CJA 620 No.

of Pages 12
11 May 2016
Chinese Journal of Aeronautics, (2016), xxx(xx): xxx–xxx
1

Chinese Society of Aeronautics and Astronautics

& Beihang University
Chinese Journal of Aeronautics
cja@buaa.edu.cn
www.sciencedirect.com

2 FULL LENGTH ARTICLE

4 Parachute dynamics and perturbation analysis

5 of precision airdrop system
6 Gao Xinglong, Zhang Qingbin, Tang Qiangang *

7 College of Aerospace Science and Engineering, National University of Defense Technology, Changsha 410073, China

8 Received 15 June 2015; revised 10 October 2015; accepted 15 January 2016

9

11 KEYWORDS Abstract To analyze the parachute dynamics and stability characteristics of precision airdrop
12
13 Airdrop system; system, the fluid–structure interaction (FSI) dynamics coupling with the flight trajectory of a
14 Flight dynamics; parachute–payload system is comprehensively predicted by numerical methods. The inflation
15 Fluid–structure interaction; behavior of a disk-gap-band parachute is specifically investigated using the arbitrary Lagrangian–
16 Parachute; Euler (ALE) penalty coupling method. With the available aerodynamic data obtained from the FSI
17 Perturbation analysis simulation, a nine-degree-of-freedom (9DOF) dynamic model of a parachute–payload system is
built and solved to simulate the descent trajectory of the multi-body dynamic system. Finally, a lin-
ear five-degree-of-freedom (5DOF) dynamic model is developed, the perturbation characteristics
and the motion laws of the parachute and payload under a wind gust are analyzed by the lineariza-
tion method and verified by a comparison with flight test data. The results of airdrop test demon-
strate that our method can be further applied to the guidance and control of precision airdrop
systems.
18 Ó 2016 Production and hosting by Elsevier Ltd. on behalf of Chinese Society of Aeronautics and
Astronautics. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

19 1. Introduction for the aerial delivery and airdrop missions. Since the 1990s, 24
the U.S. Army has developed several precision airdrop systems 25

20 Parachutes are widely used in modern smart airdrop systems by implementing a guidance, navigation & control (GN&C) 26

21 to decelerate and stabilize the payload.1–4 For the past few dec- system and smart actuator in the parachute and parafoil.5–7 27

22 ades, applications of smart technology in decelerator systems Research is still under way on methods and materials used in 28

23 were still at the exploration stage and were mainly developed parachutes and airdrop systems to guide and control para- 29
chute flight in order to achieve optimum performance to meet 30
the mission requirements.8 Based on the specific requirements 31
* Corresponding author. Tel.: +86 731 84576436. of different missions, several types of parachute–payload sys- 32
E-mail addresses: 18674853560@163.com (X. Gao), qingbinzhang@ tems have been designed and tested,9 among which the rotat- 33
sina.com (Q. Zhang), kdtqg@sina.com (Q. Tang). ing parachute–payload system stands out as a common 34
Peer review under responsibility of Editorial Committee of CJA. configurations for smart submunitions that are required to 35
perform a target maneuver operation. However, our airdrop 36
test results show that the stability of the parachute airdrop sys- 37
Production and hosting by Elsevier tem often has difficulty in target identification. 38

http://dx.doi.org/10.1016/j.cja.2016.04.003
1000-9361 Ó 2016 Production and hosting by Elsevier Ltd. on behalf of Chinese Society of Aeronautics and Astronautics.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
2 X. Gao et al.

39 The stability of the parachute system has proven to be one in the investigation of parachute related recovery problems 101
40 of the most difficult aspects of modeling parachutes because of and in assessing the performance of parachute inflation.31–37 102
41 different stability modes. A two-dimensional (2D) parachute In addition, the simplified ALE FSI method is also used to 103
42 model has been developed to compute various characteristics simulate the inflation process of a folded parachute.38,39 Com- 104
43 pertaining to the steady descent of a parachute system by pared with the space–time FSI technique, a semi-implicit 105
44 investigating the effects of wind on parachute oscillation using method for pressure-linked equations (SIMPLE) algorithm 106
45 measured wind profiles. On the basis of a typical five- was proposed to analyze the FSI and flow field characteristics 107
46 rotational-degree-of-freedom model of the parachute system, of a parachute.39 108
47 the dynamic stability problem has been theoretically and This paper first presents the analysis of aerodynamic char- 109
48 experimentally investigated.10 The results revealed that during acteristics and an FSI coupling mechanism of a parachute– 110
49 static tests a parachute with less stability vibrated with high payload system during a precision airdrop operation. The 111
50 frequency and considerable amplitudes when kept at constant 3D dynamics behavior of parachute systems during inflation 112
51 angles of attack. Thus it became a significant issue to deter- and steady descent state is specifically analyzed using the 113
52 mine the influence of the parachute’s dynamic stability, like ALE penalty coupling method within LS-DYNA nonlinear 114
53 the canopy–payload coupling, with added fluid mass compo- dynamics code. Then, a nine-degree-of-freedom (9DOF) 115
54 nents and geometrical porosity, among others.11,12 The rela- dynamic model of the parachute–payload system is developed, 116
55 tionship of aerodynamic and inertial parameters with the which can be used for the prediction of the trajectory and the 117
56 lateral stability characteristics of a gliding parachute has been stability behavior of the parachute–payload system. Good 118
57 analyzed.13 The multi-body dynamics methodology has agreement between the simulation and the airdrop test data 119
58 remarkably promoted the development of trajectory planning provides the necessary verification and validation. Finally, 120
59 and stability modeling of parachute systems, although the on the assumption that the aerodynamic velocity is constant 121
60 accuracy of these problems still mainly depends on the pro- and perturbations are sufficiently minimal, a linear five- 122
61 found insights of the aerodynamic characteristics around the degree-of-freedom (5DOF) dynamic model is developed in 123
62 parachute and payload, in both the static and dynamic the steady state. The simulation program has been developed 124
63 states.14,15 and used to remove the influence of wind gusts, and the equa- 125
64 For mission design, however, good estimates of the aerody- tions of the steady states can be applied to analyzing the des- 126
65 namics of the parachute systems are not easy goals to achieve. cent and stability characteristics of a parachute airdrop 127
66 In the past, static and dynamic experimental measurements system. The comparison results proved the efficiency of our 128
67 were employed to help the designers obtain optimal solu- method in the guide design of precision airdrop systems. 129
68 tions.16–18 In the recent years, computational simulations of
69 parachute systems have gradually played a predominant role
70 in the prediction of dynamic behaviors, and various 2. Problem formulation 130

