Document
Icing Physics Stud ies U sing the 3D SIDRM Test Article: Aerodynamic and
Supercooled Liquid Icing Analysis
Tadas P. Bartkus
Ohio Aerospace Institute , Cleveland, OH, 44142, USA
Sam Lee
HX5, LLC, Cleveland, OH, 44135, USA
Eric A. Stewart
Naval Air Warfare Center Aircraft Division , Patuxent River, MD, 20670, USA during the S IDRM tests will be used to develop and validate 3D
Abstract
computational engine icing tools, such as GlennICE, that predictively assesses the onset and growth of ice. One of the goals of the sponsoring In - flight icing is an important safety issue and is a factor that affects NASA project is to develop simulation models and tools that ca n assist aircraft design and performance. Newer regulations are driving a need in the design and certification of engines for flight in icing conditions for improvements in airframe and engine icing simulation capability.
in a cost‑effective way.
Experimental data is required for development of icing physics models and simulation validation. To that end, this paper presents the analysis
of the supercooled liquid icing data subset from tests conducted in Introduction
2022 at the NASA Icing Research Tunnel that studied both supercooled water and ice - crystal icing. The test article that was Numerous reports of turbofan engine power - loss or damage events utilized replicated 3D geometrical features of an inter - compressor duct have been attributed to the ingestion of ice crystals [1]. These event s and strut region of a turbofan engine . The surfaces of the Simulated typically occur in deep convective updraft systems and have included Inter - compressor Duct Research Model (SIDRM) can be heated to engine stall, rollback, flameout, surge, and engine component damage.
simulate the warm surfaces of the turbofan inter - compressor duct. The Mason et al. [2] hypothesized that ice crystals ingested into the engine test article is instrumented with pressure taps, heaters, heat flux undergo partial melting within the warm co mpressor system and then, gauges, and thermocouples, while a 3D laser scanner, cameras, and a as a mixed - phase water mass, accrete on surfaces within the engine scale to measure ice mass were utilized to characterize the icing core. Research efforts in understanding the physics of ice crystal icing behavior. The aim of these tests was to generate ice accretions on the ha ve grown to address this threat of engine icing. The National SIDRM test article under well - characterized supercooled liquid icing Aeronautics and Space Admi nistration (NASA) Glenn Research and ice crystal icing conditions. This paper discusses measurements Center has been conducting experiments to better understand ice related to aerodynamic test ing and supercooled liquid icing tests that crystal icing. Multiple engine tests have been conducted at the NASA were conduc ted. Aerodynamic measurements were analyzed and Propulsion Systems Laboratory (PSL) icing wind tunnel since the compared to computational simulations and were found to be in good tunnel’s spray bar install ation [3 - 5]. In addition, experimental studies agreement for the range of airspeeds (50 to 230 knots) and angles of on the fundamental physics of ice crystal icing have been conducted at attack (0 to 4°) tested . Various parametric sweeps were conducted the NASA PSL tunnel [6 - 9].
during the supercooled liquid ic ing portion of the test entry (cloud median volumetric diameter ranged from 15 to 90 μm, total air The first two test entries in this series of fundamental ice crystal icing temperature from - 3 to - 17 °C, angle of attack from 0 to 4°, and research tests were conducted in th e NASA PSL tunnel in 2016 [8] and accretion time from 5 to 20 min) . These sweeps were performed to 2018 [6, 7]. The aim of these tests was to generate ice accretions on a m easure that parameter’s impact on ice accretion size, location (icing NACA0012 airfoil under well - characterized conditions [6, 7]. Ice extent) , characteristic s (such as glaze/rime ice and shedding behavior) , shapes were generated across different flow conditions. The 2016 and test article surface temperature . Analysis of the test data showed experimental calibra tion efforts were limited to centerline that c louds composed of larger drops, colder air temperatures, smaller measurements. An instrument traversing system was developed for the angles of attack, and longer spray times were the primary parameters 2018 tests which allowed for the ability to perform flow surveys at that resulted in accretions with greater ice mass. Test article angle of various positions within the cloud at the tunnel exit plane. Ice geometry attack and cloud droplet size i nfluenced the loca tion of ice accretion as measur ements were limited to extracting 1D and 2D ice shapes from these two parameters directly impact collection efficiency . With orthogonal video images that captured the ice growth at the respect to icing characteristics, t otal air temperature dictated icing NACA0012 leading edge.
type, and smaller cloud drop size along with warmer air temperatures resulted in greater amount s of ice shedding . Surface temperature Most recently NASA conducted the next test series of fundamental increased during ice accretion from the release of latent (fusion) heat, icing physics tests at the NASA Icing Research Tunnel (IRT) [10 - 11 ] .
where t otal air temperature and cloud drop size impacted the amount These icing wind tunnel tests were conducted in early 2022. These tests of surface temperature change . The icing measurements collected utilized the Simulated Inter - compressor Duct Research Model Page 1 of 13 (SIDRM), a 3D test article whose geometry is representative of an SIDRM Aerodynamic Analysis inter - compressor duct and strut regi on of a turbofan engine. The test article has the ability to heat its surfaces to simulate the warm surfaces Test article a erodynamic analysis was performed utilizing pressure tap of the turbofan inter compressor duct, and is instrumented with heaters, measurements . The test article contains 64 pressure taps in total, split heat flux gauges, thermocouples, and pressure taps. These tests, between two rows on the upper and lower portions that run the chord however, did not fully simulate the environment within the engine. The length of the test article. Figure s 2 and 3 are schematic s showing the testing air temperatures were below freezing, and the icing clouds were location of the upper and lower pressure tap rows (labeled as P ### ).
fully glaciated before making impact with the test article surface. It is These images also show the x , y , z , coordinate system. Figure 2 also believed that ingested ice crystal clouds are pa rtially melted as they shows the location of Type - K thermocouples that are flush - mounted pass through warm engine passages before impacting internal engine on the outer surface ( labeled as T### and ST# ) . The side that can be components. In addition, this open geometry will resu lt in different seen in Fig. 2 is considered the “instrumented” side as it contains the pressur es and flow compared to the duct ed region of the compressor .
majority of the thermocouple and heat flux gauge instrumentation. Th e other side of the test article (not visible in the Fig. 2 ) contains limited The aim of these tests wa s to generate ice accretions on the SIDRM instrumentation, largely for monitoring purposes, and will be referred test article under well - characterized conditions. Icing tests were to the “non - instrumented” side in this paper. Tables A1 and A2 in the conducted under ice crystal icing and supercooled water icing appendix provide surface locations o f the Type - K thermocouples and conditions. T his paper focuses on the analysis of the supercooled liquid pressure taps respectively .
icing data subset . Supercooled liquid icing tests provide simulation validation data on a 3D geometry , in addition an estimate of collection Instrumented efficiency can be back - calculated from ice accretion geometry which Strut is also req uired for ice crystal icing simulation. The IRT is well calibrated for supercooled water icing, but not for ice crystal icing P206 P217 P216 P215 P210 P211 P212 P208 P207 P213 P214 P209 clouds. To that end, ice crystal cloud characterization tests were conducted to provide proper conditions for ice crystal icing te sts. The · · · · · · · · · · · · · icing physics data measured during these tests will be used to develop P218 and validate 3D engine icing simulation tools such as GlennICE ST2 [1 2 - 14 ]. The intent is to provide analyzed results related to ice crystal T407 ST1 x X icing in a future technical paper .
