Piloted Evaluation of Modernized Limited Authority Control Laws in the NASA-Ames Vertical Motion Simulator (VMS)

24 Pages·2003·2.8 MB·English

by  Sahasrabudhe, Vineet

Piloted Evaluation of Slodernized Limited Authority Control Laws
in the NASA-Ames Vertical Motion Simulator (VhlS)
Vi nect Cahasrabudhe, Edgar Melkers and Alexander Faynberg Chris L,. Blanken
Sikorsky Aircraft Corporation Army/NA,SA Rotorcraft Division
Stratford, CT Aerofl ightdynariics Directorate (.AMRDEC)
U.S. Army Aviation and Missile Command
Moffett Field. CA
Abstract
The UH-60 BLACK HAWK was designed in the 1970s, when the US Army primarily operated during the day in
good visual conditions. Subsequently, the introduction of night-vision goggles increased the BLACK HAWK'S
mission effectiveness, but the accident rate also increased. The increased accident rate is strongly tied to increased
pilot workload as a result of a degradation in visual cues. Over twenty years of resetarch in helicopter flight control
and handling qualities has shown that these degraded handling qualities can be reco\.ered by modifying the response
type of the helicopter in low speed flight. Sikorsky Aircraft Corporation initiatelj a project under the National
Rotorcraft Technology Center (NRTC) to develop modern flight control laws while utilizing the existing partial-
authority Stability Augmentation System (SAS) of the BLACK HAWK. This effort resulted in a set of Modernized
Control Laws (MCLAWS) that incorporate rate command and attitude command response types. Sikorsky and the
US Army Aeroflightdynamics Directorate (AFDD) conducted a piloted simulation on the NASA-Ames Vertical
h4otion Simulator, to assess potential handling qualities and to reduce the risk of subsequent implementation and
flight test of these modern control laws on AFDD's EHdOL helicopter. The sitnuladon showed that Attitude
Command Attitude Hold control laws in pitch and roll improve handling qualities in the low speed flight regime.
These improvements are consistent across a range of mission task elements and for both good and degraded visual
environments. The MCLAWS perform better than the baseline UHdOA control laws in the presence of wind and
turbulence. Finally, while the improved handling qualities in the pitch and roll axjs allow the pilot to pay more
attention to the vertical axis and hence altitude performance also improves, it is clear from pilot comments and
altitude excursions that the addition of an Altitude Hold function would further reduce workload and improve
overall handling qualities of the aircraft.
Introduction loop stability functions may need to be added as the
UCE degrades: attitude, direction. height, and
Over twenty years of research in helicopter flight position hold.
control and handling qualities has shown that as the
pilot's visual environment degrades, there is a The UH-60 BLACK HAWK is the US Army's
corresponding degradation in handling qualities for utility-class helicopter. The UH-60A was designed
near-Earth tasks. Degraded handling qualities imply in the 1970s, when the Army primarily operated in
reduced task performance and increased pilot the day in good visual conditions. Subsequently. the
workload, which contribute to reduced mission introduction of night-vision goggles increased the
effectiveness and increased accident rates. The same BLACK HAWK mission effectiveness, but the
research has also shown that these degraded handling accident rate also increased [Ref 21. This increase is
qualities can be recovered by increasing the stability strongly tied to increasej pilot workload as a result of
of the helicopter, i.e., by changing the control a degradation in visual cues from the goggles. The
response type to a higher rank of stabilization. These basic control response of the flight control system has
results and concepts have been incorporated into the not been upgraded frori the original rate-command
US Army's Aeronautical Design Standard - 33 response type. Operating rate-command response
(ADS-33). Handling Qualities Requirements for types at night (UCE>2) results in degraded handling
Military Rotorcraft [Ref. I] through the Usable Cue qualities and contributes to increased accident rates.
Eni:ironment (UCE) concept. As the UCE Improving the flight control system to help reduce
deteriorates from clear day (UCE=I), to a starlit night accident rates has been recopnized by the US Army
(VCE=2), to a moonless-overcast night (UCE=3), the Safety Center as a high priority toward reducing
helicopter control response must be improved from a accident rates. A rec:nt Safety Center accident
rate command, to an attitude command, to a investigation study states that the number one
translational rate command response type, material fix toward reducing Army aviation accidents
'especrivel y, in ordcr to maintain satisfactory is to improve the howr and low speed handling
handling qualities. In addition, the following outer- qualities. [Ref. 3J
'
Presented at the i?\mcrican Helicopter Society jYih Annual Forum, Phoenix, AZ. Xlay f4.2 003. Copyright 0 2003
h!' the American Helicopter Society International. Inc. All rights reserved.
