Category HELICOPTER FLIGHT DYNAMICS

Novel response types

The unique capability of helicopter flight in three dimensions is typified by low-speed manoeuvring close to the ground. The pilot’s task can be conceptually divided into three subtasks – navigation, guidance and stabilization. Navigation, or generally where the pilot wants to go in the long term, requires low workload and intermittent attention by the pilot to make course corrections. Guidance relates more to where the pilot wants to go in the shorter term and requires moderate levels of workload that depend principally on the speed of flight and on the level of visibility, or how many flight seconds the pilot can see ahead. In poor visibility and at low level the guidance workload can become very high. Stabilization relates to the continuous activity and workload to maintain the required aircraft attitudes. With unaugmented helicopters, the stabilization workload can be high and requires continuous pilot compensation as the helicopter is disturbed and deviates from the intended flight path. Most of the effort into augmentation for supporting precise flight path control close to the ground and obstacles has been directed towards reducing the stabilization workload by incorporating attitude hold functions, in combination with rate command, or even attitude command response types. For example, both Puma and Lynx have short-term attitude hold as part of their limited authority SCAS, triggered when the attitude falls within a small range close to zero (Lynx), or the pilot’s cyclic is stationary (Puma). With the advent of high authority digital flight control, the capability now exists for providing response types that not only remove the stabilization workload, but also directly support the guidance task. Conceptually, the pilot requires control over the magnitude and direction of the aircraft velocity vector. To date the only criterion developed for flying qualities in the guidance task has been the TRC response type required for operations in the DVE according to ADS-33 (Ref. 6.5). TRC refers to a response characteristic where constant pilot controller input leads to a proportional earth-referenced translational velocity response. Level 1 flying qualities are defined by a TRC response having a qualitative first – order shape and equivalent rise time of between 2.5 and 5 s. The lower limit is set to avoid abrupt attitude changes during TRC manoeuvres. Equivalent rise time and the Level 1 TRC control power and sensitivity boundaries are defined as shown in Fig. 6.70. The limited supporting data for TRC response characteristics are published in Ref. 6.32.

Two notable examples of the implementation of novel response types are worth highlighting – the advanced digital flight control system (ADFCS) implemented in the McDonnell Douglas experimental AH-64 Apache AV05 (Ref. 6.88) and the Velstab System designed for the production Boeing/Sikorsky RAH-66 Comanche (Refs 6.89, 6.90). The philosophy behind the ADFCS experimental system, designed and flown in the mid-late 1980s, was to provide low workload management of aircraft control for single-pilot operations. The provision of automatic moding was part of this philoso­phy, hence not ‘trading flight path with button management’. The flight path control logic implemented in AV05 comprised two selectable modes – the flight path vector system (FPVS) and aerobatic system. The control logic for the FPVS contained many

Fig. 6.70 TRC response sensitivity boundaries (Ref. 6.5): (a) Definition of equivalent rise time, Tjeq (Tyeq); (b) control/response requirement for centre-stick controllers; (c) control/response requirement for sidestick controllers

 

Fig. 6.71 Polar plot of speed/azimuth control logic for FPVS AH-64 (Ref. 6.88)

innovative features as summarized in Fig. 6.71, showing three auto-transition modes for hover, low speed and cruise. In the low-speed mode, (translational) inertial acceleration is commanded with the right-hand controller, with inertial velocity hold. This response type was selected for low-speed NoE flight to ensure that pilots would not be required to hold stick forces for long periods, as would occur with a pure TRC system. In the cruise mode, the turn rate hold feature gives the pilot the ability to maintain relative flight paths while changing speed. In both low-speed and cruise modes, the vertical axis response type of acceleration command/flight path hold simplified the task of terrain flight – the pilot could place the flight path vector symbol in the helmet-mounted display on the desired point on the terrain, e. g., hill top, to ensure clearance of vertical obstacles. In the author’s view, AV05 represented, in its day, the state of the art in a full flight envelope ACT system with novel response types. To quote from Ref. 6.88, ‘… non-pilots could command near envelope limit performance from the aircraft in the course of a one hour demonstration flight’; the present author can testify to this, as he was one of the privileged engineers to fly AV05 in exactly this fashion.

