80 CHAPTER 6 FABRICATION OF PROTOTYPE: PERFORMANCE RESULTS AND DISCUSSIONS 6.1 INTRODUCTION The proposed permanent magnet brushless dc motor has quadruplex winding redundancy armature stator assembly, permanent magnet rotor assembly and Hall sensor assembly. The design of quadruplex winding redundancy permanent magnet brushless dc motor was worked out in chapter 3 and chapter 4 as per the requirement specification of the electromechanical actuator application. Finite element based analysis are carried out in chapter 5 to validate the design parameters and to simulate the torque output. Based on the analysis results, two armature stator assemblies (48 slots armature and 60 slots armature) and a common permanent magnet rotor assembly is fabricated. The torque output and back-emf are measured experimentally to study the performances of both the motors. 6.2 SELECTION OF MATERIALS The materials and components for the stator assembly, rotor assembly and Hall sensor assembly are selected based on the operating temperature and application perspective. The material quality is classified as commercial grade, MIL grade and aerospace grade. For the proposed application, aerospace grade materials and components are used.
81 1. Permanent magnets Samarium Cobalt (Sm2Co17) 2. Rotor return ring Magnetic stainless steel 3. Bonding resin Bacon compound 4. Stator core lamination M19, 29 gage silicon steel 5. Slot insulation Kapton polyimide film 6. Copper wire NEMA standard aerospace qualified 7. Lead wires Teflon wires 8. Solder Sn96/Ag04 9. Flux Liquid flux 10. Sleeves Heat shrinkable sleeves 11. Epoxy resin Dobeckot resin and hardener 12. Potting compound Bacon compound 13. Hall sensors A3187LUA latch type 6.3 QUADRUPLEX WINDING REDUNDANCY STATOR Figure 6.1 shows the stage photos of fabrication of armature stator assembly. Figure 6.2 and 6.3 shows the prototype of integral slot and fractional slot armature stator assembly with four three phase lead wires taken out for four motors in single unit for quadruplex winding redundancy. Figure 6.1 stage photos of armature stator assembly The stator is similar to induction motor stator with three phase star connected winding housed in the slots of laminated magnetic core. The
82 quadruplex armature winding is such that each quadrant of the motor behaves as an independent motor with physical isolation between all four quadrant windings for reliability and redundancy. M19, 29 gage silicon steel electrical sheet wire-cut laminations are used for stacking the armature magnetic core. The laminations are stacked for the required length using the stacking fixture and Dobeckot resin is applied to bond the core. The magnetic core is machined for outer diameter grinding and key way milling for the interface to the actuator mechanism. Aerospace grade class C NEMA standard copper wire is used for winding the coils. All four quadrants coils are wound and interconnected separately for physical isolation to meet the four quadrant redundancy. Figure 6.2 48 slots: Potted quadruplex armature stator assembly Figure 6.3 60 slots: Potted quadruplex armature stator assembly
