Design Test Of Electromechanical Actuators For Thrust Vector . - NASA
Design and Test of Electromechanical Actuators for ThrustVector ControlJ. R. Cowan* and Rae Ann Weir*National Aeronautics and Space (AdministrationMarshall Space Flight CenterABSTRACTNew control mechanisms technologies are currently beingexplored to provide alternatives to hydraulic thrust vector control(TVC) actuation systems. For many years engineers have beenencouraging the investigation of electromechanical actuators (EMA) totake the place of hydraulics for spacecraft control/gimballing systems.The rationale is to deliver a lighter, cleaner, safer, more easilymaintained, as well as energy efficient space vehicle. In light of thiscontinued concern to improve the TVC system, the PropulsionLaboratory at the NASA George C. Marshall Space Flight Center (MSFC)is involved in a program to develop electromechanical actuators forthe purpose of testing and TVC system implementation. Through thiseffort, an electromechanical thrust vector control actuator has beendesigned and assembled. The design consists of the following majorcomponents: Two three-phase brushless dc motors, a two pass gearreduction system, and a roller screw, which converts rotational inputinto linear output. System control is provided by a solid-stateelectronic controller and power supply. A pair of resolvers andassociated electronics deliver position feedback to the controller suchthat precise positioning is achieved. Testing and evaluation iscurrently in progress. Goals focus on performance comparisonsbetween EMAs and similar hydraulic systems.INTRODUCTIONRecent studies have shown that hydraulic actuation systems costthe space program many valuable hours for tests, maintenance, andrepairs. During the typical turnaround cycle for a space shuttle orbiterand its integrated systems, many maintenance personnel inspect theentire vehicle, repairing hydraulic leaks and examining lines, while atthe same time qualifying each hydraulic unit for its next flight.Qualification alone necessitates extensive hours, or about 10 per centof the total inspection time. Estimates submit that fully electricorbiters could possibly be readied for flight ten days earlier than
hydraulic ones. These problems affecting mission readiness haveprompted investigations by NASA into alternate actuation systems foruse in existing space applications, as well as new programs solicitingheavy lift TVC technology. Some reservations of implementing electricTVC systems into these very new programs overshadow the fact thatEMAs have been in service for more than thirty years. A goodexample of an early EMA technology application is the RedstoneMissile in the 1950's. In this system, an electrical chain drive actuatedair fins for aerodynamic steering. As early as 1972, NASA engineersexpressed concern over the space shuttle's hydraulic system due todifficult maintainability and some minor inefficiencies. In the last fewyears, many advances in the fields of power electronics and motortechnology have renewed interest in the use of EMAs for both low andhigh power actuation in space applications. In 1987, the ControlMechanisms and Propellant Delivery Branch at MSFC designed andtested an electromechanical propellant valve actuator applicable to thespace shuttle main engine. It performed as well as, and in some areasbetter than its hydraulic counterpart. Therefore, realizing the potentialwhat this new science can provide for space vehicle actuation systems,MSFC has undertaken this project to explore the full potential of EMAs.After fine tuning of the hardware and system, implementation of afamily of EMAs designed for use in multiple applications will follow.DESIGNThe electromechanical thrust vector control (EMTVC) actuator wasdesigned to meet basic loading requirements necessary to supportfuture heavy lift space vehicles. These requirements are given inTable 1. The primary mechanical components of an EMTVC actuatorcan be seen in Figure 1. These components are, 1) some type ofelectric motor, 2) if needed, a gear train, and 3) a linear screw. Figure2 shows an assembly drawing of the EMTVC.Table 1.Basic Design RequirementsDynamic Load Capacity: 35,000 lbLinear Velocity: 5 in/secMaximum Stroke: /- 6 inControl: Two Channel RedundantBandwidth: 3.0 HzLinearity: 2%Accuracy: 0.050 inch
I.MotorsSeveral types of motors were considered for this application.Through careful deliberation and various trade studies, three-phase,permanent magnet (PM), brushless, direct current (DC) motorscontaining a large number of poles were chosen. A fundamentallysound design and combined overall characteristics helped finalize thischoice. PM, brushless motors have high torque-to-weight ratios andhigh torque capability at low speeds. They are able to operate at highspeeds with moderately linear output (torque versus speed) curves.The most significant feature of the brushless PM motor that provesmost beneficial to the actuator is its ability to integrate into aredundant single shaft system. In this arrangement, a shortedwinding proves to be less of a problem than would be expected. Ashorted winding in a redundant electric motor system forces the fullyoperational motor to overcome drag torque produced by the failedmotor. To overcome this drag torque, or generator effect, requiresextreme overdesign in motor sizing unless a type of clutching device isused to separate the systems. Figure 3 shows important datanecessary to the design of the entire motor scheme. As can be seen bythe curve, at higher speeds drag torque decreases sharply. Thisdecrease in drag torque occurs due to the multiplicative property thatsignal frequency has upon the inductive element in an electric motor.More poles in the motor make available a larger reactive component ofthe impedence, which increases the total impedence. Thus, a smallercurrent through the motor winding is created and, therefore, lesspower to cause drag on the system. Others have done work using thistheory and have proven it with credible results.Characteristics of the motor used are:Type: Three-phase brushless dcNo Load Speed: 9300 RPM @ 270 voltsTorque Constant: 34.6 oz-inlampeBack EMF Constant: 25.6 v/1000 rpmDimensions: 5.50 inch O.D.x 5.045 inch lengthaWeight: 17 lb.B.
