FREE INTERNET ROWING MODEL (FIRM) EXAMPLES: Coxed Eights March 25, 2015 FIRM IS RESEARCH CODE! Please check all estimates generated by the program against experimental results before committing any time or funds to your project as no liability can be accepted by Cyberiad. c 2015 Cyberiad All Rights Reserved Contents 1 INTRODUCTION 1 2 M8+: Men’s Coxed Eight (Normal Rig) 2.1 M8+ exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 4 3 W8+: Women’s Coxed Eight (Normal Rig) 3.1 W8+: Women’s Coxed Eight (German Rig) 3.2 W8+: Women’s Coxed Eight (Italian Rig) . 3.3 W8+: Women’s Coxed Eight (Rig A) . . . . 3.4 W8+: Women’s Coxed Eight (Rig B) . . . . 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 10 11 12 13 INTRODUCTION Six examples of coxed eights are included in this version of FIRM. More will be added in future versions. The men’s eight is for a real crew. The five examples for women’s eights are for an idealised crew. These examples show how to set up rowers in various arrangements such as Italian, German, and three other “zero yawing moment” configurations. 1 2 M8+: Men’s Coxed Eight (Normal Rig) The on-water trial for this example was conducted over 500 metres. The rowers named Adam, Bob, Conan and Doug row on the starboard (i.e. bow) side; Wally, Yuri, Xave and Zeke row on the port (i.e. stroke) side. Rigging details, oar angles, gate normal forces, and anthropometry were recorded and they have all been used as input to FIRM. Body angle regimes were not recorded. Average angle were estimated for the crew as a whole using an elaborate fitting process, and these were used as input to FIRM. Table 1: Summary of experimental results for this simulation: number of strokes, stroke rate, non-dimensional pull phase duration (tp /ts ), minimum hull velocity (Umin ), maximum hull velocity (Umax ), and mean hull velocity (U ). Item Nstrokes Rate (spm) tp /ts Umin (ms−1 ) Umax (ms−1 ) U (ms−1 ) Value 37.808 0.453 4.715 7.345 6.249 14 ±0.389 ±0.009 ±0.056 ±0.056 ±0.057 Table 1 summarises the main quantities relating to the simulation for this crew. Values are given ± one standard deviation. Table 2: Experimental oar-related values for this simulation: Minimum and maximum oar angles, and maximum gate normal force. Name Seat Adam Wally Bob Xave Conan Yuri Doug Zeke 1 2 3 4 5 6 7 8 Min. Angle (degrees) Port Oar Max. Angle (degrees) -55.2±0.64 31.9±0.47 931.7±61.1 -55.6±0.92 31.7±0.72 1220.3±35.4 -54.5±0.61 33.6±0.45 1006.8±39.5 -71.3±1.33 28.1±0.62 1304.2±55.5 Max. FGn (N) Min. Angle (degrees) -53.5±0.64 Starboard Oar Max. Angle Max. FGn (degrees) (N) 34.2±0.53 814.2±40.8 -58.1±0.91 34.0±0.37 1174.3±65.1 -64.0±1.20 31.2±0.55 1280.7±43.8 -50.8±1.07 35.7±0.29 988.1±71.7 A blade loss factor of kloss = 0.02 has been applied to each rower’s gate normal force to bring FIRM predictions in line with the experimental mean speed of U = 6.249ms−1 . Given the very many uncertainties, the small (2.0%) reductions seem quite acceptable. Note that the adjustments are well within the standard deviations for the maximum gate normal forces shown in Table 2. Instead of using a blade loss factor, we could have adjusted, for example, the oarhandle centres of effort, or the hull viscous form factor, however, as we have discussed previously, it is probably better to show this “fudge factor” in the (single) main input file. The hull propulsive acceleration is shown in the left panel of Fig. 1. Interestingly, there is no “dip” around t/ts ≈ 0.15 apparent in the acceleration curve. It seems that the dip is more pronounced in singles, pairs and doubles. There is good agreement between measured and predicted hull propulsive velocities in the plot at the right of Fig. 1. The forces in the equations of motion are shown in the left panel of Fig. 2. Drag components during the stroke are in the right panel of Fig. 2. Experimental oar azimuth angles shown in