355
~~~-~ -~
~ ~
'I
UH-60X aircraft. The pitch-SAS is essentially a rate ni
Rzcently the US Army has entered in:o a feedback system that augments the damping of the d)
recapitalization program to extend the -:r;.ice life of bare airframe dynamics. il;
the BLACK HAWK for decades to come (UH-60R.I). 3 I
On May 2, 2001 the Army contracted with Sikorsky The MCLAWS implement a two-mode control P'
Aircraft Corporation to build four UH-60M system. In the attitude mode the pitch and roll axes ai
prototypes. The upgrades include not only new wide have Attitude Command Attitude Hold (ACAH) type a:
chord rotor blades, but also inclusion of two neu responses. while the yaw axis has Rate Command
digital flight control computers and associated Direction Hold (RCDH) characteristics. The control 7
sensors. In the baseline plan, however, these laws switch to a rate command mode if the helicopter n
s
prototypes will fly with flight control laws that are velocities or attitudes exceed the limits shown in
fundamentally similar to existing UH-60A flight Table 1. As the name indicates, in the rate mode the C
control laws. Many years of research have been aircraft has a Rate Command (RC) response type. In
performed: from documenting the UH-60 relative to order to switch back to ACAH mode from rate mode, 1
ADS-33 [Ref. 41, to developing and evaluating new more restrictive conditions must be met which are \
control concepts [Refs. 5-81 that provide improved shown in Table 1. Note that in this study only the I
handling qualities while retaining the existing partial- inner loop SAS servos are used to implement I
authority actuation system. In 2000, Sikorsky MCLAWS. This simplifies the design and limits the I
initiated a project under the National Rotorcraft changes to the SAS computer - leaving the Flight
Technology Center (NRTC) to develop modem flight Path Stabilization (FPS) system untouched. It also
control laws while still utilizing the existing partial- allows for a set of control laws that are not inherently
authority Stability Augmentation System (SAS) of dependent on whether or not the pilot has the trim
the BLACK HAWK. Sikorsky's NRTC effort release switch depressed. This approach does not
resulted in a set of control laws that incorporate rate prec-l ude the integration of the outer loop trim servos
command and attitude command response types and at a later stage to help re-center the SAS servos in the
have provisions for translational rate command Ion, term.
response types.
When the system switches from attitude mode to rate
To advance possible implementation of these modem mode the dashed paths in Figure 1 are gracefully
control laws into the UHdOM, Sikorsky was invited removed, and the systems reverts back to a rate
by the US Army Aeroflightdynamics Directorate feedback architecture almost identical to the baseline
(AFDD) to participate in a piloted simulation on the UH-60A SAS control laws. Conversely, when the
NASA-Ames Vertical Motion Simulator (VMS). aircraft re-enters the attitude mode, these paths are
The objectives of the VMS simulation were to assess brought back in gradually. The overall objective was
potential handling qualities improvements in to retain ACAH characteristics over a useful range of
simulated degraded visual environments and to aircraft velocities and attitudes without persistently
reduce the risk of subsequent implementation and saturating the SAS.
flight test of these modem control laws on AFDD's
EH-60L helicopter. This paper will describe The MCLAWS have a model following type
development of the modern control laws, the VMS architecture, where the pilot stick input is passed
based piloted simulation evaluation, and associated through a command model to generate desired 0-
results. ideal rates and attitudes. These are compared to the
actual rates and attitudes and the feedback reduces
Modernized Control Laws the difference between the two. There are a few key
This section describes the MCLAWS architecture, differences from more conventional implementations
the control modes, and the design and analysis of the model following architecture. First, instead of
techniques. In addition, results are presented from a just using a commanded attitude both rate and
Sikorsky piloted simulation that was used as a attitude commands are generated and used. Since the
preliminary evaluation in preparation for the VMS basic aircraft responds like a rate command system,
experiment. using a commanded rate leads to some advantages.
