Professional Documents
Culture Documents
基于超声导波的航空航天结构损伤诊断成像技术研究进展 郑跃滨
基于超声导波的航空航天结构损伤诊断成像技术研究进展 郑跃滨
基于超声导波的航空航天结构
*
损伤诊断成像技术研究进展
郑跃滨 1,武湛君 1,雷振坤 1,高东岳 2,周 凯 1,张佳奇 1,杨正岩 1,邹建超 1
(1. 大连理工大学工业装备结构分析国家重点实验室,大连 116024;
2. 中山大学航空航天学院,广州 510275)
[ 摘要 ] 超声导波损伤诊断成像技术是近年来无损检测 / 结构健康监测领域的研究热点,它能够直观反映航空航
天结构上的损伤位置及严重程度。概述了超声导波诊断成像技术的一般流程,介绍了 8 种具有代表性成像算法的
研究现状,包括相控阵成像、空间 – 波数滤波器成像、逆时偏移成像、时间反转成像、延迟叠加成像、基于模型成像、
层析成像和概率成像,对比了这些成像算法的应用范围和优缺点,最后结合实际工业应用需求讨论了超声导波诊断
成像技术面临的挑战和发展方向。
关键词:超声导波;损伤诊断成像;健康监测;无损检测;航空航天结构;复合材料结构
DOI:10.16080/j.issn1671-833x.2020.18.024
现不易察觉的损伤。同时,超出最初 超声导波技术通过捕捉损伤引
设计寿命的飞行器面临着老化问题, 起的导波信号变化对结构的损伤
其结构可靠性有所下降 [1]。当前应 状况进行评估,适用于不同类型损
用新型材料(如高性能轻质合金 [2] 和 伤,包括金属结构的断裂、变形、腐
复合材料 [3])的新一代航空航天飞 蚀,复合材料结构的层间分层和基
行器处于高速发展中,如何评估材料 体开裂,热防护材料的脱粘,接头结
内部损伤仍存在挑战 [4]。因此,发展 构的连接失效等 [5,11]。采用成像技
应用于航空航天结构的损伤诊断技 术,综合考虑损伤对结构全局的影
术一直备受关注。凭借对损伤敏感, 响,能 够 直 观 地 定 位 损 伤 及 反 映
传播范围远的优点,超声导波技术成 损 伤 严 重 程 度 [12]。近年来,超声导
为无损检测(Non-destructive testing, 波损伤诊断成像技术已从实验室探
NDT)和结构健康监测(Structural 索逐步迈向大型真实结构(如全尺
郑跃滨
health monitoring,SHM)领域的一 寸 复 合 材 料 尾 翼 [13–14] 和 复 合 材 料
博士研究生,研究方向为超声导 个研究热点 [5–6]。该技术以导波作为 机身壁板 [15])应用,并取得了良好
波、结构健康监测、传 感 器 优 化 与 设 损伤信息的传递媒介,先用激励传感 效果。根据不同成像原理及计算过
计。
器在结构上激发导波,再用接收传感 程,有多种导波诊断成像技术,如相
器采集与损伤相互作用后的导波响 控阵(Phased array)成像方法、空间
航空航天飞行器在实际服役过 应信号,随后分析信号获取金属或复 – 波数滤波器(Spatial–wavenumber
程中要长期承受复杂的载荷条件,甚 合材料结构上的损伤情况,从而保证 filter)成 像 方 法、逆 时 偏 移 成 像
至可能遭受意外的冲击载荷,容易出 结构的可靠性和完整性 [7–10]。 (Reversal-time migration)方法、时间
反转(Time reversal)成像方法、延迟
* 基金项目:国家重点研发计划(2018YFA0702800);国家自然科学基金(11772081、51805068)。 叠加(Delay–and–sum)成像方法、基
24 航空制造技术·2020年第63卷第18期
∂2u
µ Δ 2
u + (λ + µ )
⋅u = ρ 2 ΔΔ
∂ 2 u ∂t
µ 2u + (λ + µ )
Δ =ρ 2
⋅ u封面文章 ΔΔ
COVER STORY ∂t 2
∂ u
µ 2u + (λ + µ )
Δ ⋅u = ρ 2 ΔΔ
1
∂t– ±兰姆
于模型成像(Model–based imaging) (1)首先将传感器预先布置于 自由边界条件,可得到瑞利
tan ( k s h ) 4k kl k s±1
2
方 法、层 析 成 像(Tomography)方 结构选定位置,组成传感器阵列或网 (Rayleigh-Lamb) = −频散方程 2
法以及基于概率的损伤图像重建 络;(2)经激励器激发信号,导波在 tantan ( ks(hk)l h ) 4k 2kk2l k−s ks2 ±1 ( )
=−
算法(Reconstruction algorithm for
结构中传播,遇到损伤等结构空间不 tantan( k(l hks)h ) =−k2 −4kk2s2kl2ks (2) ( )
tan ( kl h ) k 2 − k 2 2
probabilistic inspection of defects)。
连续处发生散射(包括反射、折射、衍
ω
2
µ 2
u + ( λ
s
+
ω µ 2
) ⋅ u = ρ (
Δ∂2u ) ΔΔ
针对超声导波诊断成像技术,本 射、模态转换等现象),经传感器采集 kl2 = − k 2 , k s2 = − k 2 ∂t 2
