N A SA T E C H N I C AL
`M E M O R A N D UM
`
`NASA TM X-3524
`
`CM
`
`CO
`
`><
`
`DYNAMICS OF HIGH-BYPASS-ENGINE
`THRUST REVERSAL USING
`A VARIABLE-PITCH FAN
`
`John W. Schaefer, David A. Sagerser,
`and Edward G, Stakolich
`
`Lewis Research Center
`Cleveland, Ohio 44135
`
`NATIONAL
`
`AERONAUTICS AND SPACE ADMINISTRATION
`
`• WASHINGTON, D. C.
`
`• MAY 1977
`
`GE-1009.001
`
`

`
`1.
`
`4.
`
`Report No.
`NASA TM X-3524
`Title and Subtitle
`DYNAMICS OF HIGH-BYPASS-ENGINE THRUST REVERSAL
`USING A VARIABLE-PITCH FAN
`
`2. Government Accession No.
`
`3. Recipient's Catalog No.
`
`5. Report Date
`May 1977
`6. Performing Organization Code
`
`7. Author(s)
`John W. Schaefer, David A. Sagerser, and Edward G. Stakolich
`
`9. Performing Organization Name and Address
`
`National Aeronautics and Space Administration
`Cleveland, Ohio 44135
`
`12.
`
`Sponsoring Agency Name and Address
`National Aeronautics and Space Administration
`Washington, D. C. 20546
`
`15.
`
`Supplementary Notes
`
`8. Performing Organization Report No.
`E-9026
`10. Work Unit No.
`505-05
`11. Contract or Grant No.
`
`13. Type of Report and Period Covered
`Technical Memorandum
`14. Sponsoring Agency Code
`
`16.
`
`Abstract
`The unique reverse-thrust performance requirements of a variable-pitch-fan propulsion system
`for future short-haul aircraft required the early full-size engine demonstration that rapid
`forward-to-reverse -thrust transients could be performed. The test program demonstrated
`that successful and rapid forward- to reverse-thrust transients can be performed without any
`significant engine operational limitations for fan blade pitch changes through either feather
`pitch or flat pitch. For through-feather-pitch operation with a flight inlet, fan stall problems
`were encountered and a fan blade "overshoot" technique was used to establish reverse thrust.
`Variable-pitch fans offer a potentially attractive means of providing rapid reverse-thrust
`capability for future short-haul aircraft.
`
`17.
`
`Key Words (Suggested by Author(s) )
`Aerodynamics
`Aircraft propulsion
`Variable pitch fan
`
`Thrust reversal
`Dynamic tests
`
`18. Distribution Statement
`Unclassified - unlimited
`STAR. Category 07
`
`19.
`
`Security Classif. (of this report)
`Unclassified
`
`20. Security Classif. (of this page)
`Unclassified
`
`21. No. of Pages
`34
`
`22. Price"
`
`A03
`
`* Fm sale by the National Technical Information Service, Springfield. Virginia 22161
`
`GE-1009.002
`
`

`
`DYNAMICS OF HIGH-BYPASS-ENGINE THRUST REVERSAL
`
`USING A VARIABLE-PITCH FAN
`
`by John W. Schaefer, David A. Sagerser, and Edward G. Stakolich
`
`Lewis Research Center
`
`SUMMARY
`
`During the past several years, there has been a concerted effort to develop the
`technology necessary to meet the unique reverse-thrust performance requirements of
`a variable-pitch-fan propulsion system for future short-haul aircraft. A significant
`portion of this effort involved the testing of a full-size, variable-pitch-fan engine to
`demonstrate rapid forward-to-reverse-thrust
`transients. This report presents the re-
`sults of this effort, which encompassed approach-power thrust reversals, in both
`through-feather-pitch and through-flat-pitch modes of operation as well as aborted-
`takeoff transients. Tests were performed with both a bellmouth and a flight inlet. This
`program has demonstrated that rapid approach-power thrust reversals can be accom-
`plished without any significant engine operational limitations for fan blade pitch changes
`through either feather pitch or flat pitch. Also, the aborted-takeoff operational mode
`has been satisfactorily demonstrated. For through-feather-pitch operation with a flight
`inlet, however, fan stall problems were encountered, but a fan blade "overshoot" tech-
`nique was used to establish reverse thrust. High fan blade vibratory stresses relative
`to forward-thrust values occurred during through-feather-pitch transient operation but
`were well within acceptable limits. In conclusion, this program has shown that variable-
`pitch fans offer a potentially attractive means of providing rapid reverse-thrust capability
`for future short-haul aircraft.