Second, unlike a full authority system, the MCLAWS
Basic striictitre have to contend with a full authority mechanical path
that has ten times the authority of the flight control
The basic structure of the Modernized Control Laws
system. Finally, the current implementation uses a
(MCLAWS) investigated in this study is shown in
unity inverse plant model.
Figure 1. The figure shows the pitch axis structure
only; the roll and yaw axes have a similar structure.
Linear Analysis
Also shown for comparison is the structure of the
current pitch axis control laws that are part of the The control laws were designed using nine and
Stability Augnentation System (SAS) on current twenty-two state linear models of the UH-60 The
nine state model represents only the rigid bod! Reconfizurable Cockpit Simulator shown in Fipe
dynamics. hile the twenty-two state model adds the 3(a) while subsequent evaluations were carried out on
\\
flap and lag dynamics and dynamic inflow. The SAS the Sikorsky Fixed-base simulator shown in Fisure
and primary servos were modeled includin? rate and 3(b).
position limits. Also modeled were computational
and filtering delays to account for implementation Figure 4 shows the results of an evaluation of the
aspects of the control laws. MCLAWS as compared to the UH-60 control laws in
the Sikorsky Fixed-base jimulator. This simulator has
The control architecture, linear airframe and other a wide field-of-view and actual H-60 cyclic and
models described above were implemented in a collective sticks. The aircraft model used was a UH-
SIMULINK model and the linear' analysis was 60A Next-GenHel model at a heavy gross weight of
carried out using standard MATLAB tools. 19.302 pounds. The intcmtion was to evaluate the
control laws at a higher gross weight that was more
The linear design and analysis of the ACAH mode representative of the weights of current and future H-
were driven by ADS-33E bandwidtwphase delay 60 variants. The mission task elements (MTEs)
requirements and the more general time domain evaluated were a subset of those described in ADS-
requirements for a "sood" attitude response. Broken 33E and the performan,:e limits imposed on the
loop stability requirements [Ref. 91 were also evaluations were those for the utility configuration.
imposed. An example of the linear analysis carried Three pilots carried out the evaluations although due
out is shown in Figure 2. The left half of Figure 2 to time constraints on14 one pilot evaluated the
shows the ADS-33E small amplitude Accel./Decel. MTE. The figure shows the pilot
bandwidthlphase-delay pitch and roll axes evaluation ratings as measured on the Cooper-Harper scale (on
for an intermediate design. The right half of the this scale numerically lower ratings correspond to
figure shows the gain and phase margins for the pitch better pilot opinion). As tie Figure shows, across all
and roll axes. the MTE's evaluated the MCLAWS earned better
pilot ratings on the Cooper-Harper handling qualities
The rate mode was designed to be similar to the rating (HQR) scale [Ref. IO], with a maximum
existing UH-60A rate feedback system. Considerable improvement of 2.3 points and an average
effort was spent in making the transition between the improvement of 1.1 points.
attitude and rate modes seamless and transparent to
the pilot. This included adjustments to the switching VMS and Test description
conditions in Table 1 and the addition of faders and The next phase of the assesment was carried out on
transient free switches. the NASA-Ames Vertical )Motion Simulator (VMS).
The VMS environment includes large-amplitude
Once the linear design had been completed, a limited motion and the ability to simulate both day and
amount of numerical optimization was carried out degraded visual environments. Handling quality
using MATLAB optimization tools. This was done evaluations of five MTEs from ADS-33 were
by making selected control system gains variables of performed with the baseline UH-60A control laws
the optimization problem and imposing the above and the MCLAWS in a simulated day environment
design specifications while minimizing actuator and a Degraded Visual Environment (DVE). In
activity. The final gains obtained from this process addition, the baseline and :he modem control laws
formed the starting point for implementation of were assessed in the presence of wind and turbulence.
MCLAWS in a nonlinear simulation model and This section describes the sLmulation facility, matrix
subsequent piloted simulation evaluation. of configurations, and the conduct of test.
Sample results j+orn recorzjiigurable and fixed-base Descriptiorz of the Facilih,
The control laws developed using the linear analysis The NASA-Ames large-amplitude motion flight
tools were implemented in the Sikorsky Next-GenHel simulator is shown in Figure 5. The VMS real-time
model of the UH-60. This was done using an mathematical model of the UH-60.4 is based upon
automated pictures-to-code framework developed at the generalized, modularized programs that make up
Sikorsky, which allowed for a short cycle time the Sikorsky General Helicopter Flight Dynamics
between control law changes and piloted simulation Simulation (GenHel) [Ref. 1 11. Off-axis corrections
evaluation. to the model, also termed iierodynamic phase lap,
were implemented to correct the low speed off-axis
Once the control laws had been successfully response [Ref. 111. The overall helicopter weight was
implemented in the nonlinear Next-GenHel model. set to 16.525 pounds.
the simulation was hosted on two separate facilities.