At the time of writing the first edition of this book, the core active flight control sys­tem (AFCS) of the RAH-66 Comanche had been demonstrated in piloted simulation to confer Level 1/ good Level 2 flying qualities for the UCE 1 ADS-33 MTEs (Ref. 6.89). In addition, the selectable control modes, giving the pilot hybrid ACVH (attitude com­mand, velocity hold) for flight in DVE, were evaluated as solid Level 1 for the ADS-33 DVE MTEs (Ref. 6.90). On the RAH-66, the DVE control system is described as the VELSTAB mode, and the characteristics relative to inertial groundspeed (V <60

Fig. 6.72 Comanche VELSTAB characteristics (Ref. 6.90)

knots)/airspeed(V > 60 knots) are illustrated in Fig. 6.72 (from Ref. 6.90). At very low speed (groundspeed within ±5 knots), in the shaded hover-hold region on Fig. 6.72, TRCPH is provided, giving the pilot a precise positioning aid. Hover-hold break-out is enabled when the pilot demands a velocity outside the threshold or applies a large cyclic demand. Below 60 knots groundspeed, the pilot flies with ACVH, with wind compensation to eliminate, as far as possible, non-uniformities in the required pilot con­trol strategies in windy conditions, and to ensure a smooth blend between ground and airspeed at 60 knots. Low-speed turn coordination combined with altitude hold allows the pilot to fly single handed with the aircraft body axis always aligned with the flight path.

ADFCS on the Apache and AFCS with VELSTAB on the Comanche are visions of things to come in helicopter flight control and flying qualities, which have been realized successfully in flight and simulation. Flight with novel ground-referenced response types is being enabled by advances in sensor and digital flight control system technologies and clearly offers the potential for significant reductions in pilot workload, particularly for flight in DVE. The military driver is to provide capabilities previously not possible, but significant safety improvements in civil operations in poor visibility or congested and/or confined areas are also likely to be realized with this technology.

Multi-Axis Response Criteria and Novel-Response Types

This section covers two areas that are relatively immature in terms of the existence of any underlying flying qualities database. The primary emphasis of all flying qualities requirements has been the division of criteria into axes over which the pilot has control. In practice, most MTEs require coordinated control inputs in all axes, and the question arises as to whether the combination of single axis criteria is sufficient to ensure pilot acceptance in multi-axis tasks. Practically all the material in the earlier sections of this chapter deals with the most conventional, rate or attitude command, response types. With the advent of fly-by-wire/light and the attendant active controls technologies, the scope for changing the way pilots fly helicopters is very broad indeed. The term novel response types is coined to classify non-attitude-based systems, and some discussion on the current status and thinking in this area constitutes the final topic in this section.

6.8.1 Multi-axis response criteria

Most of the test MTEs in ADS-33 are primarily single axis tasks, e. g., accel-decel (pitch), bob-up (heave), sidestep (roll) and hover turn (yaw). For these, at least in theory, off-axis control inputs are required only to compensate for cross-couplings. Flying qualities requirements on couplings (see Section 6.7), at least when fully developed, should ensure that aircraft are built that demand minimum compensation only. Other MTEs are in their nature multi-axis and require the pilot to apply coordinated controls to achieve satisfactory task performance, e. g., pirouette, angled approach to hover, yo-yo combat manoeuvres and roll reversals at reduced and elevated load factors. Very little research has been done, at least in recent years and hence related to modern missions, on flying qualities criteria specifically suited for combined-axis helicopter manoeuvres. ADS-33 refers only to the requirement that control sensitivities should be compatible and responses should be harmonious. Control harmony is arguably one of the most important aspects of flying qualities, but finding any formal quantification has proved difficult. An intuitive definition seems to be that harmony is a quality achieved by having similar levels of characteristic response parameters, at least in the interacting axes. At a fundamental level, harmony then implies the same response types in the different axes, e. g., rate command in pitch combined with attitude in roll would not be harmonious, perhaps even leading to degraded ratings. Harmony applies most of all to pitch and roll, normally commanded through the same right-hand controller. Manoeuvring at low speed and close to the ground, the pilot directs the rotor thrust with the right-hand controller. The author is of the view that harmony in this mode of control should, as far as possible, encompass response type, bandwidth and control power (particularly for AC response types). Then if the pilot wants to fly at 45° to

the right, he initiates and terminates the manoeuvre by moving his controller in the desired direction. This requirement is naturally met in TRC response types discussed below, but would not be for AC or RC types if the ratio of the minimum requirements of ADS-33 were maintained (e. g., ±30° pitch, ±60° roll for aggressive manoeuvring with attitude response types in UCE 1).