`
`INTRODUCTION
`
`For the past several years, the development of advanced technology for a quiet,
`clean, high-bypass-ratio turbofan engine to be used on future short-haul aircraft has
`been pursued. At the NASA Lewis Research Center, a major portion of this effort is
`the Quiet, Clean, Short-Haul Experimental Engine (QCSEE) Program. A status report
`on this program is presented in reference 1. The reduced field lengths envisioned for
`
`GE-1009.003
`
`

`
`short-haul aircraft operation have made reverse-thrust performance a critical part of
`the propulsion system's design requirements. Noise requirements for short-haul air-
`craft dictate that a low-pressure-ratio, high-bypass-ratio fan be used, especially for
`under-the-wing engine installation. For such installations, engines designed with
`variable-pitch fans for reverse thrust have been shown (refs. 2 and 3) to be superior to
`those with fixed-pitch fans and conventional reversers. The potential advantage of using
`variable-pitch fans is the elimination of the conventional heavy, high-maintenance, tar-
`get or cascade thrust-reversal hardware plus the added benefit of improved thrust re-
`sponse time (refs. 2 and 3). Obtaining reverse thrust with a variable-pitch-fan engine
`involves a new mode of engine operation. Therefore, early in the QCSEE program,
`collecting information on the feasibility of this approach as well as gaining operating ex-
`perience was considered imperative. An overview discussion of the status of reverse-
`thrust technology for variable-pitch-fan systems is presented in reference 4.
`Investigative tests were undertaken on a full-scale, high-bypass-ratio, variable-
`pitch-fan engine developed by the Hamilton Standard Division of United Technologies
`Corp. Steady-state aerodynamic and acoustical performance, including net thrust and
`external exhaust velocity profiles, was determined for both forward- and reverse-thrust
`modes, with both bellmouth and flight inlets.
`Transient tests were performed in the "through-feather-pitch" (stall) and "through-
`flat-pitch" (zero lift) operational modes with both inlets. In addition to the normal
`transient operation of approach power to reverse thrust, the aborted-takeoff transient
`operation was also investigated.
`This report discusses the results of the transient test program that was performed
`under NASA contract at Hamilton Standard and at the Lewis Research Center. Detailed
`discussions of the Hamilton Standard testing are presented in references 5 to 8.
`
`APPARATUS AND PROCEDURES
`
`The initial transient tests were conducted at the Hamilton Standard Hilltop Test Fa-
`cility as described in references 7 and 8. Transient testing was resumed later at the
`NASA Lewis Test Facility. The Lewis facility, test hardware, and experimental methods
`used are described here.
`
`Test Facility
`
`The Lewis test facility shown in figure 1 is an outdoor test facility capable of test-
`ing turbofan engines up to a maximum thrust level of 133 500 newtons. The large tripod
`structure supports the cantilevered overhead thrust-measuring engine mount with the
`
`GE-1009.004
`
`

`
`engine centerline at 2.9 meters above ground. The area around the test stand is paved
`with concrete to provide a flat, consistent ground surface. The engine support and thrust-
`measuring system can be rotated to allow the movable shelter to be rolled in place over
`the test stand.
`A 37.85-cubic-meter fuel tank is located underground adjacent to the movable
`shelter, which is positioned approximately 90 meters from the test stand.
`In the same
`area above ground is a large high-pressure gaseous nitrogen storage tank used for fire
`protection. An instrumentation vault housing signal conditioners and amplifiers is
`located near the test stand. The control room is approximately 120 meters away from
`the test stand. A digital data acquisition system is capable of recording pressures,
`temperatures, and other engine parameters and, together with a computer system, can
`print out engineering parameters in the control room.
`
`Test Hardware
`
`The variable-pitch, gear-driven fan engine that was tested is shown installed at
`the Lewis test facility in figure 2. The structure beneath the engine provided a stiff
`mounting support for the thrust cell, which resulted in improved real-time thrust re-
`sponse measurement. These measurements are discussed later. The variable-pitch-
`fan engine, described in references 5 to 8, was supplied by Hamilton Standard and in-
`cluded the fan, the gearbox, the hydraulic equipment, the computer control system, and
`the engine cowling. The NASA Lewis Research Center supplied additional inlet and noz-
`zle hardware. The fan is driven by an Avco Lycoming T55-L-11A turboshaft engine.