Preliminary engineering and piloted evaluations of The crew station. or cochpit cab. has single pilot
;1
the control laws were carried out on the Sikorsky ceat mounted in the center of the cab and four-trnaze
preheiitation "windows" to provide outside imagery Reference 14. Table 3 shows the matrix of theat
Figure 6. The cockpit cab was recently modified configurations.
[Ref. 131 into a UH-60A configured cockpit to
support the Joint Shipboard Helicopter Integration Cotidiict of the Test
Process (JSHIP). The cockpit visual display r>:. .stem
Participating in the test were five experienced rOt3r1
consisted of an array of five flat panel screens with cs
wing experimental test pilots representing the
each screen abutted against the others :n a 230-degree
Army. NASA, and Sikorsky. The pilots were allo\red
horizontal by 70-degree vertical senxircle. It was some time to practice each of the maneuvers to
masked to represent the pilot's field of view from the
acquainted with the aircraft's response and with thc
BLACK HAWK'S right seat. The visual imagery
simulator visual cues. The MTEs were first evaluated
was carefully tailored t:> contain adequate macro- in the day scene to allow maneuver training 2nd
texture (i.e.. large objects and lines on the ground) for
course cueing familiarization under good lighting
the determination of the rotorcraft position and
conditions. All evaluation runs were carried out \\it11
heading with reasonable precision. The baseline
the motion system operational. For evaluation in a
stick-to-visual delay was about 70 msec including a
DVE, the out-the-window view was severel!;
IO-msec math model cycle time. A seat shaker pro-
degraded and the pilots used NVGs to carry out the
vided vibration cueing to the pilot, with frequency
maneuvers. The simulator has more travel in one
and amplitude programmed as functions of airspeed, horizontal axis as compared to the other and the cab
collective position, and lateral acceleration. Aural was rotated to take advantage of this depending
011
cueing was provided to the pilot by cab-mounted
the maneuver being evaluated. Initial trainin2
speakers. The aural model, driven by aircraft sessions were provided prior to at least three data
parameters, provided main rotor noise, tail rotor
collection records for each maneuver. A structured
noise, engine and transmission noise. Standard pilot questionnaire was used to elicit pilot comments
helicopter instruments and BLACK HAWK grips
and a HQR was provided. Aircraft flight dynamic
were installed in the cockpit Figure 7. The cockpit
and control response parameters and task
instruments were displayed on two Cathode Ray performance data were recorded for display and l;t~cr
Tubes (CRTs) mounted on a panel directly in front of
analysis. Monitors in the VMS control room
the pilot seat. The orientation of the instruments was provided quick and easy assessment of pilot-task
the same as the actual aircraft and they were performance relative to the ADS-33 desired and
compatible with night vision devices. The VMS
adequate standards Figure 10. This information W:IS
control loading system (McFadden Systems) is
relayed to the evaluation pilot in the cockpit Lo
digitally programmed to provide realistic force-feel
confirm of his perception of task performance based
cues for the cyclic, pedal, and collective controls.
on course cueing. Evaluation sessions were limited
The overall VMS motion system capabilities are
to one hour to mitigate fatigue effects.
listed in Table 2.
Results
Matri.r of Configurations
The results below are presented in terms of HQRb
The matrix of configurations included the baseline and summary task data for each of the MTEs. Also.