In forward flight, one of the important multi-axis criteria that has received attention is the requirement for turn coordination. As a pilot rolls into a turn, two compensating controls have to be applied. Aft cyclic is required, for helicopters with manoeuvre stability, to compensate for the pitch damping moment in the turn. Into-turn pedal is required to compensate for the yaw damping in the turn. Additional compensation will be needed for any steady-state incidence or sideslip required to augment the turn performance. The requirements for manoeuvre stability have already been discussed in Section 6.3. The requirements on yaw control harmony, and on the attendant sideslip response, are more complicated as they depend on the phase between roll and sideslip in the Dutch roll lateral/directional oscillation. The turn coordination requirements in ADS-33, for example, focus on the amount of sideslip resulting from an abrupt lateral cyclic control input; the criterion also highlights the point that sideslip response is more tolerable when it obviously lags the roll response.

The requirement for cyclic control harmony in manoeuvring flight at moderate to high speed translates into the need for similar time constants for roll attitude and normal acceleration response. Fortunately, this is normally the case, with the pitch bandwidth and control power being harmonized with the correspondingly higher parameters for roll. For example, as a pilot rolls into a turn with rate command in both pitch and roll, a bank angle of 60° and load factor of 2 can be achieved in similar times (about 1.5 s for an agile helicopter). One potential problem can arise with a pure RC response type during a roll reversal manoeuvre. Flying a steady turn the pilot will be pulling back to maintain the pitch rate in the turn. As the pilot executes the roll reversal, he has to judge his control strategy carefully, making sure that the cyclic passes through the centre with zero pitch input, to avoid making a discrete change in pitch attitude. Reference 6.35 reports on a study to evaluate the relative benefits of rate and attitude command response types. One of the workload problems with RC highlighted by pilots was the care required when reversing a roll to avoid making a pitch change that inevitably led to a speed decrease or increase as the manoeuvre progressed. To overcome this problem, speed hold functions were proposed. Also, automatic turn coordination was generally preferred by pilots, at least up to a moderate level of agility, obviating the need to apply any compensating pitch or yaw inputs.

One final point on multi-axis tasks, and to make it we assume that an aircraft that has been demonstrated as Level 1 in all axes according to clinical objective criteria will also consistently achieve ‘desired performance’ in practice. It is recognized that this is a contentious issue, but for the moment we assume that the individual criteria are robust enough that failure to comply will guarantee bad flying qualities. It is well recognized that at high levels of pilot aggression or in degraded environmental conditions, an otherwise Level 1 aircraft can degrade to Level 2. This is generally accepted as being an inevitable consequence of operating helicopters in harsh environments and can apply to both military and civil operations. But this raises the question as to whether a helicopter that has degraded to Level 2 in two or more axes will still be able to meet adequate performance levels in multi-axis manoeuvres. There is some evidence to suggest that the answer is negative. In discussing combined axis handling qualities,

Hoh (Ref. 6.60) advances an advisory ‘product rule’ that predicts that an aircraft with two axes both receiving ratings of 5 on the Cooper-Harper scale will actually work out as a 7 in practice, i. e., Level 3. We shall return to this and other related issues concerning subjective pilot opinion in the next chapter.

Collective to yaw coupling

The application of collective pitch causes the rotor to slow down (or speed up) and the governor to increase (or decrease) the fuel flow, hence to increase (or decrease) the engine torque, which in turn results in a yawing moment reaction on the fuselage. Helicopter pilots learn to compensate for this effect early in their training and need to allocate a certain level of compensatory workload for harmonious inputs in pedal when applying collective. Most helicopters are built with a mechanical interlink between tail rotor collective pitch and main rotor collective lever, hence nullifying the gross effects at one particular flight condition. ADS-33 sets a limit of maximum yaw rate excursions of 5°/s following abrupt collective inputs, and also sets more complex limits on the ratio of yaw rate to vertical velocity, for which no substantiating data have appeared in the open literature; readers are referred to Ref. 6.5 for details.