`Table I lists the main design characteristics of the fan and the engine.
`Cross-sectional views of the cowling hardware are illustrated in figure 3. Several
`combinations of inlets and exhaust ducts were tested, but the standard fan nozzle was
`not used when the engine was generating reverse thrust because reverse-thrust per-
`formance is severely degraded with that configuration (refs. 7, 8)
`Instead a bellmouth-
`shaped inlet was used to simulate the open position of a variable-area fan nozzle, called
`an exlet (ref. 9). The flight inlet was tested primarily to assess its effect on reverse-
`thrust performance. Its characteristics are presented in figure 4. The flight inlet was
`designed with the same inlet lip contour to be used on the QCSEE engines (ref. 10).
`The programmable digital computer control system allowed for the changing of the
`control parameters; a simplified schematic is presented in figure 5. The different re-
`search inputs listed could be changed between test points and are defined later in the
`text. The control system interpreted the requests and controlled the fan blade pitch and
`power lever position during the transient.
`
`GE-1009.005
`
`

`
`Experimental Methods
`
`Generally, steady-state aerodynamic instrumentation was located in the planes
`identified in reference 5. Specifically, inlet wall static pressures were added for the
`Lewis Research Center tests, located as shown in figure 6. The pressures, tempera-
`tures, and other outputs were received as millivolt signals, digitized, and then trans-
`mitted to a remote data collector system. A computer then reduced the data to appro-
`priate engineering parameters.
`The following transient instrumentation was used:
`(1) Ambient pressure, PQ
`(2) Ambient temperature, TQ
`(3) Fan speed, N
`(4) Engine speed, NC
`(5) Engine torque, 3~
`(6) Power level position, PLA
`(7) Fan blade angle, /3
`(8) Engine fuel flow, WF
`(9) Engine thrust, F
`(10) Compressor inlet total pressure (ref. 5), ?„„
`(11) Compressor discharge pressure, P,™
`(12) Low-pressure-turbine inlet temperature, Tr
`(13) Fan strain gages (located 36. 2 cm from blade tip)
`Excitation of the overhead engine support structure at its natural frequency during
`thrust reversal was a significant concern. The result would be to obscure the actual
`real-time engine thrust value measured by a load cell. The amplitude of the vibration
`was significantly reduced and the natural frequency increased by using a secondary sup-
`port and incorporating a bidirectional load cell. The secondary support located beneath
`the engine is shown in figure 2. The secondary support was available and used only for
`the through-feather-pitch portion of the transient program.
`The real-time transient data signals were transmitted as millivolt signals to a re-
`mote high-speed digitizing system. The data were digitized at a rate of approximately
`5000 samples per second. The digitized signals were transmitted to a computer for ap-
`propriate calculations, and the final engineering data were presented at a rate of ap-
`proximately 50 samples per second.
`
`RESULTS AND DISCUSSION
`
`Steady-State Aerodynamic Performance
`
`Forward thrust. - To better understand,the transient tests, steady-state forward-
`
`4
`
`GE-1009.006
`
`

`
`and reverse-thrust aerodynamic performance is discussed first. Figure 7 presents the
`forward steady-state aerodynamic performance as a plot of corrected thrust versus
`corrected fan speed. Various fan-blade-angle operating lines are shown. The three
`solid symbols indicate the selected takeoff and landing approach operating conditions.
`These points were used as the starting conditions for all transients. The landing ap-
`proach thrust level of 60 percent of takeoff thrust was selected based on the short-haul
`aircraft study discussed in reference 11. For forward-to-reverse-thrust transients
`through feather pitch from the approach power setting, 100-percent fan speed was se-
`lected for good waveoff thrust response. But for some of the reverse transients through
`flat pitch, the initial fan speed was reduced to approximately 75 percent of takeoff
`speed to allow a greater fan overspeed margin. The transients were conducted with the
`bellmouth-shaped exlet and not the standard fan nozzle (fig. 3) to provide a better air-
`flow path for reverse thrust. This change had little effect on the starting thrust level
`(less than 5 percent) because the effective nozzle area of the exlet was essentially
`equal to the standard nozzle area.