UHdOA flight control laws, the modem control laws, SAS actuator saturation data and task time histor)
five ADS-33 MTEs, and variations in ambient data are presented for illustration.
conditions to include day and night, and calm or with
windslturbulence. The five ADS-33 MTEs included Overview of HQRs
Hover, Vertical Maneuver, Pirouette, Lateral
Figure 11 and Figure 12 summarize the average
Reposition, and Departure/Abort. Reference 1
HQRs for good and degraded visual condition>
provides a detailed description of these MTEs along
respectively for both the baseline UH-60A coiiti.01
with the desired and adequate performance standards.
laws and the MCLAWS. Figure 11 shows the averas
The VMS course cueing for these MTEs was
HQRs from up to three pilots across five difl'ercnc
carefully matched to the flight test set-up (Figure 8)
ADS-33 maneuvers. The patterened bars corresporld
used in Reference 3. In this way, some comparison
to baseline UH-60A CLAWS while the dark barb
could be inferred between the actual flight test results
represent MCLAWS. For day (WE) conditions.
and the UH-60A simulated day conditions on the
there was a consistent improvement in the HQRj
VMS. Ambient conditions for pilot evaluation of
across maneuvers and pilots as a result of usillr
these five MTEs were varied from the day conditions
MCLAWS instead of the baseline UH-60A cL,\bb's.
to night. In the night scene, the pilot's vision was
It is clear that for each of the five rnaneuvsih
aided by wearing night vision goggles (NVGs)
evaluated, there was between a 0.5 and 2.5 Poi'''
(Figure 9). In addition, some MTEs were evaluated in
improvement due to the MCLAWS. A\,eragelf oic"
the presence of wind and turbulence. Details of
all maneuiers the improvement was slight]? mor'
these wind and turbulence models are reported in
than 1.0 point.
for file out of the six comparisons. The RMS
For DV-E. on the other hand, several of the debiations in longit ~dinal and lateral position,
mancu\.crs were too difficult to complete with altitude, and heading were less in three out of four
adequate performance levels with either control cases with the MCLA'NS compared to the UH-60'4.
system. Pilot comments indicate that this was Looking at the maximum excursions, the lateral
because of the degraded visual eni3ronment being position, altitude, and heading were within desired
unrealistically degraded and the task cues blending performance standards for nearly all evaluations.
into the background. The result was a series of high However. maximum excursions in longitudinal
HQRs with no ability to distinguish one control position ranged from desired, to adequate. to not
systcm from the other. Pilot debriefs indicated that adequate for both control laws. The problem in
operations in such conditions are unlikely. A UCE controlling the longitudinal axis was a main factor
determination was not performed. For other toward degrading the I-IQR. As the figure shows,
maneuvers however, at least one of the pilots was only two MCLAWs cases are within adequate time (5
able to distinguish differences. with the MCLAWS 8 sec); all the rest are outside of adequate.
always being superior to the baseline CLAWS.
Fizure 12 shows the DVE ratings only for those Figure 15 shows the performance trends for the
maneuvers where cueing appeared to distinguish Vertical maneuver in GVE. There are some
bctween desired and adequate standards. All pilot interesting differences between pilots for this MTE.
ratins pairs (i.e. ratings for baseline CLAWS and For Pilot 2, all the RMS deviations in longitudinal
MCLAWS) that were numerically high and equal and lateral position, and heading were less with the
have been eliminated. The resulting average ratings MCLAWS compared tcl the UH-60A evaluations.
are shown in the figure. It can be seen that the For Pilot 1, only the RMS deviations in longitudinal
MCLAWS configuration was rated between 1.0 and position were less with MCLAWS whereas, the
1.5 points better than the baseline and that the lateral position and heading deviations were less with
average improvement across maneuvers was a little the UH-60A. Looking at the maximum longitudinal
more than one point. and lateral position excursions, evaluations included
a mixture of desired and adequate performance with
Finally shown in Figure 13 are the HQRs for only one falling to not adequate. Six of nine cases
different wind levels. The patterned group of bars with excursions into adequate were with the UH-
correspond to the baseline UH-GOA control laws with 60A. Maximum heading excursions were nearly all
the first pattern corresponding to evaluation without within desired tolerances for both MCLAWS and
any winds and the checked pattern for evaluation in UH-60A evaluations.
light winds. The shaded bars are for the MCLAWS
with the shading corresponding to different levels of The performance summary for the Pirouette MTE is
wind and turbulence (black is no winds, dark gray is shown in Figure 16. The the subplot shows both the
with winds, and light gray is with winds and time to complete the circle and time to stabilize to a
turbulence) with MCLAWS. It is immediately hover at the end of the maneuver. Both
apparent from the figure that the MCLAWS configurations were roughly equivalent. The RMS
performed well in rejecting the disturbances and did deviations show a mixturt of less deviation with
not show significant degradation as a result of the either MCLAWS or UH-t10A depending upon the
wind and turbulence. Note that in all MCLAWS task parameter and pilot. The maximum radial
cases, the HQRs in the presence of winds and excursion was within desired tolerance for nearly all
turbulence were still better than the HQRs for the evaluations. However, altitude excursions for two-
baseline CLAWS without winds and turbulence. thirds of all evaluations were within the adequate
range. Improving the height axis response with the
Peiforntarrce analysis of G VE rims addition of Altitude Hold could be beneficial.
