2.1 Morphing aileron architecture

The morphing aileron design starts from the definition of aerodynamic shapes to ensure enhanced aircraft aerodynamic performance in off-design conditions. Such target shapes shall be matched by a tailored kinematics driven by an appropriate actuation system. In accordance with (78) seat regional aircraft specifications, the morphing aileron hinge axis is located at 70% of the chord. Aerodynamic requirements (delay flow transition) and structural constraints drive the aileron external contour (Fig. 1). In fact, the shapes must satisfy the following geometrical requirements:

• • defined range for the aileron's tip deflection around the hinge axis (–7^° to 7^°);

• • continuous monotonic curvature of the morphed camber lines (no changes in the sign of the slope of aileron camber);

• • no elongations in camber line and/or skin induced by morphing.

Figure 1.

Morphing aileron aerodynamic shapes.

The aileron aerofoil was then approximated by a segmented rib architecture based on a finger-like layout properly tailored to enable aileron camber morphing upon actuation. Each rib (Fig. 2) was assumed to be segmented into three consecutive blocks (B1, B2 and B3) connected by means of hinges located on the aerofoil camber line (A, B). Block B1 is rigidly connected to the rest of the wing structure through a torsion tube enabling aileron rotation for roll control. Blocks B2 and B3 are free to rotate around the hinges on the camber line, thus physically turning the camber line into an articulated chain of consecutive segments. A linking rod elements (L) hinged on non-adjacent blocks forces the camber line segments to rotate according to specific gear ratios.

Figure 2.

Morphing rib architecture: (a) blocks and links, (b) hinges.

The linking element makes each rib equivalent to a single-DOF mechanism: if the rotation of any of the blocks is prevented, no change in camber/shape can be obtained; on the other hand, if an actuator moves any of the blocks, all the other blocks follow the movement accordingly. The rib mechanism therefore uses a three-segment polygonal line to approximate the camber of the aerofoil and to morph it into the desired configuration while keeping the aerofoil thickness distribution approximately unchanged. The ribs’ kinematic is transferred to the overall aileron structure by means of a multi-box arrangement (Fig. 3).

Figure 3.

Morphing aileron structure: multi-box arrangement.

Each box of the structural arrangement is characterised by a single-cell configuration delimited along the span by homologue blocks of consecutive ribs, and along the chord by longitudinal stiffening elements (spars and/or stringers). Upon the actuation of the ribs, all the boxes are put into movement, thus changing the external shape of the aileron; if the shape change of each rib is prevented by locking the actuation chain, the multi-box structure is elastically stable under the action of external aerodynamic loads. A four-bay (five-rib) layout was considered for an overall (true-scale) span of 1.4m; AL2024-T351 alloy was used for spars, stringers and rib plates, while C50 steel was used for the ribs’ links. Off-the-shelf airworthy components were properly selected for the bearing and bushings at the hinges and coupled to torsional springs to recover any potential freeplay. A multi-module skin was considered in conformance with the multi-box segmentation; three aluminium-alloy panels were then adopted, each panel sliding over the consecutive one in an armadillo-like configuration. Airflow leakage at the skin segment interfaces was prevented through low-friction silicone seals. As one might expect, the segmented skin architecture does not significantly impact the aileron torsional stiffness and results in a slightly higher (but on the same order) conventional aileron.

The deployment kinematics use a ‘direct-drive’ actuation based on actuation arm that is rigidly connected to the B3 block in Fig. 2. This arm rotates the one-DOF-based mechanical system and transmits the actuation torque from the actuator to the adaptive rib. The control actions aim at producing small camber variation in the adaptive aileron corresponding to a rigid rotation of a plain control surface comprised between –7° and 7° during flight.