`Reverse thrust. - A basic concern in the operation of a variable-pitch fan is the
`direction in which the fan blade pitch should be changed to develop reverse thrust. The
`two possible ways are illustrated in figure 8. Cross sections of two fan blade shown in
`their normal forward-thrust position are at the top of this figure. From this position,
`the blades can be turned through flat pitch, a condition of zero lift, to the reverse-thrust
`position as shown on the left in figure 8. Two things should be noted for this approach.
`First, adjacent blade leading and trailing edges must pass each other during the transi-
`tion through flat pitch. This requires that the blade solidity be less than 1 at all radii.
`This constraint can limit fan performance, especially at the hub. Second, while the
`blade leading edge remains the same relative to the airflow, the blade camber is wrong
`for reverse-thrust operation.
`The alternative approach is to turn the blades through feather pitch, passing through
`a separated flow or stall condition. This is shown on the right side in figure 8. In this
`case, the blade camber is correct in the reverse-thrust position, but the leading and
`trailing edges are reversed. During the transition the flow over the blades separates
`and then reattaches in reverse thrust, moving in the opposite direction relative to the
`blade. With this approach the blade solidity may exceed 1, which allows more freedom
`in the blade design. The blade twist and camber can still limit the hub solidity to some
`extent.
`In figure 9 the steady-state reverse aerodynamic performance is presented as a plot
`of corrected thrust versus corrected fan speed. Various relative reverse fan blade
`angles for through feather pitch and a single angle for through flat pitch are shown. The
`solid symbols indicate the selected reverse-thrust operating conditions. The reverse
`through-feather-pitch fan blade angle was selected based on obtaining the required
`thrust with the lowest fan speed operating line tested. Attempts to test on a lower fan
`
`GE-1009.007
`
`

`
`speed operating line (A/3 = -88°) were limited by high fan blade vibratory stresses. This
`approach results in the largest thrust increase capability for emergency operations. A
`reverse-thrust level of 35 percent of takeoff thrust has been established as the required
`reverse thrust for short-haul aircraft, as discussed in reference 12. For thrust re-
`versals with the bellmouth the fan speed was increased to maintain the 35-percent
`reverse-thrust requirement. Even though through-flat-pitch reverse thrust does not
`meet the reverse-thrust requirements for short-haul aircraft, an understanding of this
`mode of operation was considered valuable should future considerations require this
`mode of operation. A reverse-thrust fan speed of 90 percent was used to allow some
`margin for fan overspeed.
`
`Approach-Power Transient Performance
`
`Reversal through feather pitch - bellmouth inlet. - Before discussing the results of
`the through-feather-pitch transients, a discussion of the real-time variations during a
`transient is desirable. In figure 10 are presented characteristics during a typical
`reverse-thrust transient through feather pitch. The selected transient parameters are
`shown plotted as a function of time. Data obtained using the high-speed transient data
`system are shown with a solid line faired through the points. At the start of the transient
`the power lever was immediately raised to the level required for the selected ending
`condition. Fan blade angle (fig. 10(a)) was changed at approximately 130 degrees per
`second and reached the final position is less than 1 second. Fuel flow (fig. 10(b)) in-
`creased as a result of the power lever change, followed by a few rapidly damped oscil-
`lations in fuel flow. Thrust, torque, and fan speed should be viewed collectively (figs.
`10(c), (d), and (e)). The initial change in fan blade pitch increases the aerodynamic
`loads on the blades. As this happens the fan speed is lowered as the fan rotational en-
`ergy is converted into a thrust increase. As the blade pitch continues to increase, the
`fan eventually stalls and the thrust falls to zero. This stalled condition unloads the
`blades to some degree, causing the fan speed to increase. Shortly after the blades
`reach their reverse position, the flow reattaches and reverse thrust is obtained. During
`the transient the core compressor must tolerate variations in pressure and speed (figs.
`10(f), (g), and (h)). The compressor inlet pressure rises somewhat with increasing fan
`blade angle and then falls suddenly as the fan stalls while, in response to the change in
`fuel flow, the compressor speed and discharge pressure both increase. No problems
`with the core compressor were evident due to these variations.