In order to further understand handling qualities
For two-thirds of the etaluarions, the heading
improbements offered by MCLAWS, task
excursion data show Izrger deviations with
perl'ormance data were analyzed for both general
trends and to provide specific examples of MCLAWS. It should be noted that a MCLAWS-
to-GenHel implementation error was discovered
ILnprovement. This subsection presents some of
after the VMS simulation which caused the
thcse trends for the GVE runs.
MCLAWS heading hold tc nzork well for small
deviations, but saturate for larger ones.
Figure 13 shows the performance summary data for
he Precision Hover maneuver for both the baseline
Figure 17 shows the performance trends for the
[IH-60A control laws and the MCLAWS for two
Pilots. Although the times to achieve a stabilized Lateral Reposition MTE. The times to complete the
maneuver were shorter for the MCLAWS for every
were rather long for both control laws. the
comparison aith LIH-60'4 exn ihrough the times
X'ICLAWS e\.aluations were completed in less time
7
\%ere all similar in duration. For 13--out--of-- I Y performance summary data for the Vertical LITE
cases. the RMS deviations for longitudinal position. in the DVE is shown in Figure 20. All the
altitudz, and heading \yere less with hICLAWS evaluations werz completed within desired times nith
compared to UH-60% Maximum exccrsions for all of the MCLAWS cases being slightly better than
altitude and heading were nearly all within desired the UH-60A cases. With the exception of a couple of
performance standards for both MCLAWS and UH- comparisons, the &\Is deviations in position and
60A. Longitudinal excursions wer: nearly all within heading were substantially less with b.ICL,A&s
desired for Pilot 2 for both control laws. Although compared to UH-60A. In general, maximurn
Pilot 1 had mainly adequate performance for excursions in longitudinal and lateral position were
longitudinal excursions for both control laws, the all within desired performance standards with
MCLAWS evaluations were always significantly MCLAWS whereas the UH-60A cases were in the
better. adequate range. All of the maximum heading
excursions were within desired performance for both
The Normal Depart-Abort MTE performance sets of control laws.
summary is shown in Figure 18. For this MTE. all
MCLAWs cases were completed within desired Actiiutor satiiratiori
times whereas only 2 of 7 runs with UH-60A were One of the primary concerns when designing control
within desired. One UH-60A time was outside of laws for a partial authority system is actuator
adequate. Smaller RMS deviations in lateral position. saturation. In the current implementation of the
altitude, and heading are about equally split between MCLAWS actuator saturation is addressed directly
the control law comparisons. For comparisons where by switching from an ACAH response system to a
large differences in RMS exist, the MCLAWS RC system before actuator saturation occurs.
deviations were always smaller compared to the UH- However, this switching needs to be balanced with
604. Maximum altitude and heading excursions for the need to maintain ACAH characteristics over a
the two control laws were all within desired useful range of aircraft velocities and attitudes. As a
performance and roughly the same. Lateral result at the extreme edges of the ACAH envelope
excursions were generally less with the MCLAWS some saturation is expected to occur. Figure 21
compared to UH-60A. shows the percentage of time for which the pitch and
roll axis actuator authority was saturated. when
Pe formnnce analysis of DVE rims performing different MTEs.
The performance summary for the Pirouette and
Lateral Reposition MTEs in DVE is shown in Figure As the figure shows, for the lateral reposition the roll
19. Similar to the GVE evaluations, the two times axis actuator authority was saturated for
associated with the DVE Pirouette were all within the approximately 1520% of the time. Analysis of
desired tolerances and roughly the same between the individual run data shows that most of the saturation
two control laws. With the exception of an initial occured at the start and the end of lateral maneuver.
comparison, the RMS excursions in radial position The saturation in the pitch axis happened primarily
and altitude were significantly less for the MCLAWS during the terminal phase of the maneuver.
cases compared to the UH-60A. RMS heading
deviations appear larger with MCLAWS, but this is During the normal depadabort maneuver, the lateral
attributed to the previously mentioned integration axis showed almost no saturation, but there was up to
error. Looking at the maximum excursion data, 30% saturation on some runs in the longitudinal axis.
trends that mimic the RMS results are seen. It should Most of this occured when transitioning out from
be noted that nearly all of the altitude excursions ACAH to RC during the forward acceleration phase
were within the adequate performance standards. of the maneuver and was not extensively commented
Once again this points to further improvement upon by the pilots.
possible by addition of an altitude hold function.
