Document
2015-01-XXXX
I ce Particle Analysis of the Honeywell ALF502 Engine Booster
Author, co-author (Do NOT enter this information. It will be pulled from participant tab in
MyTechZone)
Affiliation (Do NOT enter this information. It will be pulled from participant tab in MyTechZone) rollbacks and in one case a dead stick landing. The current research
Abstract
includes flight testing, instrument development, ground based facility development and testing and modeling of the ice accretion process in A flow and ice particle trajectory analysis was performed for the turbo-machinery.
booster of the Honeywell ALF502 engine. The analysis focused on two closely related conditions one of which produced an icing event Recently a set of HIWC tests on the Honeywell ALF502 engine were and another which did not during testing of the ALF502 engine in the conducted in the newly completed Ice Crystal Engine Test Chamber Propulsion Systems Lab (PSL) at NASA Glenn Research Center. The at the Propulsion Systems Laboratory (PSL) at NASA Glenn flow analysis was generated using the NASA Glenn GlennHT flow Research Center [3]. The tests were able to repeat icing events from solver and the particle analysis was generated using the NASA Glenn flight tests of the engine. The data from these tests is being used to LEWICE3D v3.63 ice accretion software. The inflow conditions for develop and verify computational models with varying levels of the two conditions were similar with the main differences being that fidelity. One topic of interest is determining the conditions that the condition that produced the icing event was 6.8 K colder than the produce icing events and those that do not.
non-icing event case and the inflow ice water content (IWC) for the non-icing event case was 50% less than for the icing event case. The Of particular interest from the recent tests are a pair of closely related particle analysis, which considered sublimation, evaporation and test points one of which produced an icing event while the other did phase change, was generated for a 5 micron ice particle with a sticky not. To further understand the difference between these two impact model and for a 24 micron median volume diameter (MVD), conditions a high fidelity flow and icing analysis of the ALF502 low 7 bin ice particle distribution with a supercooled large droplet (SLD) pressure compressor was conducted using the NASA Glenn GlennHT splash model used to simulate ice particle breakup. The results from flow solver [4] and the NASA Glenn LEWICE3D ice accretion the analysis showed that the amount of impingement for the software [5]. These tools use a 3D, steady mixing plane approach to components were similar for the same particle size and impact model analyze flow and icing in turbomachinery.
for the icing and non-icing event conditions. This was attributed to the similar aerodynamic conditions in the booster for the two cases.
The particle temperature and melt fraction were higher at the same
Numerical Method
location and particle size for the non-icing event than for the icing event case due to the higher incoming inflow temperature for the Grid and Flow Calculations non-event case. The 5 micron ice particle case produced higher impact temperatures and higher melt fractions on the components downstream of the fan than the 24 micron MVD case because the The GridPro grid generation software was used to develop the three- dimensional grids for the geometry [6] and the GlennHT flow solver average particle size generated by the particle breakup was larger than 5 microns which yielded less warming and melting. The analysis [4] was used to generate the flow solutions for the analysis. The also showed that the melt fraction and wet bulb temperature icing GlennHT code is a three-dimensional, finite volume based, Reynolds Averaged Navier-Stokes flow solver. The code computes flow on criterion developed during tests in the Research Altitude Test Facility (RATFac) at the National Research Council (NRC) of Canada were complex propulsion system configurations using multi-block body useful in predicting icing events in the ALF502 engine. The fitted grids. The method employs a “mixing - plane” procedure to pass boundary condition data between grid blocks for the steady state flow development of an ice particle impact model which includes the effects of particle breakup, phase change, and surface state is analysis of turbomachinery. The code supports parallel computing necessary to further improve the prediction of ice particle transport and supports several turbulence models.
with phase change through turbomachinery.