A self-contained morphing device made of links, hinges and joints to alter the inner geometry is developed with the purpose of providing a standard hinged control surface with an added functionality that may improve aircraft off-design points such as cruise or climbing. However, as with any promising technology to be integrated in aircraft, an accurate estimation of its weight loss or weight gain with respect to the conventional configuration is crucial. To date, this benefit can be only grossly computed or preliminarily assessed. On the one hand, according to Breguet's formulae, aircraft range strictly depends on aircraft aerodynamic efficiency and the ratio between the maximum take-off weight and the burned fuel weight. On the other hand, it is evident that the benefits associated with morphing will be great enough to compensate for the drawback of possible weight penalties. Therefore, in order to gain competitive advantages through morphing devices, it is necessary that:

(1) $$\begin{equation} \Delta W_{fuel}^E > \Delta W_{fuel}^{{M_{TOW}}} \end{equation}$$

Where ΔWE fuel indicates the saved fuel weight percentage due to the incremental aerodynamic efficiency for the effect of the morphing device, and ΔW MTOW fuel represents the overall aircraft structural weight penalty due to the use of the morphing aileron.

The weight of the morphing aileron designed for a 78-seat aircraft was about 25kg. Since the aircraft maximum weight is around 20tons, we see that the morphing aileron is only 5% of the entire aircraft weight. It is obvious that the weight penalty could be easily offset by a fuel savings of 3% to 6%^{(Reference Bolonkin and Gilyard18)} provided by such a morphing technology From the manufacturing standpoint, the developed concept consists of many standard pieces and requires careful assembly procedures to support operators. This may affect its industrial applicability. Efforts are currently being pursued to simplify the design using topology optimisation methodologies to reduce the number of parts.

2.1.1 Aerodynamic loads evaluation

The VLM method was adopted to evaluate aerodynamic pressure distribution along the aileron corresponding with each considered flight attitude (wing angle-of-attack, flight altitude and speed) and aileron geometrical configuration. 3D flat-panels mesh was generated in correspondence of the outer wing segment; the mesh was constituted by 6 macro-panels (Fig. 4(a)) respectively representative of the outer wing root and tip portions (panels P1 and P3), of the wing box including wing leading edge (panel P2) and of the three aileron's segments (panels P4, P5 and P6). Each panel was further subdivided in a convenient number of boxes. For each flight attitude and aileron shape, the lifting pressure (P_,i ) acting along each box (b_i ) was calculated according to the following equation:

(2) $$\begin{equation} {P_i} = q({P_{0,i}} + \alpha {P_{\alpha ,i}} + \gamma {P_{\gamma ,i}}), \end{equation}$$

where:

• • q = 0.5×ρ × V ² _∞ is the dynamic pressure, ρ the air density at the flight altitude and V _∞ the airspeed;

• • α is the wing angle of attack;

• • P_{0, i} is the pressure arising on b_i corresponding to unitary dynamic pressure at α, γ equal to zero (aerofoil baseline camber effect);

• • P_{α , i} is the pressure on b_i due only to unitary adue only to unitary l baseline camber effect); at

• • Pγ_{, i} is the pressure on b_i due only to unitary γ at unitary dynamic pressure (morphing effect).

Figure 4.

Aerodynamic mesh (a) and pressure distribution along the aileron span (b).

Thanks to Equation (1), P_0,i, P_{α , i ,} P_{γ , i} are calculated only once for all the boxes and then combined according to the flight attitude parameters (α, q) and aileron morphed shape (γ) to be investigated.

The combination of α,q, and γ leading to the most significant pressure levels along aileron segments was then determined and used as design operative condition for structural sizing purpose. Span-wise pressure distributions at the design operative condition (α = 2°, q = 4,425N/m², γ = 7°) are plotted in Fig. 4(b).

2.2 Actuation system design

Five different distributed actuation arrangements were specifically developed for the morphing aileron. The distributed actuation design consists of a number of actuators potentially enabling a redundant and fault-tolerant operation of the adaptive ribs. In this work, the following solutions were investigated:

• • linear guide with rollers with arm linked to the first movable rib block (B2);

• • precision linear slide with recirculating and non-recirculating ball carriages driving the second movable rib block (B3);

• • cam follower with arm linked to the second movable rib block (B3);

• • cylindrical ball-bearing guide mechanism driving the second movable rib block (B3).