11.G e a r SystemTo satisfy linear velocity requirements, a speed reduction isneeded to the output shaft of approximately 9:l. The two pass gearsystem utilizes a design such that backlash is nearly eliminated (Figure4).Spur gears transmit high torques necessary to drive the system ineither direction. A two piece idler shaftlgear allows for on assemblyadjustment to aid in minimizing rotational play. Characteristics of thegear system are shown below:eeeeeType of Gearing: SpurTeeth: Involute, 20 degFace Width: 0.50 in. (1st pass)0.75 in. (2nd pass)Material: Steel alloy 8620Lubricant: Molybdenum disulfide GreaseInternal gear stresses and tooth contact stresses were calculated,which directly affected structural sizing of the gears.111.Linear ScrewRotational motion is converted to linear motion using a rollerscrew. Roller screws are high efficiency linear devices which provide arobust means of transmitting very high loads with considerableaccuracy. They consist of a threaded screw shaft and a nut whichhouses contacting rolling elements (Figure 5). Triangular threads, withan included angle of 90 degrees, are machined onto the main screwshaft. Thread pitch may range from 0.015 inch to as much as 1.250inches with 4, 5, or 6 starts. The rollers housed in the nut aremachined with a single start triangular thread. Contact is madebetween the nut and shaft by the rollers. A barrelled thread formprovides a large contact radius for high load carrying capacity andrigidity.Two critical areas of highest concern were:1) Dynamic load capacity for the given geometric envelope2) Shock load capabilityThe Dynamic load rating in an application depends on the typeand magnitude of the load applied, and the life of the screw in millionsof revolutions. In a TVC system, the maximum dynamic load is only
experienced at very short intervals during a flight. This characteristicduty cycle aids in compacting the actuators' geometry which, of course,is much to the advantage of the overall system design. Extreme shockloads and adverse environments may also be encountered on amission. Transient shocks much larger than those loads experiencedunder normal continuous operation may be experienced by a TVCactuator at engine start up. Documentation shows that roller screwsare best suited for these conditions. Data for the roller screw used areas follows:.00Material: Shaft - 4140, Nut - 52100Lead: 0.40 incheslrevShaft Dia: 1.89 inchesLubrication: Molybdenum disulfide GreaseOther linear actuation devices were considered for thisapplication, although none proved as worthy, based on all literatureand performance data, as the roller screw.IV.Electronic ControllerThe Control Electronics Branch of the Information and ElectronicsSystems Laboratory was responsible for the design and fabrication ofthe analog controller (270V, 100A, 27kW) for the TVC actuator system.A switching regulator in the controller pulse width modulates (PWM)the 270 volt power source (currently provided by a battery bank).The modulated source is then passed through a coupling inductor toprovide current for a three-phase, six transistor, six step, bridgenetwork. The six transistors in this network aided by two additionaltransistors used in the PWM process, as well as the regenerativecircuitry are insulated gate bipolar transistors (IGBT) rated at 500V,200A. This bridge assembly provides the correct commutation ofcurrent to the motor windings. Synchronization is achieved byutilizing output applied to the commutation logic from motor HallEffect devices. This logic also protects the circuit by ensuring that bothtransistors of a phase are not turned on simultaneously. Current issensed out of the inductor by a separate Hall Effect device for currentfeedback to the controller. Position feedback is provided by theresolver at the output of the gear train. This signal is thencompensated for the difference between measured position and theactuator position (at the output of the roller screw) before it is usedfor feedback to the controller. For redundancy purposes, a controllerfor each motor will be built.