71 approaches and numerical methods have been developed to

72 model and perform the simulation of parachute aerodynamics The spatial motion of a precision airdrop system is chaotic and 131
73 and fluid–structure interaction (FSI) behavior. During both complicated. Upon payload ejection, the canopy will quickly 132
74 the inflation and steady descent stages, the parachute dynamics inflate into a hemisphere shape. Under the effect of the aerody- 133
75 are governed by a coupling between the structural dynamics of namic pressure on the surface of canopy, the system is deceler- 134
76 the parachute and the surrounding fluid flow. As such, the sys- ated and guided into a steady state with the payload spinning 135
77 tem must be treated as coupled to gain a proper representation at a constant rate for target identification. This paper mainly 136
78 of a holistic dynamic system. focuses on the forming phase from the opening of the canopy 137
79 Ongoing research has yielded software that improves the to the steady scanning of the payload. The parachute is a 138
80 accuracy of computational fluid dynamics (CFD) and compu- scaled disk-gap band (DGB) parachute (as shown in Fig. 1); 139
81 tational structure dynamics (CSD), and the aerodynamic char- the construct diameter of the parachute Dc = 7.5 m, the vent 140
82 acteristics as well as the response of the structure can be diameter Dv = 0.0738Dc, the width of gap Hg = 0.0424Dc, 141
83 comparatively studied, which is beneficial for the trajectory the width of band HB = 0.1209Dc, the width of band on gore 142
84 and stability computation of parachute coupling systems. On B = 0.02 m, the length of suspension lines L = 1.713Dc, and 143
85 the basis of the deforming spatial domain/stabilized space– the number of gores N = 24. The payload is constructed by 144
86 time (DSD/SST) technique,19,20 which was applied to three- a conical head, cylindrical body and six wrap-around fins. 145
87 dimensional (3D) computations soon after its develop-
88 ment,21,22 FSI modeling of several kinds of parachutes was
89 carried out, including ram-air parachutes,23 solid round
90 parachutes,24 and complex solid parachute designs.25 With
91 the new generation of DSD/SST formulations and space–time
92 FSI techniques,26 many additional 3D computations presented
93 by parachute FSI were addressed,27,28 including ringsail para-
94 chutes and reefed ringsail parachutes,29 and the evaluation of
95 the stability characteristics of a parachute based on
96 aerodynamic-moment calculations.30 The explicit finite ele-
97 ment method is also an efficient tool to replicate the FSI
98 dynamics of parachute systems. With the algorithmic enhance-
99 ments of the arbitrary Lagrangian–Euler (ALE) penalty cou-
100 pling method in LS-DYNA, considerable efforts were made Fig. 1 Schematic of disk-gap band parachute.

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
Parachute dynamics and perturbation analysis of precision airdrop system 3

146 The parachute–payload system is airdropped from an airship 3. Governing equations 182
147 at approximately 1 km altitude.
148 Before modeling the subsystems, the following simplifying 3.1. Multi-body dynamic model 183
149 assumptions are needed:
The mass of parachute and the added mass are combined into 184
150 (1) The initial state of the parachute is folded in the axial
a general mass matrix, mp , and a general moment matrix of 185
151 direction, the influence of fabric interaction is ignored,
parachute Ip ; the increments of aerodynamic force and 186
152 and the payload is a rigid body.
moments induced by the unsteady motion of parachute are 187
153 (2) The parachute may yaw, pitch, or roll relative to the
represented by the apparent mass, thus mp and Ip are 188
154 payload. 189
155 (3) The behavior of surrounding airflow is fundamentally mp ¼ diagðmc þ a11 ; mc þ a33 ; mc þ a33 Þ ð1Þ 191
156 time-dependent and unsteady. 192
157 (4) The trajectory of the parachute system is represented by Ip ¼ diagðIx þ a44 ; Iy þ a66 ; Iz þ a66 Þ ð2Þ 194
158 the movement of the joint connecting the parachute with
159 the payload. where mc represents the mass of parachute, (Ix , Iy , Iz ) the axial 195

160 (5) The ground is considered to be flat, non-rotating and moments of parachute on the joint O, and (a11 , a33 , a44 , a66 ) the 196

161 fixed in space by taking consideration of the earth- apparent mass of parachute. 197

162 fixed reference system as an inertial system. Using principle multi-rigid-body dynamics theory or 198

163 method of Dohher and Schilling,40 we can describe the para- 199

164 As shown in Fig. 2, the following coordinate systems are chute–payload system as 200
201
165 used: (A) earth-fixed reference Od xd yd zd with origin Od ; (B) d d
166 parachute-fixed reference Op xp yp zp with origin O; (C) mp ðVp þ Xp  Lp Þ þ mb ðVo þ Xp  Lp Þ
dt dt
167 payload-fixed reference Ob xb yb zb with origin O. The motion
¼ ðmp g þ mb gÞ þ Fp þ Fb ð3Þ 203
168 of the system is described by the system translational velocity
204
169 V0 , the angular velocity of parachute Xp and the payload Xb .
d d
170 In addition, use the pitching angle hp , yawing angle wp , rolling ðIb Xb Þ þ mb Lb  Vo ¼ Mb þ Lb  mb g ð4Þ 206
dt dt
171 angle cp to describe the transformation relation from 207
172 parachute-fixed reference Op xp yp zp to earth-fixed reference d d
ðIp Xp Þ þ mp Lp  Vo ¼ Mp þ Lp  mp g ð5Þ
173 Od xd yd zd and denote the transformation matrix as Bpe ; simi- dt dt 209