Particle Transport Calculations
Introduction
The LEWICE3D V3.63 grid based icing tool [5,7], which incorporates droplet trajectory, heat transfer and ice shape calculation A large amount of research is currently being conducted to quantify, model and simulate the High Ice Water Content (HIWC) threat [1,2]. into a single computer program, was used for the particle transport analysis. This program has several features which allow the analysis The HIWC environment, which contains large ice crystals (> 100 of turbomachinery subject to HIWC or SLD environments. These microns) in large concentrations (> 2 g/m ) at high altitudes (~40,000ft), has been responsible for over 150 incidents including features include a particle splash and bounce algorithm, a geometry Page 1 of 15 handling scheme which allows complex mirroring, transformation particle flux rate is greater than the free stream level. The average and relative motion of input grid blocks and an algorithm which collection efficiency is defined as calculates zone to zone collection efficiencies using a mixing plane 𝑛=𝑁 approach.
∑ 𝛽 × 𝐴 𝑛 𝑛 𝑛=1 𝛽 = (1) 𝑎𝑣𝑔 𝐴 𝑤𝑒𝑡𝑡𝑒𝑑
Results
Where, N, is the number of surface elements with nonzero The ALF502 analysis included the calculation of flow and ice particle impingement and β , A are the collection efficiency and area of n n transport properties. The particle analysis results are presented for a 5 surface element, n, respectively. The wetted area of the element is the micron particle using a sticky impact model and a 24 micron MVD, 7 sum of the area of the elements which have non-zero impingement bin distribution using an SLD splash model to simulate ice particle for which we have the equation; breakup. The particle analysis was carried out for two conditions one of which generated an icing event and the other which did not (PSL 𝑛=𝑁 test point DP0443 and DP0256 respectively). The inflow conditions 𝐴 = ∑ 𝐴 (2) 𝑤𝑒𝑡𝑡𝑒𝑑 𝑛 for the two conditions were similar with the main differences being 𝑛=1 that the condition that produced the icing event (DP0443) was 6.8 K colder than the non-icing event case (DP0256) and the inflow ice The impingement rate for a surface is defined as: water content (IWC) for the non-icing event case was 50% less than for the icing event case.
𝐼𝑅 = 𝛽 × 𝐿𝑊𝐶 × 𝑉 × 𝐴 (3) 𝑎𝑣𝑔 ∞ ∞ 𝑤𝑒𝑡𝑡𝑒𝑑 The grid and surface model used for the flow and particle analysis is shown in Figures 1-2. The grid contained 33 structured, abutted grid Where, 𝐿𝑊𝐶 , is the free stream liquid water content and, ∞ blocks with a total of 1,596,897 nodes. Steady, inviscid flow 𝑉 , is the free stream speed. The free stream catch fraction or ∞ solutions were generated for DP0256 and DP0443 test conditions.
scoop factor (SF) is defined as the ratio of the mass impinging Figure 3 depicts the elements of interest for the compressor. The flow on a component divided by the mass available in the free analysis for the DP0256 and DP0433 cases are documented in a stream for an area equal to the area bounded by the highlight companion paper[8].