ST2 · · ST1 T401 T301 · · · · · · · · · · · · · T303 T302 T306 T311 T312 T313 T314 T315 T304 T305 T310 z Z This paper provides analysis results and discussions from the · · T207 T201 supercooled liquid icing and the test article aerodynamic Leading T307 characterization portion s of the test entry . This paper , along with an Edge accompanying conference paper by Stewart et al. [1 5 ] build upon t he P118 work by Bartkus et al. [10]. These testing efforts are part of NASA’s Flow · · · · · · · · · · · · · P113 P116 P115 P112 P114 P117 P107 P108 P109 P110 P111 P106 Advanced Air Transport Technology (AATT) Project activities to improve understanding of the ice growth physics and advance engine aero thermodynamic modeling tools to predictively ass ess the onset and growth of ice in current and future engines during flight.
Simulation tools like GlennICE will allow industry to design and certify engines for flight in icing conditions in a cost - effective way.
Figure 2. Schematic of pressure tap and thermocouple locations on SIDRM .
Experimental Data Analysis and Discussion
The opposite side (non - instrumented side) is not shown but the pressure tap locations mirror the side that is show n (instrumented side) .
A limited description of the experimental set - up is provided here . For the interested reader, a more detailed description is provided by Bartkus et al. [10] elsewhere . These tests utilized the SIDRM test article, which is shown in Fig. 1. SIDRM aerodynamic analysis and P220 supercooled liquid icing analysis along with discussion are provided in P205 P204 the following two subsections.
P203 P219 P202 P201 21.7 in x y P101 P119 P102 71.8 in 62.6 in P103 Flow P104 P120 P105 Figure 3 . Schematic of pressure tap locations on SIDRM staggered at the Figure 1. SIDRM test article schematic and dimensions with floor mounting leading edge. Flow direction is into the paper.
plate at the bottom .
Page 2 of 13 AoA = 0°, U = 150 knots -3.0 Aerodynamic characterization was performed varying test article Lower Instr.
angle s of attack ( AoA = 0, 1, 2, 3, and 4 ° ) and airspeed ( U = 50, 100, -2.5 Lower Non-Instr.
150, 200, and 230 knots). Table 1 shows the aerodynamic conditions Upper Instr.
-2.0 that were run. The pressure coefficient, C p , was calculated from Upper Non-Instr.
CFD experimental measurements. Equation 1 shows the pressure coefficient -1.5 expression where P is the surface pressure as measured by the tap, tap P s is the freestream static pressure , and P 0 is the total pressure. Total -1.0 and static pressure m easurements were made by the North and South -0.5 pitot - static probes located at the entrance of the IRT test section.
Pressure Coefficient 0.0 Table 1. Tabulation of aerodynamic test conditions that were conducted.
0.5 AoA (°) 0 0 0 0 0 1 2 3 4 4 4 4 4 1.0 U (knots) 50 100 150 200 230 150 150 150 50 100 150 200 230 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 z/c 𝑃 − 𝑃 𝑡𝑎𝑝 𝑠 𝐶 = (1) Figure 4 . Pressure coefficient plotted against normalized chord - length for a test 𝑝 𝑃 − 𝑃 0 𝑠 run at 150 knots and 0 ° A o A . Experimental values are depicted in lines with symbols, w hile the CFD prediction is shown as a solid line. Note that the bottom axis values are negative because of the coordinate convention used.
Two - dimensional (2D) c omputational fluid dynamics (CFD) simulations were performed to compare with e xperimental aerodynamic pressure measurements . With the struts located far from the pressure taps and not impacting pressure measurements, performing 2D sim ulations was sufficient for comparison with pressure measurements. SIDRM’s main body (no struts) was meshed inside the IRT’s tunnel geometry using a 2D mesh that was 1 - cell deep at a total cell count near 60,000. SIDRM’s high flow blockage causes the prefe rence of a rectangular tunnel grid over a circular far - field grid.
The inlet and outlet of the IRT was extended out about 12 chord lengths Flow from the test article. Angle of a ttack changes were created by rotating the test article within the IRT geometry and creating a new mesh. The m esh was composed of using a hybrid mesh of prism layers (with y+ ~1) at the surface and unstructured tetrahedral in the rest of the z/c - 0.90 - 0.80 volume grid. Refinement was placed around the test article and downstream to capture the wake. CFD simulations were r u n in ANSYS Fluent using the k - Omega SST t urbulence model with viscous heating Figure 5 . Image of tufts on the back half of the test article showing where flow bec ame detached . Flow is left to right.
and curvature correction enabled. V elocity inlet and pressure outlet boundary condition s w ere applied, with a no - slip surface condition on Figure 6 shows measured pressure coefficients plotted against the test article. T he t unnel walls were simulated with an inviscid normalized chord - length for various airspeeds at 4 ° A o A, which is in boundary condition. Static pressures and total temperatures were kept the same for all angle of attack runs. SIDRM’s separation near the good agreement with CFD predictions shown in Fig. 7 . There is generally good agreement between experiment and CFD regarding t he trailing edge require d running thousand s of iterations until the C spike on the suction side near the leading edge for these 4 ° A o A coefficient s of lift and drag plateaued reac hing steady - state p convergence . tests. The continuously decreasing value of C p (more negative) at z/c = - 0. 4 with increasing velocity is due to Reynolds number effect 6 6 (Re = 1 x 10 to 5 x 10 ) [ 16 ] . Greater momentum from faster airspeeds Figure 4 shows that there is good agreement between CFD predictions pushes the location of separation farther downstream , which in turn (b lack solid curve) and experimental measurements ( curves with aids airspeed and affects C upstream at the model’s thickest width.