ω 2 c 2
l ω 2 c
2
s
文在分析和总结近 20 年国内外研究 得到; (3)利用相应信号处理技术, k 2 =∂ 2 u ω − k 2 , k 2 = ω − k 2(3)
µ 2u + (λ + µ ) ⋅ u = lρkl2 =2cl − k 2 , s k s2 =cs − k 2
Δ ΔΔ
成果的基础上,梳理出其中应用广泛 提取散射波中如能量、波速、相位、频 ∂t cl cs ±1
(4)将所提取信
的成像技术并做简要综述。 谱等信号变化特征; λ + 2µ tan ( ks hµ) 4k k k
2
i =1
p
− k−2
i =1 i < j
p p −
i =1 i < j j < k
( p p )p − p∑ i
S n =0
∑ Sn =0
i =1N −1 N
• 去噪
z (t ) = ∑ wn f n (t − ∆ Ptn=)∏ pi
• Hilbert
2 2 • FFT nN
=0−1 i =1
肋条
ω ω
kl2 = − k 2 , k s2 = − k 2 N
z (t ) =
• SAFT、DWT、CWT ∑
wn f n (t − ∆tn )
N −1 n = 0
缺陷 cl cs P = ∏ pi z (t ) = ∑ wnAf n (t − ∆tnP) = 1 N
板
i =1
n =0
f n (t ) = e
j (ω t − kn ⋅rn )
N
∑ pi
i =1
rn
λ + 2µ µ 1 N A j (ωt − kn ⋅rn )
f n (t ) = e
cl =
ρ
, cs =
ρ
P= ∑ pi
A r ∑ Sn =0
N
诊断图像 数据融合
i =1
算法生成 f n (t ) = enj (ωt − kn ⋅rn )
rn r − rn
∂ 2u ( n) ∑ Sn =0 N −1
µ (n) Δ 2 ( n)
u (
+ λ ( n) +µ ( n)
) ( Δ Δ ⋅u ( n)
∂t 2
) =ρ ( n)
r−−rzr(nt ) = ∑ wn f n (t − ∆tn )
(n =1, 2, , N )
r
∆tn = n
c n =0
N −1 r − rn
( )
A ω , k , λ ( n ) , µ ( n ) , h( n ) z (t=) 0= ∑ wn f n (t − ∆tn )
r rS−
∆tn = n
rn
n =0
c
r −rn f n (t ) =
A j (ωt − kn⋅rn )
∆t =
n r ≈ rn
e
N N −1 j N − 2 N −1 N N c rn
P = ∑图1pi −主动式超声导波诊断成像技术流程图
∑ ∑ pi pimaging
j − ∑ ∑ ∑ pi p j pk − ∏ pi r S
n
Fig.1 Flowchart of activei =damage
1 idiagnostic
=1 i < j i < j j < k based onAultrasonic
i =1 approach i =1 j (ωt − kguided
n ⋅rn ) waves φ = rr
f n (t ) = e r Sn
r2020年第63卷第18期·航空制造技术
n r ≈ rn 25
∆t = φ ⋅ Sn r − rn
N
r ≈ rnn c
P = ∏ pi φ= r r
i =1
r − rn DAi + DiS r − rn
λ + 2µ µ
cl = , cs =
ρ ρ
封面文章 COVER STORY
2 (n)
u
) ∂层各向异性
2 ( n)
u (
+ λ ( n ) +带入方程 ) (
µ ( n) (5)
Δ Δ
) ⋅u中,
( n ) 得到= ρ ( nN
∂t 2
(n =1, 2, 器布置方式可扩展至整个结构形成
, N) 率 – 波 数 联 合 分 析。 近 年 来,稀 疏
的复合结构板的频散方程为 网络。网络中两个相邻传感器的间 表示理论中的匹配追踪(Matching
距较大,通常为分米级,称为稀疏的 pursuits,MP)算 法 已 被 用 来 对 导
( )
A ω , k , λ ( n ) , µ ( n ) , h( n ) = 0 (6)
分布式传感器网络。 波 信 号 进 行 去 噪 和 特 征 提 取 [40]。
通常采用转换矩阵法、全局矩阵 3 信号处理与特征提取 Raghavan 等 [41] 改进 MP 算法,弥补
N 法、 N −1 j
有限元及半解析有限元求解方程 N − 2 N −1 N N 由 于 导 波 的 多 模 态 和 频 散 特 频散效应,分离多模态反射波,从而
P = ∑ pi − ∑ ∑ pi p j − ∑ ∑ ∑ pi p j pk − ∏ pi [31]
i =1
(6), 此外需特别注意复合材料结构
i =1 i < j i =1 i < j j < k i =1
性 ,加 上 噪 声、环 境 变 化 以 及 边 简化了导波信号的特征提取难度。
[11]
中导波的衰减问题 。 界反射等多种因素,造成传感器所 此外,多重信号分类(Multiple signal