Fig. 3 have been shifted so they are referenced to the centre of the pin. Values for starboard-side rowers are shown in the right side panel of the figure; port-side rowers are at the left. Curves are values used as input to FIRM. See the oarangles.csv input files for each rower. Note the scatter for some of the oar angles for Seat 3 (Bob) and Seat 4 (Xave). The faulty data was removed for calculations of means and standard deviations, but it was left in the plots to illustrate that on-water signals are not only scattered because of normal variations, but they can also be the result of equipment glitches and failures. Gate normal forces are shown in the two parts of Fig. 4. Curves are values used as input to FIRM. See the gateforces.csv input files for each rower. Blade propulsive forces for the starboard-side and port-side rowers are shown in the two parts of Fig. 5. It is fairly obvious who the greatest contributors to propulsion are from these graphs, however, they are also among the heaviest rowers in the crew and so they contribute more to the total drag. Deciding on whether there are better candidates for a crew is best done by experienced coaches who can take into account intangibles, such as commitment, grit and esprit de corps, factors that no computer program can ever hope to model. 2 0.8 7.5 7 0.4 6.5 a (g) U (ms-1) 0 6 -0.4 5.5 M8+: AWBXCYDZ Exp. Exp. Mean ± SD Pred. Crew -0.8 -1.2 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 M8+: AWBXCYDZ Exp. Exp. Mean ± SD Pred. Crew 5 4.5 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 Figure 1: Hull propulsive acceleration and crew cg acceleration (left); hull velocity and crew cg velocity (right). 3000 800 M8+: AWBXCYDZ Fprop Fboat Fcrew -Fdrag Fsys 2000 M8+: AWBXCYDZ Air Viscous Wave Total 700 600 500 Drag (N) Force (N) 1000 0 400 300 200 -1000 100 -2000 0 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 Figure 2: Equation of motion forces (left) and drag components (right). The oarlever ratios in Fig. 6 include the effect of variations in the location of the OBCP during the stroke, which is why they are not constant, as they are in most other rowing models. For this example, body angles were not recorded. The regimes shown for the two rowers in Fig. 7 are identical, as they are for all other members of the crew. The body angles were found through a fitting process (not described in this manual) that minimised the difference between measured and predicted values of the hull propulsive acceleration and hull velocity. In Fig. 8, yawing moment lever arms and yawing moments are shown only for the drive phase. FIRM convention is that positive yawing moments tend to turn the hull bow to port (i.e. the stroke side). Thus, the total yawing moment for this (conventionally-rigged) 8+, with these particular rowers, tend to move the boat to starboard (i.e. the bow side) during the early part of the pull phase. Note how Zeke, the rower in Seat 8 (i.e. the “stroke” seat) tends to pull the hull bow to starboard during the early part of the stroke, and then to port for the remainder of the stroke. Kleshnev reported the same phenomenon in his newsletter of Nov. 2009 [?]. Trajectories of the OBCP are almost identical for all rowers, as shown in Fig. 9. The minimum and maximum values for the vertical oar angles were chosen so that oars would enter the water near the catch, and exit at the release. As discussed previously, vertical angles do not play a significant role in FIRM performance predictions. Training a crew to display such exquisite blade harmony would be a considerable challenge for coaches! The trajectories of the OBCP for all eight oars are shown in the two parts of Fig. 10. The traces of each oar are relative to the individual rowers’ ankles (or heel cups). The path of Zeke’s oar is quite different to those of the other crew. The on-water trial for this crew was conducted at an early stage of their campaign, before they had become fully coordinated. The OBCP trajectories are shown on the correct side of the hull and relative to the boat in Fig. 11. 