CN104699947A - Method for simulating aircraft rock movement with RANS/LES (Reynolds average numerical simulation)/LES (large eddy simulation) mixing technique - Google Patents

Abstract

The invention discloses a method for simulating aircraft rock movement with an RANS/LES (Reynolds average numerical simulation)/LES (large eddy simulation) mixing technique, and aims to solve the problem about self-excitation rock movement under the condition that wide-range separated flow of an aircraft cannot be accurately simulated in the prior art. The method mainly comprises steps as follows: a length scale of a turbulence model is appropriately corrected on the basis of a conventional unsteady RANS method, and DES (detached eddy simulation) methods are constructed; tightly coupled solution of a flow field governing equation and a rigid body motion equation is realized in the dual-time-step pseudo time iteration process with adoption of a prediction-correction algorithm, and the changing course of aerodynamic parameters and motion parameters of the aircraft with time is obtained. On the premise that certain computational efficiency is guaranteed, self-excitation rock motion of the aircraft under the complicated flowing condition is more accurately simulated, and an effective means is provided for nonlinear coupling property evaluation of aerodynamic motion/movement of the aircraft.

Description

A kind of method adopting RANS/LES hybrid technology simulated flight device rock and roll to move

Technical field

What the present invention relates to is flight vehicle aerodynamic analogy method, especially a kind of method adopting RANS/LES hybrid technology simulated flight device rock and roll to move.

Background technology

Rock and roll (Rock) be modern combat aircraft and missile armament often can run into pneumatic/sports coupling phenomenon, it is the vibratory movement of aircraft in rolling direction.In some cases with rolling vibration, yaw direction also can vibrate, and is called Dutch roll (Dutch-roll).Along with the angle of attack increases, Dutch roll develops into the rock and roll motion based on rolling gradually.Rock and roll motion normally occurs with limit cycles oscillations (Limit Cycle Oscillation, LCO) form.Limit cycles oscillations means that the energy variation of moving in one-period is zero, and the energy namely absorbed is equal with the energy of dissipation, and vibration is neither dispersed and also do not restrained, and forms constant amplitude constant periodic oscillation.In theory, under perturbation action, all can enter rock and roll from arbitrary initial state, finally be stabilized in this limit cycle state, but some limit cycles oscillationss are unstable, now may there is other limit cycle in phase plane, and system is meeting saltus step between limit cycle under perturbation action, enters chaotic motion.Rock and roll motion is breakneck state, if do not understand characteristic and the flow mechanism of occurred rock and roll motion, control system does not have suitable control method to go to revise, and just enters flight out of control so theoretically.

At present, for the predictions and simulations of aircraft rock and roll motion, carry out mainly through wind tunnel test and numerical evaluation two kinds of methods.

Wind tunnel test can the rounded flowing pattern of simulated flight device more truly, and the motion of simulation rock and roll has certain accuracy.But there is many-sided interference such as uneven incoming flow, wind tunnel wall interference, support interference, bearing friction in wind tunnel test, and the motion of the self-excitation rock and roll of aircraft is comparatively responsive to inlet flow conditions, makes wind tunnel simulation result and Live Flying condition there is certain difference.

Numerical evaluation mainly divides the spatial spreading grid of aircraft Flow Field outside, and be coupled flow field control equation and rigid block element, carry out unsteady flo w time stepping method and solve.There is not flow field interference in this method, close to Live Flying condition.

Unsteady flo w RANS method is generally adopted at present in the iterative computation of flow field.Its thought is based on Reynolds average, and information of flow (speed, pressure etc.) is resolved into two parts, and Part I is the permanent part irrelevant with the time, and another part is pulsatile portion.Therefore the eddy stress item produced, usually adopts based on Boussinesq eddy viscosity hypothesis, is converted into the Solve problems of eddy viscosity, thus solves by turbulence model.Turbulence model conventional in engineering has an equation S-A model, Shuan Fangcheng, model etc.

More accurate result can be obtained in the prediction that RANS method is flowing in attachment flowing or little separated flow in attachment, but when flowing exists large area separation, RANS method is owing to introducing larger dissipation, inhibit the parsing that microvortex moves, therefore along with the increase of the angle of attack, simulation precision sharply declines, and proposes certain challenge to the motion of accurate simulation rock and roll.