174 larly, use the Euler angle hb , wb and cb to describe the transfor- where Fp , Mp , Fb , Mb denote the aerodynamic forces and 210
175 mation relation from payload-fixed reference Ob xb yb zb to moments of parachute and payload, Vp denotes the velocity 211
176 earth-fixed reference Od xd yd zd and denote the transformation of parachute, mb and Ip denote the mass and moment matrix 212
177 matrix as Bbe ; the transformation relation from payload-fixed of body, Vo represents the velocity of joint. Further develop 213
178 reference Ob xb yb zb to parachute-fixed reference Op xp yp zp is the above equations as follows: 214
179 denoted as Bpb . Lp is the vector from joint to the mass center
 215
180 of canopy centroid and Lb the vector from joint to the mass _ b  Lb þ Xb  Xb  Lb Þ
mb ðV_ o þ Xp  Vo Þ þ Bpb ðX


181 of center of payload body. þ mp ðV_ o þ Xp  Vo Þ þ ðX _ p  Lp þ Xp  Xp  Lp Þ
¼ Fb þ Fp þ mb g þ mp g ð6Þ 217
218
_ b þ Xb  Ib Xb þ mb Lb  ðV_ 0 þ Xp  V0 Þ
Ib X
¼ Mb þ Lb  mb g ð7Þ 220
221
_ p þ Xp  Ip Xp þ mp Lp  ðV_ 0 þ Xp  V0 Þ
Ip X
¼ Mp þ Lp  mp g ð8Þ 223

An anti-symmetric matrix is introduced for simulation as 224

2 3 225
0 Lpy Lpz
6 7
~ p ¼ 6 Lpy
L 0 Lpx 7 ð9Þ
4 5
Lpz Lpx 0 227
228
2 3
0 Lby Lbz
~b ¼ 6
L 4 Lb
y
0
7
Lbx 5 ð10Þ
Lbz Lbx 0 230

Then the dynamic equations can be further developed. Next 231

introduce the generalized mass matrix of parachute system 232

Fig. 2 Schematic of parachute–payload airdrop systems. Amass and generalized force matrix Bforce . 233
234

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
4 X. Gao et al.
2 3 8 280
mb E33 þ mp ~b
Bpb mb L ~p
mp L > q0i ¼ cos w2i cos h2i cos c2i  sin w2i sin h2i sin c2i
>
>
6 ~p 7 >
>
Amass ¼4 mb L Ib 033 5 ð11Þ >
< q1i ¼  cos wi cos hi sin ci  sin wi sin hi cos ci
236 mp L~p 033 Ip
2 2 2 2 2 2
ð20Þ
>
> q2i ¼  cos w2i sin h2i sin c2i  sin w2i cos h2i cos c2i
>
>
237
2 3 >
>
ðmb E33 þ mp ÞðXp  Vo Þ : q ¼  cos wi sin hi cos ci  sin wi cos hi sin ci
282
6 7 3i 2 2 2 2 2 2
Bmass ¼ 4 Xb  ðLb Xb Þ þ mb Lb  ðXb  Bbp Vo Þ 5
Subscript i denotes ‘‘p” or ‘‘b” for the equation of para- 283
Xb  ðLb Xb Þ þ mp Lb  ðXp  Vo Þ chute and payload. Then the dynamic equation of parachute
2 3 284
mb Bpb ðXb  Xb  Lb Þ þ mp ðXp  Xp  Lp Þ system can be solved. 285
6 7
4 0 5
0 3.2. Fluid–structure interactions model 286
2 p 3 2 3
Bb Fb þ Fb mb g þ mp g
6 7 6 7 3.2.1. Structure dynamics 287
þ4 Mb 5 þ 4 Lb  mb g 5
Parachute components are mainly flexible and continuous 288
Mp Lp  mp g media. Let Xs be the spatial domain where superscript ‘‘s ” 289
239 ð12Þ implies the structure, and let @Xs denote the boundary of Xs . 290
The governing equation of the structure system is 291
240 where E is unit matrix and 0 zero matrix. Then, the vector 292
241 form of parachute–payload system dynamic equation is @u
qs ¼ rs ðuÞ þ qs  fs on Xs ð21Þ
242
2 3 @t 294
V_ 0
6_ 7 1 where qs is the material density, u the velocity vector of the 295
4 Xb 5 ¼ Amass Bforce ð13Þ
structure media, rs the Cauchy stress tensor, and fs the external 296
244 X_ p body forces acting on the structure. 297

245 Introducing the vector R from connection point to original Considering the large deformation and non-linear dynamic 298

246 point, the Euler angle’s differential equation of system is characteristics of the canopy, a special membrane element for- 299
247 mulation with a constitutive material model is better suited to 300
249 V0 ¼ Bpe R_ ð14Þ it. For thin fabrics, buckling (wrinkling) can occur with the 301
250 associated inability of the structure to support compressive 302
2 3 2 3
c_ p 1  tan hp cos cp tan hp sin cp stresses. The membrane is a 2D shell suited for a three- or 303
6 _ 7 6 0 cos c = cos h  sin cp = cos hp 7 four-node element, and the stress–strain relationship of the
4 wp 5 ¼ 4 p p 5Xp ð15Þ 304
membrane is given by 305
252 h_ p 0 sin cp cos cp 8 306
>
> e1 ¼ E11 ðr1  t1 r2 Þ
253
2 3 2 3 >
<
c_ b 1  tan hb cos cb tan hb sin cb
6_ 7 6 7 e2 ¼ E12 ðr2  t2 r1 Þ ð22Þ
4 wb 5 ¼ 4 0 cos cb = cos hb  sin cb = cos hb 5Xb ð16Þ >
>
>
: 2e ¼ 1 s þ as3
255 h_ b 0 sin c b cos cb 12 G12 12 12 308

256 When h ¼ 90 , the singularity will appear in Eqs. (15) and where r1 , t1 and E1 are the longitudinal stress, Poisson’s ratio, 309

257 (16). Then the Quaternion is introduced into the flight dynamic and elastic modulus, respectively; r2 , t2 and E2 are the traverse 310

258 model to prevent the singularity. The Quaternion is a kind of stress, Poisson’s ratio and elastic modulus, respectively. s12 is 311

259 hypercomplex number composed by four real argument the shear stress, G12 the shear elasticity and a a non-linear coef- 312

260 including one real unit and three imaginary units i, j and k, ficient that can be measured by a stress–strain relationship test. 313

261 which can be represented as Additionally, the suspension lines are made of rope and are 314
262 acted on by drag, gravitational forces, and aerodynamic forces 315
264 Q ¼ q0 þ q1 i þ q2 j þ q3 k ð17Þ during the inflation process. Therefore, the influence of damp- 316