TESTINGI,FacilityA rigorous testing program is currently under way in a full scalehydraulic test facility at MSFC. Component and subsystemdevelopment from concept through flight qualification can beperformed. This facility, originally constructed for Apollo and SpaceShuttle TVC systems, contains operational test facilities such as fluidpumping systems, flow test benches, static load application testfixtures, and dynamic inertia simulators which are fully supported byinstrumentation, data acquisition, and analysis equipment. Dataacquisition equipment used for all tests consisted of a 200 Hz, 12 bit, 8channel computer operated system. Data analysis was performedusing MATLAB based program.This facility will allow comparisons to be made of the EMA tosimilar hydraulic provisions. A broad spectrum of capabilities areavailable.Fluid power requirements for the entire facility are provided byseveral pumping systems. Two large units, when combined, have aflow capability of 800 gallmin at 5000 psig. Four small pumpsprovide fluid power to the smaller test fixtures. These units, rated 30gallmin at 3500 psig, may be combined to supply 120 gallmin totalcapacity. A separate pumping network rated 15 gal/min at anoperating pressure of 8000 psig supports a high pressure prototypesystem.Hydraulic flow benches deliver directional flow and operationalcontrol to various test panels. Fluid manifolds and test blocks areavailable to interface with all standard servovalve and high flowdeflector jet types. Pressure can be regulated from start-up to fullsystem capacity. Flow and pressure instrumentation is available inreal-time.A dynamic load simulator (Figure 6), originally configured tosimulate structural compliance, inertia, and mounting provisions forthe TVC system on the Solid Rocket Booster (SRB) of the Space Shuttlefacilitates testing of the EMA. Tests such as frequency response,stability, and step response are discussed in more detail in the nextsection.
Load stand characteristics are:Pendulum mass: 5000 lbmMoment arm: 65 inDynamic spring rate: 140 MlblinPower: '27 kwTesting Scheme/ResultsThe first phase of testing has been completed. Data analysis forthis series of tests was performed and documented and will be used toupdate mathematical models of the system. Tests include stepresponse, discrete sine dwells, frequency sweep response, andlinearity. In addition, actual flight duty cycles were performed by theEMA and hydraulic systems. These tests determined the performanceparameters of the actuator for comparison against design parameters.All design parameters were verified with the exception of piston rateunder maximum load. Rate-vs-Load tests were omitted due toinadequacies encountered with the load fixtures. These tests will beperformed during the second phase of the testing program.All tests were executed at full power (270V, 100A) and aresummarized below. Peak power reached 33.75 kW when the currentspiked to 125 amps. Five channels of data were acquired relating thefollowing: command signal, actuator position, load position, motorcurrent, and supply current.Figure 7 shows a plot of the frequency response of the actuator.The envelope around the data exhibits the current SSME requirement.The bandwidth is 4 Hz with 20-25 degrees of phase lag at 1 Hz (anSSME requirement). Resonance with the load structure occurs between8 and 9 Hz. The peak in response magnitude corresponds to a criticaldamping ratio between 0.5 and 0.6. Figure 8 shows both a small andlarge excursion step response. The small step (0.25 inches) falls withinthe requirement envelope of SSME specifications. A 15 percentovershoot is seen. Data from the large step (5.0 inches) demonstratesthat the actuator exceeds the design requirement velocity of 5 in/sec,and is actually capable of 6.8 in/sec under inertia load. The overshootcorresponding to the large step is approximately 14 percent.Overshoot associated with these step responses relate to the dampingratio determined by the frequency response. During the next phase oftesting, a piston velocity of 5 in/sec will be verified with the actuatorunder rated load. Linearity tests on the actuator produced excellentresults. Position data for a large excursion (5.0 inches) resulted in anerror less than 0.030 inch which exceeds the 0.050 inch requirement.11.