`Before attempting to analyze the through-feather-pitch transient operation, it is
`desirable to define several descriptive terms, as shown in figure 11. The primary
`means of reviewing specific transient sequences is by the thrust response time, which
`is defined as the time from the request to change the engine thrust level until 95 percent
`
`GE-1009.008
`
`

`
`of the final thrust is achieved. Cutback position is a temporary level to which the power
`lever on the fuel controller is initially moved before the power lever is set to its final
`reverse-thrust position. Blade travel time is the time it takes the fan blade pitch to
`change from the initial forward-thrust angle to the reverse-thrust angle. Flow reat-
`tachment time is the time it takes the air to reattach to the fan blade after the fan blade
`has reached the reverse-thrust angle. Flow reattachment was assumed to occur when
`there was a rapid decrease in blade vibratory stress level. This assumption was sup-
`ported when the rapid decrease in blade stress level was compared with other measured
`engine parameters, such as presented in figure 10. Flow reattachment was determined
`in this manner because, during the transient, oscillatory motion of the engine support
`structure precluded accurate measurement of engine thrust.
`If the fan speed is insuf-
`ficient to generate the required reverse thrust when the flow reattaches to the fan blade
`(point A), a delay time occurs while the fan accelerates.
`The thrust response time for thrust reversal through feather pitch is a summation
`of the blade travel time, the flow reattachment time, and the delay time, as shown in
`figure 11. The effect of blade travel time on thrust response time is presented in fig-
`ure 12. Specifically, thrust response time varies directly with blade travel time, and
`blade travel time has no significant effect on flow reattachment time. Figure 13 pre-
`sents the possible delay time that should be added if the fan speed at flow reattachment
`is not sufficient to generate the required reverse thrust.
`In general, the fan speed at
`flow reattachment is reduced by greater amounts of cutback for longer durations. Fig-
`ure 14 presents flow reattachment time, which is the final controlling item of thrust
`response time, as a function of reverse fan blade angle. Flow reattachment appears to
`be significantly affected only by the reverse fan blade angle and the fan duct geometry.
`Flow reattachment time decreases as the reverse fan blade angle of attack is reduced
`(i. e., A/3 decreases). Flow reattachment time is also reduced by shortening the over-
`all length of the fan inlet and exhaust ducts.
`When considering figure 14, note that transient operation with blades moving directly to
`the selected steady-state relative reverse fan blade angle (A0 = -90°) could not be per-
`formed because the air would not reattach to the fan blade. A method called "overshoot-
`ing" was conceived that allowed transient operation to the selected fan blade angle. With
`this technique, the fan blades are moved beyond the selected reverse fan blade angle,
`held there for a short period of time (dwell time), and then returned to the selected fan
`blade angle. This overshoot temporarily reduces the angle of attack on the blades, al-
`lowing the separated airflow to reattach and reverse flow to be established. Figure 15
`illustrates this operational technique. The dashed line shows that no reverse thrust, A,
`is generated if the fan blade angle is rotated directly to the selected reverse-thrust
`angle. If the fan blade angle is rotated directly to some angle beyond "optimum" so that
`reverse thrust is achieved, the level of reverse thrust, B, is less than desired, as
`shown by the dot-dashed line. The solid line shows how the "overshoot" technique is
`
`GE-1009.009
`
`

`
`performed such that flow reattachment occurs and the required reverse thrust, C, is
`generated.
`The results of using the overshoot technique are shown in figure 16. Presented is
`a plot of flow reattachment time as a function of degrees of fan blade angle overshoot.
`The thrust response time for these transients varied only with flow reattachment time
`since the delay time was zero and the blade travel time was approximately constant
`(1. 09±0. 05 sec). For a constant dwell time the flow reattachment time can be signifi-
`cantly reduced, approximately 0. 5 second, by increasing the degrees of overshoot. In-
`creasing the dwell time for a constant degree of overshoot will also reduce the flow
`reattachment time but only by approximately 0. 2 second.
`Reversal through feather pitch - flight inlet. - Testing with the flight inlet in con-
`trast to the bellmouth inlet showed that with the fan stalled, the inlet can produce back-
`pressure on the fan, which tends to prevent flow reattachment to the stalled fan blades
`and hence the establishment of reverse thrust. This can occur when the fan is started
`from rest with the blades initially in a reverse position or, more importantly, during
`a forward-to-reverse-thrust transient through feather pitch. This effect can be ob-
`served with the aid of tufts, as shown in figure 17. When the fan is in stall, the flow is
`primarily tangential, tending to rotate with the fan. When the fan is unstalled and pro-
`ducing reverse thrust, the flow is nearly axial.