6.7.2 Sideslip to pitch and roll coupling

The remaining cross-coupling effect to which we give some attention is the attitude response to sideslip. Pitch response to sideslip is a peculiar helicopter phenomenon that can lead to control problems in uncoordinated manoeuvres when it is required to point the fuselage off the flight path. The rotor downwash field can affect the horizontal stabilizer giving a powerful nose-up pitching moment in zero sideslip conditions. As sideslip builds up, the wake washes off to one side, exposing the tail to free air and leading to pitch down moments in both port and starboard manoeuvres. Figure 6.69, taken from Ref. 6.87, illustrates the various contributions to the pitching moment due to sideslip on the UH-60 helicopter. The canted tail rotor contributes 50% of the strong cross-coupling and the horizontal stabilizer, contributes 25%. The overall value for the derivative Mp is about 2000 ft lb/° and is equally as powerful as main rotor cyclic

Fig. 6.69 Contribution of aircraft components to the pitching moment due to sideslip –

UH-60 (Ref. 6.87)

control, therefore requiring significant pilot compensation. A strong pitch response to sideslip can exacerbate pilot disorientation problems following tail rotor failures. To the author’s knowledge there are no published data defining handling qualities boundaries for sideslip to pitch effects; it remains a topic for future research.

Roll-sideslip coupling is defined in ADS-33 as part of the forward flight criteria for lateral/directional oscillatory characteristics (Ref. 6.5). It follows the fixed-wing format and is expressed in terms of the ratio of the oscillatory to the average component of roll attitude response following a lateral cyclic control input. The assumption with this type of format is that the roll oscillations are caused by sideslip excursions in a roll manoeuvre. The Level 1/2 boundary for this parameter depends on the phase angle between roll and sideslip.

Two final points need to be made on cross-coupling in general. First, it should be stated that most helicopters are designed with mechanical interlinks, or control couplings, that minimize the initial coupled motions following abrupt control inputs. In both the Sikorsky CH-53E and UH-60A, for example, application of collective lever couples to all the remaining controls through mechanical interlinks. This is a relatively simple and effective way of reducing some of the primary effects, but does nothing about the rate and velocity couplings. Second, there is evidence that the maximum level of acceptable coupling is a strong function of on-axis response characteristics – the poorer the on-axis handling qualities, the less tolerant pilots are of coupling. This is consistent with the intuitive rule that the presence of more than one degrading handling influence will lead to a combined handling worse than the average of the individual characteristics. This is bad news of course, but shouldn’t come as a surprise to any pilot who has tried to ‘tighten-up’ on the controls with an unstabilized or partially stabilized helicopter, or tried to define a Level 1 roll axis response boundary, with a configuration having Level 2 pitch or cross-coupling characteristics. But future helicopters with active control technology will have low levels of couplings by design. What does need consideration with these highly augmented aircraft is the level of

mission criticality and even flight criticality of the coupling augmentation. Future pilots may not be as well trained to fly cross-coupled helicopters, and loss of augmentation may be analogous to engine or tail rotor failure in today’s helicopter operations. The central issue then becomes one of system integrity, particularly relating to sensors, and sufficient integrity/redundancy needs to be incorporated so that the risk of loss of coupling augmentation is remote.

Pitch-to-roll and roll-to-pitch couplings

Подпись: and Подпись: Mp Mq

Pitch-roll and roll-pitch cross-couplings can be powerful and insidious. The natural sources of both are the gyroscopic and aerodynamic moments developed by the main rotor and, in dynamic manoeuvres with large attitude excursions, the uncommanded and sometimes unpredictable off-axis motion can require continuous attention by the pilot. Generally, the magnitude of the pitch-to-roll couplings are more severe than roll to pitch, due to the large ratio of pitch to roll moment of inertia, but are, arguably, more easily contained by the pilot, at least at low to moderate frequencies. Roll-to-pitch coupling effects can have a much stronger impact on flight path and speed control and hence handling qualities in moderate to large manoeuvres. From the results of a piloted simulation study on the NASA Flight Simulator for Advanced Aircraft (FSAA), Chen and Talbot (Ref. 6.80) hypothesized that the critical cross-coupling handling qualities parameters were the ratios of short-term steady-state roll (pitch) to pitch (roll) angular rates, approximated by the ratios of aerodynamic derivatives