Both linearity and position error data may be seen in Figure 9. Theposition error for the small excursion was below the noise level of thedata. For an initial comparison between the hydraulic and EMAsystems, an actual STS flight profile was commanded to the actuator.Response to the STS-44 SRB command profile may be seen in Figure10. Actuator position response error in relation to the command signalis shown in the bottom plot. Note though, that while the EMA showedless error than the hydraulic unit, the test was run on the inertiasimulator only and lacked the flight loads the hydraulic unitexperienced. The capability to apply flight type loads for testingpurposes is included in future plans for the test facility.Since this testing was performed, a new single pass gear systemhas been designed for the actuator. A rate loop has also been addedand is now undergoing tuning such that a l l design parameters areaccomodated. Following the completion of this task, an additional testseries will be run. Testing the actuator in a redundant configurationwill be the next milestone. These tests will prove helpful in futureredundancy studies as well as the introduction of Vehicle HealthManagement (VHM) to these systems.SECOND GENERATION EMAA second generation high power EMTVC actuator has beendesigned and is currently being assembled at NASA, MSFC. Thisactuator incorporates features that will increase performance, reduceweight, and provide a more compact package' than that of the firstgeneration discussed earlier. The primary mechanical scheme of bothactuators are relatively the same, yet the second generation EMAutilizes features that enhance the entire component design. Thisactuator also incorporates length, stroke, and power capabilitiesrequired to support TVC system testing for heavy lift vehicles.Basic Design Requirements:oDynamic Load Capacity: 45,000 1boNull Length: 47.330 inchesoMaximum Stroke: /- 6 inoControl: Four Channel RedundantoLinear Velocity: 5 in/secoBandwidth: 9 4.2 HzFour high speed low inertia motors configured in a torquesumming arrangement power the system. Since inertia is a majorconcern due to the EMA's nature of operation, optimization ofhorsepower, RPM, and motor rotor diameter is imperative. The
amount of inertia seen by the controller governs the power requiredfor each cycle of the actuator. Calculations show that the amount ofinertia created by the rest of the system is essentially negligible whencompared to the inertia created by the cyclic action of each rotorinertia.The motors deliver torque to a single pass gear reduction system.The gear system transmits the necessary torque to a roller screw shaftwhich has been hollowed to decrease inertia. Torque to the rollerscrew shaft is converted to linear movements by the roller screw nut.As the nut moves, precise gimbal outputs are translated to the outputpiston.A small harmonic drive (Figure 11) has been added to aid inposition control. It provides a reduction mechanism such that themoving resolver race will not rotate greater than 360 degrees.High strength aluminum (7075) is used in more parts to reduceoverall weight. Figure 12 shows the second generation EMAconception. Component specifications are shown in Table 2.Table 2.Component specificationsMotors:oType: Three phase DC brushless permanent magnetoNo load speed: 20,000 RPM @ 230 voltsoTorque constant: 16.8 oz-inlampoBack EMF constant: 13.8 /1000RPMoDimensions: length-9.875 india-2.380 inoWeight: Approx. 6 lbGearing:oType of gears: spuroTeeth: Involute, 20 degoFace width: ,625 inoMaterial: Steel alloy 8620oLubricant: Dry filmLinear Screw:(same specifications as 1st generation EMAdiscussed earlier)
CONCLUSIONConclusive system data concerning the EMA is forthcoming. Finaltesting goals will soon be fulfilled for this phase of MSFC's EMAprogram. These results are expected to supply greater confidence inthe capabilities of EMTVC systems for future comparisons to hydraulicequivalents.Following the final phase of testing, the 25 hp EM actuator will beshipped to Kennedy Space Center, Florida, where it will be used tofamiliarize personnel with operational issues associated with EM TVCsystems. Design changes are being incorporated into the gear trainand motors to better suit data criteria and EMA performance. Problemissues arising from the operations area may then be taken into accountand incorporated into future actuator requirements and designs. Usinginformation gained from experience with the first generationprototype EMA, MSFC will utilize the second generation, 4-motor, 45hp EMA to further provide insight into topics such as load sharing(between channels), full redundancy implementation, and starttransient load capabilities. MSFC test facilities will undergo upgradingto support this effort with additions of flight programmable loads tothe inertia simulators, a possible VHM test platform, and othermodifications to existing facilities to better handle specificrequirements associated with the EMA. After the completion ofextensive component testing, the actuator will undergo TechnologyTest Bed (TTB) qualification . TTB data will include vibration analysis,EMIJEMC, as well as thermal, shock, and acoustic information. The TTBhot fire tests will support the NASA Electrical Actuation TechnologyBridging program.NASA, MSFC plans to investigate all avenues of this technologysuch that optimization is accomplished. Upcoming EMA designs willaid in establishing a continued learning process to integrate test dataand hardware for development of a proven EMTVC system.
Shigley, Joseph Edward, and Charles R. Mischke. Standard Handbook ofMachine Design, McGraw-Hill Book Company, 1986Shigley, Joseph Edward, and Larry D. Mitchell. Mechanical EngineeringDesign, McGraw-Hill Book Company, 1983Weir, R. A., and W. N. Myers. "Electromechanical Propellant ControlSystem Actuator." AIAA 90-1946CatalogsHarmonic Drive, Cup and Pancake Component Gear Sets, 1989.SKF, SKF Planetary and Recirculating Roller Screws, 1990.