`For the swirling flow in the stalled condition to be exhausted out the smaller diam-
`eter throat of the flight inlet, the tangential components of velocity of the swirling flow
`must increase to conserve angular momentum. Since the static pressure at the front of
`the inlet is ambient, a higher than ambient pressure at the fan is implied. This higher
`pressure (called backpressure) at the fan tends to prevent flow reattachment. The mag-
`nitude of the backpressure will depend on the inlet geometry and the forward velocity
`(ram pressure).
`Test data showing this effect are presented in figure 18. Ratios of wall static pres-
`sure to ambient pressure are compared for the extended bellmouth and the flight inlet
`for both stalled and unstalled conditions at the same fan speed (76 percent) and com-
`parable fan blade angles (6° from the reverse fan blade stall angle). Of primary inter-
`est is the static pressure at the fan face. As can be seen from figure 18, in a stalled
`condition a higher pressure does exist with the flight inlet.
`Successful forward-to-reverse-thrust transients through feather pitch were per-
`formed with the flight inlet by using the technique of fan blade overshoot.
`In this case,
`the fan blade overshoot angle was necessary to unstall the fan and to establish reverse
`flow, and the final reverse fan blade angle was necessary to obtain the required re-
`verse thrust. The transients were performed with the maximum overshoot mechanical
`capability of the equipment. But this by itself was not large enough to consistently
`establish reverse flow, and augmentation from three air jets was also used. The pur-
`pose of the air jets was to help promote flow reattachment to the fan blades by blowing
`
`8
`
`GE-1009.010
`
`

`
`into the blade channels in the reverse flow direction. Even though the engine customer
`bleed would have been sufficient for the air jets, a facility air supply was used to sim-
`plify the test setup. The three air jets were installed approximately 6. 35 centimeters
`behind the fan rotor and positioned 1/2 blade spacing apart, as shown in figure 19. The
`fan blades are shown at a forward-thrust pitch angle. A survey was performed to de-
`termine the best radial position "and jet flow angle for the air jets. They were finally
`located 5. 08 centimeters into the stream with the 2. 54-centimeter by 0. 64-centimeter
`slots oriented 68° in the feather-pitch direction from the fan blade design angle. The
`total air jet mass flow rate was 0. 59 kilogram per second. If more overshoot capability
`could be made available by modifying the pitch change mechanism, the air jets would
`probably not be needed. Table II presents results of the fastest successful forward-
`to-reverse-thrust transient with the flight inlet. This thrust response time of 1. 84
`seconds is not optimum since the thrust response can be improved by decreasing the
`blade travel time by increasing the fan blade pitch change rate. Also the excessive
`flow reattachment time could have been reduced by a minor mechanical redesign to in-
`crease the fan blade overshoot capability.
`Reversal through flat pitch. - For reverse-thrust applications where approximately
`20 percent of takeoff thrust is adequate, operation through flat pitch can be considered.
`Through-flat-pitch operation is significantly noisier than through-feather-pitch opera-
`tion (ref. 4). A typical time history for through-flat-pitch operation is presented in fig-
`ure 20. All the data points are shown, and a curve was faired through them. In this
`transient, the blade pitch (fig. 20(a)) was changed at a rate of approximately 100 degrees
`per second, with the power lever and hence the fuel flow (fig. 20(b)) held constant.
`Thrust (fig. 20(cj), presented as a percentage of measured takeoff thrust, responded
`to the blade angle change and fell off smoothly. The final reverse-thrust level was
`reached in somewhat less than 1 second. The fan during this time accelerated quickly
`(fig. 20(e)) as the load on the fan blades was reduced. As the blade loading increased
`again in reverse thrust, the fan speed peaked and then converged on the final reverse-
`thrust value. During this same time period, there occurred a slight dip in torque
`(fig. 20(d)). Torque started to decrease due to the reduction in the fan blade loading,
`which was followed by an increase as the blade loading increased in reverse thrust.
`Core compressor speed (fig. 20(f)) and compressor discharge static pressure ratio
`(fig. 20(h)) remained practically constant since there was no change in fuel flow. The
`compressor inlet total pressure ratio (fig. 20(g)) fell off to a level slightly higher than
`a similar through-feather-pitch transient.