Pilot HQRs were consistently awarded in the Level 2 area for values of the ratios greater than about 0.35. When the revision to MIL-H-8501 was initiated in the early 1980s, the

NASA results were initially used as the basis of new pitch-roll criteria. The derivative ratios clearly took no account of the control couplings, however, and were also difficult to measure with accuracy. After some refinement, the criteria adopted in ADS-33 were based on a time domain formulation, in terms of the ratio of the peak off-axis response to the desired on-axis response after 4 s following an abrupt step input, in the form

roll step^, pitch step^<±0.25 (Level 1), ±0.6 (Level 2) (6.35)

A series of additional piloted simulation and flight trials, conducted in the late 1980s at the Ames Research Center (Refs 6.81,6.82), confirmed the importance of the derivative ratios, but argued that the new ADS-33 criteria did not cater for the higher frequency control coupling effects, or the interaction with on-axis characteristics. With regard to control coupling, the data in Refs 6.81 and 6.82 suggested that equivalent rotor control phase angles of about 30° would lead to Level 3 handling, confirming the RAE results reported earlier in Ref. 6.17. The relationship between the ADS-33 criteria and the equivalent linear system parameters can be illustrated using the simple first-order rate response formulation (Ref. 6.83), given by the equations

Подпись: / 0 pk VT Подпись: Lp  IMmc Mq) V L r/1c Подпись: Mp Mq Подпись: "41c Подпись: (6.38)

The relationship between the ADS-33 criteria and the parameters in eqns 6.36 and 6.37 can be reduced to the form

indicating that the pitch attitude coupling in a roll manoeuvre is dependent on both the cross-damping ratio of Ref. 6.80 (Mp/Mq) and the control sensitivity ratio scaled by the ratio of roll to pitch damping. Even with a zero value for the rate coupling Mp, control couplings can give rise to similar levels of pitch attitude excursion. The ADS time domain parameter in this simple model is therefore linearly related to the derivative ratios Mp/Mq and Mmc/Lmc. Given the roll axis control sensitivity and bandwidth, the importance of the control coupling is therefore inversely proportional to the pitch attitude bandwidth, Mq, hence emphasizing the importance of pitch axis effectiveness in cancelling coupling effects. Contours of equal ADS response are therefore given as shown in Fig. 6.67.

Our understanding of the handling qualities effects of roll-pitch coupling has been significantly extended by the series of flight/simulation experiments conducted by the US Army/DLR in the early 1990s, with support from DRA. The work to date is reported in Refs 6.84 and 6.85 and has focused on evaluating handling qualities in forward flight MTEs typified by the lateral slalom. In Ref. 6.84, couplings are classified into three types – those due to rate and control effects and the so-called washed-out coupling effects, more typical of augmented rotorcraft. Reference 6.85 concludes that the current ADS format is adequate for discriminating against unacceptable characteristics in the

Fig. 6.67 Contours of equi-response on cross-coupling chart

first two categories, but not the washed-out effects, which appear to be frequency dependent. However, data are presented in Ref. 6.85 which suggest a modification to the ADS Level 1/2 boundary as shown in Fig. 6.68(a), where the acceptable level of coupling has been reduced to 0.1. A new frequency domain criterion is proposed in Ref. 6.85 which appears to give a more consistent picture for all three types of coupling. The general form of the criterion is presented in Fig. 6.68(b), where the key parameters are the magnitudes of the frequency response functions between pitch (roll) and roll (pitch) rates, evaluated at the bandwidth of the off-axis attitude response. This format, therefore, again reflects the importance of the response characteristics in the coupled axis. Strictly, the data from Ref. 6.85 will define only the vertical portions of the boundaries in Fig. 6.68(b): the author has hypothesized the upper, horizontal boundaries, which would be defined for pitch axis tasks, and the curved boundary between, reflecting the additional degradation in multi-axis tasks, when couplings in both axes are present.