The actuation system design included the worst-case design in terms of operative loads and room available for the kinematics. For this reason, the third aileron rib (Rib 3) was considered for the structural sizing of the actuation architecture (Fig. 5). Due to their small size, Ribs 4 and 5 were considered passive and their movements slaved to Rib 3.

Figure 5.

Aileron inner structure with rib identification.

2.2.1 Mathematical model

The actuation concept is an enhanced release of the oscillating glyph, already addressed in literature on a morphing trailing-edge device^{(Reference Dimino, Flauto, Diodati, Concilio and Pecora3)}. The analytical sketch of the concept is shown in Fig. 6.

Figure 6.

Oscillating glyph connected to the second rib segment of the morphing aileron.

By assuming that the system is perfectly rigid and there is no friction among the components, the mechanical advantage of the mechanism (MA) can be written as:

(3) $$\begin{equation} {\rm{MA}} = \frac{{{\rm{LOAD}}}}{{{\rm{DRIVER}}}} = \frac{{{{\rm{M}}_{{\rm{rib}}\# 2}}}}{{{{\rm{M}}_{{\rm{att}}}}}} = \frac{{{\rm{F}}{{\rm{B}}_{\rm{L}}}}}{{{\rm{F}}{{\rm{B}}_{\rm{R}}}}} = \frac{{{{\rm{B}}_{\rm{L}}}}}{{{{\rm{B}}_{\rm{R}}}}}, \end{equation}$$

where the ${M_{rib\# 2}}$ is the operational torque due to aerodynamic loads acting on the third rib segment, while M_att is the actuation torque provided by the actuator in order to equilibrate the system. Furthermore, F is the force that the crank produces by means of the cursor, B_L is the force arm and B_R is the crank projection along the guide. From Equation (3), it follows that the mechanical advantage only depends upon the geometry of the system. In particular:

(4) $$\begin{equation} \cot \varphi = \frac{{\rm{L}}}{{{\rm{R}}\sin \beta }} - \cot \beta \end{equation}$$

Getting previous assumptions concerning no friction-based concept, MA and the actuator rotation (β) can also be calculated for the mechanical system shown in Fig. 7. In this case, the mechanical advantage assumes the form:

(5) $$\begin{equation} {\rm{MA}} = \frac{{{\rm{LOAD}}}}{{{\rm{DRIVER}}}} = \frac{{{{\rm{M}}_{{\rm{rib}}}}}}{{{{\rm{M}}_{{\rm{att}}}}}} = \frac{{{\rm{F}}{{\rm{B}}_{\rm{L}}}}}{{{\rm{FR}}\sin }} = \frac{{{{\rm{B}}_{\rm{L}}}}}{{{\rm{R}}\sin }} \end{equation}$$

and

(6) $$\begin{equation} \beta = {\sin ^{ - 1}}\left[ {\left( {\frac{{{\rm{BL}} + {\rm{X}}}}{{\rm{R}}}} \right)\sin } \right] \end{equation}$$

Figure 7.

Analytical scheme of the cylindrical bushing.

The mechanical advantage is a crucial feature which characterises the actuation concept. From Equations (2) and (4) it is possible to scale the external aerodynamic moment acting on the ribs to obtain the balancing torque. In addition, it follows that the resulting mechanical advantage drastically decreases with shorter B_L. This aspect is very important because it reduces the applicability of the second concept even if the the beam excursion range angle is wider than in the previous case. Equations (3) and (5) allow calculating the actuator shaft rotation (β) needed to achieve a given morphing angle (φ) of the rib block and hence of the entire mechanism. Finally, the force F is determined to verify the stress occurring in the carriage moving into the rail that will not to exceed its design value. Therefore, the actuation rod is subjected to the simultaneous action of the force F and the external moment ${M_{rib\# 2}}$ , both producing bending stress. This indicates that actuation system design requires a trade-off between the mechanical advantage and the geometrical constraints limiting the actuator shaft rotation and L/R ratio.