3 40 20 20 0 0 ψxy (degrees) ψxy (degrees) 40 -20 -40 -20 -40 M8+: AWBXCYDZ (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke -60 -80 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 M8+: AWBXCYDZ (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug -60 -80 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 Figure 3: Oar azimuth angles. Port side: seats 2,4,6,8 (left); Starboard side: seats 1,3,5,7 (right). 1400 1200 M8+: AWBXCYDZ (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug 1200 1000 1000 800 800 FGn (N) FGn (N) 1400 M8+: AWBXCYDZ (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke 600 600 400 400 200 200 0 0 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 Figure 4: Gate normal forces. Port side: seats 2,4,6,8 (left); Starboard side: seats 1,3,5,7 (right). The thick grey lines are the OBCP trajectories of Seats 1 and 2 during the previous stroke. It is obvious that the crew have managed to “clear the puddles” at this stroke rate and hull speed. If they had not, then the OBCP trajectories of Seats 7 and 8 would have intersected the previous OBCP traces of Seats 1 and 2. 2.1 M8+ exercises Exercise M8+ 1.0: The on-water trial was held in winter when the air and water temperatures were about 8◦ C. The Olympics in Brazil will be held during the middle of August in 2016. Look up the expected water temperature for Lagoa Rodrigo de Freitas in Appendix ?? and make an estimate of the predicted mean hull velocity for the temperature expected for mid-August. How does that compare with the world’s best time for the M8+ class shown in Appendix ??? Exercise M8+ 1.1: This on-water trial was held over 500 metres. It is unlikely that the crew could sustain the same effort over a 2000 metre course. Reduce each rower’s maximum gate normal force by 5%. Would they still achieve a world’s best time in the warm water expected for Rio 2016? Exercise M8+ 1.2: What would be the expected mean boat velocity if the lagoon will be filled with salt water instead of fresh? 4 350 300 M8+: AWBXCYDZ (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug 300 250 250 200 200 FBx (N) FBx (N) 350 M8+: AWBXCYDZ (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke 150 150 100 100 50 50 0 0 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 Figure 5: Blade propulsive forces. Port side: seats 2,4,6,8 (left); Starboard side: seats 1,3,5,7 (right). 2.4 2.4 M8+: AWBXCYDZ (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke 2.39 2.39 2.38 Dynamic Oarlever Ratio 2.38 Dynamic Oarlever Ratio M8+: AWBXCYDZ (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug 2.37 2.36 2.37 2.36 2.35 2.35 2.34 2.34 2.33 2.33 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 Figure 6: Dynamic oarlever ratios. Port side: seats 2,4,6,8 (left); Starboard side: seats 1,3,5,7 (right). 180 150 M8+: AWBXCYDZ Adam Knee Hip Neck Shoulder 150 120 Joint Angle (degrees) 120 Joint Angle (degrees) 180 M8+: AWBXCYDZ Wally Knee Hip Neck Shoulder 90 60 30 90 60 30 0 0 -30 -30 -60 -60 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 Figure 7: Body angle regimes for seat 2 (left) and seat 1 (right). 5 0.5 t/ts 0.6 0.7 0.8 0.9 1 8 4 2 M8+: AWBXCYDZ (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke Total 1000 Yawing Moment (Nm) 6 Yawing moment lever arm (m) 1500 M8+: AWBXCYDZ (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke 0 -2 500 0 -500 -4 -1000 -6 -8 -1500 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0.2 0.2 0.1 0.1 zobcp (m. above water) zobcp (m. above water) Figure 8: Yawing moment lever arms (left); yawing moments (right). 