`As with the through-feather-pitch transients, it is desirable to define several
`terms, which are illustrated by figure 21. Cutback position and blade travel time main-
`tain the same definitions as previously described. Duration is the time that the power
`lever is maintained at the cutback position. If the initial fan speed is too great for the
`power lever angle selected, the fan will overspeed as demonstrated in the illustration
`
`9
`
`GE-1009.011
`
`

`
`by comparing the solid and short-dashed lines. K the higher initial fan speed is re-
`quired, careful selection of power lever cutback position and duration are required to
`prevent a fan overspeed from occurring.
`The thrust response time for thrust reversal through flat pitch is a function of the
`fan acceleration time and the blade travel time. Fan acceleration time is a function of
`the cutback position and the initial fan speed. For a reduced initial fan speed (76 per-
`cent), the effect of cutback position and duration are shown in figure 22. The fan accel-
`eration time varies directly with duration and inversely with cutback position.
`When starting from a higher initial fan speed, the fuel flow can initially be cut back
`to reduce the available engine power during the transient such that when the reverse
`blade position is reached the fan speed is at the desired level and no overspeed occurs.
`Figure 23 presents test data that show the effect of initial fan speed on forward- to
`reverse-thrust through-flat-pitch operation. When a high initial fan speed was used, the
`fuel flow and fan blade pitch change rate needed to be coordinated to prevent a fan over-
`speed condition, which limits further reductions in thrust response time (see fan speed
`versus time, long-dashed curve in fig. 21). However, with a reduced initial fan speed
`and proper coordination of fuel flow and fan blade pitch change rate, the thrust response
`time could be reduced to the blade travel time without a fan overspeed.
`Fan blade stress. - Of significant interest is the structural response of the fan
`blades during the forward- to reverse-thrust transition. Figure 24 presents real-time
`traces of the fan blade vibratory stress. The transient was initiated at time zero. For
`through-flat-pitch transitions (fig. 24(a)), the vibratory stress gradually increased
`throughout the transient, leveling off at approximately twice its initial value. The
`through-feather-pitch transition (fig. 24(b)) was significantly different. During the
`transient, the fan blade stresses built up and peaked as the blade stalled. A second
`peak, generally somewhat higher than the first, occurred as flow reattached to the
`blades in the reverse direction. These stress peaks, while high relative to forward-
`thrust levels, did not limit the transient tests.
`The peak vibratory stress is important, but it alone does not present the entire
`picture. The time duration of the high vibratory stress is also important, and together
`they affect the fatigue life of the fan blade if the high stress exceeds the infinite life al-
`lowable stress of the fan blade material. The effect of flow reattachment time on the
`fan blade relative fatigue life is presented in figure 25. The absolute value of fatigue
`life used per transient is primarily a function of fan blade material selection and is not
`important for this discussion. Reducing the flow reattachment time provides for an in-
`crease in fatigue life by reducing the time period of high vibratory stress.
`
`10
`
`GE-1009.012
`
`

`
`Aborted-Take off Transient Performance
`
`Also of significant interest is the aborted-takeoff emergency operation in which the
`propulsion system must generate the required reverse thrust with the engine initially at
`takeoff power. Figure 26 presents the results of the aborted-takeoff transient operation
`through feather pitch. The results are consistent with the approach power transients,
`in which the thrust response time is a function of blade travel time and flow reattach-
`ment time. As a result of the high initial power lever position, the delay time is zero.
`For aborted-takeoff transient operation through flat pitch, only a few transients
`were conducted, but this mode of operation was satisfactorily demonstrated. For ex-
`ample, the thrust response time was 1. 3 seconds for a transient in which fan speed
`started and finished at 100 percent and the fan blade pitch change rate was 110 degrees
`per second. To avoid fan overspeed, the fuel flow must be cut back and hence, in this
`case, the power lever was cut back to idle for 0. 5 second.
`
`SUMMARY OF RESULTS
`
`This test program, conducted to develop the technology necessary for a Quiet,
`Clean, Short-Haul Experimental Engine (QCSEE) type of propulsion system, has shown
`that variable-pitch fans offer a potentially attractive means of providing rapid reverse-
`thrust capability for fu