83 The individual coils are wound using winding machine shown in Figure 6.4.a. and inserted in the appropriate slots as per the winding pattern. The coils are interconnected using the soldering station shown in Figure 6.4.b. Teflon lead wires are used to take lead wires of three phase windings. The stator assembly is encapsulated with epoxy potting compound for better thermal conductivity and mechanical protection. (a) (b) Figure 6.4 (a) Winding machine (b) Soldering station 6.4 PERMANENT MAGNET ROTOR The rotor assembly has high grade rare earth permanent magnets housed on the magnetic return ring in alternate polarity to provide required airgap flux for the motor output. The rotor assembly design conforms to the quadruplex armature in which each quadrant armature and permanent magnet rotor performs independently as separate motor. The potted permanent magnet rotor assembly is shown in Figure 6.5. The rotor assembly houses magnetic stainless steel wire-cut ring and high energy Samarium Cobalt permanent magnets. Sixteen radially oriented magnets are arranged in alternate polarity on the slots of the magnetic ring. The magnets are numbered and marked for its polarity to ensure alternate polarity fixing. Epoxy resin is used to bond the
84 magnets with the magnetic return ring. The permanent magnet rotor assembly is potted to protect the magnets. Figure 6.5 Potted permanent magnet rotor assembly 6.5 HALL SENSOR ASSEMBLY The Hall sensor assembly has three sets of three Hall sensors separated by 120 degree electrical from each other for triplex redundancy. Triplex redundancy design is adopted as per the design requirement for improving the reliability of the system. The Hall sensors are positioned and adhesively fixed in the sheet. The assembly is potted to form a separate sensor assembly ring. Figure 6.6 shows the latch type Hall sensors and sensor assembly. The ring is aligned with respect to phase windings and adhesively fixed with the stator assembly. The sensors signals are used to commutate the windings for six step commutation logic to drive the motor. Figure 6.6 Triplex redundancy Hall sensor assembly
85 6.6 BACK-EMF AND TORQUE MEASURING SETUP The performance of the prototype motors are evaluated using the back- EMF and torque measurement setup. The back-emf measuring set up is shown in Figure 6.7. Figure 6.7 Back-EMF measuring setup The quadruplex winding armature stator and permanent magnet rotor are assembled in the back-emf testing fixture. External drive induction motor shaft in the test fixture is coupled to the prototype motor. The external drive is controlled to vary the speed of the prototype motor. The back-emf voltage generated by the winding phases at various speeds is measured to calculate the back-emf constant. The measurements are taken at three different speeds for all four quadrants windings for performance comparison. True RMS voltmeter and digital storage oscilloscope are used to measure the readings. The torque measuring setup shown in Figure 6.8 has torque testing fixture, MODIA torque pickup, speed transducer and power supplies. The static torque is measured using this pickup. The quadruplex winding armature stator is fixed to the test bench and the permanent magnet rotor is coupled to
86 the torque pickup. The windings are energized with the peak load current of 13 Ampere. The stall torque is measured for all four quadrant motors. Figure 6.8 Torque measuring setup 6.7 IDENTIFICATION OF PHASE LEADS COLOUR Motor leads Phase A Phase B Phase C Red White Black Sensor leads Ha Brown Hb Orange Hc Yellow Supply Blue Ground Green
