on characterizing the particle size distribution of the icing cloud as a function of simulated atmospheric conditions. These results will aid in upcoming ice crystal and supercooled liquid icing tests in PSL. Measurements acquired with the Phase Doppler Interferometer and High Speed Imaging instruments are presented. Experimental results indicate that the particle size distribution is primarily a function nozzle air and water pressures, and that air speed is not a significant effect for ice crystal clouds in PSL and both thermodynamic conditions and air speed are not significant effects for supercooled liquid water clouds in PSL.

This paper presents results from a study of the fundamental physics of ice-crystal ice accretion using a NACA 0012 airfoil at the National Research Council of Canada (NRC) Research Altitude Test Facility in August 2017. These tests were a continuation of work which began in 2010 as part of a joint collaboration between NASA and NRC. The research seeks to generate icing conditions representative of those that occur inside a jet engine when ingesting ice crystals. In this test, an airfoil was exposed to mixed-phase icing conditions and the resulting ice accretions were recorded and analyzed. This paper details the specific objectives, procedures, and measurements which included the aero-thermal and cloud measurements. The objectives were built upon observations and hypothesis generated from several previous test campaigns regarding mixed-phase ice-crystal icing. The specific objectives included (A) ice accretions under different wet-bulb temperatures, (B) investigations of steady-state ice shapes previously reported in the literature, (C) total water content variations in search of a threshold for accretion, and (D) probe characterization related to measuring melt fraction which is important to characterize the mixed-phase condition. The resulting ice accretions and conditions leading to such accretions are intended to help extend NASA’s predictive ice-accretion codes to include conditions occurring in engine ice-crystal icing.

This paper describes plans and preliminary results for using the NASA Propulsion Systems Lab (PSL) to experimentally study the fundamental physics of ice-crystal ice accretion. NASA is evaluating whether this facility, in addition to full-engine and motor-driven-rig tests, can be used for more fundamental ice-accretion studies that simulate the different mixed-phase icing conditions along the core flow passage of a turbo-fan engine compressor. The data from such fundamental accretion tests will be used to help develop and validate models of the accretion process. This paper presents data from some preliminary testing performed in May 2015 which examined how a mixed-phase cloud could be generated at PSL using evaporative cooling in a warmer-than-freezing environment.

This paper builds on previous work that compares numerical simulations of mixed-phase icing clouds with experimental data. The model couples the thermal interaction between ice particles and water droplets of the icing cloud with the flowing air of an icing wind tunnel for simulation of NASA Glenn Research Centers (GRC) Propulsion Systems Laboratory (PSL). Measurements were taken during the Fundamentals of Ice Crystal Icing Physics Tests at the PSL tunnel in March 2016. The tests simulated ice-crystal and mixed-phase icing that relate to ice accretions within turbofan engines.

This paper will describe two recent modifications to the Glenn ICE software. First, a capability for modeling ice crystals and mixed phase icing has been modified based on recent experimental data. Modifications have been made to the ice particle bouncing and erosion model. This capability has been added as part of a larger effort to model ice crystal ingestion in aircraft engines. Comparisons have been made to ice crystal ice accretions performed in the NRC Research Altitude Test Facility (RATFac). Second, modifications were made to the run back model based on data and observations from thermal scaling tests performed in the NRC Altitude Icing Tunnel.

This paper describes ice accretion measurements from experiments conducted at the National Research Council (NRC) of Canada's Research Altitude Test Facility during 2012. Due to numerous engine power loss events associated with high altitude convective weather, potential ice accretion within an engine due to ice crystal ingestion is being investigated collaboratively by NASA and NRC. These investigations examine the physical mechanisms of ice accretion on surfaces exposed to ice crystal and mixed phase conditions, similar to those believed to exist in core compressor regions of jet engines. A further objective of these tests is to examine scaling effects since altitude appears to play a key role in this icing process.

This paper describes a numerical model that simulates the thermal interaction between ice particles, water droplets, and the flowing air applicable during icing wind tunnel tests where there is significant phase-change of the cloud. It has been previously observed that test conditions, most notably temperature and humidity, change when the icing cloud is activated. It is hypothesized that the ice particles and water droplets thermally interact with the flowing air causing the air temperature and humidity to change by the time it reaches the test section. Unlike previous models where the air and particles are uncoupled, this model attempts to explain the observed changes in test conditions by coupling the conservation of mass and energy equations. The model is compared to measurements taken during wind tunnel tests simulating ice-crystal and mixed-phase icing that relate to ice accretions within turbofan engines. The model simulates trends that were experimentally observed, but does not fully explain the measured values. Some possible explanations for this discrepancy are offered. This model, written in MATLAB, is based on fundamental conservation laws and empirical correlations.