Dynamic Landing Loads on Combat Aircraft with external Stores using Finite Element Models

Wolfgang Luber, Eugen Dill, Karl Kainz European Aeronautic Defence and Space Company - EADS Military Air Systems OPES 81663 Munich, Germany Wolfgang.Luber@eads.com

ABSTRACT Modern combat aircraft are designed for multi role military missions, performing air superiority, and reconnaissance and ground attack. Therefore a large number of different external stores must be integrated during development and production phase of the aircraft. During the landing impact, the external stores generate high dynamic loads on the undercarriage, the landing gear attachment points and on the aircraft structure. A method has been developed for prediction of the dynamic loads, taking into account ground and flight test results. The computations are based on a general dynamic model including a non-linear model of the landing gear and the tire. The aircraft model can be used with and without external stores. In the first step dynamic loads are generated due to the landing impact on the main wheels and used as design and clearance loads of the undercarriage. In the second step the attachment loads of the undercarriage are used as input for the excitation of the aircraft structure. Symmetric landing impacts as well as crosswind landings with different sink rate and external store carriage have been investigated and will be discussed in the paper. Loads of all different ground manoeuvres have been analyses. Besides the normal landing in a symmetric carriage configuration, also heavy asymmetric external store carriage has been investigated. Rejected take off loads, arrested landings loads, towing, turning and symmetric and asymmetric braking loads have been analyzed for hot, normal and cold days. Dedicated flight tests have been performed to validate the theoretical predictions. The test results were used to update model parameters like damping coefficient and stiffness. The updated model is used to predict the dynamic loads on the undercarriage, aircraft structure and external stores for design and clearance. Generally speaking, one aim of this methodology and

the resulting quality of the aircraft model is to provide the store manufacturers with accurate preliminary specifications early in the store integration process in order to minimize the number of ground and flight tests, the external stores design modifications and finally the overall integration costs. Meanwhile the in service fatigue issues are taken care of thanks to the accuracy and robustness of the specifications.

INTRODUCTION Modern combat aircraft are designed for multi role military missions, performing air superiority, and reconnaissance and ground attack. Therefore a large number of different external stores must be integrated during development and production phase of the aircraft. During the landing impact, the external stores generate high dynamic loads on the undercarriage, the landing gear attachment points and on the aircraft structure. A method has been developed for prediction of the dynamic loads, taking into account ground and flight test results. The computations are based on a general dynamic model including a non-linear model of the landing gear and the tire. The aircraft model can be used with and without external stores. In the first step dynamic loads are generated due to the landing impact on the main wheels and used as design and clearance loads of the undercarriage. In the second step the attachment loads of the undercarriage are used as input for the excitation of the aircraft structure. Symmetric landing impacts as well as crosswind landings with different sink rate and external store carriage have been investigated and will be discussed in the paper. Loads of all different ground manoeuvres have been analyses. Besides the normal landing in a symmetric carriage configuration, also heavy asymmetric external store carriage has been investigated. Rejected take off

loads, arrested landings loads, towing, turning and symmetric and asymmetric braking loads have been analyzed for hot, normal and cold days. Results from loads analysis on repaired runway with minimum single spacing mats of 130 m will be shown and the difference on loads during wet and try runway landing impacts.

Aircraft Description The Demonstrator aircraft is a lightweight single seat; double sonic, twin engine advanced jet fighter. It is designed up to 9-g manoeuvring flight. It is fitted with provisions for up to eleven external store stations including four on each wing and three under the fuselage and a fixed tip pod on each wing tip station. Figure 1 illustrates the two side view of the aircraft. The external store stations can carry an assortment of hardware including fuel tanks, electronic counter measure pods, sensor pods, precision guides bombs, air to ground missile, air-to-air missile, and other devices. The delta wing structure is a carbon fibre design and the wing box is sealed and carries fuel. There are flaps and slats on the wing, as well as a fore plane.