BATTERIESIMOTOR 1hr LINEAR SCREW ZQEARINQIFigure 1.Basic EMA SchematicFigure 2. EMA Assembly Drawing11 \I0IInland Motor #RBE-0300s2I4I6I1810Shaft Speed ( R P M x 1000)Figure 3. Motor Drag Torque-vs-RPMFigure 4. View of Gear Arrangement
Figure 5. SKI? Linear Roller ScrewFigure 6. Dynamic Load Simulator (schematic)361
F R E Q U E N C Y (HzNFFI 2848- OVERLAP 1824HSFC EHAF R E Q U E N C Y (Hn 1Figure 7. Frequency Response With SSME Envelope Requirements362
HSFC TEST: t2-3981 STEP RESPONSE 8 . 1 5 HZ 8 . 2 5 INCH 9#81#9z12.412.612.81313.213.413.613.814L?TIHE ( S E C )n S F C TEST: t2-8981 STEP RESPONSE 8.15 HZ 5 INCH 9/81/92Figure 8. Small and Large Step Response
Position Command (Inch)Figure 9. Large Excursion Linearity and Position Error
2 5 HP YUC/Efih,.B E R S U R E D POSITIOH: STS-44-SUB. . . . . . . . . . . . . . . . . . . .TELEMETRY.,. . . . . . . . . . . . . . . . . . . . . . .TiREICSECIR E S O V E R S I G A L U l T l l 25 ltz LOU P A S S F I L T E HTime ( s e c )Figure 10. STS-44Flight Profile Comparison365
lnstaned RelationshipFigure 11. Harmonic DriveFigure 12. Second Generation EMA Assembly366
Laboratory at the NASA George C. Marshall Space Flight Center (MSFC) is involved in a program to develop electromechanical actuators for the purpose of testing and TVC system implementation. Through this effort, an electromechanical thrust vector control actuator has been designed and assembled.
Circuit symbols - A description of symbols to BS2917 and ISO1219 The design, operation and application of pneumatic actuators and valves - including: compact actuators ISO actuators VDMA actuators rod-less actuators cushioning actuators manual valves mechanical valves pilot valves electrically operated valves
Electromechanical Feeders The high-capacity performers Syntron MF Heavy-Duty Electromechanical Feeders are the heavy-weights of bulk material handling and are used for higher capacity requirements. The ten heavy-duty models handle capacities from 600 to 4,000 tons per hour.* Syntron Heavy-Duty Electromechanical Feeders combine
There are two versions of linear actuators: Double-acting actuators Single-acting actuators (with mechanic spring return) In double acting actuators - depending on the direction required - compressed air flows to the outside or to the inside of the two pistons. This can be easily controlled by using 5 -way directional control valves (5 .
Sensors and Actuators employed in Robotics 13 Sensors/Actuators Trends Worldwide sales of sensors/actuators are forecast to grow 14% to a high of 9.9 billion in 2014, followed by a 16% increase in 2015 to 11.4 billion Between 2013 and 2018, the sensors/actuators market is projected to rise
Merced et al. / Sensors and Actuators A 196 (2013) 30-37 31 The family of thermal actuators can be divided according to the phenomenon caused by the difference in temperature. Among all the types of mechanical actuators, shape memory alloy (SMA) actuators offer the highest strain energy density [7]. They have been
Actuators - Category 1 0 27 2 3 32 30 960 . 7 PS2/GS2/L110 EMA ACTUATOR Actuators - Category 2 37 24 2 3 66 22 1452 8 C25 MRC ACTUATOR Actuators - Category 3 57 38 3 3 101 2 202 9 CUS EGC ACTUATOR, ,C25 EGC ACTUATOR Actuators - Category 4 57 38 3 3 101 12 1212 10 CUS AVR MRC ACTUATOR Actuators - Category 5 57 38 3 3 101 4 404 .
Dr. Adel Gastli MCTE3210: Electromechanical Systems & Actuators 3 Introduction Most of the electrical machine in service are AC type. DC machine are of considerable industrial importance. DC machine mainly used as DC motors and the DC generators are rarely used. DC motors provides a fine control of the speed which can not be attained by AC motors.
image as a guide. Fill in a few petals using satin stitch, and add . The secret is to hold the working thread taut while you pull it through. Come up at point 1, then wrap the thread around the needle twice. Holding the working thread with your non-dominant hand, bring the needle down at point 2 (close to point 1, but not the same hole) and pull slowly until the knot is formed. 01 01 01 .