2 传感器布置 接收信号波形复杂。信号处理技术 classification,MUSIC)方法可以有
N
传感器是P NDT/SHM = ∏ pi 中数据采 对诸如去噪、数据压缩、特征提取等 效去噪并提取特征信号 [42],该方法
集 系 统 的 关 键 组i =件。 1 有 多 种 类 型 数据预处理过程至关重要 [32]。常用 基于特征值分解技术,其中大特征值
的 传 感 器 可 实 现 超 声 导 波 的 激 励 信号处理方法包括可用于提取包络的 指向声源信号,小特征值对应噪声
或 接 收,包 括:锆 1钛 N 酸铅压电陶瓷 希尔伯特变换(Hilbert Transform)[33]、 可 源。He 等 [43] 将 MUSIC 与时间反转
P = ∑ pi [19]
(Piezoelectric,PZT) N i =1传感器 、光纤 用 于 去 噪 的 快 速 傅 里 叶 变 换(Fast 结合,实现了高定位精度的多损伤
光栅(Fiber grating)传感器 [20]、聚偏二 fourier transform)[34]、可 用 于 时 频 成像。
氟乙烯(Polyvinylidene difluoride)传 分析的短时傅里叶变换(Short-time 4 损伤成像算法
[21] ∑ Sn =0
感器 、磁致伸缩(Magnetostrictive) Fourier transform,STFT)和小波变 常用损伤成像算法主要有 8 种,
[22] [35]
传感器 、空气耦合(Air-couple)传 换(Wavelet transform) 。 其 中, 包括相控阵、空间 – 波数滤波器、时
N −1
感器 [23]、脉冲激光仪(Pulsed-laser) [24]
小波变换克服了 SFTF 窗口大小不 间反转、延迟叠加成像、逆时偏移成
z (t ) = ∑ wn f n (t − ∆tn )
和激光多普勒测振仪 n =0 (Scanning laser 随频率变化的缺点,可实现多分辨 像、基于模型成像、层析成像以及基
doppler vibrometer,SLDV)[25]。前 率处理,是信号时频分析的重要工 于概率成像,根据传感器布置密度可
3 种类型可用于在线监测,其中 PZT 具。 小 波 变 换 分 为 离 散 小 波 变 换 分为密集阵列成像与稀疏阵列成像,
由于成本低、 功耗小, A j且兼具激励和 (ωt − kn ⋅rn ) (Discrete wavelet transform,DWT) 根据算法求解过程可分为反演求解
f n (t ) = e [19]
接 收 信 号 功 能,因 rn 此 应 用 广 泛 。 和连续小波变换(Continuous wavelet 与正向求解。
PZT 可通过粘贴方式安装在结构表 transform,CWT),Yu 等 [36] 采 用 4.1 密集阵列成像与稀疏阵列成像
面,也可以嵌入至复合材料结构的内 DWT 对 信 号 去 噪,Qiu 等 [37] 使 用 采用密集传感器阵列的成像算
[26–27] [35,38]
部 。后 4 种类型属于离线测量, r − rn CWT 提取单频分量。Pai 等 对比 法能实现导波回波信号的聚焦,提高
其中 SLDV 的时空分辨率高,但操作 并讨论了 STFT、希尔伯特 – 黄变换 信噪比,比如相控阵 [44]、空间 – 波数
复杂,能量转化率低 [6]。 (Hilbert-Huang transform)和共轭对 滤波器(相控阵的变种形式)[45]、时
r − rn
导波中主要有两种激励 ∆tn = / 接收 分解(Conjugate-pair decomposition) 间反转 [46] 以及逆时偏移成像 [47] 等
c
形式:脉冲回波(Pulse-echo)[28–29] 和 等时频域分析方法。除时频域分析 方法。密集传感器阵列可以方便地
单发单收(Pitch-catch) [28,30]。以 PZT 外,Yu 等 [39] 使用二维 FFT 得到波 对结构中不可达区域实现损伤检测。
r Sn
为例,如图 2 所示,图 2(a)中是脉 数域信息,对损伤引起的入射波、反 由于阵列相对损伤是非对称分布,采
冲回波形式,传感器均布置在损伤的 射波和模态转换波实现了空间 – 频 集不到完整的散射波信息,若反射回
r ≈ rn
同一侧,选择一个激励器(Actuator)
激 发 导 波 信 φ号, = r 其r 他 接 收 传 感 器 Sensor Sensor
(Sensor)采集回波信号。阵列形式
不局限于线型,也可以是其他形状的 φ ⋅ Sn Damage Damage
∆tn =
阵列。阵列中两个相邻传感器的间 c
距较小,通常为毫米级, 是一种密集
DAi + DiS Actuator/Sensor Actuator Sensor
的传感器阵列。图 2(b)中是单发
τ p ( x, y ) =
单收形式,或单发多收,损伤在传感 cg
(a)密集阵列,采用脉冲回波的 (b)稀疏网络,采用一发一收的
器围成的凸包区域内部,选择一个激 激励 / 接收形式 激励 / 接收形式
励器(Actuator)激发导波信号, N