87 6.8 48 SLOTS 16 POLES TEST RESULTS 6.8.1 48 Slots: Resistance and Inductance Measurement Table 6.1 48 slots stator: Resistance and inductance measurement S.No. Winding Set 1-1 2-2 3-3 4-4 Winding Sequence Winding Resistance ) Winding Inductance (mh) Red - White 2.816 9.047 White - Black 2.820 9.960 Black Red 2.822 10.15 Red - White 2.812 8.792 White - Black 2.812 9.901 Black - Red 2.812 10.04 Red - White 2.825 8.883 White - Black 2.831 9.949 Black - Red 2.822 10.13 Red - White 2.815 9.020 White - Black 2.821 9.921 Black - Red 2.817 10.06 Table 6.2 48 slots stator: Insulation resistance measurement S.No. Terminals Insulation 1 2 3 4 Between stator body & winding together () Between stator body & quadrant 1 winding Between stator body & quadrant 2 winding Between stator body & winding 3 and winding 4 >220 M @250 V DC >220 M @250 V DC >220 M @250 V DC >220 M @250 V DC
88 6.8.2 48 Slots: No-Load Speed at 75V DC Table 6.3 48 slots stator: No-load speed measurement S. No. Winding energized 1 1 2 2 3 3 4 4 Hall sensor energized 5-1 -2-3 6-2 -2-3 7-3 -2-3 8-4 -2-3 9 10 11 12-2-3 Set - 1 Set - 2 Set - 3 Dir Speed Current in RPM (A) CW 990 0.4 CCW 990 0.4 CW 994 0.4 CCW 989 0.4 CW 1004 0.4 CCW 993 0.4 CW 1003 0.4 CCW 1002 0.4 CW 1006 0.4 CCW 1003 0.4 CW 1016 0.4 CCW 1005 0.4 CW 992 0.4 CCW 991 0.4 CW 994 0.4 CCW 991 0.4 CW 1004 0.4 CCW 993 0.4 CW 981 0.4 CCW 980 0.4 CW 984 0.4 CCW 982 0.4 CW 994 0.4 CCW 984 0.4 CW 993 0.4 CCW 990 0.4 CW 1006 0.4 CCW 1003 0.4 CW 994 0.4 CCW 990 0.4 CW 985 0.4 CCW 981 0.4 CW 1022 0.4 CCW 1019 0.4 CW 1019 0.4 CCW 1019 0.4 CW 1021 0.4 CCW 1020 0.4 CW 1032 0.4 CCW 1021 0.4
89 6.8.3 48 Slots: Back-EMF Measurement Table 6.4 48 slots stator: Back-EMF measurement Speed (rpm) 1000 500 300 Quad 1 Quad 2 Quad 3 Quad 4 Sequence Back- Back- Back- Back- EMF EMF EMF EMF Voltage Voltage Voltage Voltage Red White 53.8 54.2 54.6 54.4 White Black 54.0 54.1 54.6 54.5 Black Red 53.9 54.0 54.5 54.5 Red White 26.9 27.1 27.4 27.3 White Black 27.0 27.1 27.3 27.2 Black Red 27.2 27.0 27.5 27.3 Red White 16.1 16.3 16.4 16.6 White Black 16.1 16.3 16.6 16.6 Black Red 16.1 16.3 16.5 16.7 6.8.4 48 Slots: Stall Torque Measurement Table 6.5 48 slots stator: Stall torque measurement S. Current Torque Coil Hall Sensor Direction No (A) (Nm) 1 Quad -1 Set-1-2-3 CW 12.86 7.4 ACW 13.46 7.34 2 Quad -2 Set-1-2-3 CW 12.67 7.2 ACW 13.22 7.2 3 Quad -3 Set-1-2-3 CW 12.98 7.3 ACW 13.23 7.34 4 Quad -4 Set-1-2-3 CW 12.89 7.5 ACW 13.17 7.4 5 Quad Set-1-2-3 CW 50.6 30.9 ACW 52.6 26.9 6 Quad Set-1 CW 50.6 30.9 ACW 52.5 27 7 Quad Set-2 CW 50.7 30.9 ACW 52.5 26.9 8 Quad Set-3 CW 50.6 30.9 ACW 52.5 26.9
90 6.9 60 SLOTS 16 POLES TEST RESULTS 6.9.1 Resistance and Inductance Measurement Table 6.6 60 slots stator: Resistance and inductance measurement S.No. Winding Set 1-1 2-2 3-3 4-4 Winding Sequence Winding resistance ) Winding inductance (mh) Red - White 2.700 8.548 White - Black 2.699 9.038 Black Red 2.697 9.205 Red - White 2.708 8.480 White - Black 2.708 8.970 Black - Red 2.704 9.105 Red - White 2.720 8.409 White - Black 2.714 8.869 Black - Red 2.709 9.112 Red - White 2.700 8.398 White - Black 2.698 8.877 Black - Red 2.697 8.999 Table 6.7 60 slots: Insulation resistance measurement S.No. Terminals Insulation measurement 1 2 3 4 Between stator body & winding together () Between stator body & quadrant 1 winding Between stator body & quadrant 2 winding Between stator body & winding 3 and winding 4 >220 M @250 V DC >220 M @250 V DC >220 M @250 V DC >220 M @250 V DC