of the inlet lip. The scoop factor is then; The LEWICE3D ice particle analysis required several cloud input 𝐼𝑅 conditions and modeling parameters. The ice particle analysis 𝑆𝐹 = (4) assumed an inflow relative humidity of 100% and a particle 𝐼𝑅 ∞ 3 3 concentration (IWC) of 2.0 g/m for case DP0443 and 1.0 g/m for case DP0256. The ice particles were assumed to be completely frozen The free stream impingement rate, 𝐼𝑅 , is defined as the rate ∞ and at the ambient temperature of the surrounding air at the inflow at which the particles pass through an area traced out by the boundary. Two particle conditions were simulated in the analysis for highlight of the inlet lip ( 𝐴 ) traveling at the free stream each flow condition. A 5 micron particle size with a sticky impact ∞ speed ( 𝑉 ) with an average collection efficiency of 1 and an model was chosen because it has been successful in predicting icing ∞ risk in simpler, lower fidelity analysis where it was used to represent LWC matching that of the free stream ( 𝐿𝑊𝐶 ). The average ∞ the ice particles resulting from the particle breakup in the fan. A 24 collection efficiency for a surface is then: micron MVD, 7 bin particle distribution was chosen to match the cloud generated by the PSL spray system. An SLD splash model was 𝐼𝑅 × 𝐴 𝐴 ∞ ∞ 𝛽 = = 𝑆𝐹 × (5) chosen for the 24 micron MVD distribution to simulate the ice 𝑎𝑣𝑔 𝐼𝑅 × 𝐴 𝐴 ∞ 𝑤𝑒𝑡𝑡𝑒𝑑 𝑤𝑒𝑡𝑡𝑒𝑑 particle breakup because a model was not available and it was thought that the breakup characteristics of an ice particle were similar to that of a similarly sized water droplet. Although the SLD splash The collection efficiency results for case DP0256 and case DP0443 model generates more impingement on a surface than an ice particle for both particle sizes are shown in Figure 4. The results show that impact model due to the stickier nature of liquid water versus ice it is the collection efficiency on the spinner and fan are larger for the 24 useful because it can give an indication of the location and state of micron MVD, SLD cases than for the 5 micron cases due to the larger the impacting particles which is useful for assessing regions which inertia of the larger particles. The collection efficiency on the are at risk for icing. The ice particle calculations were made from the components downstream of the fan are smaller for the larger inflow boundary through the compressor and out the compressor exit particles. The reason for the smaller collection efficiency for the boundary.
larger particle cases is threefold. First, the fan removed more mass for the larger particles making less mass available for the downstream It is worthwhile to report the definitions and equations used for the components. This can be seen in Tables 1-4 and Figure 5 where we particle analysis. These include collection efficiency or impingement can see that the scoop factors are much greater for the larger efficiency ( β ), average collection efficiency ( β ), impingement rate ave particles. Second, the larger particles are less able to negotiate the ( IR ), and scoop factor ( SF ). Impingement efficiency is a non- flow into the inner core due to the flow curvature. This can be seen in dimensional measure of the mass flux for a surface and is dependent the Figure 6 which shows axial mass flux for particles through the upon the amount of convergence or dispersion of particles in a flow compressor. Thirdly, the reduction in collection efficiency on the and the orientation of the surface relative to the particle paths. An components downstream of the fan is due in part to the reduction in impingement efficiency of one means the surface particle flux rate is average particle size from the particle breakup for the larger SLD equal to the free stream particle flux rate. A value less than one splash model cases. Figure 7 shows that the average particle size for means the surface particle flux rate is less than the free stream the cases decreases as the particles pass through the compressor. At particle flux rate and a value greater than one means that the surface Page 2 of 15 the fan exit the average particle size for the 24 micron MVD, SLD particle becomes fully melted as can be seen in Figure 11 for the 5 splash model cases is approximately 18 microns. At the inflow to micron cases.
IGV #1 the average particle size is approximately 9 microns.
The melt fraction results are shown in Figures 12,13. From the The average particle size for all cases decreased through the low figures one can see that there is no appreciable melting for any of the pressure compressor. For the 5 micron cases, which did not involve a cases until the outflow of rotor #1. The melt fractions are higher for breakup model, the reduction in particle size was small and was the the warmer cases at the same particle size. The average melt fractions result of sublimation and evaporation. The average particle size at the compressor exit are also higher for the 5 micron mono-disperse change for the 24 micron MVD, SLD splash model cases, which cases than for the SLD splash model cases because the average were predominantly due to particle breakup, were much larger than particle size is larger for the SLD splash model cases. The average for the 5 micron cases. From Figure 8 we can see the mass loss due to melt fraction for the DP0443, SLD splash model case was 0.68 while evaporation and sublimation was greater for the warmer cases the particles for the other three cases were fully melted at the duct (DP0256) and was largest for the 24 micron MVD, SLD splash exit. From the contour plots of the melt fraction one can see a large model case (28%). The 24 MVD micron, DP0256 SLD splash model increase in melt fraction below the mid span location for EGV #2 and breakup case produced smaller particles which were subject to more aft of the trailing edge of the EGV #2 on the outer duct wall. This is sublimation and evaporation than the 5 micron DP0256 case. This is due to the impact of much smaller particles which are more readily because particle sublimation and evaporation increases with warmed and melted than the larger particles in the surrounding increased temperature and with the increased surface area of the regions (Fig. 14).