其他 图2 传感器布置方案
接收传感器 P(Sensor) (
( x, y ) = ∑采集, ) ( x, y )
Si τ p 这种传感 Fig.2 Configuration of sensor arrangement
i =1
26 航空制造技术·2020年第63卷第18期
M
yi = ∑ x j D{x(t ), Ti , p j , Ri } + ei
j =1
l ρ ρ
λ + 2µ ρ µ s ρ
cl = , cs =
ρ ρ ∂ 2u ( n)
Δ
(
µ ( n) 2u ( n) + λ ( n) + µ ( n) ) (
Δ Δ
∂2 2 u( n()n )
)
⋅u ( n ) = ρ ( n)
∂t 2
(n =
(( )) (( ))
( n) 2 ( n) + ( n) + ( n) ( n )
⋅u( n ) = ρ( n) ∂ u 封面文章( n )
µ u λ µ (n =1, 2, , N )
Δ Δ Δ
µ ( n) 2u ( n) + λ ( n) + µ ( n) ⋅uCOVER =ρ
2 ( n )STORY∂2t
2 (n =1, 2, , N )
Δ Δ Δ
(
µ ( n) 2u ( n) + λ ( n) + µ ( n)
Δ
) ( )
⋅u ( n ) = ρ ( n)
Δ Δ ∂ u
( n (
=1, 2, , N ) )
∂At ω , k , λ ( n ) , µ ( n ) , h( n ) = 0
(( ))
2
, λ( n()n∂) ,tµ( n()n ) , h( n ) = 0
[58]
波特征不明显,则成像结果往往不够 相控阵、延迟叠加以及基于概率 A据ωP , k的通用公式可分别表示为 :
A ω , k , λ , µ , h( n ) = 0
理想。 的成像算法在求解时通常不涉及到 ( n ) ( n )
稀疏传感器网络适用于带有加 复杂的波动方程反演,属于正向算 (
A ω , k , λ , µ , h( n ) N = 0 N −1 j )
P = ∑ pi − ∑ ∑ pi p j − ∑ ∑ ∑ pi p j pk −
N − 2 N −1 N
N N −1 j N − 2 N −1 N N
∑∑ ∑∑∑∑ ∑∑∑∑∑∑ ∏
i =1 i =1 i < j i =1 i < j j < k
筋肋等特征的复杂结构,经扩展后, 法。相控阵可看作是密集阵列的延 P = N pi −N −1 j pi p j −N − 2 N −1 N pi p j pk − N pi
可以覆盖整个待测区域 [12]。采用一 N
迟叠加,二者均仅涉及简单的时移和
P =N −1i =1jpi − i =1 iN<−jp2iNp−j1 −N i =1 i < j j < kpiNp j pk − i =1pi ∏
发一收形式,获取散射信号后,采用
P=
求和运算,计算效率高, 实现简单。 ∑
pi − i =1 pi pi =j1−i < j ∑∑ ∑ ∑∑ ∏
i =1 i < j j < k
pi p j pk − pi i =1
N
(7)
i =1 i =1 i < j i =1 i < j j < k i =1
如时间反转 [48]、延迟叠加 [49]、基于模 概率成像方法背后的物理模型是,损 N
P= pi ∏
∏
N i =1
型成像 [50] 以及基于概率成像 [12] 等 伤距离传感路径越近,则损伤对该传 P= p
方法均能够有效地实现损伤诊断。 感路径及其散射信号影响越大,使用
N
P = i =1pi i ∏ (8)
P= pi i =1
∏ 1 N
若在单个凸包的包围线(图 2(b)虚 概率量化这种影响,映射于离散化结 i =1 P =
N i =1
pi ∑
1 NN
∑∑
线)上布置足够多传感器,能够采集 构作为网格节点幅值,待融合所有路 P =1 p (9)
P = N i =1pi i
到近乎全向的损伤散射信息,那么采 径后作归一化处理,所得每个网格节 1 N N i =1
用层析成像方法可以获得高精度的 点处的图像幅值大小可有效反映该 N
P= p ∑
式 中,iN 为 数 据 总 条 数,p i 代Sn表 =0第 ∑
i =1
诊断图像 [51]。但随着传感器数量增 处存在损伤的概率 [53]。该方法计算 i 条路径的感知信息。析取融合可
∑∑ S =0
Sn n=0
多,对导波系统软硬件带来的负担也 复杂度为 O(n),因此计算效率高, 能“悲观地”夸大了损伤的存在可
越大。 且在复杂结构中避免了对信号传播 能性,而合取融合可能
∑ Sn =0
z ( t )
“乐观地”=
N −1
∑ wn f n缩(t − ∆tn )
N −1
z (t ) = ∑ wn f n (t − ∆tn )
N −1 n =0
4.2 反演求解与正向求解 历程进行解释。 小了损伤的存在可能性。折衷融合
z (t ) = ∑
N −1则平衡了所有传感路径的贡献,
w f (t − ∆tn )
n =0 n n
算法求解过程是诊断成像中的 采用时间反转方法进行成像时, 提
z (t ) = ∑ wn f n (t − ∆tn )n =0