Fig. 1: Demonstrator Aircraft used for landing impact process

Ground Landing Physics Ground based military aircraft operation in service implies hitting the ground, which can be flat or bumpy and being stopped in a very short distance. Therefore the amount of energy to be dissipated on this short period is excessively high. The ground landing can be considered into five distinct events with respect to time and landing case. Generally speaking, the main landing gear hits the ground first, followed by the nose landing gear impact, very shortly afterwards followed by the breaking phase with and without braking chute of the landing run. In case of emergency all military aircraft have to be designed that the aircraft withstand an arrester hook landing, where the loads are different compared with the normal landing course. Beside this different landing procedures an additional case, the repaired runway case have to be considered and is part of the structural design criterias of combat aircraft Figure 2 shows the different phase of impact during landing impacts. The two major physical phenomena are the main landing gears impact and the time elapsed between the right and left main wheels rotation motion starts or vice versa depending of the direction of the crosswind. This time is due to the aircraft bank angle and to the local contact conditions between ground and the tires. The consequences are a very high-energy shock correlated with the wheels rotation motion start and the strut elastic drawback, which is known as spin up and spring back and inboard outboard motion. The elastic drawback of most combat aircraft occurs at a natural frequency between 10 and 25 Hz, depending on the vertical speed of the aircraft. This includes low frequency excitation on the aircraft structure, internal avionic equipment and external stores depending on the coupling between their own dynamic behaviour and the initial input as well as the associated damping coefficients. The connection between the vertical reaction force (Fz) with the local friction coefficients between each tire and the concrete ground produces a rearward reaction force (Fx) and a lateral reaction force (Fy). This forces can applied on the connection point (ground loads) as well as on the wheel axle (axle loads) of each wheel. It should be mentioned that the only forces can be measured are the axle loads and not the ground loads. The aircraft bank angle connected with the friction coefficient includes a time lapse in the reaction force application between each main wheel.

Main gear impact

General Methodology The loads to apply to external stores and internal equipment depend on the dynamic coupling between the aircraft structure and the stores and or equipment. The couplings depend on the spectral density content of the input load and on the dynamic behaviour of the coupled stores and aircraft structure system. For each external store configuration an overall dynamic model is built from the basic finite element model and the different external store models as dynamic super elements. The modal damping coefficients are initially set up on the basis of past experiences of integration of external stores featuring similar dynamic behaviour, like weight, centre of gravity, stiffness and boundary conditions. This value is adjusted for each configuration with measured results from ground vibration and flight test.

Nose wheel impact

Braking

Chute landing impact

Fig. 3: Principle system of main and nose Landing gear mechanic Arrester hook impact Fig. 2: Different Landing Cases In summary, three types of dynamic models were applied: Validated general finite element model of the aircraft. The validation is performed with the

GRT results and applying the error localization and updating methods. Non-linear models for the landing gears and the arrester hook for calculating loads on the undercarriage itself and on the aircraft structure External stores dynamic super elements adjusted to the baseline aircraft model to analyze the dynamic response on store, u/c and the aircraft during landing impact.

antisymmetric boundary conditions applied to the symmetry plane. In advanced development it turned out that the half models cannot describe very well asymmetric carriage of external stores, therefore a complete model was generated for this purpose.

The principle forces on the main and nose landing gear are summarised in Fig. 3

Fig. 5: Finite Element Model Aircraft, dynamic coarse mesh Definition of fuselage generalized points For each fuselage section, a point was required, such that its motion would be representative of the section as a whole. Thus for a series of such points distributed along the length of the fuselage, the reduced stiffness and mass matrices could be obtained sand hence the mode shapes and frequencies. The generalized point for the section was located at the centre of gravity of the set of grid points to which it was connected. This generalized point was then connected using an NASTRAN RBE3 element, to this set of points, each of which was a local point and had stiffness in all three directions at least. The RBE3 element defines the motion as a weighted average of this setoff points on the section. This method eliminates the local eigenmodes, which are part or inside the fine grid system. Fuselage Equipment Points Special grid points were created at the centre of gravity of large item (greater than 30 kg) of non-structural mass, like engines, gun or avionics etc. where inertia loads where applied or mass and inertia properties allocated. Each equipment point was connected back to the structure using either elastic or rigid elements, in a manner representative of the actual installation.