smaller sized particle cloud. The particle breakup resulted in an approximate reduction of 33% in the average particle size at the Although the calculated transport data does not produce ice shape inflow to IGV #1 and approximately 25% at the exit of the transition predictions it can be examined using icing sensitivity parameters duct (Fig. 7). The majority of the reduction in particle size due to developed previously to better understand the potential for ice growth breakup occurs in the fan and the entrance to the inner core. This is and significant performance losses in the compressor. Previously because the SLD splash model breakup model generates less ejected researchers have isolated two parameters which are useful in mass and larger ejected particles as the impacting particle size is assessing the potential for an icing event [9-12]. These are the melt reduced. It is also due in part to the lower impingement rates and fraction and the wet bulb temperature. It is known that ice crystal impact speeds for the smaller particles which more readily adapt to icing requires some amount of water for the ice to adhere and that the changes in the flow about the engine components. The particle wet bulb temperature of the flow be less than several degrees above distributions at various axial locations in the compressor are shown freezing for the ice to grow. If the particles are 100% ice they bounce for the SLD splash model cases in Figure 9. From the figure we can or breakup and are ejected from the surface. If the wet bulb see that there are some differences in the particle distributions for the temperature is too much above freezing the convective heat load is intermediate stages but that the distributions at the fan outflow and at too high and ice growth cannot be sustained. Tests of several the duct exit are similar for both the DP0256 and DP0443, 24 MVD geometries in the NRC RATFac showed that ice build-up was micron, SLD splash model cases. The particle breakup results in a produced for a range of melt fractions of 0.05-0.32 and at wet bulb o reduction in the average particle size along with a reduction in the temperatures below 5.5 C[11]. From Figure 15 we can see that for minimum and maximum particle size in the distribution. For both the warmer DP0256 cases that the range of melting fractions between cases the maximum particle size in the distribution was reduced from 0.05-0.32 occurs upstream of EGV #2. The wet bulb temperature is 67 microns to 12 microns while the minimum particle was reduced to above 278.65 K in this range which would indicate that ice buildup 2 microns from 7 microns.
should not occur. For the DP0443 case we have a wet bulb temperature of less than 278.65 K in the region where the melt In general the mass transport properties through the compressor were fraction is between 0.05-0.32 indicating that icing is possible.
similar for case DP0256 and case DP0443 for the same particle size.
This is due to the similarity in the flow conditions for the two cases.
Although the average values of particle temperature and melt The mass transport properties of the particles are predominantly fractions are useful in assessing relative risk they do not give dependent on the flow velocity and density and particle size which information as to the location of the icing risk. An examination of can be characterized by the modified inertia parameter. For both flow the local values of surface temperature, pressure, velocity and particle cases the rotational speed was the same and the inlet pressure, impact concentration, temperature and melt fraction are required to temperature and velocity were similar which yielded a similar flow.
deduce this. If we look at the particle impact melt fraction and The modified inertia parameter for the 5 micron particle was 0.01773 temperature surface contour plots for the DP0443, 24 micron MVD, for case DP0256 and 0.01747 for case DP0443. For the 24 micron SLD splash model case (Fig. 10,12) we see that the conditions near MVD, SLD splash model cases, the modified inertia parameters at the intersection of EGV #2 and the outer duct wall indicate ice the MVD drop size of the distribution (24 microns) were 0.02181 and accretion based on the above criterion. If we look at the same region 0.02149 for DP0256 and DP0443 respectively.
for the warmer DP0226, SLD splash model case we see that the melt fraction is too high (> .4) indicating that icing would probably not The average particle impact temperatures for the low pressure occur.
compressor are shown in Figure 10. The particle temperature increases as it transits the warming environment of the compressor.