p symbols) for a test run where airspeed was 150 knots, and the test The pressure taps are too coarsely distanced to capture the exact article was at a 0 ° angle of attack . The pressure coefficients are plotted location of most negative C p around z/c = - 0.4 as can be seen in the against the normalized chord length ( z represents the chord direction, CFD predictions. Also, of note here is that at an angle of attack (in this c is the c h ord length). The pressure tap measurements are plotted as case 4 ° A o A ), the lift generated at the front half of the test article is four separate sections of the model to visualize any variation bet ween countered by negative lift on th e aft half. In the forward half, the upper sides . The upper and lower pressure tap rows are split between the two curves represent the suction side, the lower curves represent the sides of the model ( instrumented and non - instrumented sides). At 0 ° pressure side. Integrating the difference between the two provides the A o A, all four experimental measurements are expected to collapse on amount of lift generated by the forward half. In the aft half, the upper top of each other as flow is symmetric and n o variation is expected cur ves represent the pressure side, the lower curves the suction side.
between the upper and lower rows . The simulation predicted flow The integrated difference now produces negative lift (a downward separation beyond z/c = - 0.8 where the Cp flattens to the trailing edge force) on the aft half . While the SIDRM geometry produces lift poorly and is in close agreement with experiment . Separation beyond at non - zero angle s of attack, the intent for varying AoA is to investigate z/c = - 0.8 was also visually observed when utilizing tufts as shown in its impact on local collection efficiency , including shadow and Fig. 5 .
concentration regions during icing tests .
Page 3 of 13 AoA = 4°, Lower Row Taps U = 150 Knots, Lower Row -4.0 -2.5 U = 50 knots -3.5 AoA = 0 Deg U = 100 knots -2 AoA = 1 Deg -3.0 U = 150 knots AoA = 2 Deg U = 200 knots -1.5 -2.5 AoA = 3 Deg U = 230 knots -2.0 AoA = 4 Deg -1 -1.5 -0.5 -1.0 -0.5 0.0 Pressure Coefficient 0.5 Pressure Coefficient 0.5 1.0 1.5 1.5 -1 -0.5 0 0.5 1 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 z/c Surface Distance from Leading Edge (in) Figure 6. Measured pressure coefficients plotted against normalized chord - Figure 8 . Pressure coefficient measurements against surface distance near the length for various airspeeds at 4 ° AoA .
SIDRM leading edge at various angle of attacks at 150 knots.
Supercooled Liquid Icing Analysis -2.4 Twenty - eight (28) supercooled liquid icing test run s were conducted -2.2 during the test entry and are discussed here . Various parameter sw eeps were performed to measure that parameter’s impact on ice accretion -2.0 size, location, characteristics (such as glaze/rime ice and shedding -1.8 behavior) , and test article surface temperature, and is organized in this
Suction
-0.08 -0.04 section accordingly. Primary parameter va riation sweeps included cloud median volumetric diameter ( MVD ), total air temperature ( T ), angle of attack , and accretion time .
Neg Lift
Table 2 shows the target test conditions at the test section and is
Pos Lift
ordered in blocks according to the 7 parameter sweeps (Blocks A – G).
Suction
Three MVD sweeps (Blocks A – C) were conducted at varying airspeeds, angle s of attack, and accretion times. A total air temperature Pressure Coefficient sweep was conducted that ranged from glaze to rime ice (Block D). An angle of attack sweep (Block E) was performed along with two accretion time sweeps for rime ice (Block F) and glaze ice (Block G) conditions. Positive a ngle of att ack values mean that the test article
B
was rotated in the direction that made t he instrumented side the
- - - - -
pressure side. U nique test runs are identified by their test ID
z/c
(UG####) . In the table , liquid water content is noted a s LWC . The Figure 7 . CFD predictions of p ressure coefficients plotted against normalized static relative humidity is approximated to be saturated at the test chord - length for various airspeeds at 4 ° A o A. The inset in the CFD simulation section for this atmospheric ic ing wind tunnel as an equilibration spray provides a zoomed in view of the suction side spike near the leading edge.
was performed at the start of each day of testing . Five test points were repeated to provide a measure of repeatability (e.g., UG3540 and Figure 8 shows a z oom ed in view of pressure coefficient measurements UG3547 in Block B) . Five test runs are duplicated in different near the SIDRM leading edge at various angle s of attack at 150 knots .
parameter sweeps in this table (e.g., UG3540 in Blocks B and E) .
The pressure coefficients are plotted against surface wrap distance, s , where the test article’s geometric leading edge is at 0 inches at 0 ° AoA .
The red curve represent s 0 ° A o A and is not symmetric around 0 inches (if the stagnation line is at 0, then Cp symmetry is expected) . The plotted curves suggest that the stagnation point is between 0 and 1 ° AoA. This could mean there is a small flow angularity, the model itself is not perfectly symmetrical, the turntable is not perfectly at 0 ° A o A (even though measurement says it i s 0 ° ) , or some combination of these factors . This is to say that when SIDRM is set to 0 ° A o A, it may more accurately be about 0.3 ° A o A , for exam ple (rotated such that the instrumented side is the slight suction side ). Future tests will investigate this observed slight asymmetric flow more closely.
Page 4 of 13 Table 2 . Tabulation of test conditions and ice accretion measurements for t hree clean 3D scan of the SIDRM test article was subtracted from the 3D MVD sweeps , a total air temperature sweep , and angle of attack sweep , and two scans of the accreted ice. The center 8 - in span of the 3D scan was used accretion time sweeps for rime and glaze ice conditions . Values denoted by in the volume calculation, with a cut in the xy - plane at z = - 0.5 inches asterisks refer to test runs where measurement of ice mass on the struts were deep into the model leading edge. The volume of ice is shown in the affected by ice sheds.
bottom right of Fig . 9 and is quantified in the Column 10 of Table 2.
Target Test Conditions Ice Accretion Measurements Col 1 Col 2 Col 3 Col 4 Col 5 Col 6 Col 7 Col 8 Col 9 Col 10 Accretion Both Struts Main Body LE Test ID Time T U AoA MVD LWC Ice Mass Ice Mass Ice Vol.
3 3 (#) (min) (°C) (knots) (°) (μm) (g/m ) (g) (g) (cm ) Block A: MVD Sweep at 150 knots, 0 ° AoA, and 10 min accretion time 8 in UG3513 10 -17 150 0 15 0.45 66 172 56.3 UG3514 10 -17 150 0 18 0.45 70 221 62.3 UG3518 10 -17 150 0 30 0.45 79 407 78.8 UG3515 10 -17 150 0 50 0.45 79 548 87.1 UG3516 10 -17 150 0 90 0.45 83 781 93.7 The gray semi-circle is 10 in ° Block B: MVD Sweep at 150 knots, 4 AoA, and 5 min accretion time the SIDRM test article UG3541 5 -17 150 4 18 0.45 33 105 35.5 for this 4° AoA test run.