Fig. 4: Finite Element Model Aircraft, static fine mesh

ANALYSES METHOD - STRUCTURE

Finite Element Model of the Structure The aircraft was dynamically modelled by a 6 degree of freedom finite element model, which fully represents the total aircraft stiffness and mass. The clean aircraft was split up into different substructures, namely, fore plane, wing with flaps and slats, fuselage, fin and rudder. All substructures stiffness matrices were calculated with MSC NASTRAN, starting with a very fine static finite element model Figure 4, applying a dynamic condensation to a coarse dynamic model, see Figure 5. For the fuselage generalized and equipment points were generated. Port and star port of the structure is assumed to be symmetric which allows creating a half model with symmetric and

Some details of the model:

The dynamic FE-model used for Modal approach represents a half model and a full model. That means that also asymmetric carriage can be analyzed with this set of models. Pylon flexible structure is included in FEM via NASTRAN GENEL elements Nonlinear structural behaviour due to backlash effects at connection parts on pylon, wing, missile are not modelled and not considered with any reduced stiffness values Missile structure is assumed to be rigid Fuel status for wing internal and external tank as well as fuselage tank are model by concentrated masses

gear system. Figure 7 shows the dynamic spring curve of the NLG. The Nose landing gear shock absorber is also a two-stage absorber with less oleo travel in comparison with the MLG. In figure 8 the tire dynamic spring curves for MLG and NLG is depicted. These nonlinear elements are introduced into the analysis model to simulate the landing impact.

Aerodynamic Model The unsteady aerodynamic forces used in the dynamic model calculation shall be represented in a conservative manner. The magnitude (modulus) of the unsteady forces of the flexible modes shall be predicted to represent a realistic value of Mach numbers and incidences. Aerodynamic loads are calculated with linear potential aerodynamic in-house codes applicable for subsonic and supersonic flight conditions.

Fig. 6 MLG Shock Absorber Static spring curves

Landing gear models Concerning the external forces, the steady as well as the unsteady aerodynamic coefficients corresponding to the rigid aircraft are applied on the overall aircraft model. Beside these external forces also the flight control laws were introduced in the model in order to simulate also the manoeuvres and the real landing impact. During the landing phases, due to high-energy shocks generated in the aircraft structure the right hand terms of the equation becomes very significant. Knowing that the solicited structures, like main undercarriage and nose undercarriage have a nonlinear behaviour, the dedicated models are generated with ADAMS and NASTRAN. The main structural parts are represented via standard linear finite elements of the aircraft. On the other side non-linear elements were introduced into the analysis model to simulate the real undercarriage. In figure 6 the static spring curve for the MLG is presented for the maximum oleo travel in the shock absorbers for three different landing conditions. The difference is tremendous between hot and cold day of the landing

Fig. 7 NLG Shock Absorber Static spring curves Following the above figures, the wheel is represented via a nonlinear element that considers the contacts with the ground. The shock absorber itself is modelled in principle with two different elements. A linear beam which represents the bending stiffness of the sliding struts and the absorber, which represents the oleo lamination and the friction which is direct proportional to the sinking speed. The dynamic behaviour and the dynamic loads provided by the models have been

validated with a couple of campaigns of flight tests for high sink trials with different aircraft configurations.

desig purposes only the linear curve of the expected landing speed will be used as the first assumption. All friction coefficient curves show the same trend that at a slip ratio between 0.1 and 0.15 the maximum of the ground friction coefficient is reached and after this point the value decreases to zero.

MAIN ANALYTICAL RESULTS The model description allows to predict, early in a development program or in the clearance procedure, the level of loads acting on the landing gear, the aircraft structure and the external stores during landing impact phases.