The present analysis although useful in understanding the relative In all cases the average particle temperature lags the local static effect of particle breakup and phase change falls short as an accurate temperature and the lag in temperature increases with increased assessment of the ice particle transport with phase change through particle size due to the increased thermal mass of the larger particles.
booster. A more accurate analysis requires the development of an ice The DP0256 cases produced higher final particle temperatures than particle impact model which includes the effects of particle breakup, those for DP0443 due to the higher inflow temperature. The particle phase change, and surface state.
temperature approaches the wet bulb temperature of the flow as the Page 3 of 15 7. Bidwell, C., “A Lagrangian Parcel Based Mixing Plane Method
Conclusions
for Calculating Water Based Mixed Phase Particle Flows in Turbo- machinery,” DOI 10.1007/s40571 -015-0033-z, Journal of The GlennHT flow solver and the LEWICE3D icing software were Computational Particle Mechanics, February, 2015.
used to analyze the ALF502 low speed compressor for two closely 8. Rigby, D., Bidwell, C., ”Three Dimensional Navier -Stokes related conditions one of which generated an icing event and one Simulation of Flow in an Axial Low Pressure Compressor at which did not during testing in the PSL engine icing test facility. The Engine Icing O perating Conditions”, SAE 15ICE -0139, June main differences between the two cases were the inflow temperature 2015.
which was 6.8 K cooler for the condition that generated the icing 9. Tsao, J., Struk, P., Oliver, M., “Possible Mechanisms for event and the inflow IWC which was 50% lower for the non-icing Turbofan Engine Crystal Icing at High Altitude,” AIAA 2014 - event case. The collection efficiency results at the same particle size 3044, 2014.
for the two cases were similar due to the similarity in the 10. Jorgenson, P., Veres, J., “Modeling Commercial Turbofan aerodynamic conditions. The collection efficiency results showed that Engine Icing Risk With Ice Crystal Ingestion,” AIAA 2013 - the collection efficiencies were larger for the larger SLD splash 2679, 2013.
model cases upstream of the splitter lip but smaller for the SLD 11. Wr ight, W., Jorgenson, P., Veres, J., “Mixed Phase Modeling in splash model cases downstream of the splitter lip due to the removal GlennICE with Application to Engine Icing,” AIAA 2010 -7674, of mass by the fan and spinner, the reduction in particle size caused 2010.
by impact with the fan and spinner and due to the flow curvature 12. Currie, T., Fuleki, D., Mahallati, A., “Experimental Studies of limiting the amount of larger particles transiting into the core. The Mixed-Phase Sticking Efficiency for Ice Crystal Accretion in Jet particle breakup generated a reduction in the average particle size of Engines,” AIAA 2014-3049, 2014.
approximately 33% for the larger SLD splash model cases. The majority of the breakup (25%) occurred in the spinner, fan and splitter lip regions. The thermal analysis showed that the larger particles warmed less and produced less melting than the smaller particles due to their increased thermal mass. The larger SLD splash model cases produced more mass loss than the smaller particle cases because they produced a large amount of smaller particles during impact (< 5 microns) which were more susceptible to sublimation and evaporation. The icing risk criterion developed during the NRC tests for melt fraction (0.05 > melt fraction < .32) and wet bulb o temperatures (< 5.5 C) was useful in predicting the icing risk for the ALF502 low pressure compressor for the two test points selected for this analysis. The criterion showed that the icing event case (DP0443) was susceptible to icing in the EGV #2-outer duct intersection region and that the particle melt fractions and temperatures were too high to generate icing for the warmer non- icing event (DP0226). These results show that the GlennHT/LEWICE3D steady, mixing plane approach can be useful for predicting icing risk in turbomachinery. The development of an ice particle impact model which includes the effects of particle breakup, phase change, and surface state is necessary to further improve the utility of these tools in the prediction of ice particle transport with phase change through turbomachinery.