[58]
关键步骤,可以分为反演和正向求 通常须与其他成像算法结合。例如,n =0高了数据鲁棒性和可靠性 A,但缺
ωt − kn ⋅rn )
解。从反演波动方程出发的逆向求 与延迟叠加成像结合 [37],或与概率 点在于有时会出现明显的伪像 f (t ) = e[59]j (。
n
A j (ωt − k ⋅r ) rn
成像结合 [54],均属于正向求解。需 Lu f (t ) = [60]
A ej (ωt − kn⋅nrn )n
解,多 数 情 况 下 的 解 不 唯 一,而 正 f n n(t等) = 在数据融合时结合逻辑运
rne
A 消除了边界反射波造成的伪像。
f n (t ) = e j (ωt − krnn⋅rn[61]
向时解是唯一的,二者差异主要体 注意的是,成像前有两种实现时间反 算, )
( ) ( )
λ + 2 µ µ ∂ u c = f ( t ) =
c
= e Nr−1
cc(ln⋅=u cs= 1, 2, , N ) ρ ∑ SN r=0
N
µ (n) 2 ( n) +
λ ( n) + µ ( n) µ cl2 u=⋅u+((Pnλ)=+ρ∑ µ=) p,ρ ) ρ l n s
u s ∑=∑ )( n2∑
in=1 ∂ρ2 u ( n )z (t ) = n w f (t − ∆t )
∑1, 2,n n , N ) n
Δ ΔΔ Δ
ρ∂tp∂2it 2p jµ−( n∑ u ( n )∑ + pλi (pnj) p+k µ−( n∏ ( ) ( )
ΔΔ
i − Δ ) pi ⋅u ( n ) =rnnρ ( n)
Δ Δ
(n =
分为两类:一类是反演求解;另一类 i =1 iλ=1 +i <2j µ
技术被开发出来用于雷达定位和跟 i =1µi < j j < k i =1 是目标 P(r, 2 ( n ) θ) ∂ 2
处振子的位移振幅。
t n =0
cl = ,c = ∂ u
是正向求解。在初期的导波诊断成 踪,并很快在声纳技术、 µρ( n ) 2 us( n ) + ρλ ( n) + 医学影像、
Δ
µ ( n)( 地 ) (Δ Δ
)
⋅u ( n ) = ρ( n)
当前阵元 r− N −r1∑ S2 =0 相对于阵列原点的波程 (n =1, 2, , N)
(( ) )
像技术研究中,研究人员通过借鉴医 A ω , k , λ ( n ) c ( n=)震学、海洋学以及工业
, µ , h
λ + 2 µ
,=∂±c12s0=
u
µ
( n ) µ ( n ) 2 uNDT ( n ) + 等领域
λ
Δ ( n ) + A (ω
µ ( ( ) ( )
nz)(,tk), =
( n ) n∂t(
λ ∑⋅,uµw( nnr) f
Δ Δ n )n
− r ) 2 (n)
,nh=((tnρ−) ( n∆)=t∂n 0) u r (−n r= n 1, 2, , N )
) (
l
⋅u ( n) 2 (=nρρ) ( n) n = 0r − rn ,则时间延迟
N
( n) 2 ( n) + ( n) + ( n) ρ(n =1, 2, , N ) ( n ) 差即为 n 2 Δt A n 为:
µ u λ µ( ks h ) 4得到快速发展。近年来, k kl k s P∂t=2 ∏ pi 2 基于导波的 ( ) ∂ t
Δ Δ Δ
j (ωt − kn ⋅rn )
学影像及地震勘探技术, tan
发展了导波 A ω , k , λ , µ ( n )
, h = 0 f ( t ) =
e
= − ( n) (n) r (N− n )−1rn n
( n ) i==1 ( n ) ∂ u
的层析成像和逆时偏移成像等反演 µ ( n ) tan u ( kl h )λ ( n ) + µk相控阵成像技术已成功实现了多种
2 ( n) + Δ 2 ( ( ) () )
( n ) − k(sn )
2 2 ⋅u
Δ Δ
ρ N ( n =N1,−12,j∆
( tzn (,t=)N=) ∑ N − (2wnN
A ω , k ,cλ r −,rµr− r,nh( n ) ∆t= 0=
(n )
)n−f1n (
N t − ∆
) t n ) Nr r− n rn
( )
2
2P (=n )∑ pi − ∑ ∑ pi p j − ∑
n = ∑ ∑
A ω , k , λ 结构的损伤识别, , µ , h( n ) = 0 包括各向同性金属 ∂ t pip j pk − ∏ pc(12) n
∆−tn1∆t=
算法。此类涉及波动方程的反演算 n = 0 n i
) ∂N u Ni =−11 j = i =1 i < jN − 2 N
N N −1( n )j 2 ( n ) + (N
µ u λ n ) −+2 N( n−)1 N
Δ
µ ( ) ( ⋅ )
u ( n ) = N( nNΔ Δ
1 ρ ( n 1, 2, ,
N ) NiA
= 1 i < c jj ( ω cj t
< − k k n ⋅rn )
N i =1
∑ ∑∑ ∑ ∑∑ ( −,∑ p∑ i0=1 ∑ ∏) pi p j − ∑ f ( t
r ∑