91 6.9.2 60 Slots: No-Load Speed at 75V DC Table 6.8 60 slots stator: No-load speed measurement S. No. Winding energized 1-1 2-2 3-3 4-4 Hall sensor energized 5-1 -2-3 6-2 -2-3 7-3 -2-3 8-4 -2-3 9 10 11 12-2-3 Set - 1 Set - 2 Set - 3 Dir Speed Current in RPM (A) CW 1059 0.5 CCW 1058 0.5 CW 1062 0.5 CCW 1061 0.5 CW 1061 0.5 CCW 1063 0.5 CW 1046 0.5 CCW 1045 0.5 CW 1047 0.5 CCW 1049 0.5 CW 1048 0.5 CCW 1050 0.5 CW 1049 0.5 CCW 1048 0.5 CW 1051 0.5 CCW 1052 0.5 CW 1051 0.5 CCW 1050 0.5 CW 1057 0.5 CCW 1056 0.5 CW 1058 0.5 CCW 1069 0.5 CW 1059 0.5 CCW 1063 0.5 CW 1063 0.5 CCW 1052 0.5 CW 1048 0.5 CCW 1048 0.5 CW 1051 0.5 CCW 1058 0.4 CW 1058 0.5 CCW 1059 0.5 CW 1086 0.5 CCW 1089 0.5 CW 1085 0.5 CCW 1085 0.5 CW 1086 0.5 CCW 1090 0.5 CW 1086 0.5 CCW 1093 0.5
92 6.9.3 60 Slots: Back-EMF Measurement Table 6.9 60 slots stator: Back-EMF measurement Speed (rpm) 1000 500 300 Quad 1 Quad 2 Quad 3 Quad 4 Sequence Back- Back- Back- Back- EMF EMF EMF EMF Voltage Voltage Voltage Voltage Red White 51.0 51.5 51.1 51.1 White Black 51.2 51.6 51.1 51.2 Black Red 51.1 51.6 51.1 51.3 Red White 25.6 25.6 25.4 25.3 White Black 25.7 25.6 25.4 25.4 Black Red 25.5 25.8 25.5 25.5 Red White 15.2 15.3 15.0 15.0 White Black 15.2 15.4 15.0 15.1 Black Red 15.1 15.2 15.0 15.2 6.9.4 60 Slots: Stall Torque Measurement Table 6.10 60 slots stator: Stall torque measurement S. Current Torque Coil Hall Sensor Direction No (A) (Nm) 1 Quad -1 CW 13.26 7.1 Set-1-2-3 ACW 12.83 6.9 2 Quad -2 CW 13.27 7.1 Set-1-2-3 ACW 12.70 7 3 Quad -3 CW 13.39 7.1 Set-1-2-3 ACW 12.92 6.9 4 Quad -4 CW 13.32 7.2 Set-1-2-3 ACW 12.88 6.9 5 Quad CW 52.7 28.9 Set-1-2-3 ACW 50.9 26.9 6 Quad CW 52.6 28.8 Set-1 ACW 50.9 26.9 7 Quad CW 52.6 28.8 Set-2 ACW 50.4 26.9 8 Quad CW 52.6 28.8 Set-3 ACW 51 26.9
93 6.10 HIGHLIGHTS OF THE TEST RESULTS 1. The line to line resistance and inductance value (Spec: 2.4±10%, 9.0± 30% mh). 48 slots armature stator: R= 2.8 and L= 9.0 mh 60 slots armature stator: R= 2.7 and L= 9.0 mh 2. Insulation resistance (Spec: >100M at 250 V DC) The insulation value is greater than 220 M at 250 V DC between all lead wires and body for both 48 slots and 60 slots armature stator. 3. No-load speed at 75 V DC: (Spec: 1000 rpm at 75 V DC) The 48 slots motor runs around 990 rpm when one quadrant is energized and around 1020 rpm when all four quadrants and three Hall sensor sets are energized together. The 60 slots motor runs at 1060 rpm during one quadrant excitation and 1090 rpm when excited by all four quadrants together with three Hall sensor sets. Both clockwise and counter clockwise speeds are almost same. 4. Back-EMF reading: (Spec: 0.7 ± 7 % V/(rad/sec)) 48 slots motor: 54V at 1000 rpm, 0.72 V/(rad/sec) 60 slots motor: 51V at 1000 rpm, 0.68 V/(rad/sec) 5. Stall torque: (Spec: 8 Nm per quadrant, 32 Nm for all four quadrants) 48 slots stator: 7.4 Nm per quadrant 30.9 Nm when all four quadrants energized 60 slots stator: 7.1 Nm per quadrant 28.8 Nm when all four quadrants energized.