Fig. 6.68 Comparison of ADS-33C and Pausder-Blanken criteria for roll-pitch coupling requirements (Ref. 6.85); (b) proposed frequency domain format for roll-pitch-roll coupling

(based on Ref. 6.85)

6.7.1 Collective to pitch coupling

At high speed, the application of main rotor collective pitch can generate powerful pitch and roll moments on the fuselage. Experiments to quantify the effects of collective to pitch coupling on handling qualities were reported in Ref. 6.86. The results were not conclusive but did indicate the powerful degrading effect, sending HQRs across the full span of the Level 2 range as the coupling parameter increased. ADS-33 reflects the limited dataset for collective to attitude couplings and sets limits on the pitch attitude change occurring within 3 s of an abrupt collective input. The limits are set as the ratio of pitch attitude change to the corresponding change in normal acceleration and take the form

where the negative value in eqn 6.40 corresponds to down collective inputs. The above criteria apply to forward flight. In low-speed flight the emphasis is more on the collective to yaw couplings.

Cross-Coupling Criteria

Helicopters are characterized by cross-couplings in practically every axis-pairing, and the ubiquitous nature of cross-coupling constitutes one of the chief reasons why piloting this type of aircraft requires such high skill levels developed through long training programmes. Satisfying the direct, or ‘on-axis’, response characteristics, described in previous sections for roll, pitch, heave and yaw, is necessary but not sufficient to guarantee good helicopter flying qualities. Any helicopter test pilot would be quick to confirm this and might even advise that fixing the off-axis, cross-coupled response, was a higher priority for conferring Level 1 on-axis handling. Ideally, a designer would like to eliminate all sources of coupling. This is not only impossible (with only four controls), but probably also unnecessary, and one focus of the efforts in handling research has been to establish the maximum level of tolerable coupling. As with on – axis response criteria, this has proved to be task specific and particularly task-gain, or task-bandwidth, dependent. In very general terms, the low frequency/trim coupling effects are driven by the velocity couplings; the moderate frequency effects are reflected in the angular rate couplings and the higher frequency effects are dominated by the control couplings, in either sustained or washed-out form. Pilot subjective opinion of the degrading influence of coupling will therefore depend on the task, e. g., precise positioning, rapid slalom or target tracking. Many of the physical sources of cross­coupling have been described and discussed in Chapter 4 of this book. Here, we shall review the major types of couplings, and the database of results relating to handling qualities criteria and discuss what more needs to be done to set quality requirements. In the following subsections, the use of the condensed descriptor, e. g., pitch to roll, refers to the roll response due to pitch; any distinctions between control and motion couplings will be made as appropriate.

Trim and quasi-static stability

We have already discussed many of the issues relevant to lateral/directional trim and quasi-static stability in the section on roll axis response characteristics, particularly the need for positive trim control gradients. One additional handling criterion that fits best in this category is the requirement on heading (or roll and pitch attitude) hold functions as defined in ADS-33. With heading hold engaged and activated by the release of the yaw control device, the reference heading should be captured within 10% of the yaw rate at release. In addition, following a disturbance in yaw, the heading should return to within 10% of the peak excursion within 20 s for UCE 1, and within 10 s for UCE > 1.

As discussed in the introduction to this section, at hover and low speed, particularly close to the ground, the helicopter creates a disturbed aerodynamic environment in which the tail rotor is required to work. When the powerful main rotor vortex wake strikes the tail, particularly from the port side, or in the form of the ground vortex in rearward flight (see discussion on interactional aerodynamics in Chapter 3), large yawing moment disturbances can make it difficult for the pilot and even for the simple automatic hold functions to perform well. The problem is intimately associated with the open tail rotor; while generally more efficient than an enclosed fan or jet thrust device in clean aerodynamic conditions, the tail rotor tends to be more sensitive to wind strength and direction, particularly when positioned close to the vertical stabilizer. This sensitivity manifests itself as a non-uniform distribution of lift over the tail rotor, giving rise to large collective and power requirements in critical flight conditions. Reference 6.76 discusses the merits of tail rotor cyclic control in this context, which could be scheduled with collective to provide the optimum lift distribution in all flight conditions.