265 It is apparent that if q1 ¼ q2 ¼ q3 ¼ 0, the Quaternion will ing and the non-linear characteristics of the rope should be 317

266 degenerate into real number. If q2 ¼ q3 ¼ 0, the Quaternion considered, and thus the dynamic governing equation can be 318

267 will degenerate into imaginary number. And the normative rewritten as 319

equation of Quaternion is  320

268 0 e0
269 F¼ ð23Þ
271 q20 þ q21 þ q22 þ q23 ¼ 1 ð18Þ pðeÞ þ C_e e > 0 322

272 The rotating motion of body represented by Quaternion is where pðeÞ represents the non-linear tensile function of the 323
273
2 3 2 3 ropes, and C is the damping coefficient. 324
q_ 0 q1 q2 q3 2 3
6 q_ 7 1 6 q q xx
6 17 6 0 3 q2 776 7 3.2.2. Fluid dynamics 325
6 7 ¼ 6 74 xy 5 ð19Þ
4 q_ 2 5 2 4 q3 q0 q1 5 The parachute opening velocity in an airdrop process is usually 326
xz
275 q_ 3 q2 q1 q0 relatively low, and the fluid field can be regarded as an incom- 327
pressible viscid flow. Let Xf and ð0; TÞ be the spatial and tem- 328
276 where xx , xy , xz represent the axial rotating velocity of body.
277 Then Eqs. (18) and (19) are a group of non-singular linear dif- poral domains, and @Xf denote the boundary. By introducing 329

278 ferential equation, and the relationship between the compo- the ALE formulation, the finite mesh can be freely moved. The 330

279 nents of Quaternion and Euler angles is fluid particle coordinates are Xj ðtÞðj ¼ x; y; zÞ, where t 2 ð0; TÞ, 331

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
Parachute dynamics and perturbation analysis of precision airdrop system 5

332 thus the Navier–Stokes governing equations for incompress- Considering the porosity of the canopy fabric, the pressure 373
333 ible flow in the reference coordinates are of porous media can be derived from the Ergun equation of the 374
334
@v 1 shell as41 375
376
þ ðv  wÞ  rv  r ¼ g ð24Þ
336 @t qf dP
¼ aðl; eÞ  vrel þ bðq; eÞ  v2rel ð27Þ 378
dr
337 where v and w are fluid particle velocity and material mesh
338 velocity, respectively, in reference coordinates and fluid where P is pressure, r the normal direction of the shell, and e 379

339 density. Eq. (20) is the Euler formulation, and if v = w, the the porosity of the material; the coefficient aðl; eÞ is the recip- 380

340 Lagrange formulation can be applied. Thus, the ALE formula- rocal permeability of the porous shell or viscous coefficient, 381

341 tion contains both the Euler formulation and Lagrange formu- and bðq; eÞ the inertia coefficient. 382

342 lation. Solid elements with the momentum advection By using the explicit dynamic integral method, the velocities 383

343 advantage are suitable for solving the Navier–Stokes equations and aerodynamic pressure can be effectively solved satisfying 384

344 for a fluid; the second-order van Leer monotonic upstream- the continuity constraint of Eq. (27); then, the external forces 385

345 centered scheme for conservation laws (MUSCL) scheme is act on the subsystems and dimensionless aerodynamic coeffi- 386

346 used to calculate the values of the solution variables in the cients can be measured to solve the multi-body dynamic 387

347 transport fluxes to attain accurate second-order monotonic equations. 388

348 results. This algorithm is accurate, stable, conservative and

349 monotonic. To improve the computational efficiency, the 3.3. Integration 389
350 single-point integral of the ALE multi-material method is cho-
351 sen, instead of the total volume integral. There is no easy way to directly couple the FSI and multi-body 390
dynamic models. The flowchart of this procedure is depicted in 391
352 3.2.3. Penalty coupling Fig. 3, and the procedure to simulate the inflation and steady 392

353 In FSI problems, the computation of the coupling interface is a descent phases is shown as follows. 393

354 key technique for the conversion of energy. Utilizing the pen- First, the initial velocity is loaded on the joint of the para- 394

355 alty coupling algorithm, coupling force can be applied in oppo- chute–payload system and the FSI computation begins. Unlike 395

356 site directions of the FSI interface. If d n represents the penalty the constant flow velocity of the system with fixed payload in 396

357 depth of structural nodes at time step t ¼ tn , it is incrementally infinite mass inflation, in this paper the payload is freely mov- 397

358 updated as ing in FSI simulation. 398

359
Second, at each explicit dynamic time step Dt, the aerody- 399
361 d nþ1 ¼ d n þ vnþ1=2
rel  Dt ð25Þ namic forces and moments of the system can be written into 400

362 where vrel is the reference velocity for the master and slave the transfer data files. With input data from the transferred 401

363 nodes, and the slave node velocity is vs . The master node veloc- data file to the 9DOF parameters, for the flight dynamic 402

364 ity can be viewed as a fluid particle within a flood element, with model, the trajectory simulation will then be triggered. 403

365 the mass and velocity interpolated from the fluid element Third, after the data files of aerodynamic forces and 404

366 nodes using finite element shape functions, thus moments are obtained, the multi-body dynamic equation starts 405
367 to compute and the velocity magnitude of the joint is then 406
nþ1=2
369 vrel ¼ vnþ1=2
s  vnþ1=2
f ð26Þ updated and reloaded on the parachute–payload system for 407
FSI simulation at the next time step. The iteration loop then 408
370 The penalty occurs only if ns  d < 0, where ns is build up
n
continues. During the inflation distance, the canopy is gradually 409
371 by averaging normals of structure elements connected to the
pressurized until it reaches a steady hemisphere profile when it is 410
372 structure node.
fully inflated, just at this moment tf denotes the inflation time. 411

Fig. 3 Flowchart of integration model.

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
6 X. Gao et al.