94 6.11 FREQUENCY RESPONSE TEST Frequency response at Actuator level and System level The motor assembly is subjected to frequency response test at actuator level and system level. The signal input to the motor is shown in red colour and the response is in blue colour curve. Slow Sine: Actuator level: QM 06 (60 slots): Amplitude 0.5-0.5 114 118 122 Time, ms Figure 6.9 60 slots: Actuator level slow sine response Slow Sine: System Level: QM 06 (60 slots) 0.5 Amplitude -0.5 28 30 32 Time, ms Figure 6.10 60 slots: System level slow sine response
95 Slow Sine: Actuator level: QM 05 (48 slots): 0.5 Amplitude -0.5 7 9 11 Time, ms Figure 6.11 48 slots: Actuator level slow sine response Slow Sine: System level: QM 05 (48 slots) 0.5 Amplitude 0-0.5-1.0 42 44 46 Time, ms Figure 6.12 48 slots: System level slow sine response Step response at 10%: Actuator level QM 05 (48 slots) 0.5 Amplitude -0.5 9 13 Time, ms Figure 6.13 48 slots: Actuator level step response
96 Step response at 10%: System level QM 05 (48 slots) 0.5 Amplitude -0.5 9 13 Time, ms Figure 6.14 48 slots: System level step response Step response at 10%: Actuator level QM 06 (60 slots) 0.5 Amplitude Amplitude -0.5 43 47 Time, ms Figure 6.15 60 slots: Actuator level step response Step response at 10%: System level QM 06 (60 slots) 0.5-0.5 44 48 52 Time, ms Figure 6.16 60 slots: System level step response
97 6.12 PERFORMANCE STUDY AND IMPROVEMENT The 60 slots motor configuration has (5-7%) lower back-emf and peak torque compared to 48 slots motor configuration for the common permanent magnet rotor. But the 60 slots configuration without skew has significantly lower cogging torque compared to 48 slots configuration having one slot pitch skew. Even though the motor level performance of 60 slots motor is lower than 48 slots motor, the slow sine and step response is better than the 48 slots configuration at both actuator level and system level as shown from Figure 6.9 to 6.16. Hence the 60 slots armature stator having significantly lower cogging torque, meeting the overall requirement specification of the quadruplex torque motor looks to be a better choice for this high precision application of electromechanical actuator for space mechanism. The improvement in the torque/back-emf constant within the same 60 slots configuration is studied further. 6.13 PERFORMANCE IMPROVEMENT IN 60 SLOTS STATOR The seven percent performance improvement is achieved by increasing two turns in the stator coils. The double wire strategy is adopted in the coil to keep the phase to phase resistance within the specification (2.4 ± 10%) and to accommodate additional conductors in the existing slot area. The coils with two wires are machine wound from two spools for compactness and consistancy. The no-load performance of new fractional slot stator confirms the improvement of seven percent in the back-emf constant and the terminal
98 resistance and inductance are also well within the specification. The improvement details are given below QM 06 (60 slots) New 60 slots 1. Number of turns per coil 26 28 2. Winding wire SWG 23 SWG 26 * 2 3. Armature resistance(line to line) 2.7 2.5 4. Armature inductance 9 mh 9mH 5. No load speed 1080 rpm 998 rpm The improved quadruplex armature fractional slot stator is fabricated and tested. The optimal design of magnetic loading and electrical loading is achieved within the volume constraints and with M19, 29 gage silicon steel lamination material (Saturation 1.9 Tesla) for magnetic core. 6.14 IMPROVED 60 SLOTS STATOR TEST RESULTS 6.14.1 Resistance, Inductance and Back-EMF Table 6.11 Phase to phase resistance ( ) for improved 60 slots stator Resistance, Phase A-B B-C C-A 1 2.492 2.488 2.490 2 2.496 2.493 2.493 3 2.495 2.496 2.494 4 2.506 2.505 2.502
99 Table 6.12 Phase to phase Inductance (mh) for improved 60 slots stator Inductance, mh Phase A-B B-C C-A 1 8.939 9.529 9.733 2 8.87 9.518 9.702 3 8.911 9.56 9.731 4 8.849 9.57 9.706 Table 6.13 Back-EMF for improved 60 slots stator Back-EMF at 1000 rpm Voltage, V Phase A-B B-C C-A 1 54.70 54.82 54.79 2 54.77 54.82 54.81 3 54.02 54.14 54.08 4 54.54 54.52 54.47 The resistance, inductance and back-emf are measured on the improved fractional slot motor for all four quadrants. Table 6.11, 6.12 and 6.13 shows the values of resistance measured across phase to phase winding leads at room temperature, inductance and back-emf respectively. The back- EMF readings are measured by coupling the magnet rotor with external drive motor rotated at 1000rpm. The values are almost equal in all four quadrants eliminating the circulating current flow in the circuit. Figure 6.17, 6.18, 6.19 and 6.20 are three phase back-emf waveforms captured for all four quadrants when the rotor driven at 1000 rpm. The back-emf constant is 0.72 V/(rad/s) and almost same in all four quadrants and meets the requirement specification.