Finally, during the precision hover maneuver there
The performance summary data for the Lateral was minor saturation in the longitudinal axis but none
Reposition in DVE shows all times were adequate in the lateral axis. The vertical and pirouette MTEs
and about the same for both configurations. The RMS showed no pitch or roll axis saturation for any of the
deviations for longitudinal position, altitude, and runs with the MCLAWS.
heading are all less with MCLAWS compared to UH-
60A cases. Maximum altitude and heading The general trends shown in Figure 21 compare
excursions were nearly all within the desired range favorably to the results in (Ref. 7). For the precision
for both the lCICLAWS and UH-60X cases. On the hover maneuver. the MCLAWS longitudinal
other hand, maximum longitudinal excursions were saturation is senerally at a much lower level. For the
nearly all in thz adequate range for both control laws. lateral reposition and departlabort maneuvers. it must
be pointed out that the more asgressive ccout/attack
290
versions of the performance requirements were u. orkload and improi e overall handling qualities
imposed during the Ref. 7 evaluation and some of the aircraft.
allou~ancem ust be made for this.
Acknowledgements
Ilbtstmrive examples Technical tasks described in this document include
In order to illustrate the improvements due to the tasks supported with shared funding by the U.S.
MCLAWS as compared to the baseline UHdOA rotorcraft industry iind government under the
control laws, Figure 22 compares two individual RITANASA Cooper; tive Agreement No. NCC2-
Precision Hover cases, one for each control system. 9019. Advanced Rotorcraft Technology, January 01,
The top row of subplots shows cyclic stick movement 2001. The authors would also like to thank the pilots
for 30 seconds after the pilot declares a hover who participated in different phases of the study and
capture, while the bottom row shows the task contributed valuable comments and insight.
performance in terms of lateral and longitudinal drift.
It is immediately apparent that for the MCLAWS run Re.feerences
the pilot was expending a smaller effort (as reflected 1. Anon., “Handling Qualities Requirements for
in stick movement) to obtain better position keeping Military Rotorcraft.” Aeronautical Design
(as reflected in the small drift) when compared to the Standard-33 (ADS-33E-PRF), US Army
case with the baseline UH-60A control laws. Aviation and Missile Command, March 21,
2000.
Finally, shown in Figure 23 is a comparison of the 2. Key, D.L., “Analysis of Army Helicopter Pilot
lateral reposition MTE using the two control laws. Error Mishap Data and the Implications for
The top half shows the ground trace of typical runs Handling Qualities,” NASA TM-1999-208797,
for each type as executed by Pilot 1. The figure USAAMCOM AFDIYTR-99-A-006, September
shows that the performance in terms of longitudinal 1999.
drift was much smaller with the MCLAWS. The pilot 3. Hicks, J.E., “Army P,viation Safety Investment
activity as measured by lateral stick Power Spectral
Strategy,” presented at American Helicopter
Density (PSD) (calculated using CIFER, Ref. 15) was
Society 58th Annual Forum, Montreal, Canada,
also lower for the MLCAWS case than for the case
June 11-13,2002.
with baseline UH-60A control laws.
4. Blanken, C.L., Cicolani, L., Sullivan, C.C., and
Arterburn, D.L., “Evaliiation of ADS-33 Using a
Conclusions
UH-60A Black Hawk Helicopter,” presented at
Sikorsky Aircraft Corporation developed modernized
American Helicopter Society 56th Annual
flight control laws that incorporate rate and attitude
Forum, Virginia Beach, Virginia, May 2-4, 2000.
command response types, while utilizing the existing
5. Mitchell, D.G., Aponso. B.L., Atencio, A., Key,
partial authority SAS actuators of the BLACK
D.L., and Hoh, R.H., “Increased Stabilization for
HAWK helicopter. These control laws were
UH-60A Black Hawk Night Operations,”
evaluated in a VMS simulation effort, where the
objectives were to assess potential handling qualities USAVSCOM TR-92-A-00’7, November 1992.
improvements in simulated degraded visual 6. Hoh, R.H., Mitchell, D.G., Baillie, S.W., and
environments and to reduce the risk of subsequent Morgan, J.M., “Flight 1.ivestigation of Limited
implementation and flight test of these modern Authority Attitude Command Flight Control
control laws on AFDD’s EH-60L helicopter. System Architectures for Rotorcraft,” NASA CR
196707, US.4d4TCOh$T F: 97-.4-008? July 1997.