UG3540 5 -17 150 4 30 0.45 35 176 40.0 An xy -plane cut made at UG3547 5 -17 150 4 30 0.45 33 184 39.4 z = -0.5 inches from the UG3537 5 -17 150 4 50 0.45 34 N/A 41.1 SIDRM leading edge.
UG3546 5 -17 150 4 50 0.45 33 249 42.6 UG3542 5 -17 150 4 90 0.45 36 382 44.9 Block C: MVD Sweep at 200 knots UG3520 7 -17 200 0 18 0.45 59 239 59.4 UG3521 7 -17 200 0 30 0.45 63 376 68.1 UG3548 7 -17 200 0 30 0.45 64 377 72.8 UG3535 7 -17 200 0 50 0.45 65 574 84.4 Figure 9 . Ice accreted on the SIDRM test article on the left (Test Run UG3536 7 -17 200 0 90 0.45 64 772 86.4 ID UG3534) , with the 10 - in wide scanned region shown in the upper right , and Block D: T Sweep the 8 - in span length used to calculate the accreted ice volume at the leading UG3531 20 -3 150 4 25 0.50 139* 395 147.8 edge shown i n the bottom right .
UG3545 20 -3 150 4 25 0.50 132* 393 147.9 UG3530 20 -6 150 4 25 0.50 247 477 136.6 Ice Accretion Size UG3544 20 -6 150 4 25 0.50 234 480 137.5 UG3532 20 -9 150 4 25 0.50 239 526 156.6 UG3533 20 -12 150 4 25 0.50 238 551 170.0 Multiple parameters impacted accretion size. According to the main UG3534 20 -17 150 4 25 0.50 224 595 172.6 body ice mass measurements ( Column 9 in Table 2 ) , larger cloud Block E: AoA Sweep MVD, colder air temperatures, smaller angle s of attack , and longer UG3538 5 -17 150 0 30 0.45 33 208 43.9 spray times were the primary parameters tha t resulted in greater ice UG3539 5 -17 150 2 30 0.45 33 205 42.7 UG3543 5 -17 150 3 30 0.45 35 188 39.8 mass. This trend is consistent even when the ice masses from the struts UG3540 5 -17 150 4 30 0.45 35 176 40.0 and the main body are combined. In all 3 MVD sweeps, l arger MVD UG3547 5 -17 150 4 30 0.45 33 184 39.4 clouds resulted in greater ice mass . L arger drops behave more Block F: Accretion Time Sweep (Rime Ice) ballistically while smaller drop s follow the air flow more closely. In UG3541 5 -17 150 4 18 0.45 33 105 35.5 turn, the larger drops resulted in greater collection efficiency which UG3519 10 -17 150 4 18 0.45 73 208 66.5 Block G: Accretion Time Sweep (Glaze Ice) directly impacts accreted ice mass. For the conditions tested in the total UG3549 10 -3 150 4 25 0.50 85 221 65.0 air temperature sweep , c older air temperatures resulted in greater ice UG3531 20 -3 150 4 25 0.50 139* 395 147.8 mass measurements . Greater amounts of s hedding occur red in the UG3545 20 -3 150 4 25 0.50 132* 393 147.9 inclined area where the strut is located during the warmer total air temperature test runs. The greater shedding amounts may be a result of The primary measurements to assess ice accretion size for these tests weaker i ce cohesion strength compared to colder accretions and were 3 D laser scans of the final ice accretion and weighing of the feather accretions experiencing more drag due to accretion shape and accreted ice. The accreted ice geometry was measured after each test size. This shedding difference accounts, in part , for the difference in run utilizing a 3D laser scanner . Hexagon ’s Romer Absolute SI 7530 final measured ice mass. Additional testing is required to determine if scanner was used and has a repeatability accuracy of +/ - 0.003”. The other mechanisms are contributing to ice mass differences. For the center 10 - in ( 0.25 - m) span was scanned from the leading edge to the conditions tested in the angle of attack sweep , increased angle s of extent of ice on either side of the test article, including the struts . Figure attack resulted in smaller ice mass. An increase in angle of attack 9 depicts the area that was scanned with resulting 3D image o f the ice result ed in more shadow regions where the cloud did not impact the accretion in the upper right of the figure . After scanning, t he center 8 - test article surface, reducing ice accretion mass. Finally, a l onger spray in ( 0.20 - m ) span of accreted ice was carefully cut from the leading time resulted in greater ice mass . For both the glaze and rime condition edge to the extent of icing on both sides , removed, and weighed. Ice accretion time sweeps , doubling spray time result ed in nearly twice the accreted on the struts was weighed separately. The combined ice mass accreted ice mass on the main body ( slightly less than double for the from both struts is provided in Column 8 of Table 2. The ice mass as glaze condition test run) .
measured from the 8 - in wide strip on the main body is provided in Column 9 of Table 2. It should be noted that the accreted ice on the Figures 10 through 1 6 show cross section cuts at the test article suction side stru t had shed towards the end of the icing test for Test midspan for the corresponding 7 sweeps tabulated in Table 2 . These ID# UG3531 and UG3545 of the total air temperature sweep, thus fig ures provide a n overall picture of where the ice accreted on the test resulting in low strut ice mass measurements. From the 3D laser scans, article. Insets are provided to show greater ice accretion detail at the the volume around the leading edge (LE) was calculated to provide an leading edge and both junctions where the struts meet the main body .
estimate of how various parameter sweeps impacted ice accretion. A Page 5 of 13 The ice accretion size t r ends in the se figures are in agreement with the main body ice mass measurement trends from Table 2. These 7 figures show that larger ice cross sections resulted when tests were conducted with larger MVD clouds, colder air temperatures , smaller test article angle of attack, and longer accretion times. Flow is fro m top to bottom in F igs. 10 – 16. For test run s conducted at a n on - zero angle of a ttack, the left side of the test article exp erie nced pressure forces and th e right side exp erie nced suction forces in F igs. 10 – 16 .
Figure 12. Cross sections at the midspan of the SIDRM test article for the MVD parameter sweep at 200 knots (Block C ).
Figure 10 . Cross sections at the midspan of the SIDRM test article for the MVD parameter sweep at 150 knots, 0 ° AoA , and 10 min accretion time (Block A) .