412 The steady descent state begins after tf, and the system and the rigid payload was simplified as a 3D rigid body with- 444
413 velocity converges to a steady value. If t equals to the terminal out considering the structural deformation. 445
414 time, which means the parachute system lands on the ground, A column of the fluid field was meshed by a 3D hexahedron 446
415 the computation of the multi-body dynamic and FSI models is element. Thus both the structure and fluid domain were 447
416 ended and the time history is saved. Thus the FSI and multi- meshed independently (as depicted in Fig. 5). The summarized 448
417 body dynamic models are loosely coupled and an integration finite element informant is presented in Table 1. 449
418 model is built.
419 For the FSI simulation, the input k-files are rewritten at 4.2. Material parameters 450
420 each time step and the computation can be continued with
421 the restart capability enabled by the LS-DYNA code. The Furthermore, the canopy is constructed using the fabric mate- 451
422 changing velocity is updated at each time step with the key- rial model that is commonly used for an airbag simulation 452
423 word *CHANGE_VELOCITY, which indicates a small deck model with LS-DYNA. The fabric material consists of ortho- 453
424 restart of the LS-DYNA code. Then, the integrated simulation tropic composites with permeable, large translation, and non- 454
425 program is developed, tested and used to investigate the FSI linear mechanical characteristics. Table 2 lists the material 455
426 and trajectory of the parachute–payload system. parameters of the MIL-C-7020 III nylon fabric and suspension 456
lines. 457
427 4. Numerical model
5. Numerical results and discussion 458
428 4.1. Finite element model
5.1. Parachute inflation dynamics 459
429 In comparison with the parachute aerodynamics model used in
430 solving the multi-body model as proposed in the literature, the In order to predict the motion of the parachute–payload sys- 460
431 improvement of the simulation model in this paper is the intro- tem, the aerodynamics of parachute and payload must be 461
432 duction of a flexible parachute and the influence of the FSI known. The finite element model of the parachute–payload 462
433 phenomenon, which can be achieved by employing the fabric system is implemented to evaluate the FSI simulation, which 463
434 and cable material models based on LS-DYNA nonlinear is performed by LS-DYNA, version R7.0.0. Fig. 6 first depicts 464
435 dynamics analysis code. To accomplish this objective, the finite the opening process of the parachute canopy; on the top are 465
436 element model of the parachute–payload system was built as the images from the airdrop tests in different inflation states 466
437 shown in Fig. 4. To practically reproduce the initial state of and the bottom is the simulation results of the canopy corre- 467
438 the parachute after stretching out from ejection is theoretically sponding to the tested states. It can be seen from the compar- 468
439 difficult; here, the folded canopy is simplified as a conical ison results that the numerical prediction has captured the 469
440 shape with an encompassed space to allow the inflow (as
441 shown in Fig. 4(a)). The fully inflated model of the parachute
442 is shown in Fig. 4(b). The parachute finite element model was
443 constructed by a tetrahedral shell and discrete beam elements,

Fig. 5 Mesh model of FSI (parachute with payload).

Table 1 Statistical information of parachute system simula-

tion model.
Component Element Material Part
Type Number
Canopy Tetrahedral shell 14,888 FABRIC Part1
Ropes Discrete beam 1872 CABLE Part2
Fluid Hexahedral solid 649,440 IDEAL_GAS Part3
Payload Hexahedral solid 64 RIGID Part4
Fig. 4 Mesh model of parachute-bomb system.

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
Parachute dynamics and perturbation analysis of precision airdrop system 7

Table 2 Parachute material properties.

Item Membrane Cable Band
Canopy Suspension line Seam & edge reinforcement Aperture reinforcement
Thickness (m) 0.0001 4  106 0.001 0.001
Density (kg/m3) 5880 5840 6800 6800
Young’ modulus (Pa) 4.309  108 1.2  1012 4.309  108 5.309  108
Poisson’s ration 0.3 0.14 0.14

Fig. 6 3D shape of canopy during parachute inflation process.

470 change of parachute during inflation well; in particular, the imum drag area, and the growing vortexes move upwards in 481
471 ‘‘lamp” shape of the canopy at initial inflating state is vividly the air flow direction, with the asymmetry of the conical vor- 482
472 simulated. Fig. 7 illustrates the results of the projected area texes appearing. When the canopy fully opens and reaches 483
473 of the canopy during inflation. It can be seen from the figures steady state, the conical vortexes move separately and break 484
474 that the parachute is fully inflated at t = 0.83 s and then into several small vortexes. In addition, the wake trailing the 485
475 reaches the steady state shortly with a steady drag area. opening canopy is moving close to the speed of the load. As 486
a result, when the load undergoes its maximum deceleration, 487
476 5.2. FSI dynamics of parachute–payload system the wake contacts the apex of the canopy. The recontacting 488
wake results in a negative differential pressure that indents 489

477 Fig. 8 shows the fluid distribution around the parachute, which the apex of the canopy. 490

478 indicates the unsteady characteristic of fluid during the infla- Figs. 9 and 10 show the stress and strain distribution on the 491

479 tion state. As the continuous airflow passes the parachute, canopy. It should be noted that the stress levels near the apex 492

480 the dimension of the canopy gradually expands up to the max- and upper area are remarkably higher than those in the skirt 493
region. However, as a whole, the structural integrity of canopy 494
keeps well, and shows no appearance of stress concentration. 495
Thus the simulation results of the employed FSI model can 496
provide more realistic coupling behavior of the parachute–pay- 497
load system. 498

5.3. Aerodynamics coefficients of parachute–payload system 499

When the parachute reaches the steady descent stage, the FSI 500
calculation continues and the aerodynamic characteristics of 501
the parachute and payload can be investigated. To further 502
comparatively acquire the aerodynamic coefficients of the 503
parachute system at different angles of attack, a group of 504
FSI simulations during steady descent state were performed 505
with different initial attitudes of the parachute–payload sys- 506
tem. The initial inflated model of the parachute system for 507
Fig. 7 Numerical results of projected area. steady simulation can be adopted from the terminal results 508

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
8 X. Gao et al.

Fig. 8 Fluid velocity distribution around parachute.

Fig. 9 Stress field of canopy.

Fig. 11 Aerodynamic coefficients of parachute.

Fig. 10 Strain field of canopy.