0 -0.1 M8+: AWBXCYDZ Waterplane (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke -0.2 -0.3 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 0 -0.1 M8+: AWBXCYDZ Waterplane (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug -0.2 -0.3 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 Figure 9: OBCP trajectories in the yz-plane. Port side: seats 2,4,6,8 (left); Starboard side: seats 1,3,5,7 (right). 3.5 3.5 Direction of Direction of Boat Travel Lateral distance from hull centreline (m) Lateral distance from hull centreline (m) Boat Travel 3 Release 2.5 Catch 2 M8+: AWBXCYDZ (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke 1.5 -3 -2.75 3 Release 2.5 Catch 2 M8+: AWBXCYDZ (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug 1.5 -2.5 -2.25 -2 -1.75 -3 x (m) -2.75 -2.5 -2.25 -2 x (m) Figure 10: OBCP trajectories in the xy-plane. Port side: seats 2,4,6,8 (left); Starboard side: seats 1,3,5,7 (right). 6 -1.75 4 Lateral distance from hull centreline (m) 3 2 M8+: AWBXCYDZ (Seat 2) Wally (Seat 4) Xave (Seat 6) Yuri (Seat 8) Zeke (Seat 1) Adam (Seat 3) Bob (Seat 5) Conan (Seat 7) Doug 1 0 -1 -2 -3 -4 0 2 4 6 8 10 12 14 x (m) Figure 11: OBCP trajectories in the xy-plane. Thick grey lines are possible location of puddles formed by seats 1 and 2 during the previous stroke. 7 M8+: AWBXCYDZ Rate 37.8 spm Speed 6.25 m/s Work on Oarhandles A. MUSCULAR 4585 W EFFORT 100 % Dead Mass 154.0 kg Moving Mass 778.0 kg Total Mass 932.0 kg Net Kinetic Energy B. HANDLES 3654 W B/A E. SYSTEM 931 W NOTE: B+F=D+H and C+E=D+G MOMENTUM 80 % E/A 20 % Blade Efficiency Mom. Efficiency C/B = 80.5 % F/E = 61.5 % C. F. PROPULSION 2942 W FOOT BOARDS 572 W (External) C/A 64 % F/A 12 % D. DRAG Propelling Efficiency D/(D+H) = 83.2 % H. BLADE 712 W LOSSES H/A 16 % Lost to water Work done on shell 3514 W D/A 77 % Transferred to air and water I=D+G+H. TOTAL 4585 W LOSS I/A 100.0 % Net Efficiency D/(D+H)-G/A = 75.3 % Figure 12: Power flow chart. 8 Air 11 % Visc. 83 % Wave 6 % Velocity Efficiency 1-G/A = 92.2 % G. BODY FLEX 359 W (Internal) G/A 8% Lost as heat, breath etc. 3 W8+: Women’s Coxed Eight (Normal Rig) To run this example double-click on the icon for the batch file w8 normal.bat. This example, and all other W8+ examples, use clones of a rower named “Arale”. Her rigging details, anthropometry, and her body angle, oar angle and gate normal force regimes are identical, except that four of the clones row on the port side, and four on the starboard side. 8 4 2 M8+: Arale Clones (Seat 1) Star Clone (Seat 2) Port Clone (Seat 3) Star Clone (Seat 4) Port Clone (Seat 5) Star Clone (Seat 6) Port Clone (Seat 7) Star Clone (Seat 8) Port Clone Total 1000 Yawing Moment (Nm) 6 Yawing moment lever arm (m) 1500 M8+: Arale Clones (Seat 1) Star Clone (Seat 2) Port Clone (Seat 3) Star Clone (Seat 4) Port Clone (Seat 5) Star Clone (Seat 6) Port Clone (Seat 7) Star Clone (Seat 8) Port Clone 0 -2 500 0 -500 -4 -1000 -6 -8 -1500 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 Figure 13: Yawing moment lever arms Lyaw (left); yawing moments Myaw (right). Yawing moment lever arms and yawing moments are shown in the two parts of Fig. 13. The plot at the right shows that the total yawing moment is negative for the first two thirds of the drive phase. This means (in our convention) that the bow will tend to be pushed to the staboard side of the hull. During the last third of the drive phase, the total yawing moment is positive, so that the bow will be pushed back towards the port side. The nett effect is that a small rudder correction will be required to keep the boat going on a straight course. Table 3: Oar arrangements used in W8+ examples. Seat 1 2 3 4 5 6 7 8 Normal S P S P S P S P German P S P S S P S P Italian P S S P P S S P Rig A P P S S S S P P Rig B S P P S P S S P Table 3 summarises the oar arrangements used in this example and those to follow. A “P” indicates that the oar is on the port side; an “S” is for a starboard side oar. The last four cases are examples of “zero nett yawing moment” oar arrangements recently investigated by Barrow [?] although they have been known for several decades. As we have already seen in the case of Men’s Fours in Section ??, zero nett yawing moment is difficult to achieve when the crew have different body sizes and each rows slightly differently. The only zero yawing moment arrangement for the 4- class is the Italian Rig in Section ??. For the 8+ class, there are four zero yawing moment arrangements. 