Late strut ice sheds for both T = - 3 ° C tests Figure 1 3 . Cross sections at the midspan of the SIDRM test article for the air temperature parameter sweep (Block D). Note that the ic e shed from the right strut late in the test for both test runs at T = - 3 °C .
Figure 1 1 . Cross sections at the midspan of the SIDRM test article for the MVD parameter sweep at 150 knots, 4 ° AoA and 5 min accretion time (Block B) .
Page 6 of 13 Ice on strut shed at ~18 min and ~19.5 min for the 20 min test runs, respectively Figure 14. Cross sections at the midspan of the SIDRM test article for the angle Figure 16. Cross sections at the midspan of the SIDRM test article for the of attack sweep (Block E).
accretion time sweep at glaze ice conditions (Block G).
The trend that greater ice mass accretions occurred with larger MVD clouds, colder air temperatures, smaller test article angle of attack, and longer accretion times is consistent with the leading edge volume calculations in Table 2, except for the total air temperature sweep. The volume trend in n ot monotonic with total air temperature . The LE volume measured at T = - 3 °C was greater than at T = - 6 °C, but as 0 0 temperature continued to decrease transitioning towards rime ice, the measured LE volume increased. Figure 13 provides a visual of the chan ging LE ice accretion with decreasing temperature. Despite glaze ice accretion impacting LE volume such that it is not monotonic with decreasing total air temperature, the overall mass on the entire test article consistently increased monotonically with de creasing total air temperature.
There were 5 pairs of repeated test runs. Comparing the mass and volume measurements in Table 2, there was good repeatability between identical test runs. The average variation between identical test runs for strut ice mass, main body ice mass , and leading edge volume was 4%, 2%, and 3% respectively. Figures 11 through 13 (Blocks B, C, and D) contain these 5 repeated cases and provide a visual of the repeatability in ice accretion. This repeatability provides a good baseline and confidence in trends that are observed within Figure 1 5 . Cross sections at the midspan of the SIDRM test arti cle for the individual parameter sweeps.
accretion time sweep at rime ice conditions (Block F) .
Ice Accretion Location Figures 10 through 16 provide a visual reference for where icing generally occurred, indicating regions of greater collection efficiency.
For any individual test run, greater amount s of ice accreted at the main body leading edge, the struts leading edges , and just upstream of the struts on the curved ra mp section on the main body. Lower amounts of ice accretion occurred on the flat stretches aft of the main body leading edge (from z = - 1 to 11 in ches) and aft of the struts.
The most prominent parameters that impacted the location of ice accretion were a ngle of attack and cloud MVD. Test article angle of attack directly influence d collection efficiency along the chord of the test article. Non - zero angles create d shadow regions on the suction side where the incoming cloud d id not impact the test article. T his can be Page 7 of 13 seen in Fig. 1 7 where A o A varied from 0 ° (left), 2 ° (middle), and 4 ° (right). Increasing the A o A pushed the initial impinging area (aft of the leading edge) f a rther back on the suction side. These images represent 3 of the 5 tests shown earlier in Fig. 14 and tabulated in Table 2 for the AoA sweep (Block E) . Figure 1 8 is a zoomed in view of the suction side strut junction for the AoA sweep shown in Fig. 14 . Th ese cross sections quantify the extent of icing on the main body. In addition, Fig.
18 s hows the impact of AoA on the ice accretion as shadow and concentration regions can be seen on the strut. The shadow region is noticeable near the strut junction near y = 5 in. There is thicker ice accreted at the strut around y = 6 inches compared to y = 8 inches for the 4° AoA, suggesting cloud concentration caused by flow upstream around the test article leading edge.
Figure 1 9 . Images showing extent of ice on the suction side of the test article where tests were conducted at 4 ° AoA, and MVD varied from 18 μm (left) to 90 μm (right). From Block B in Table 2, Test Run IDs UG3541 and UG3542 respectively.
Figure 1 7 . Images showing extent of ice on the suction side of the test article where AoA varied from 0 ° (left), 2 ° (middle), and 4 ° (right). From Block E in Table 2, Test Run IDs UG3538, UG3539, and UG3540 respectively.
Main Body Strut Leading Edge Main Body Strut Leading Edge Figure 20 . Cross sections at the SIDRM midspan, zoomed in on the suction side strut junction, for the MVD sweep at 150 knots a nd 4 ° AoA shown in Fig. 1 1 (Block B) .
In the aerodynamic analysis, data suggested that flow symmetry was not at the leading edge when the test article was set to 0° AoA, but Figure 1 8 . Cross sections at the SIDRM midspan , zoomed in on the suction side perhaps off by about 0.3° . All accretions appear symmetric and no strut junction , for the AoA sweep shown in Fig. 14 (Block E) .
discernable lean in accretion bias is observed in Figs. 10 and 12, the two MVD sweeps conducted at 0° AoA (Blocks A and C).
Cloud MVD impact ed ice accretion along the test article chord resulting in different icing limits downstream of the leading edge . The Ice Accretion Characteristic s images in Fig. 1 9 show how t he smaller MVD test on the left ( 18 μm) impinge d f a rther downstream on the suction side as compared to the Two primary parameters impacted icing characteristics (such as larger MVD test on the right ( 90 μm). These images represent 2 of the glaze/rime ice and shedding behavior ) : total air temperature and cloud 6 tests shown earlier in Fig. 1 1 and tabulated in Table 2 for the MVD MVD. Supercooled liquid freezes more readily upon impacting the sweep conducted at 150 knots and 4 ° AoA (Block B) . Figure 20 is a surface with colder air temperatures, while ice accretions in warmer air zoomed in view of the suction side strut junction for the same MVD temperatures result in more runback before freezing. Colder air sweep shown in Fig. 1 1 . These cross sections quantify the extent of temperature test s result ed in rime ice accretions which were icing on the main body. In addition, Fig. 20 shows the impact of MVD characterized by opaque , white ice with more streamlined accreted on the ice accretion as shadow regions near the strut junction are more geometries . The warmer air temperature tests resulted in glaze ice prominent for smaller MVD clouds .
Page 8 of 13 characterized by more transparent ice with horn shapes on leading Test Article Surface Temperature edges . The images in F ig. 2 1 show the qualitative difference between a glaze ice test ( T = - 3 ° C) on the left and rime ice test ( T = - 17 ° C) 0 0 Surface temperature was measured by flush mounted thermocouples on the right. These images represent 2 of the 7 tests shown earlier in during the ice accr etion process. The location of these thermocouples Fig. 1 3 and tabulated in Table 2 under the total air temperature sweep on the SIDRM test article are shown in Fig. 2 with nominal location (Block D) .
provided in Table A1 of the appendix. Analysis of surface temperature data revealed that total air temperature and cloud MVD impacted sur face temperature during ice accretion . A small sample of the many thermocouples is presented, which is representative and demonstrative of the overall surface temperature picture during the ice accretion process.