2.2.2 Linear guide with rollers

The first concept is based on the adoption of a compact linear guide characterised by a slider and a steel rail with a C-shaped cross section (Fig. 8). The slider is equipped with radial bearing rollers in alternating contact with both sides of the raceway. Radial bearings enable the guide to withstand high forces normal to the sliding line (on the order of 800N).

Figure 8.

Main components of the linear guide with bearings, rail and slider.

As shown in Fig. 9, such a device is fastened to the actuation steel rod of the morphing aileron driving the morphing rib kinematics through the control of B2 position. On the other side, the system transforms the actuator rotation in actuation force by means of the actuation leverage made of a crank.

Figure 9.

Actuation system composed by linear guides with internal bearings.

2.2.3 Precision linear slide

A precision linear slide with recirculating and non-recirculating ball carriages are investigated (Fig. 10) as an alternative solution to the linear guide. The former is made of a lightweight and compact linear-motion rolling guide comprising a U-shaped slip table and a stainless-steel track rail obtained by precision forming. The latter is made of a synthetic resin retainer used to host the balls while preventing their contact noise. The actuation architecture is shown in Fig. 11.

Figure 10.

Precision linear slide with recirculating (left) and non-recirculating (right) ball carriages.

Figure 11.

Actuation system using a precision linear slide mechanism.

2.2.4 Cam followers

Cam followers provided with bearings have stud in which needle rollers are assembled in a thick outer ring. They exhibit a small friction coefficient and excellent rotating performances with a high radial load capacity. The inner components are shown in Fig. 12.

Figure 12.

Cam follower.

In order to maximise the mechanical advantage (L/R increase) with respect to the previous configurations and to prevent potential mechanical plays arising during the manufacturing process, it was decided to connect the actuation rod to the rib block B3. The installation layout conceived for this solution is shown in Fig. 13.

Figure 13.

Cam follower-based actuation system.

It should be noticed that the cam follower is positioned along the beam longitudinal axis inside the grinding surface. In order to transmit the actuator torque, the cam is in contact with the upper (morphed-down) or lower (morphed-up) sides of the grinding surface. At the same time, rotation and sliding must be ensured during deflection. This means that the beam cross-section is sized on the base of the cam diameter; however, due to high load, the contact surface between the cam and the beam may be subjected to excessive stresses because of its low thickness. On the other hand, the increased distance between the actuator shaft and the morphing pivot results in a higher mechanical advantage at the expense of a largest beam excursions during operation.

2.2.5 Ball bearing guide mechanism

A new architecture based on a cylindrical bushing that slides along a cylindrical beam is here considered. The device is sketched in Figure 14. This concept represents a more compact solution leading to smaller (upper and lower) excursions of the actuation beam during operation but with limited mechanical advantage. The system architecture as shown in Fig. 15 details the morphed-down configuration and the kinematic components such as beam, cylindrical bearing, fork and actuator crank.

Figure 14.

Main components of the ball bearing guide mechanism.

Figure 15.

Ball bearing guide-based actuation system.

3.1 Numerical modelling

The structural sizing of the precision linear guide with non-recirculating ball carriages was addressed by means of linear static analyses carried out in MSC-PATRAN/NASTRAN^® environments. The resultant aerodynamic moment on each actuated rib was calculated from the prescribed pressure distribution (2.2) and applied to the actuation chain as concentrated load. A specific FE model was realised to validate the actuation kinematics by using concentrated load and rigid elements which transfer the load to the structure. The beam was modelled through TET10 elements⁽¹⁹⁾. For each rib, the aerodynamic moment was transmitted to the structure through rigid elements (RBE2) and the reaction force acting on the linear guide was computed. The calculated force was multiplied by the crank arm to get the torque to be balanced by the actuator.

FEA results are reported below in terms of total displacements (maximum value: 2mm at beam tip) and load transmitted to the guide (239N). Since the allowable reaction force of the linear guide is equal to 232N⁽²⁰⁾, the single linear guide solution was considered structurally inadequate.