) =
∑nrpi Np− e
j pk − ∏ pi
[44,64–66]
P=
法具有较高的精度, pi − 但计算效率有所 pi p j 2− 结构 A ω , kp2,iPλp(=以及各向异性复合材料
n ) P( =
j ,N pµk∑ n ) p p 2−
hi( n )∂it = n
S N
ω ω
[67–71]
N N i−=11 ji < 若
n 3(a) 2 N −1 iN r r S−n ,有
p p p − rn p
欠缺。随后研究人员发展了能够用 结构 。 i = 1 i = 1 i i < j j j < k图 中=1
k −1 N−( nk) ( n ) i =N1 P = ∑ pi − ∑ ∑ pri p
(t ) = Nj A−S∑
r Sn ∑ ∑
ne j (ωt − kn ⋅rn ) i j k ∏ i
i =1 2 2
i =1 ik<lNj = N−1−jki =,1 ik<s jN=−j< 2 2N 2
于实时成像的正向算法, P = ∑ 如延迟叠加 cl ∑
pi − ∑ i j ∑ ∑ p p − A ωc , (
s ∑ i j k ( n∏
k ,
波束成型基于以下假设: λ p , µ
p p , h − ) =p i
0 ( 1) )
i =1 阵 f
i =1rni <≈j r
P =n ∏ , 时 i =1 i < j j < k
认 为 检 测 目 标处 i =1
i =1 i =1 i < Nj N −i1=1 j i < j j < k N − 2N −1 Ni =1 N 此 rnpi r ≈ r 于远
pi p j −∑∑
成像和基于概率成像方法, 大大提高 Sn(n=0,
=0 1, N)位
P = ∑ pi −列中所有阵元 ∑ ∑ ∑ ∑ pi p…, j pk − ∏ pi 场
N [72] ir=1≈r
r
, ≈ r
每个阵元到目标的方向矢量 n
n r − rn
iN=1P = ∏
r − nrn ∆t =
了诊断效率和灵活性,只是成像精度 iN=1
N 于同一平面上, Ni =−11 i <j j Ni − Ni <−1j jN< k
=12并作为点状的全向 φ= p ri r N
n c
近似平行,
波阵面可近似为平面, φ= r r 成
Pλ =+ ∑ 2µ p − ∑N∑
= 射 源p和
µ i p j − ∑ ∑ ∑ pi p j pk − ∏ pi i = 1
∏
往往不如反演算法。面对不同的结 φP=N= rr ∏ r pi
cl = P =ρi =1 , icPps发 接N 收 −1 器; (2)不 考 虑 导 波 为图φ3(b) =1 r
i=i =1∏ ρzjp(it ) = i =w1 if< j (tj <−k ∆t ) P = φ ⋅r∑ 。经过阵列原点处可以
构特点和损伤形式,除了要权衡成像 的频散和多模态, i<
∑ N n n波在指定方向上
i =1
S− −pi r=i1
n nn r S n
i =1 i =1 n
∆t∆ntn= =
N
1得到一条垂直于偏转方向的伪激励 N i = 1 φ ⋅ Sn
精度和计算效率,还需要明确各成像 以恒定波速 P n=c=0∏ p
传播i 。当检测目标 [36] P = ∑ pi c cφ ⋅φS⋅ Sn ∆tn =
iN=1 N线, 图 3(b)
i =1 ∆tn =n 1 N
∆t =可看出伪阵元位置是当 n c
方法所采用的模型、适用范围及优缺 处于近场时,
r − r c r ≈ rn
( n) 2 ( n) + ( n) + ( n)
( ⋅ ( n ) 1= ( n ) ∂Pu=
) (
N 2 (
)
n
∏
)
传播波阵面为曲面,
p =
此
∆ ∑
t
P
D S=
=
=0 + ∑c
D
n p i
µ u λ
点,从而选取合适的成像算法。 1 P =时需要为每个超声元件分配单独的 µ u N ∑ pi ∂t 2 iA=1N j (ωt −k⋅r )
ρ ( n 1, 2, , N ) 前阵元 r 在伪激励线的投影,
S N 因此
Δ Δ Δ
i n n
Ai n ci =1 iS
τ p ( x, y ) = D + D
1 相控阵 P= N
pi f n (t ) = 1 e
i = 1
∑ n n
∑ S =0
在偏转方向单位矢量 cD
D
+
Ai +
τ
D D ( x, yφ)==r r 的投
iS
Ai iS
r∑ i τ pτ(pNxr,(−1yx≈r,) y
N =1( n方向矢量 Prn=,如图 p
3(a) 。阵列原点 n ) = g Ai p
iS
cg
A ω , k ,i λ )
(
, µ(n ) , h( n ) =N0niN=1 ) r=n Sn 相对于阵列原点的波 cg
相控阵是一种由多个独立换能
1 影长度即为
P z (t ) = ∑ wn f n (t − ∆tn )
∑ n g S =0c
器元件(通常为 PZT 元件)组成的阵 ∑ Sn =0 , = ∑ pi
P在各阵元对波束指向处
Ni=1 N − 1 程差, N
根据波程差和传播速度可计算 ∆ t =
φ ⋅ Sn
( )
n = 0