Figures 23 and 24 show transient temperatur e profiles for the total air temperature sweep (Block D) and MVD sweep (Block B) respectively.
The three panels in each figure represent different midspan thermocouples measurements : A) T301 at the leading edge, B) T303 at the flat section of the test article 5 inches downstream of the leading edge , and C) T306 at the base of the curved section 12.5 inches downstream of the leading edge . The flat section refers to the linear segment o f the SIDRM cross - section profile from z = - 0.6 to - 10.5 inches. The curved section refers to the curved segment from z = - 10.5 to 16 inches of the SIDRM profile, just upstream of the strut.
Icing spray time is the bottom axis , and the dashed vertical line s represent the start and end of the spray. The plots provide test article Figure 21 . Images of ice accretions from the air temperature sweep with a glaze surface temperature 30 seconds prior to and after cloud activation.
ice accretion on the left ( T = - 3 ° C) , and rime ice accretion on the right ( T = - 17 ° C) . From Block D in Table 2, Test Run IDs UG3531 and UG3 534 respectively.
Review of time - lapsed video showed that smaller MVD clouds resulted in greater amounts of feather shedding towards the top of the steep ramp on the test article. Images of final ice accretions from the MVD sweep at 150 knots and 0 ° AoA (Block A) generated from clouds of MVD = 15 μm (left) and MVD = 90 μm (right) are shown in Fig. 22 . The red boxes indicate the areas where there were differences in shedding behavior. There was continuous shedding from 4 minutes into the accretion until the end of the test (10 min) for the accretion on the left . In contrast, there was v irtually no shedding for the image on the right. A lso, a s mentioned earlier, greater amounts of shedding occurred during warmer air temperature runs compared to colder conditions. The differences in shedding can be seen in Fig. 21 in the inclined section where the strut is located.
Figure 22 . Images of ice accretions from the MVD sweep at 150 knots and 0 ° AoA where the cloud MVD equaled 15 μm on the left and 90 μm on the right . From Block A in Table 2, Test Run IDs UG3513 and UG3516 respectively. The red boxes highlight the differences in shedding behavior.
Page 9 of 13 0 -10 18 μm -3 °C -2 30 μm -3 °C -11 30 μm -6 °C -4 50 μm -6 °C C) ° C) -12 ° 50 μm -9 °C -6 90 μm -12 °C -13 -17 °C -8 -10 -14 -12 -15 Surface Temperature ( -14 Surface Temperature ( -16 -16 A T301 - Leading Edge A T301 - Leading Edge -17 -18 -50 0 50 100 150 200 250 300 350 -200 0 200 400 600 800 1000 1200 1400 Spray Time (s) Spray Time (s) -10 -2 -11 -4 C) ° C) -12 ° -6 -13 -8 -10 -14 -12 -15 Surface Temperature ( Surface Temperature ( -14 -16 -16 T303 - Flat Section B T303 - Flat Section B -17 -18 -50 0 50 100 150 200 250 300 350 -200 0 200 400 600 800 1000 1200 1400 Spray Time (s) Spray Time (s) 0 -10 -2 -11 -4 C) ° C) -12 ° -6 -13 -8 -10 -14 -12 -15 Surface Temperature ( Surface Temperature ( -14 -16 -16 C T306 - Base of Curve T306 - Base of Curve C -17 -18 -50 0 50 100 150 200 250 300 350 -200 0 200 400 600 800 1000 1200 1400 Spray Time (s) Spray Time (s) Figure 24. Transient surface temperature profiles for the MVD sweep (Block Figure 23. Transient surface temperature profiles for the total air temperature B) grouped by thermocouple lo cation: A) T301 at the leading edge, B) T303 at sweep (Block D) grouped by th ermocouple location: A) T301 at the leading the flat section of the test article, and C) T306 at the base of the curved section.
edge, B) T303 at the flat section of the test article, and C) T306 at the base of the curved section .
Figure 23 shows that there is an initial increase in test article surface temperature for all thermocouples as the spray c loud is activated and at every total air temperature. This increase in surface temperature is related to the release of latent heat as the supercooled liquid droplets freeze during the accretion process [17 - 18] . The increase in surface temperature is great e r for the colde r total air temperature test run s because these tests have a higher freezing fraction (rime > glaze) which release the latent heat in a shorter period of time. T he supercooled Page 10 of 13 droplets are approximated to be around the static wet - bulb temperature less mass loss) . Smaller angle s of attack resulted in fewer shadow upon impact, where total air temperature is a good approximation. The regions leading to more ice accretion. Longer spray times provide increase in surface temperature is greatest at the leading edge greater time to accrete more ice.
thermocouple because the collection efficiency is g reatest there, and lowest at the flat section of the test article. Greater collection efficiency Test article angle of attack and cloud MVD impacted the location of is directly related to greater fusion heat release and hence great er ice accretion. Test article angle of attack directl y influence d collection temperature increase. The overall trend is that surface temperature efficiency along the chord of the test article creating shadow regions.
deceases over time aft er the initial increase. Ice is a good insulator and Cloud MVD impact ed co llection efficiency along the test article chord a s ice thickness increases less heat conducts towards the test article resulting in different impingement limits downstream of the leading surface from the latent heat release of freezing supercooled droplets at edge as larger droplet s are more ballistic and smaller droplets follow the ice - air interface. As the spray cloud is turned off, th e test article the flow.
surface temperature decreases as no more heat is supplied from latent heat release of fusion. These transient temperature profiles indicate Total air temperature and cloud MVD were the primary parameters that there is good repeatability as there is good agreement between the impacting icing characteristics. Warmer air temperatures resulted in two pairs of repeat test runs conducted at T = - 3 °C and T = - 6 °C.
0 0 glaze ice while colder conditions resulted in rime ice accretio ns.