100 Figure 6.17 Quad 1: 3 back-emf Figure 6.18 Quad 2:3 back-emf Figure 6.19 Quad 3:3 back-emf Figure 6.20 Quad 4:3 back-emf
101 6.14.1 No-Load Speed (improved 60 slots stator) Table 6.14 No-load speed of improved 60 slots stator S. No. Winding energized 1 1 2 2 3 3 4 4 Hall sensor energized 5-1 -2-3 6-2 -2-3 7-3 -2-3 8-4 -2-3 9 10 11 12-2-3 Set - 1 Set - 2 Set - 3 Dir Speed Current in RPM (A) CW 947 0.6 CCW 949 0.6 CW 949 0.6 CCW 951 0.6 CW 956 0.6 CCW 953 0.6 CW 950 0.6 CCW 951 0.6 CW 951 0.6 CCW 954 0.6 CW 953 0.6 CCW 954 0.6 CW 960 0.6 CCW 957 0.6 CW 952 0.6 CCW 954 0.6 CW 960 0.6 CCW 961 0.6 CW 957 0.6 CCW 960 0.6 CW 967 0.6 CCW 962 0.6 CW 957 0.6 CCW 960 0.6 CW 950 0.6 CCW 953 0.6 CW 952 0.5 CCW 954 0.6 CW 958 0.6 CCW 958 0.6 CW 949 0.6 CCW 947 0.6 CW 996 0.6 CCW 996 0.6 CW 997 0.6 CCW 997 0.6 CW 1005 0.6 CCW 998 0.6 CW 998 0.6 CCW 996 0.6
102 6.15 SUMMARY The quadruplex winding redundancy permanent magnet brushless dc motor is developed based on the design and simulation results. The integral slot stator, fractional slot stator and the common permanent magnet rotor are fabricated for performance comparison of the torque output. The quadruplex redundancy armature stator assembly and permanent magnet rotor assembly is assembled in the testing fixture to measure the back-emf generated and torque output. Phase to phase resistance, phase to phase inductance and insulation values are measured and tabulated for both the configurations. The no-load speed at 75 V DC, back-emf at 1000 rpm and stall torque output of the motor are measured for all four quadrants of the motor and the results are tabulated. The test data shows that the performances of all four quadrants are identical and meets the requirement of reliability. The motors are assembled in the actual mechanism and the response is plotted for step input and slow sine input at actuator level and system level. The 60 slots 16 poles motor configuration has better response characteristics both at actuator level and system level compared to 48 slots 16 slots motor configuration. But the motor level performance of the 60 slots stator is lower than the 48 slots stator. The electrical loading is further increased in 60 slots stator to match the performance of 48 slots stator. The improved 60 slots stator is fabricated and tested for the performance. The no-load speed and back-emf are measured for the improved 60 slots stator and the results are tabulated. The back-emf is improved by 7% and the torque improvement is seen in no-load speed. The electrical loading and magnetic loading of the motor are optimised for the given volume constraint. Further the study on stator core material property is carried out. The Cobalt Iron alloy material has higher saturation flux density limit than the conventional M19 silicon steel lamination material. Hence the performance output of the 60 slots motor configuration is simulated with Cobalt Iron alloy material for stator core lamination.