Small amplitude/low to moderate frequency: dynamic stability

At high forward speed, helicopters typically suffer from the same, so-called, nui­sance mode as fixed-wing aircraft – the Dutch roll, exacerbated by weak weathercock stability and strong dihedral effect. The theory for this coupled mode has been presented in Chapter 4. In response to a doublet pedal input, the aircraft motion will soon be dom­inated by a weakly damped oscillation, comprising strongly coupled yaw, roll, sideslip and, for helicopters, pitch motions. The two fundamental parameters are the natural frequency and damping, and it is not surprising that efforts to define handling qualities related to the stability characteristics of this mode should have been focused on the corresponding two-parameter chart, or classical frequency/damping plane. The quality boundaries defined in ADS-33 are derived largely from the considerable database for fixed-wing aircraft (Ref. 6.6), with slightly relaxed stability requirements. A compar­ison of military and civil requirements (for single pilot IFR) is shown in Figs 6.65(a) and (b). The variety of boundaries drawn in Fig. 6.65(a) once again reflects the mission orientation of the military requirements. The comparison between civil and military requirements highlights several aspects already met in previous criteria – chiefly the greater demands made on designers of military aircraft. Another noticeable difference is shown in the low-frequency damping requirements. Both are based on minimum total damping at zero frequency. The more stringent military Level 1/2 boundary is set at time to half amplitude of 0.69 s, while the civil boundary is set at time to double amplitude of 9 s. The evidence supporting the minimum rnn boundary for military

Fig. 6.65 Lateral/directional oscillatory requirements: (a) military (Ref. 6.5)

helicopters in Fig. 6.65(a) is thought to be fairly limited; to the author’s knowledge, no supporting data for these boundaries relevant to helicopters have appeared in the open literature since the publication of ADS-33. It is interesting to note that the criterion for the stability of long-period pitch and roll modes (frequencies less than 1 rad/s) lies within the rnn >1.0 contour (see Fig. 6.45). One aspect raised here is the impor­tance to handling qualities of the separation of frequencies between the modes with low stability. Modes with overlapping frequencies can cause additional pilot workload, especially when strong cross-couplings are present.

The reader is referred to the analysis of Dutch roll stability and response in Chapters 4 and 5 where the results of the AGARD WG18 study on Dutch roll stability are discussed (Ref. 6.79). Most military and civil helicopters have autostabilization in the yaw axis to improve the Dutch roll damping and, generally, augmented rate damping is sufficient to achieve Level 1. In some cases, design efforts are successful in improving the natural aerodynamic stability in yaw. Reference 6.63, for example, describes how modifications to the fin of the SA332 Super Puma significantly improved the Dutch roll characteristics of this aircraft compared with the original Puma. Figure 6.66, taken from Ref. 6.61, illustrates the marked improvement in Dutch roll damping through the fitting of end-plates on the BK117 helicopter; the new design met the FAA requirement without autostabilization.

Two final points need to be made about the lateral/directional oscillatory mode. First, at high speed, the frequencies are encroaching on the range appropriate to small amplitude tracking, and the requirements on damping should be seen as supplementing the bandwidth criterion. Second, a very important handling characteristic associated with Dutch roll motion is the phase and relative amplitude between roll and yaw; separate criteria address these issues under the heading yaw-sideslip response to lateral cyclic, and we shall address these when discussing cross-coupling, in Section 6.7.

Small amplitude/moderate to high frequency: bandwidth

The heading response bandwidth requirements are presented in Figs 6.64(a) and (b). The higher performance required for tracking tasks is common to both hover/low – speed and forward flight MTEs, e. g., Level 1 boundary at 3.5 rad/s. Such high values of bandwidth do not occur naturally in helicopters; typically, the yaw axis has very low damping, particularly at low speed, with rise times of the order of 2 s (see Chapter 4). The results of Refs 6.77 and 6.78 have already indicated the levels of damping that pilots feel are appropriate for aggressive yaw tasks. Bandwidths of 3.5 rad/s and higher are more consistent with rise times of the order 0.5 s and hence require some form of response quickening control augmentation.