(r,θ)的响应 z (t ) = P∑ ( x,wyn) f=nφ(∑ t r−rS∆
= ≈irtnrτn)pN(Δt x, y ) nN
c
N − 2 N −1 N ∑S
列,可通过电子控制系统实现各元件 f (t)
N −1 j
S =0 ∑
n n =0 上首先施加对应 得到时间延迟 P为: ( x , y ) (
= ∑ )
Si τ p ( x , y )
(( ))
N N i =1 N −1 N
=∑ ( yxt ,) )y ) i =1
n
= ∑ pi − ∑ ∑ pi np j N−的时间延迟 −1∑ ∑ ∑ pΔt r −p,再乘以权系数
r w n, P (zP
n = 0
x(,(t )yx,)=y=)∑ Si (tτ(−px,∆
jn k n− ∏ pi ∑
的精准时序控制, P形成新的波阵面, i p w S n1i fτ np n DAi + DiS
实现缺陷检测 。主要原理是将各 [63] i =1 i =1 i < zj (t ) = ∑i =w 1 n if<nj (tj <−k ∆
N 个阵元叠加后成为相控阵输出为 ∑
N −1
S
tn ) n =0
i =1
z M ∆ φtA φ
= rn==i =0r1 n
j
i
(
⋅ =
ω
S
t − k ⋅ r ) τ p ( x, y ) =
(13)
n =0 yi = ∑ f ( t )
n x j D{x (= n e
t), T , cp j , Ri } + ei
n n
M cg
单元发射超声波至某聚焦点或扫查 (t),上述过程可表示为 z (t ) = ∑ wrn f−n (rtn− ∆tn ) j =A
M i
1 Mj (ωt −rnkn ⋅rn ) y i = ∑ x j D{x(t ), Ti , p j , Ri } + ei
∑∑
∆ t = f n (t ) = y y=i e= 如果阵列是图 x j{Dx {(tφx),(⋅tTS),n,T p, jRj=,1}R+i }e+ ei
方向的波程差转为相位差,相位差再 N −1 Nnn=−10 c xjD pi ,j 3(b)所示的线
∑ j =1∆tn =D
N i +i D i i
由时序控制弥补,从而控制总声场的 z (t ) = wnPf =n (∏ t −zp(i∆ t ) t=n ∑ ) wn f n (t − ∆tn ) (10) τ阵, r
(
n j 则这样的投影过程可解释为将阵
f =1
x ,
( t y
) ) = =
A Ai iS
ce j (ωt − kn ⋅rn )
N
A j (ωt −nkn=⋅0rn )
( )
p n
声束焦点和轴线。相控阵波束成形 n = 0f n (t ) = i =1 e r Sn N 列旋转至伪激励线。基于平行射线
N r ×n
M g
c P ( x, y )N= ∑ Si τ p ( x, y )
y ∈R D∈R D + D
rn A j (ωt − kn⋅rn ) r − rn yAi∈ R N iSe ∈ R D ∈ i =1 N × M
R
可以看作是一种时空滤波过程,由一 f (t ) = e y 近似原理,
∈ τNR
N
( x , y ) =伪激励线到远场目标的波 NR× M
N ×M e∈R
1 N n (11)
y ∈r R− r
p D∈R c D ∈
e ∈ e R∈NR N
系列来自不同方向的波经过相移加 P = ∑ pi rA ≈rn rn 程一致, n N
因此所有伪阵元在偏转方向 g M
权后聚焦,使各阵元激励的波阵面发
N i =1f n (t ) = e j (ωt − kn ⋅rn )
P ( x, y ) =2r ∑
min上的波阵面将发生相干干涉,x-Gx i+=1rλ −n yrn
− Sr i ( τ )
p (yxi , =y )∑ x j D{x(t ), Ti , p j , Ri } + ei
其他形
A 式中, j (rω− t −rknn 为波数,⋅rn )φ = rrn rω 为角频率,rn 为阵 x ∆tn= 2 cN 21 min
j =1
x - Gx
2
+λ y 1
生同相叠加(In-phasefstacking) n (t ) = , 可 e r − min rn x-Gx
状阵列同理。波束成型后的合成信 2 + λ y 2
x
( y
)
x
∆tn = Pmin M(xx, y ) = ∑2Si τ
x - Gx + 2 λ p (1x, y ) 1
显著提高信噪比。1940 年代相控阵 rn ∑ 元 Sn至目标 =0 P(r, r − rnθ) 的方向矢量,
A 号再根据波速映射为扫查区域的诊
y= =x1∑ c x D{ x i =t1),
( r T−,rpn , R } + e
x i + r xj + S + x
i M j iN i
r − rn
∆ t =φ ⋅ Sn
1
j =1
∆2 tn =n y∈ xR = x1 + x2D+∈ R+N ×xMM e
∆tn = 目标 r − rn c
n x = x + x c+ +1 x 目标
c r − rn 伪阵元位置 r x S =
N 1 n 1 2N