For the aircraft rock and roll motion under accurate simulation At High Angle of Attack, complex flowfield condition, need to simulate comparatively accurately the flow field vortex structure in each moment.And RANS/LES hybrid technology, combine the advantage of RANS method and LES method, be widely used in the calculating of high reynolds number, large separated flow.The basic thought of RANS/LES hybrid technology is that stream field adopts multidomain treat-ment, adopts RANS method that calculated amount can be born, adopt LES method to make flow field vortex structure meticulousr in separated flow near wall flowing.

Current RANS/LES hybrid technology is mainly used in model unsteady flow separation that is static or designated movement mode and flows in calculating, also do not find application in pneumatic/sports coupling calculates, in the former, the result of calculation of previous moment only affects the flow field evolution of later point, and also can impact the model sport of later point in the latter simultaneously.And RANS/LES mixed method will introduce more gas information, compared with traditional unsteady flo w RANS method, speed of convergence is slack-off, the error of calculation increases, and the error of calculation accumulates in time and is easy on the impact of the evolution of subsequent flows field structure and the attitude motion of aircraft the result that causes simulating mistake.Therefore, need on the one hand to adopt less time step to carry out time Exact Solution, and carry out the comparatively validate of different time step-length, the coupling needing the higher tight coupling method of employing time precision to carry out pneumatic/motion on the other hand calculates.

Summary of the invention

Object of the present invention, be exactly for the deficiency existing for prior art, and a kind of technical scheme of the method adopting RANS/LES hybrid technology simulated flight device rock and roll to move is provided, the program adopts the motion of RANS/LES hybrid technology simulated flight device rock and roll, under the prerequisite ensureing solution efficiency, resolve the vortex structure in complex separations flowing more subtly, thus predictions and simulations is more accurately made in the aircraft rock and roll motion leading to whirlpool motion.

This programme is achieved by the following technical measures: a kind of method adopting RANS/LES hybrid technology simulated flight device rock and roll to move, is characterized in that: comprise the following steps:

Step one, on traditional unsteady flo w RANS method basis, length dimension in turbulence model is replaced with the length dimension of DES class methods, make the turbulent flow near wall continuation RANS method simulation small scale in the solution process of flow field, and adopt similar Smagorinsky sub-grid scale model to simulate vortex structure in Disengagement zone, thus obtain the flow field calculation method of simulated flight device;

Step 2, set up the rolling single freedom rigid body equation of motion, and the flow field calculation method that integrating step one obtains, adopt combined interior homotopy, construct pneumatic/motion tight coupling method for solving;

The rolling single freedom rigid body equation of motion is as follows:

$\left\{\begin{array}{c}\frac{{\mathrm{d\ω}}_x}{\mathrm{dt}}=\frac{M_x}{I_{\mathrm{xx}}}\\ \mathrm{d\γ}/\mathrm{dt}={\mathrm{\ω}}_x\end{array}\right.$

In formula, ω _xfor model angular velocity in roll, M _xfor model is by rolling moment, I _xxfor the moment of inertia in model rolling direction, γ is model roll angle, and t is for there being the dimension time;

N-th actual time step calculate start time, before use, the aerodynamic force that calculates is predicted the angular velocity in roll of current time and roll angle the moment, and prediction algorithm is:

$\left\{\begin{array}{c}{\mathrm{\ω}}_x^n={\mathrm{\ω}}_x^{n-1}+\frac{\mathrm{dt}}{12}\·[23{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-1}-16{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-2}+5{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-3}\\ {\mathrm{\γ}}^n={\mathrm{\γ}}^{n-1}+\frac{\mathrm{dt}}{12}\·[{23\mathrm{\ω}}_x^{n-1}-{16\mathrm{\ω}}_x^{n-2}+{5\mathrm{\ω}}_x^{n-3}]\end{array}\right.$

In formula, subscript n represents that the n-th actual time walked;

Freezing the current true moment subsequently, carrying out pseudo-time iteration and solve in the process in flow field, use correcting algorithm to correct current kinetic parameters, correcting algorithm is:

$\left\{\begin{array}{c}{{\mathrm{\ω}}_x}^n={{\mathrm{\ω}}_x}^{n-1}+\frac{\mathrm{dt}}{12}\·[5{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n,\mathrm{new}}+8{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-1}-{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-2}\\ {\mathrm{\γ}}^n={\mathrm{\γ}}^{n-1}+\frac{\mathrm{dt}}{12}\·[{5\mathrm{\ω}}_x^{n,\mathrm{new}}+{8\mathrm{\ω}}_x^{n-1}-{\mathrm{\ω}}_x^{n-2}]\end{array}\right.$

In formula, subscript " new " represents the value obtained by the pseudo-time iteration of a nearest step;

Step 3, for given aircraft profile grid division;

Step 4, the computing method adopted in step 2, for the space lattice that step 3 obtains, carry out calculating, obtain rudders pneumatic power parameter and kinematic parameter course over time.

Preferred as this programme: the scale of the grid divided for given aircraft profile in step 3 is a bit larger tham RANS and is calculated required grid scale.

The beneficial effect of this programme can be learnt according to describing of such scheme, due in this scenario on traditional unsteady flo w RANS method basis, length dimension in turbulence model is replaced with the length dimension of DES class methods, make the turbulent flow near wall continuation RANS method simulation small scale in the solution process of flow field, the rock and roll motion of turbulence model can be simulated more exactly, can under the prerequisite ensureing solution efficiency, resolve the vortex structure in complex separations flowing more subtly, thus predictions and simulations is more accurately made in the aircraft rock and roll motion leading to whirlpool motion.

As can be seen here, the present invention compared with prior art, has substantive distinguishing features and progress, and its beneficial effect implemented also is apparent.

Accompanying drawing explanation

Fig. 1 is multi-wall interference mode shape schematic diagram.

Fig. 2 is that the space lattice of multi-wall interference model divides schematic diagram.

Fig. 3 is the model roll angle curve map over time adopting RANS method to obtain.

Fig. 4 is the model roll angle curve map over time adopting DDES method to obtain.

Fig. 5 is with the form of angular velocity in roll-roll angle phasor the wind-tunnel free rolling test provided and the Comparative result figure adopting DDES method to carry out free rolling numerical simulation.

Embodiment

All features disclosed in this instructions, or the step in disclosed all methods or process, except mutually exclusive feature and/or step, all can combine by any way.

Arbitrary feature disclosed in this instructions (comprising any accessory claim, summary and accompanying drawing), unless specifically stated otherwise, all can be replaced by other equivalences or the alternative features with similar object.That is, unless specifically stated otherwise, each feature is an example in a series of equivalence or similar characteristics.

The present invention includes following steps:

Step one, on traditional unsteady flo w RANS method basis, length dimension in turbulence model is replaced with the length dimension of DES class methods, make the turbulent flow near wall continuation RANS method simulation small scale in the solution process of flow field, and adopt similar Smagorinsky sub-grid scale model to simulate vortex structure in Disengagement zone, thus obtain the flow field calculation method of simulated flight device;

Step 2, set up the rolling single freedom rigid body equation of motion, and the flow field calculation method that integrating step one obtains, adopt combined interior homotopy, construct pneumatic/motion tight coupling method for solving;

The rolling single freedom rigid body equation of motion is as follows:

$\left\{\begin{array}{c}\frac{{\mathrm{d\ω}}_x}{\mathrm{dt}}=\frac{M_x}{I_{\mathrm{xx}}}\\ \mathrm{d\γ}/\mathrm{dt}={\mathrm{\ω}}_x\end{array}\right.$

In formula, ω _xfor model angular velocity in roll, M _xfor model is by rolling moment, I _xxfor the moment of inertia in model rolling direction, γ is model roll angle, and t is for there being the dimension time;

N-th actual time step calculate start time, before use, the aerodynamic force that calculates is predicted the angular velocity in roll of current time and roll angle the moment, and prediction algorithm is:

$\left\{\begin{array}{c}{\mathrm{\ω}}_x^n={\mathrm{\ω}}_x^{n-1}+\frac{\mathrm{dt}}{12}\·[23{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-1}-16{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-2}+5{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-3}\\ {\mathrm{\γ}}^n={\mathrm{\γ}}^{n-1}+\frac{\mathrm{dt}}{12}\·[{23\mathrm{\ω}}_x^{n-1}-{16\mathrm{\ω}}_x^{n-2}+{5\mathrm{\ω}}_x^{n-3}]\end{array}\right.$

In formula, subscript n represents that the n-th actual time walked;

Freezing the current true moment subsequently, carrying out pseudo-time iteration and solve in the process in flow field, use correcting algorithm to correct current kinetic parameters, correcting algorithm is:

$\left\{\begin{array}{c}{{\mathrm{\ω}}_x}^n={{\mathrm{\ω}}_x}^{n-1}+\frac{\mathrm{dt}}{12}\·[5{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n,\mathrm{new}}+8{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-1}-{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-2}\\ {\mathrm{\γ}}^n={\mathrm{\γ}}^{n-1}+\frac{\mathrm{dt}}{12}\·[{5\mathrm{\ω}}_x^{n,\mathrm{new}}+{8\mathrm{\ω}}_x^{n-1}-{\mathrm{\ω}}_x^{n-2}]\end{array}\right.$

In formula, subscript " new " represents the value obtained by the pseudo-time iteration of a nearest step;

Step 3, for given aircraft profile grid division;

Step 4, the computing method adopted in step 2, for the space lattice that step 3 obtains, carry out calculating, obtain rudders pneumatic power parameter and kinematic parameter course over time.

Be a bit larger tham RANS for the scale of the grid of given aircraft profile division in step 3 and calculate required grid scale.

Apply the present invention to the free rolling motion simulation of multi-wall interference under ° condition of M=0.6, α=35.

Calculate and adopt profile and moment of inertia with reference to certain multi-wall interference free rolling model in wind tunnel, model total length is 504mm, and length is 281.4mm, and fuselage diameter is 63mm, and the moment of inertia of winding mold type x-axis (fuselage axis) is 0.003675kgm2.

The present embodiment, based on Spalart-Allmaras Equation Turbulence Model, revises wherein length dimension d _wfor the length dimension of DDES , structure obtains DDES method.

This example based on multi-wall interference model, defined basis grid, grid cell total amount about 5,920,000.As shown in Figure 1 and Figure 2.

Design conditions are M=0.6, α=35 °, model reference length L _refget fuselage maximum gauge 63mm, area of reference S _refget fuselage maxi mum cross sectional area 3117.245mm2, based on the reynolds number Re of model reference length _l=1 × 10 ⁶.

In order to make model enter limit cycle rock and roll motion state sooner, in numerical simulation, initial for model roll angle is set as 5 °, and the degree of freedom in releasing theory rolling direction.

Select different time step-length, and adopt RANS method and DDES method respectively, in conjunction with rigid block element, carry out tight coupling and calculate, obtain model roll angle, angular velocity in roll, rolling moment coefficient equivalence course over time.

According to the numerical simulation result of different calculation methods, make model roll angle curve over time respectively, as shown in Figure 3, Figure 4.Adopt RANS method, model roll angle there is vibration and very rapid convergence to equilibrium position; Adopt DDES method, model enters the motion of limit cycle rock and roll fast.Concerning two kinds of computing method, different time step-length does not have essential distinction to analog result.

According to numerical simulation result and results of wind tunnel, make model angular velocity in roll-roll angle phase diagram, as shown in Figure 5.Adopt DDES method acquired results result corresponding to wind tunnel test to coincide better, because wind tunnel test exists more disturbing factor, therefore in result, the randomness of different cycles is larger, and the randomness of result of calculation is relatively little.

Wind tunnel test and different calculation methods as shown in table 1 to multi-wall interference rock and roll motion simulation result:

Table 1 wind tunnel test and numerical simulation are to multi-wall interference rock and roll motion simulation result

Analogy method	Amplitude (°)	Frequency (Hz)
			Wind-tunnel free rolling is tested	35.271	7.418
Numerical simulation (RANS)	Motion convergence	7.993
			Numerical simulation (DDES)	35.732	7.621

As shown in Table 1, the simulate effect adopting traditional RANS method to move to wing body rock and roll is poor, and adopt DDES method can obtain with wind tunnel test closer to result.