S maller MVD clouds and warmer air temperatures resulted in greater The transient temperature profiles of the MVD sweep in Figure 24 amounts of feather shedding on the steep inclined area of the test ( conducted at T 0 = - 17 °C) are similar to profiles in the total air article .
temperature sweep in Figure 23. All three thermocouples for all MVD test runs show ed an initial increase in surface temperature followed by Finally, total air temperature and cloud MVD impacted surface a slow decrease until when the spray cloud was turned off whe re temperature during ice accretion. Surface temp erature increased at the surface temperature decreased even further. The same physics onset of accretion due to the latent heat release as supercooled liquid described in the previous paragraph applies for this transient trend. A froze. A greater amount of instantaneous latent heat is released for l arger surface temperature increase occurred for larger MVD sprays as supercooled droplets in colder air temperatures, resulting in a greater the larger droplets were more ballistic and resulted in greater collection surface temperatur e increase. Larger MVD droplets are more ballistic efficiency and by extension greater latent heat release. As in the total and contribute to a high collection efficiency which in turn results in air temperature sweep, the smallest sur face temperature rise occurred greater amounts of latent heat release and temperature increase.
at the flat section of the test article ( T303 ). Of note is that the largest surface temperature increase occurred at the base of the curved section
References
(T306) during the largest MVD spray (90 μm). Collection efficiency played a r ole in this outcome as this sweep was conducted at 4° AoA, which slightly reduced the collection efficiency at thermocouple ID 1. Bravin, M., and Strapp, J.W., “A Continuing Investigation of T301 and increased collection efficiency at T306 as compared to a 0° Diurnal and Location Trends in an Ice Crystal Icing Engine Event Data Base,” SAE International Conference on Icing of Aircraft, AOA run . Again, there is reasonably good agreement in measured transient surface temperature between the two pairs of repeat test runs Engines, and Structures , SAE, Minneapolis, MN, 2019, SAE where MVD = 30 and 50 μm. Technical Paper 2019 - 01 - 1964.
2. Mason, J. G., Strapp, J. W., and Chow, P., “The Ice Particle Threat to Engines in Flight,” 44th AIAA Aerospace Sciences
Conclusion
Meeting and Exhibit , AIAA, Reno, NV, 2006, AIAA - 2006 - 206.
3. Flegel, A. B., “Ice Crystal Icing Investigation on a Hone ywell An icing test entry was conducted with the SIDRM test article installed Uncertified Research Engine in an Altitude Simulation Icing at the NASA IRT icing wind tunnel in early 2022. The data measured Facility,” Proce e dings of the ASME Tu rb o Expo 2020 , London, during these tests will be used to develop and validate 3D engine icing England, 2020, GT2020 - 14714.
simulation tools such as GlennICE. This paper discussed 4. Flegel, A. B., and Oliver, M. J., “Preliminary Results from a measurements related to aerodynamic tests and supercooled liquid Heavily Instrumented Engine Ice Crystal Icing Test in a Ground icing tests that were conducted.
Based Altitude Test Facility,” 8th AIAA Atmospheric and Space Environments Conference , 2016, AIAA - 2016 - 3894 .
Test article surface pressure was measured utilizing pressure taps that 5. Oliver, M. J., “Validation Ice Crystal Icing Engine Test in the ran the length of the chord. Experimental measurements were Propulsion Systems Laboratory at NASA Glenn Research compared to computational CFD simulations and were in good Center ,” 6th AIAA Atmospheric and Space Environments agreement for various angle s of attack and airspe eds.
Conference , AIAA, Atlanta, GA, 2014, AIAA - 2014 - 2898 .
6. Struk, P. M., Agui, J. H., Bartkus, T. P., Tsao, J. - C., King, M. J., Various parametric sweeps were conducted during the supercooled and Ratvasky, T. "Ice - Crystal Icing Accretion Studies at the liquid ice accretion portion of the test entry. Primary measurements NASA Propulsion Systems La boratory," SAE 2019 International included accreted ice mass, 3D scans to capture the geometry of the Conference on Icing of Aircraft, Engines, and Structures , SAE, accreted ice, and test article surface temperature. These sweeps were Minneapolis, MN, 2019, SAE Technical Paper, 2019 - 01 - 1921 .
performed to measure that parameter’s impact on ice accretion size, 7. Bartkus, T. P., Tsao, J. C., and Struk, P. M. "Analysis of location, characteristic s , an d test article surface temperature . Experimental Ice Accretion Data and Assess ment of a Thermodynamic Model During Ice Crystal Icing " SAE International Conference on Icing of Aircraft, Engines, and Larger cloud MVD, colder air temperatures, smaller angle s of attack , Structures , SAE Technical Paper, 2019 - 01 - 2016 .
and longer spray times were the primary parameters that resulted in 8. Struk, P. M., Ratvasky, T. P., Bencic, T., Van Zante, J. F., King, greater ice mass . Larger droplets are more ballistic and result ed in M. C., Tsao, J. - C., a nd Bartkus, T. P. "An Initial Study of the greater colle ction efficiency. Colder air temperatures resulted in less Fundamentals of Ice Crystal Icing Physics in the NASA shedding compared to warmer air temperature accretions (and hence Propulsion Systems Laboratory," 9th AIAA Atmospheric and Page 11 of 13 Space Environments Conference , AIAA, Denver, CO, 2017, acknowledge the efforts made by Jordan Salkin and Quentin Schwinn AIAA - 2017 - 4242 . in collecting and processing the 3D scanned icing geometry data that 9. Struk, P. M., Tsao, J - C., and Bartkus, T. B., “Plans and is used extensively in this technical paper.
Preliminary Results of Fundamental Studies of Ice Crystal Icing Physics in the NASA Propulsion Systems Laboratory,” 8th AIAA
Appendix
Atmospheric and Space Environments Conference , 2016, AIAA - 2016 - 3738 .
Tables showing the location of built - in instrumentation on the SIDRM 10. Bartk us, T. P. , Lee , S. , Potapczuk, M. G. , and Flack. C. A., surface are provided in the appendix.
“Description of Cloud Characterization and Icing Tests for a 3D Heated Test Article at the NASA Icing Research Tunnel,” AIAA Table A1. Table providing the nominal location of the flush mounted Type - K AVIATION 202 2 FORUM, AIAA, Chicago, IL , 202 2 , AIAA - thermocouples .
2022 - 3700.
11. Bartkus, T., Potapczuk, M., Lee, S., Stewart, E., and Chen, R. - C., TYPE-K THERMOCOUPLES “Plans for Ice Crystal Icing Tests Using a 3D Heated Test Article TC Name x (in) y (in) z (in) at the NASA Icing Research Tunnel,” AIAA AVIATION 2021 INSTRUMENTED MAIN BODY AND STRUT FORUM, AIAA, Virtual Event, 2021 , Oral Presentation.