9 3.1 W8+: Women’s Coxed Eight (German Rig) To run this example double-click on the icon for the batch file w8 german.bat. 8 4 2 M8+: Arale Clones (Seat 1) Port Clone (Seat 2) Star Clone (Seat 3) Port Clone (Seat 4) Star Clone (Seat 5) Star Clone (Seat 6) Port Clone (Seat 7) Star Clone (Seat 8) Port Clone Total 1000 Yawing Moment (Nm) 6 Yawing moment lever arm (m) 1500 M8+: Arale Clones (Seat 1) Port Clone (Seat 2) Star Clone (Seat 3) Port Clone (Seat 4) Star Clone (Seat 5) Star Clone (Seat 6) Port Clone (Seat 7) Star Clone (Seat 8) Port Clone 0 -2 500 0 -500 -4 -1000 -6 -8 -1500 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 Figure 14: Yawing moment lever arms Lyaw (left); yawing moments Myaw (right). Yawing moment lever arms and yawing moments are shown in the two parts of Fig. 14. This oar arrangement has been used for several years by very successful German Men’s 8+ crews. 10 0.7 0.8 0.9 1 3.2 W8+: Women’s Coxed Eight (Italian Rig) To run this example double-click on the icon for the batch file w8 italian.bat. 8 4 2 M8+: Arale Clones (Seat 1) Port Clone (Seat 2) Star Clone (Seat 3) Star Clone (Seat 4) Port Clone (Seat 5) Port Clone (Seat 6) Star Clone (Seat 7) Star Clone (Seat 8) Port Clone Total 1000 Yawing Moment (Nm) 6 Yawing moment lever arm (m) 1500 M8+: Arale Clones (Seat 1) Port Clone (Seat 2) Star Clone (Seat 3) Star Clone (Seat 4) Port Clone (Seat 5) Port Clone (Seat 6) Star Clone (Seat 7) Star Clone (Seat 8) Port Clone 0 -2 500 0 -500 -4 -1000 -6 -8 -1500 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 Figure 15: Yawing moment lever arms Lyaw (left); yawing moments Myaw (right). Yawing moment lever arms and yawing moments are shown in the two parts of Fig. 15. 11 0.7 0.8 0.9 1 3.3 W8+: Women’s Coxed Eight (Rig A) To run this example double-click on the icon for the batch file w8 riga.bat. 8 4 2 M8+: Arale Clones (Seat 1) Port Clone (Seat 2) Port Clone (Seat 3) Star Clone (Seat 4) Star Clone (Seat 5) Star Clone (Seat 6) Star Clone (Seat 7) Port Clone (Seat 8) Port Clone Total 1000 Yawing Moment (Nm) 6 Yawing moment lever arm (m) 1500 M8+: Arale Clones (Seat 1) Port Clone (Seat 2) Port Clone (Seat 3) Star Clone (Seat 4) Star Clone (Seat 5) Star Clone (Seat 6) Star Clone (Seat 7) Port Clone (Seat 8) Port Clone 0 -2 500 0 -500 -4 -1000 -6 -8 -1500 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 Figure 16: Yawing moment lever arms Lyaw (left); yawing moments Myaw (right). Yawing moment lever arms and yawing moments are shown in the two parts of Fig. 16. 12 0.7 0.8 0.9 1 3.4 W8+: Women’s Coxed Eight (Rig B) To run this example double-click on the icon for the batch file w8 rigb.bat. 8 4 2 M8+: Arale Clones (Seat 1) Star Clone (Seat 2) Port Clone (Seat 3) Port Clone (Seat 4) Star Clone (Seat 5) Port Clone (Seat 6) Star Clone (Seat 7) Star Clone (Seat 8) Port Clone Total 1000 Yawing Moment (Nm) 6 Yawing moment lever arm (m) 1500 M8+: Arale Clones (Seat 1) Star Clone (Seat 2) Port Clone (Seat 3) Port Clone (Seat 4) Star Clone (Seat 5) Port Clone (Seat 6) Star Clone (Seat 7) Star Clone (Seat 8) Port Clone 0 -2 500 0 -500 -4 -1000 -6 -8 -1500 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 t/ts 0.6 Figure 17: Yawing moment lever arms Lyaw (left); yawing moments Myaw (right). Yawing moment lever arms and yawing moments are shown in the two parts of Fig. 17. 13 0.7 0.8 0.9 1
© Copyright 2026 Paperzz