References
1. M ason, J; Strapp, W and Chow, P; “The Ice Particle Threat to Engines in Flight”, 44th AIAA Aerospace Scienc es Meeting, v4, 2006, pp2445-2465.
2. M azzawy R.S., Strapp J.W.; “Appendix D – An Interim Icing Envelope; SAE 2007-01-3311; SAE 2007 Aircraft and Engine Icing International Conference ; Seville, Spain, November 2007.
3. Oliver, M., “Validation Ice Crystal Icing Engine Test in the Propulsion Systems Laboratory at NASA Glenn Research Center,” AIAA 2014 -2898, 2014.
4. Steinth orsson, E., Liou, M., and Povinelli, L., ” Development of an Explicit Multiblock/Multigrid Flow Solver for Viscous Flows in Complex Geometries ,” AIAA‒93‒2380 (NASA TM‒106356), 1993.
5. Bidwell , C., Potapczuk, M., “Users Manual for the NASA Lewis Three-Dimensi onal Ice Accretion Code (LEWICE3D),” NASA TM-105974, December 1993.
6. Program Development Corporation. , “ GridPro GUI Manual Version 1.0,” September 24, 2014 .
Page 4 of 15 MVD median volume diameter, m
Nomenclature
2 SF scoop f actor A area, m SLD supercooled large droplet collection efficiency BETA V velocity, m/s D ice particle diameter, m impingement efficiency DIAMAVG particle diameter, m DPMFRAC particle melt fraction DPTEMP particle temperature, K exit guide vane EGV I GV inlet guide v an e IR i mpingement rate, g/s ice water content, g/m IWC LWC l iquid w ater c onte n t, g/m Page 5 of 15
Appendix
Table 1. Transport statistics for case DP0256, 5 micron mono-disperse, sticky impact.
Element Impingement Rate (g/s) Scoop Factor D ( m) T (K) Melt Fraction avg avg avg Inlet Capture 201.162 1.0000 5.00 260.10 0.000 Spinner 0.005 0.0001 4.95 267.70 0.000 Fan Blade 11.001 0.1094 4.99 267.83 0.000 Splitter Lip 0.412 0.0041 4.97 273.15 0.006 IGV #1 2.047 0.0204 4.96 273.15 0.018 Rotor #1 1.205 0.0120 4.93 273.15 0.086 EGV #1 0.528 0.0053 4.91 273.15 0.189 EGV #2 0.905 0.0090 4.86 273.15 0.468 EGV #2+EGV #3+wall 0.168 0.0017 4.88 273.15 0.401 Inner Core Exit 2.606 0.0259 4.58 286.87 1.000 Table 2. Transport statistics for case DP0443, 5 micron mono-disperse, sticky impact.
Element Impingement Rate (g/s) Scoop Factor D ( m) T (K) Melt Fraction avg avg avg Inlet Capture 191.528 1.0000 5.00 253.30 0.000 Spinner 0.009 0.0001 4.96 261.62 0.000 Fan Blade 23.872 0.1246 5.01 263.03 0.000 Splitter Lip 0.589 0.0031 5.00 268.75 0.000 IGV #1 4.026 0.0210 4.99 268.10 0.000 Rotor #1 2.558 0.0134 4.97 271.92 0.000 EGV #1 1.099 0.0057 4.96 273.15 0.011 EGV #2 1.869 0.0098 4.93 273.15 0.156 EGV #2+EGV #3+wall 0.357 0.0019 4.94 273.15 0.111 Inner Core Exit 5.397 0.0282 4.69 281.75 1.000 Table 3. Transport statistics for case DP0256, 24 micron MVD, 7 Bin, SLD splash model.