509 of the inflation process. Fig. 11 shows the aerodynamic coeffi-

510 cients of the steady drag force coefficients CpD and pitch Fig. 12 Aerodynamic coefficients of payload.
511 moment coefficients Cpm of the canopy at different spatial
512 angles of attack; the aerodynamic coefficients of the payload
513 are also depicted in Fig. 12, where Cbx and Cby are the drag coef- decreasing. It is obvious that the aerodynamic coefficients of 523

514 ficients of the payload, Cbm

is the pitch moment coefficient of the parachute are still symmetric with the opposite angle of 524

515 the payload. attack, which exhibits good stability of the parachute. For 525

516 At zero angle of attack, the drag coefficient CpD is a mini- the payload, the aerodynamic coefficients vary linearly during 526

517 mum, approximately 0.62, with no pitch moment. When the the rigid rotation process. 527

518 parachute oscillates, both CpD and Cpm first climb with the
5.4. Experimental method and trajectory analysis 528
519 increase of angle of attack owing to the asymmetric distribu-
520 tion of aerodynamic forces on the canopy; after reaching the
521 summit at the critical value of angle of attack, CpD and Cpm Under suitable weather with fewer crosswinds, the airdrop 529

522 begin dropping and the drag performance of the canopy starts experiments were launched eight times from the platform of 530

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
Parachute dynamics and perturbation analysis of precision airdrop system 9

531 a remote-controlled helium airship at an altitude of 1000 m ually to a steady value; this value is simply the steady descent 562
532 above the ground. The measurement system was the gyro on velocity of the parachute–payload system, which is a typical 563
533 the payload and the ground optical system. The CS41A-2 type design parameter for parachutes. 564
534 single-axis angular rate gyro was installed on the axis line of
535 the payload to record the rolling angle rate in real time. The 5.5. Perturbation dynamics analysis 565
536 ground optical measurement system was composed of three
537 computers and two sets of camera systems; the positions of The precision airdrop system requires constant angular rota- 566
538 the optical measurement systems are illustrated in Fig. 13. tion rate with very small pitch, yaw, and roll oscillations dur- 567
539 The parameters of trajectory and flight information were
540 obtained by imaging processing later, and the spatial 3D posi-
541 tions of parachute–payload system were recorded in real time.
542 Further, upon the release of main canopy, the payload
543 would rotate at a constant angular rate; this phenomenon
544 was also validated in flight tests. The trajectory of the precision
545 airdrop system mainly depends on the aerodynamic forces and
546 moments that both the parachute and payload experience dur-
547 ing the operation. From the results of aerodynamic coefficients
548 from the FSI simulation, the 9DOF model of the parachute–
549 payload system can be solved. Predicted altitude versus time
550 plots is shown in Fig. 14, along with the airdrop measure-
551 ments. A comparison of lateral displacement versus time for
552 the parachute–payload joint is shown in Fig. 15. It can be seen
553 from the deviation of lateral displacement that there are still
554 some crosswind influences. Good agreement between the
555 actual and simulated motion is observed. The vertical velocity
556 versus time plot for the parachute–payload system is shown in Fig. 15 Lateral displacement versus time for joint O of
557 Fig. 16. From the variation trend of the vertical dropping parachute–payload system.
558 velocity of the system we can know the variation of the cano-
559 py’s drag area; during the inflation process, the canopy quickly
560 inflates to decelerate the parachute–payload system. When the
561 canopy’s area attains steady state, the velocity also slows grad-

Fig. 13 Schematic of optical measuring system’s position. Fig. 16 Velocity versus time for joint O of parachute–payload
system.

Fig. 14 Altitude versus time for joint O of parachute–payload

system. Fig. 17 Pitch angle versus time for payload.

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
10 X. Gao et al.

ing the steady descent state; as a result, the influence of oscil- 2 32  3

568 mb þ mp mb Lb mp Lp v
76 7
y
lations on the variation of steady descent velocity and the roll 6 6
0 54 x 7
569
4 mb Lb Ib
570 rate is inevitable. However, for the purpose of linearization, b 5
571 this influence is ignored, considering the yaw and roll perturba- mp Lp 0 Ip xp 
572 tions are sufficiently small; we assume that the system descent 2 qS Cb ðv þx L ÞV þqCp V ðv þx L ÞV 3
 p D y b b 0 D 0 y b b 0
573 velocity and the roll rate remain constant, thus allowing the 6 2
7 ð29Þ
574 simplifications: ¼64
qS L V ðv þx L ÞCb
 b b 0 y2 b b m
7
5
575
577 sinhb  hb ; coshb  1 ð28Þ qV20 S0 R0 Cpm
2 3
578 Substituting these values into the 9DOF model of the para- ðmb þ mp ÞV0 xp  ðmb þ mp Þghp
6 7
579 chute–payload system arrives at the following form: þ4 mb Lb xb V0  mb Lb ghb 5
580
mp Lp xp V0  mp Lp ghp 582

Eq. (29) is the 5DOF model of the parachute–payload sys- 583

tem, where CbD and CpD are drag force coefficients of parachute 584
and bomb, xb and xp are rotating velocities of parachute and 585
bomb, vy is the velocity of joint and y direction, Sb is the drag 586
area of bomb, and the R0 is the radius of bomb’s section. This 587
equation contains five unknown quantities: hb , xb , hp , xp and 588
vy . 589
According to the above equation and our airdrop test 590
results, for a stable parachute system the pitch angle of the 591
payload can be written as 592
593
hðtÞ ¼ h0 expðasys tÞ ð30Þ 595

where h0 is determined by wind gust and asys determined by 596

parachute–payload system parameters. 597
As shown in Fig. 1, only after real-time acquisition of the 598

Fig. 18 Scanning distance versus time for payload. terrestrial scanning point is finished by the sensors on the pay- 599
load, can the position of the projective point be calculated. 600

Fig. 19 Euler angles of parachute during oscillation under earth-fixed reference Od xd yd zd .

Fig. 20 Euler angles of load during oscillation under earth-fixed reference Od xd yd zd .