Present invention employs DES class methods and numerical simulation is carried out to the motion of aircraft rock and roll, fine resolution can be made to the complicated whirlpool motion of large area separated flow, thus simulation obtains the rock and roll more consistent with results of wind tunnel moves.

The present invention is not limited to aforesaid embodiment.The present invention expands to any new feature of disclosing in this manual or any combination newly, and the step of the arbitrary new method disclosed or process or any combination newly.

Claims (2)

1. the method adopting RANS/LES hybrid technology simulated flight device rock and roll to move, is characterized in that: comprise the following steps:

Step one, on traditional unsteady flo w RANS method basis, length dimension in turbulence model is replaced with the length dimension of DES class methods, make the turbulent flow near wall continuation RANS method simulation small scale in the solution process of flow field, and adopt similar Smagorinsky sub-grid scale model to simulate vortex structure in Disengagement zone, thus obtain the flow field calculation method of simulated flight device;

Step 2, set up the rolling single freedom rigid body equation of motion, and the flow field calculation method that integrating step one obtains, adopt combined interior homotopy, construct pneumatic/motion tight coupling method for solving;

The rolling single freedom rigid body equation of motion is as follows:

$\left\{\begin{array}{c}\frac{d{\mathrm{\ω}}_x}{\mathrm{dt}}=\frac{M_x}{I_{\mathrm{xx}}}\\ \mathrm{d\γ}/\mathrm{dt}={\mathrm{\ω}}_x\end{array}\right.$

In formula, ω _xfor model angular velocity in roll, M _xfor model is by rolling moment, I _xxfor the moment of inertia in model rolling direction, γ is model roll angle, and t is for there being the dimension time;

N-th actual time step calculate start time, before use, the aerodynamic force that calculates is predicted the angular velocity in roll of current time and roll angle the moment, and prediction algorithm is:

$\left\{\begin{array}{c}{\mathrm{\ω}}_x^n={\mathrm{\ω}}_x^{n-1}+\frac{\mathrm{dt}}{12}\·[23{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-1}-16{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-2}+5{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-3}]\\ {\mathrm{\γ}}^n={\mathrm{\γ}}^{n-1}+\frac{\mathrm{dt}}{12}\·[23{\mathrm{\ω}}_x^{n-1}-16{\mathrm{\ω}}_x^{n-2}+5{\mathrm{\ω}}_x^{n-3}]\end{array}\right.$

In formula, subscript n represents that the n-th actual time walked;

Freezing the current true moment subsequently, carrying out pseudo-time iteration and solve in the process in flow field, use correcting algorithm to correct current kinetic parameters, correcting algorithm is:

$\left\{\begin{array}{c}{{\mathrm{\ω}}_x}^n={{\mathrm{\ω}}_x}^{n-1}+\frac{\mathrm{dt}}{12}\·[5{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n,\mathrm{new}}+8{\left(\frac{M}{I_{\mathrm{xx}}}\right)}^{n-1}-{\left(\frac{M_x}{I_{\mathrm{xx}}}\right)}^{n-2}]\\ {\mathrm{\γ}}^n={\mathrm{\γ}}^{n-1}+\frac{\mathrm{dt}}{12}\·[5{\mathrm{\ω}}_x^{n,\mathrm{new}}+8{\mathrm{\ω}}_x^{n-1}-{\mathrm{\ω}}_x^{n-2}]\end{array}\right.$

In formula, subscript " new " represents the value obtained by the pseudo-time iteration of a nearest step;

Step 3, for given aircraft profile grid division;

Step 4, the computing method adopted in step 2, for the space lattice that step 3 obtains, carry out calculating, obtain rudders pneumatic power parameter and kinematic parameter course over time.

2. a kind of method adopting RANS/LES hybrid technology simulated flight device rock and roll to move according to claim 1, is characterized in that: be a bit larger tham RANS for the scale of the grid of given aircraft profile division in described step 3 and calculate required grid scale.