Fig. 8 MLG/NLG tyre dynamic spring curves

Fig. 10 Three side view of dynamic wheel centre loads of star port main landing gear at asymmetric impact

Fig. 9 Different friction parameter for landing fatigue cases In figure 9 the tire to ground friction parameters are shown for different aircraft speeds during landing. On the right side of this diagram the friction coefficient is shown at that point when the tire is full skidding on different landing speeds. It is very difficult in the stage of analysis to choose the right coefficient. During the development of the landing gear a fixed friction coefficient was chosen, but for the fatigue design of the landing gear different friction coefficients must be used in order to get a more realistic simulation of the practical landing procedures for aircraft in service. In MIL-8863C the standard values are explained, which is also the basic of the structural design criteria's. For

Fig. 11 Three side view of dynamic wheel centre loads of port main landing gear at asymmetric impact

The landing gear model combined with the aircraft model enables to analyse the influence of physical parameters on the dynamic response of the aircraft carrying external stores and provide load envelopes for clearances or specifications for design. Main results which can predict are the dynamic coupling between the landing gear and the elastic aircraft structure and between external store elastic modes with aircraft elastic modes. In case of new external store development the store can be modified early in the program to avoid restrictions in the flight envelope and damaging coupling of the system. On the aircraft side different parameters act on the frequency of elastic modes and therefore influence the dynamic response behaviour of the vehicle. For instants the amount of fuel can change the fuselage modes and the different weight of the external stores can also change the dynamic characteristic of the elastic aircraft. For a new store development and the landing gear itself a robust load specification must be delivered.

structure and the landing gears. To shorten this work a couple of mass key configurations were chosen. In principal the heavy weight of an aircraft influences the design of the landing gear due to the high loads during impact. The shock absorber as well as the landing gear must be tuned together to withstand the high peaks of the landing impact. In practice, the sink rates for aircraft are lower in heavier weight configuration. The aircraft structure itself has a vice versa behaviour. Due to the heavy masses on external pylons the aircraft structure shows lower excitation during the landing impact. Figure 10 depicts in a three side view the loads on the star port gear during an asymmetric impact. In this case the star port gear touches the ground first. The blue curves show the loads for the three axis of the centre wheel, the red curve is the summary of the three blue curves, which shows the resulting loads after the impact. Figure 11 shows the loads after the aircraft for the port gear with the same explanation of figure 10. It is worth to mention, that the loads are higher on the port gear,

spin inboard

axial

spring b

Fig. 12 Time history of port and star port main landing gear at asymmetric landing impact During the development of an aircraft a lot of configurations must be considered to design the aircraft which is the second impact a as on the star port gear which is the first impact on an asymmetric landing. These loads will be transformed to the attachment point

of the landing gear. The elastic gear as well as the nonlinear shock absorber will be taken into account at the transformation procedure. This attachment forces or loads will be used as an inner excitation of the aircraft structure itself. Figure 12 shows the complete landing in the time domain for port and star port wheels for the main landing gear. The important phases are marked, as spin up and spring back movement and the inboard and outboard movement of the star port and port landing gear.

Fig. 14 Load Envelopes at pylon interface

CONCLUSION This methodology enables to integrate and to certify on each external aircraft pylon stations very different types of stores in terms of weight, stiffness and centre of gravity and in the early development of new weapons to create a very robust store specification in terms of dynamic response. In case of store development the process have to start at the very beginning of the store development in order to provide for the Design department reliable specifications. In case of certification of the store, the data can be used to reduce ground and flight testing due to precise predictions. Therefore the integration of a new external store can be done in a minimum of time and cost. This program is under constant improvement due to a series of test campaigns, which improves the level of confidence in the prediction.

Fig 13 Comparison of test and analysis for forward wing tip z-acceleration

Figure 13 compares the vertical acceleration of the right forward wind tip pod. The black line shows the test results and the green line the calculation. This figure demonstrates that the dynamic system is well described because the frequency and the magnitude of the dynamic response are well predicted. This data is used to update the dynamic system and to predict in case of store integration the dynamic loads for the specifications. Using the flight test results to update the dynamic model and transform the ground or axle loads from main landing gear wheel to the attachment points of the gear. With this input a dynamic response analysis will be performed. Figure 14 shows the result of a store integration analysis. The calculated loads (red area) are compared with the allowable load envelopes (Ale's green area). This example discusses the loads on the interface of a pylon with a heavy store mounted. As it can seen in Figure 14 the Mx versus My exceeds the allowable load envelope. in this particular case the sink rate for this external store configuration must be restricted or the store must be tuned to keep the loads inside the allowable loads envelope.