Element Impingement Rate (g/s) Scoop Factor D ( m) T (K) Melt Fraction avg avg avg Inlet Capture 201.162 1.0000 24.00 260.10 0.000 Spinner 4.178 0.0415 37.73 264.80 0.000 Fan Blade 35.521 0.3532 22.09 265.39 0.000 Splitter Lip 0.119 0.0012 12.54 271.60 0.000 IGV #1 0.249 0.0025 9.57 273.15 0.007 Rotor #1 0.173 0.0017 9.49 273.15 0.064 EGV #1 0.179 0.0018 11.28 273.15 0.088 EGV #2 0.061 0.0006 6.19 273.15 0.465 EGV #2+EGV #3+wall 0.086 0.0009 12.15 273.15 0.130 Inner Core Exit 0.177 0.0018 5.94 277.54 1.000 Table 4. Transport statistics for case DP0443, 24 micron MVD, 7 bin, SLD splash model.
Element Impingement Rate (g/s) Scoop Factor D ( m) T (K) Melt Fraction avg avg avg Inlet Capture 191.528 1.0000 24.00 253.30 0.000 Spinner 8.270 0.0432 37.97 257.86 0.000 Fan Blade 75.016 0.3917 23.07 258.89 0.000 Splitter Lip 0.145 0.0008 11.82 265.02 0.000 IGV #1 0.307 0.0016 9.27 266.03 0.000 Rotor #1 0.208 0.0011 9.33 269.36 0.000 EGV #1 0.195 0.0010 10.07 272.02 0.000 EGV #2 0.080 0.0004 7.06 273.15 0.134 EGV #2+EGV #3+wall 0.062 0.0003 10.98 273.15 0.045 Inner Core Exit 0.211 0.0011 6.84 273.15 0.678 Page 6 of 15 Figure 1. Grid structure.
Figure 2. Surface model.
Figure 3. Element description.
Page 7 of 15 a) DP0256, 5 micron b) DP0443, 5 micron c) DP0256, 24 micron MVD, SLD splash model d) DP0443, 24 micron MVD, SLD splash model e) DP0256, 5 micron f) DP0443, 5 micron g) DP0256, 24 micron MVD, SLD splash model h) DP0443, 24 Micron MVD, SLD splash model Figure 4. Collection efficiency for booster.
Page 8 of 15 Figure 5. Scoop factor for compressor components.
Figure 6. Mass flux rates for the compressor Figure 7. Average particle size for the compressor.
Page 9 of 15 Figure 8. Particle mass loss for the compressor due to sublimation and evaporation.
Figure 9. Compressor particle distributions for 24 Micron MVD, SLD splash model cases.
Page 10 of 15 a) DP0256, 5 micron b) DP0443, 5 microns c) DP256, 24 micron MVD, SLD splash model d) DP0443, 24 micron MVD, SLD splash model e) DP0256, 5 micron f) DP0443, 5 micron g) DP0256, 24 micron MVD, SLD splash model h) DP0443, 24 micron MVD, SLD splash model Figure 10. Particle impact temperature distributions for compressor.
Page 11 of 15 Figure 11. Flow and particle temperatures for compressor.
a) DP0256, 5 micron b) DP0443, 5 micron c) DP0256, 24 micron MVD, SLD splash model d) DP0443, 24 micron MVD, SLD splash model Figure 12. Particle impact melt fraction distributions for compressor.
Page 12 of 15 e) DP0256, 5 micron f) DP0443, 5 micron g) DP0256, 24 micron MVD, SLD splash model h) DP0443, 24 microns MVD, SLD splash model Figure 12 concluded. Particle impact melt fraction distributions for compressor.
Figure 13. Average particle impact melt fractions for compressor.
Page 13 of 15 a) Impact particle melt fraction b) Average impact particle diameter Figure 14. Compressor impact particle distributions for DP0256, 24 micron MVD, SLD splash model.
Page 14 of 15 Figure 15. Compressor wet bulb temperature and melt fraction distributions for 24 micron MVD, SLD splash model cases.
Page 15 of 15