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
Parachute dynamics and perturbation analysis of precision airdrop system 11

601 However, the airdrop tests show that the atmospheric turbu- simulation tool for the design of precise parachute–payload 659
602 lence at low altitude causes a regularly larger fluctuation of airdrop systems. 660
603 the scanning point of the payload under the disturbance of a
604 wind gust. The motion characteristics of the parachute–pay- Acknowledgements 661
605 load system affected by the wind gusts at 4 m/s and 6 m/s were
606 respectively simulated in this study. Fig. 17 illustrates the pitch This study was co-supported by Research Project of Chinese 662
607 angle versus time for the payload, and Fig. 18 illustrates the National University of Defense Technology (No.: JC13-01- 663
608 scanning distance versus time for the payload. Both two figures 04) and the National Natural Science Foundation of China 664
609 show that after the disturbance of a wind gust, the payload (Nos.: 51375486 and 11272345). And also thanks for the found 665
610 first oscillates significantly under the effect of the lateral aero- support from China Scholarship Council (CSC). 666
611 dynamic forces; then, affected by the drag forces of the para-
612 chute, the payload gradually becomes stable in a fixed time. References 667
613 The flight tests also validated this phenomenon.
614 The Euler angles of the parachute and payload affected by 1. White FM, Wolf DF. A theory of three-dimensional parachute 668
615 the wind gust at 6 m/s are illustrated in Figs. 19 and 20. Under dynamic stability. J Aircraft 1968;5(1):86–92. 669
616 the small disturbance of wind we know that the attitude of the 2. Wolf D. Dynamic stability of a nonrigid parachute and payload 670
617 parachute shows good stability; the wind causes a larger oscil- system. J Aircraft 1971;8(8):603–9. 671
618 lation of the roll motion than those of the other two directions. 3. Edward F. Parachute dynamics and stability analysis of the Queen 672

619 For the attitude of the load, the payload swings fiercely at the Match Recovery System11th Aerodynamic decelerator systems 673

620 beginning under the act of wind force, but combined with the technology conference. Reston: AIAA; 1991. 674
4. Anthony C. An analysis of aerodynamic control for direct fire 675
621 aerodynamic drag forces from the parachute, the payload
spinning projectilesAIAA guidance, navigation, and control confer- 676
622 tends to keep steady and decays in the rule of index, especially ence and exhibit. Reston: AIAA; 2001. 677
623 for the roll and pitch angles. The payload shows periodicity in 5. Brown G, Haggard R, Almassy R, Benney R, Dellicker S. The 678
624 the yaw motion. affordable guided airdrop system (AGAS)15th CEAS/AIAA 679
aerodynamic decelerator systems technology confer- 680
625 6. Conclusions ence. Reston: AIAA; 1999. p. 8–11. 681
6. Benney R, Barber J, McGrath J, McHugh J, Noetscher G, Tavan 682
S. The joint precision airdrop system advanced concept technology 683
626 There exists a significant potential for improving the precision demonstration18th AIAA aerodynamic decelerator systems tech- 684
627 accuracy of a parachute–payload system through the imple- nology conference and seminar. Reston: AIAA; 2005. p. 23–30. 685
628 mentation of dynamics analysis. Combined with the ALE pen- 7. Wegereef J, Jentink H. Parafoil characterisation tests with 686
629 alty method and multi-body dynamics, a monolithic coupling SPADES8th AIAA aerodynamic decelerator systems technology 687
630 method is developed to numerically study the dynamic behav- conference and seminar. Reston: AIAA; 2005. p. 22–8. 688
631 ior of the parachute–payload system. 8. Hattis PD, Appleby BD, Fill TJ, Benney R. Precision guided 689
airdrop system flight test results14th AIAA aerodynamic deceler- 690

632 (1) The computational model employs FSI simulation of ator systems technology conference. Reston: AIAA; 1997. p. 3–5. 691

633 incompressible flows coupled with a thin shell structure 9. Maydew RC, Peterson CW. Design and testing of high-performance 692
parachutes. Paris: AGARD; 1991, Report No.: AGARD-AG-319. 693
634 finite element model. The FSI simulations can reproduce
10. Doherr K. Theoretical and experimental investigation of para- 694
635 both the 3D shape of canopy inflation and the unsteady chute-load-system dynamic stabilityProceedings of 5th AIAA 695
636 flow characteristics around the canopy and structural aerodynamic decelerator and balloon technology confer- 696
637 properties during small oscillation. ence. Reston: AIAA; 1975. 697
638 (2) Selecting the aerodynamic coefficients of parachute and 11. Eaton JA. Added mass and the dynamic stability of parachutes. J 698
639 payload as inputs, the 9DOF dynamic model of the pre- Aircraft 1982;19(5):414–6. 699
640 cision airdrop system was solved and compared favor- 12. Cockrell DJ, Haidar NIA. Influence of the canopy–payload 700
641 ably with the flight measured values. coupling on the dynamic stability inpitch of a parachute sys- 701

642 (3) A perturbation analysis of the precision airdrop system temAerospace design conference. Reston: AIAA; 1993. 702

643 was conducted to eliminate the influence the wind gusts; 13. Crimi P. Lateral stability of gliding parachutes. J Guidance Control 703
Dyn 1990;13(6):1060–3. 704
644 the data show that pitch oscillations of the parachute–
14. Guglieri G, Quagliotti F. Validation of a simulation model for a 705
645 payload system could damp out under the effect of aero- planetary entry capsule. J Aircraft 2003;40(1):127–36. 706
646 dynamic drag from the parachute and a final equilib- 15. Avanzini G, Guglieri G, Torasso A. Multibody analysis of 707
647 rium of the subsystems can be achieved. terminal phase for a reentry vehicle: a comparative study. J 708
648 (4) Good agreements prove the accuracy of this prediction Aircraft 2012;49(6):1940–52. 709
649 method. The successful airdrop test demonstrates that 16. Shpund Z, Levin D. Static and dynamic coefficients of a cross-type 710
650 our coupling method can be potentially applied in the parachute. J Aircraft 1994;31(1):132–7. 711
651 guidance of precision airdrop systems. 17. Guglieri G, Quagliotti F. Low speed dynamic tests on a capsule 712

652 configuration. Aerosp Sci Technol 2000;4(6):383–90. 713

653 In summary, good prediction of parachute dynamics and 18. Levin D, Shpund Z. Dynamic investigation of the angular motion 714
of a rotating body-parachute system. J Aircraft 1995;32(1):93–9. 715
654 stability of parachute–payload system can significantly
19. Tezduyar TE, Behr M, Liou J. A new strategy for finite element 716
655 improve the maneuvering ability of an airdrop system. The computations involving moving boundaries and interfaces-The 717
656 monolithic method proposed in this paper for the analysis of deforming-spatial-domain/space-time procedure: I. The concept 718
657 FSI and perturbation behavior of the parachute–payload sys- and the preliminary numerical tests. Comput Methods Appl Mech 719
658 tem is a complete theoretical method that can be a high-fidelity Eng 1992;94(3):339–51. 720

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003
 CJA 620 No. of Pages 12
11 May 2016
12 X. Gao et al.