The conclusions of this investigation are: 7. Whalley, M.S. Howitt, J., and Clift, S.,
(1) Attitude Command Attitude Hold control laws in “Optimization of Partial Authority Automatic
pitch and roll improve handling qualities in the Flight Control Systems i‘or HoverLow Speed
low speed flight regime. These improvements are Maneuvering in Degraded Visual
consistent across a range of MTEs and for both Environments,” presented at American
GVE and DVE. Helicopter Society 55th Annual Forum,
(2) The MCLAWS perform better than the baseline Washington D.C? May 1999.
UH-60A control laws in the presence of wind 8. Key, D.L. and Heffley, R.K., “Piloted
and turbulence.
Simulation Investigation of Techniques to
(3) The improved handling qualities in the pitch and
Achieve Attitude Command Response with
roll axis allow the pilot to pay more attention to
Limited Authority Servo,’ NASA CR 2002-
the vertical axis and hence altitude performance
2 1 139 1, USAAMCOM i\FDI)lTR-02-A-003,
also improves. However, it is clear from pilot
January 3002.
comments and altitude excursions, especially
during the Pirouette MTE. that the addition of an
Altitude Hold function would further reduce
29 1
9. hlIL-STD-9-190. General Specification for Flight
Control Systems Design. Installations. and Test
of Piloted Aircraft.
10. Cooper, G.E. and Harper, R.P., “The Use of P;!ot
Rating in the Evaluation of Aircraft Handling
Qualities,” NASA TN D-5153, April 1939.
I 1. Howlett. J.J., “UH-60 Black Hawk Engineering
Simulation Program: Volume 1 - Mathematical
Model,” NASA Contractor Report CR 166309,
1951.
12. Schulein, G. J. , TiscnIer: M. B., Mansur, M. H.,
Rosen, A., ”Validation of Cross-Coupling
Modeling Improvements for UH-60 Flight
Mechanics Simulations.’’ Journal of the
American Helicopter Society, Vol. 47, No 3, pg.
209-2 13, July, 2002.
13. Sweeney, C., and Nichloson, R., “Using
Dynamic Interface Modeling and Simulation to
Develop a Launch Recovery Flight Simulation
for a UH-60A Black Hawk,”
Interservice/Industry Training, Simulation and
Education Conference, Orlando, Florida,
November 200 1.
14. Lusardi, J.A., Blanken, C.L., and Tischler, M.B.,
“Piloted Evaluation of a UH-60 Mixer
Equivalent Turbulence Simulation Model”,
presented at American Helicopter Society 59th
Annual Forum, Phoenix, AZ, May 2003.
15. Tischler, M.B. , Cauffman, M.G., “Frequency-
Response Method for Rotorcraft System
Identification: Flight Applications to BO- 105
Coupled RotorFuselage Dynamics,” Jocirnal of
the American Helicopter Society, Vol. 37, No 3,
P ~ 3S-1 7, July 1992.
i
293
7
.
Figure 1: Architecture of MCLA'CVS compared to baseline UH-60A SAS
Pitch Bandwith Pitch GainlPhase Margins
0.4
-
80
-
0.3
60
Y> 3
m sP)
5 0.2
-0 I 40
al n
2VI
0.1 20
n
nv 0
0 2 4 0 10 20
GM[db]
Bandwidth [radisec]
Roll Bandwith Roll GainIPhase Margins
0.4
-
80
f
0.3
60
Y -tOI
>
zm sQ)
0.2
7 I 40
n
P)
2VI
0.1 20
a
n 0
U 0 10 20
0 2 4
GM[db]
Bandwidth [radisec]
Figure 2: Linear analysis for ,\.ICLACVS pitch and roll axes
294