For this reason, two linear guides per each actuated beam were considered in order to improve their capability to withstand loads induced by aerodynamic pressure. The double-guide structural arrangement is depicted in Fig. 18. The actuator torque is here transmitted through a fork-shaped crank.

Figure 17.

Beam displacement contour (left); guide reaction load of 239N (right).

Figure 18.

Structural integration among the actuation system and the aileron with kinematic chain detail.

The FE results for the new actuation architecture are reported in Fig. 19; in this case the reaction forces normal at the guides result respectively equal to 177N and 179N.

Figure 19.

Beam displacement contour (left); guide reaction loads of 177N and 179N (right).

Figure 20.

Comparison between Analytical and CAD values for actuator shaft rotation and slider stroke during aileron morphing.

The total actuation torque required to deflect the morphing aileron was then evaluated. Considering previous reaction loads, M_att is equal to 7.96Nm. The slider stroke and the actuator rotation are compared with the CAD model value and represented in two diagrams. As expected, FE simulations confirm that at the maximum morphing deflection (7°) the guide slider has a stroke of 12mm while the crank rotation is equal to 37.30.

3.2 Structural analysis of the overall system

Once the optimal actuation arrangement was selected, detailed structural analyses of the overall system composed by morphing structure and actuation chains were addressed. A very accurate finite element model was then implemented for such a purpose (Fig. 21 and Table 2). All the relevant structural components (segmented ribs, spars, stringers, skin panels, actuation system leverages) were covered as well as the (screwed or riveted) joints between different parts.

Figure 21.

FEM overview: (a) overall mesh, (b) ribs and spars, (c) detail of the actuation leverage.

Table 2

FEM: mesh general data and adopted materials

Solid elements (CTETRA) were used for the mesh of the primary structure and the actuation leverage, while beam elements (CBEAM) were used for bolts, fasteners and pins.

Hinges were modelled referring to the usual scheme of rigid body connections. At each of the two sides of the hinge housing, a master node was placed at the centre of the circular hole; nodes on the edge of the circular hole were then slaved to it through an RBE2 connection. Master nodes belonging to the two sides of the hinge housing were finally joined through a CBUSH element showing low stiffness about the hinge axis⁽¹⁹⁾. In Fig. 22, the hinges connecting rib block 1 with 2 and rib block 2 with rib block 3 are shown.

Figure 22.

Modelling approach for the hinged connection.

Static analysis was performed with respect to the design load condition reported in paragraph 2.1.1. Specific load fields were applied to the upper and lower skin in order to reproduce the aerodynamic pressure distribution along each segment of the aileron (Fig. 4, paragraph 2.1.1).

Rigid rotation of the fork-shaped crank was constrained, thus simulating a typical condition characterised by locked actuators’ shafts. Aileron rotation around hinge axis was prevented as well. Positive margins of safety (Table 3) were found with respect to local plasticisation at design loads and with respect to failure at ultimate loads (design loads multiplied by 1.5).

Table 3

Margin of safety in the most stressed components

Table 4

Value and location of the maximum stress in the most relevant components

No relevant torsion around the hinge axis was detected (Fig. 23); in spite of ribs and skin segmentation, the conceived multi-box layout was shown to be adequately stiff in torsion with practically no impact on roll control effectiveness. A moderate yet undesirable elastic rotation of the last aileron segment about its hinge axis was observed; instead of adding stiffness to the structure, it was considered wiser to investigate the feasibility of recovering the rotation using actuator torque. The torque required to restore the undeflected configuration was then calculated by means of a dedicated linear analysis carried out on the deformed shape with enforced motion of the actuator shafts. The obtained torque successfully resulted in compliance with the performances of several commercially available actuators; moreover, the recovery of the undeflected configurations occurred without any local increase of stress.

Figure 23.

Morphing aileron elastic deflection under design loads (total displacements contour).