721 20. Tezduyar TE, Behr M, Mittal S, Liou J. A new strategy for finite 33. Coquet Y, Bordenave P, Capmas G, Espinosa C. Improvements in 767
722 element computations involving moving boundaries and inter- fluid structure interaction simulations of parachutes using LS- 768
723 faces-The deforming-spatial-domain/space-time procedure: II. DynaÒ21st AIAA aerodynamic decelerator systems technology 769
724 Computation of free-surface flows, two-liquid flows, and flows conference and seminar. Reston: AIAA; 2011. 770
725 with drifting cylinders. Comput Methods Appl Mech Eng 1992;94 34. Gao XL, Zhang QB, Tang QG. Transient dynamic modeling and 771
726 (3):353–71. analysis of complex parachute inflation with fixed payload. J 772
727 21. Tezduyar TE, Aliabadi S, Behr M, Johnson A, Mittal S. Parallel Aerosp Eng 2015;28(4):1–14. 773
728 finite element computation of 3D flows. Computer 1993;26 35. Gao XL, Zhang QB, Tang QG, Yang T. Fluid-structure interac- 774
729 (10):27–36. tion simulation of parachute in low speed airdropWorld congress 775
730 22. Tezduyar TE, Aliabadi SK, Behr M, Mittal S. Massively parallel on engineering 2013. London: IAENG; 2013. 776
731 finite element simulation of compressible and incompressible 36. Gao XL, Zhang QB, Tang QG, Li JH. Numerical study on fluid- 777
732 flows. Comput Methods Appl Mech Eng 1994;119(S1-S2):157–77. structure interaction of slot-parachute’s inflation process. Acta 778
733 23. Kalro V, Tezduyar TE. A parallel 3D computational method for Aeronaut Astronaut Sin 2013;34(10):2265–76 [Chinese]. 779
734 fluid-structure interactions in parachute systems. Comput Methods 37. Gao XL, Zhang QB, Tang QG. Reliability assessment of slot- 780
735 Appl Mech Eng 2000;190(S3-S4):321–32. parachute inflation based on Bayes theory. J Stat Comput Sim 781
736 24. Stein K, Benney R, Kalro V, Tezduyar TE, Leonard J, Accorsi M. 2014;84(6):1159–72. 782
737 Parachute fluid-structure interactions: 3-D computation. Comput 38. Yu L, Cheng H, Zhan Y, Li S. Study of parachute inflation 783
738 Methods Appl Mech Eng 2000;190(s 3-4):373–86. process using fluid–structure interaction method. Chin J Aeronaut 784
739 25. Sathe S, Benney R, Charles R, Doucette E, Miletti J, Senga M, 2014;27(2):272–9. 785
740 et al. Fluid–structure interaction modeling of complex parachute 39. Yu L, Shi XL, Ming X. Numerical simulation of parachute during 786
741 designs with the space-time finite element techniques. Comput opening process. Acta Aeronaut Astronaut Sin 2007;28(1):52–7 787
742 Fluids 2007;36(1):127–35. [Chinese]. 788
743 26. Tezduyar TE, Sathe S. Modeling of fluid–structure interactions 40. Doherr KF, Schilling H. Nine-degree-of-freedom simulation of 789
744 with the space-time finite elements: solution techniques. Int J rotating parachute systems. J Aircraft 1992;29(5):774–81. 790
745 Numer Methods Fluids 2007;54(1):855–900. 41. Ergun S. Fluid flow through packed beds. Chem Eng Prog 1952;48 791
746 27. Tezduyar TE, Sathe S, Pausewang J, Schwaab M, Christopher J, (2):89–94. 792
747 Crabtree J. Interface projection techniques for fluid–structure 793
748 interaction modeling with moving-mesh methods. Comput Mech Gao Xinglong is a Ph.D. candidate at College of Aerospace Science and 794
749 2008;43(1):39–49. Engineering, National University of Defense Technology. He received 795
750 28. Tezduyar TE, Sathe S, Schwaab M, Pausewang J, Christopher J, his B.S. degree from Dalian University of Technology, and M.S. 796
751 Crabtree J. Fluid–structure interaction modeling of ringsail degrees in Aerospace Science and Engineering from National 797
752 parachutes. Comput Mech 2008;43(1):133–42. University of Defense Technology in 2009 and 2012 respectively. His 798
753 29. Tezduyar TE, Takizawa K, Moorman C, Wright S, Christopher J. area of research includes parachute recovery system and multi-body 799
754 Space-time finite element computation of complex fluid–structure dynamics. 800
755 interactions. Int J Numer Methods Fluids 2010;64(64):1201–18. 801
756 30. Takizawa K, Tezduyar TE, Boswell C, Tsutsui Y, Montel K. Zhang Qingbin received his B.S. degree and M.S. degrees from 802
757 Special methods for aerodynamic-moment calculations from National University of Defense Technology, and then became a tea- 803
758 parachute FSI modeling. Comput Mech 2015;55(6):1059–69. cher there. His main research interests are multi-body dynamics of 804
759 31. Ben T, Taylor AP. The use of LS-DYNA to simulate the inflation aircraft. 805
806
760 of a parachute canopy18th AIAA aerodynamic decelerator systems
761 technology conference and seminar. Reston: AIAA; 2005. Tang Qiangang is a professor and Ph.D. supervisor at College of 807
762 32. Ben T, Taylor AP, Berland J, Gargano B. The use of LS-DYNA to Aerospace Science and Engineering, National University of Defense 808
763 assess the performance of airborne systems North America Technology. He received the Ph.D. degree from the same university. 809
764 Candidate ATPS main parachutes18th AIAA aerodynamic decel- His current research interests are hypersonic aircraft and multi-body 810
765 erator systems technology conference and seminar. Reston: AIAA; dynamics. 811
812
766 2005.

Please cite this article in press as: Gao X et al. Parachute dynamics and perturbation analysis of precision airdrop system, Chin J Aeronaut (2016), http://dx.doi.org/
10.1016/j.cja.2016.04.003