T201 -2.500 0.000 0.000 12. Wright, W. B., Porter, C. E. , Galloway E. T., and Rigby D. L, T301 0.000 0.000 0.000 “ GlennICE 2.1 Capabilities and Results ,” AIAA AVIATION 202 2 T401 2.500 0.000 0.000 FORUM, AIAA, Chicago, IL , 202 2 , AIAA - 2022 - 3309.
T302 0.000 -0.842 -2.500 13. Porter, C. E., “ A Comparison of Trajectory Refinement Schemes T303 0.000 -1.283 -5.000 for GlennICE ,” AIAA AVIATION 202 2 FORUM, AIAA, Chica go, T304 0.000 -1.723 -7.500 IL , 202 2 , AIAA - 2022 - 3692.
T305 0.000 -2.164 -10.000 14. Wright, W., Porter, C., Galloway, E., and Rigby, D. “An T306 0.000 -2.762 -12.500 Automated Refinement Process for Particle Trajectory Methods T207 -2.500 -4.196 -15.000 in GlennICE,” AIAA AVIATION 2021 FORUM, AIAA, Virtual T307 0.000 -4.196 -15.000 Event, 2021, AIAA - 2021 - 2631.
T407 2.500 -4.196 -15.000 15. Stewart. E. A., and Bartkus, T. P., “ Computational Icing Analysis ST1 0.000 -5.34 -15.978 on NASA’s SIDRM Geometry to Investigate Collection ST2 0.000 -5.74 -15.978 Efficiency,” SAE International Conference on Icing of Aircraft, T310 0.000 -9.207 -21.480 Engines, and Structures , SAE, V ienna, Austria, 2023 (submitted T311 0.000 -10.546 -23.980 for consideration).
T312 0.000 -10.841 -26.480 16. Brunner, C. E., Kiefer, J., Hansen, M. O., and Hultmark, M., T313 0.000 -10.830 -31.480 “ Study of Reynolds number effects on the aerodynamics of a T314 0.000 -10.715 -36.480 moderately thick airfoil using a high‑pressure wind tunnel ,” Exp T315 0.000 -7.716 -50.940 Fluids 62, 178 (202 1) .
NON-INSTRUMENTED MAIN BODY 17. Bartkus, T. P., Struk, P. M., Tsao, J. C., and Van Zante, J. F., BT303 0.000 1.283 -5.000 “Numerical Analysis of Mixed - Phase Icing Cloud Simulations BT307 0.000 4.196 -15.000 in the NASA Propulsion Systems Laboratory,” 8th AIAA BT311 0.000 10.546 -23.980 Atmospheric and Space Environments Conference , AIAA, BT314 0.000 10.715 -36.480 Washington D.C., 2016, AIAA - 2016 - 3739.
BT315 0.000 7.716 -50.940 18. Bartkus, T. P., Struk, P. M., and Tsao, J. C., “Development of a Coupled Air and Particle Thermal Model for Engine Icing Test Facilities,” SAE International Journal of Aerospace, Vol. 8, No.
1, 2015, pp. 15 - 32, SAE - 2015 - 01 - 2155.
19. Hindmarsh, J., Russell, A., and Chen, X., “Experimental and numerical analysis of the temperature transition of a suspended freezing water droplet ,” International Journal of Heat and Mass Transfer 46: 1199 - 1213, 2003.
Contact Information
Tadas P. B artkus, Ph.D.
Work phone: (216) 433 - 6915 E - mail: tadas.p.bartkus@nasa.gov Affiliation: Ohio Aerospace Institute
Acknowledgments
The authors wish to acknowledge the financial support for this work by the Power and Propulsion sub - project of the NASA Advanced Air Transport Technology p roject (AATT) under NASA's, Advanced Air Vehicles P rogram (AAVP) . In addition, t he authors would also like to Page 12 of 13 Table A2. Table providing the nominal location of the surface pressure taps.
PRESSURE TAPS Tap Name x (in) y (in) z (in) LOWER TAPS P101 -16.750 0.000 0.000 P102 -17.000 -0.165 -0.030 P103 -17.250 -0.289 -0.100 P104 -17.500 -0.377 -0.190 P105 -17.750 -0.437 -0.300 P106 -17.250 -0.953 -3.130 P107 -17.250 -1.505 -6.260 P108 -17.250 -2.770 -12.520 P109 -17.250 -7.127 -18.780 P110 -17.250 -9.506 -21.910 P111 -17.250 -10.777 -25.040 P112 -17.250 -10.840 -28.170 P113 -17.250 -10.831 -31.300 P114 -17.250 -10.658 -37.550 P115 -17.250 -9.907 -43.810 P116 -17.250 -8.080 -50.070 P117 -17.250 -4.794 -56.330 P118 -17.250 0.000 -62.590 P119 -17.250 0.289 -0.100 P120 -17.750 0.437 -0.300 P121 -17.250 0.953 -3.130 P122 -17.250 1.505 -6.260 P123 -17.250 2.770 -12.520 P124 -17.250 7.127 -18.780 P125 -17.250 9.506 -21.910 P126 -17.250 10.777 -25.040 P127 -17.250 10.840 -28.170 P128 -17.250 10.831 -31.300 P129 -17.250 10.658 -37.550 P130 -17.250 9.907 -43.810 P131 -17.250 8.080 -50.070 P132 -17.250 4.794 -56.330 UPPER TAPS P201 16.75 0.000 0.000 P202 17.000 -0.165 -0.030 P203 17.250 -0.289 -0.100 P204 17.500 -0.377 -0.190 P205 17.750 -0.437 -0.300 P206 17.250 -0.953 -3.130 P207 17.250 -1.505 -6.260 P208 17.250 -2.770 -12.520 P209 17.250 -7.127 -18.780 P210 17.250 -9.506 -21.910 P211 17.250 -10.777 -25.040 P212 17.250 -10.840 -28.170 P213 17.250 -10.831 -31.300 P214 17.250 -10.658 -37.550 P215 17.250 -9.907 -43.810 P216 17.250 -8.080 -50.070 P217 17.250 -4.794 -56.330 P218 17.250 0.000 -62.590 P219 17.250 0.289 -0.100 P220 17.750 0.437 -0.300 P221 17.250 0.953 -3.130 P222 17.250 1.505 -6.260 P223 17.250 2.770 -12.520 P224 17.250 7.127 -18.780 P225 17.250 9.506 -21.910 P226 17.250 10.777 -25.040 P227 17.250 10.840 -28.170 P228 17.250 10.831 -31.300 P229 17.250 10.658 -37.550 P230 17.250 9.907 -43.810 P231 17.250 8.080 -50.070 P232 17.250 4.794 -56.330 Page 13 of 13