Fig. 6.64 Yaw axis bandwidth/phase delay boundaries (Ref. 6.5): (a) (low speed) target acquisition and tracking – (forward flight) air combat (yaw); (b) general MTEs

Moderate to large amplitude/low to moderate frequency: quickness and control power

Following the formats adopted for the roll and pitch handling criteria, Fig. 6.61 shows the heading quickness boundaries for hover and low-speed MTEs, and Fig. 6.62 shows the minimum control power requirements for rate response types. It can be seen that the requirements for quickness are as demanding as for roll response, placing a particularly strong emphasis on yaw moment capability. For example, the ability to

Fig. 6.61 Yaw axis quickness – hover and low-speed flight (Ref. 6.5): (a) target acquisition and tracking; (b) general MTEs

achieve a yaw rate of 40°/s in a discrete 20° heading change requires a maximum acceleration of about 2 rad/s2. For an aircraft like Lynx this corresponds to generating a tail rotor thrust perturbation of about 1000 lbf. Overlaid on Fig. 6.61 is the boundary of maximum quickness values measured on Lynx performing precision hover turns, with heading changes from 30° through to 180°. The ADS-33 requirements for target acquisition and tracking are fairly demanding and call for a powerful tail rotor, or fantail in the case of the aircraft for which ADS-33 was developed, the RAH-66. The high quickness levels were partly established through simulation trials conducted on the VMS. Reference 6.77 presents results from simulation trials that included target

Fig. 6.62 Minimum yaw control power requirements – rate response type (Ref. 6.5)

acquisition and tracking MTEs. Experimental variables under investigation included the yaw damping, weathercock stability and response shape. Figure 6.63, taken from Ref. 6.77, shows the apparently very limited region fit for Level 1 handling in air-to-air target engagement, on a damping/response-shape diagram. The high levels of yaw damping required to achieve Level 1 for this kind of operation could not normally be produced without significant artificial response augmentation. Similar results were reported in Ref. 6.78 for forward flight MTEs. The tracking phase of an aerial combat engagement is more concerned with the higher frequency, small amplitude behaviour, and once again, the authors of ADS-33 turned to bandwidth to discern quality.

Fig. 6.63 Short-term yaw response requirements in air-to-air tracking task (Ref. 6.77)

Yaw Axis Response Criteria

As we turn our attention to the fourth and final axis of control, the reader may find it useful to reflect on the fact that of all the ‘control’ axes available to the pilot, yaw is, arguably, the most complex and the one that defines the greatest extent of the flight envelope boundary, both directly or indirectly. Figures 6.60 (a) and (b), for example, show the SA330 Puma control limits for the forward flight sideslip envelope, bounding the envelope at higher speeds, and for hovering in a wind from the starboard side, bounding the low-speed envelope. Excursions beyond these boundaries can lead to

Fig. 6.60 Puma sideslip and sideways flight limits: (a) sideslip envelope in forward flight; pedal margin for hover in wind

loss of control or structural damage. Within these constraints, the pilot may feel able to command yaw motion in a relatively carefree manner. However, the pilot is not provided with a cue as to the magnitude of the loads in the tail rotor critical components. The tail rotor can absorb up to 30% of the total engine power, and in some flight conditions, tail rotor torque transients can lead to damaging loads. The pilot is also not provided with precise knowledge of sideways velocity or sideslip angle, but will typically fly at low level with primary reference to ground cues, oblivious to velocities relative to the air mass, and relying on tactile cues through control position for information on the proximity to aerodynamic limits. Our discussion suggests that yaw control is far from carefree and any handling deficiencies can contribute significantly to pilot workload for both civil and military operations.

Yaw control functions can be grouped into the following categories:

(1) balance of powerplant torque reaction on the fuselage, in steady state and manoeuvring flight;

(2) control of heading and yaw rate in hover and low-speed flight, giving all-aspect flight capability;

(3) sideslip control in forward flight, giving fuselage pointing capability;

(4) balancing or unbalancing manoeuvres, to increase or decrease turn rate.

Of these, the control functions in category (2) have probably accounted for by far the greatest range of yaw handling problems, stemming largely from the effects of main rotor wake-tail rotor-rear fuselage-empennage interactions (Refs 6.74-6.76).

In ADS-33, criteria for yaw handling are defined in much the same formats as for roll and pitch. These will be reviewed briefly.