M1 x + 1 x + 2 + x M
M
N −1 P(r, θ) yi p= (∑ yrx, ≈y x)jrD {x(t ), T , p , R } +P(r,
n= ∑ DI iiWi [jRNi (ix, y ) ]i
e θ) N
∑
当前阵元位置 y
z (t ) = ∑ t
波束聚焦点 ∆ = P ( x , y ) =
w f ( t − ∆ n t ) N r S
P( xN, ×NyM) = ∑ pi ( x, y ) N=2∑ DI iWi [ Ri ( x,
n n n c 当前阵元位置 i N
DAi + DiS 波束偏转
n
j = 1 n
N N
其他阵元位置 =r0 S
nτ ( x, y )r=− rn yPxr∈( iyx≈=R,)1yr= ) = ∑
n∑ (pxD i ,(∈
i=
x,)1R y= )=∑ DI Wie[x∈
1ii [ R (R-xiGx =]1λ y 1
,( yx,) ]2yi)+
R
r − rn p∆tn = c 其他阵元位置 P ( ,
φ = i
p
=r1
i
r
y
方向
∑iDI = 1
iiW=min
x i
r Scn g
i = 1
r ≈≈ rn
i = 1
r ≈ rn φ = r r N β − Ri
y ∈ R φD ,R
∈ Ri N<×Mβ e∈βxR−N+ Ri + x
rn
f n (t ) = e
A j (ω t − k
nr n S
⋅ r ) Nn
W i [ R min
i ] =
∆
tx - βGx
= −
1 β
⋅2
S +−β
n λR − R
y i
x 1 = x1 +
2 , Ri <Mβ
−φ =rrnn r P( x,ry)≈=r∑
r 曲面波阵面 φ = r r
i , RWi i<[iRβi ] = β − 1 , R < β
( ) x 2 1
n
Si τ p ( x , y ) W [nR 0, ] = R c ≥ β β
− 1
Wφφi[⋅ R Siin] =i β i − 1 0, R N≥ β
r
∆tnd =
r ≈ ri =1 ∆ tn =
d 平面波阵面 N
β pi (x, y ) = i∑ DI iWi [ Ri (
Sn c φ = r r n S n x = x + x min c x - Gx
D 2
2P(0,
0, +
φ ⋅+
+ Rx
λ
D
S
, R
≥
yi) ≥
i nx 1 y =
β ∑
1 x1 ∆ t2nAi+ = iS M i =1 i =1
x φ ⋅ SnM
∆tn =y = x D{x(t ), T , p , R } + e τ p ( x
, y ) =
D + D c x
N r − rni c ∑φ = rj r φ ⋅ i j i i D + DiS cg
Ri=(NxN, yAi) =伪激励线
Ai
N
iS
DAi + DiS
r Sn S τ p( x , y ) =px1( x+, yx)2D
DAi+ +
RMiSiD([iSxR, y( x) ,=y)β] −
(a)近场
j =1
∆ tn =
c
n
P( x, y ) =x(b) ∑ 1 远场
R Rcixi g,( yx,) y=) ==i∑
(
D +DAi DI iD
Ai +
xW DiiS i
R
Dii , Ri < β
DAi + DiS φ ⋅ S i =τ1 i ( x, y ) = i =1 D iWi [ Ri ] = β − 1
NDci g
p
τ p ( x, y )r=− rn 图3 ∆tn 相控阵示意图 = n
N
( )
∑
N
∆tn = y ∈ R N g c D∈cAiR + N ×D M iS P( xN, y ) N= ∑ pi ( x, y ) = ∑ DI)iWi [ Ri ( x, y ) ]0, Ri ≥ β
P ( x , y ) = S i τ p ( x , y
r ≈ rn cτ p ( x, y ) =
Fig.3 Illustration D of phased array e ∈ R i = 1
c
g
P ( x, y ) = ∑ Si =i 1 τ p (xβ, y−)Rii =, 1R < β ( ) i
28 航空制造技术·2020年第63卷第18期 DAi + DiS W
i=M1i [ Ri ] = β − 1
N
r Sn p ( ) N τ x , y = P ( x
∑ j βi0,=i1−RRi ≥ , y ) = ∑ (S τ )( x+, ey ) DAi + DiS
φ = r P(rx, y) = ∑ Simin τ p (xx-,Gx (
y ) 2 +cgλ y ) M
y i =
j = 1
x D { x ( t
), T , pi
j β
i
,
,
R
R
p
i
i
}
< Rβi (i x, y ) =
Di
∑
N 2 1
i =1 x y x D { x W( t ), [ R
T ,] p = , R β } − 1e
r ≈ rn P ( x, y ) = ∑ Si τ p ( x, y ) i j =1 j ( =
) i ii
M
j
i +
i
i =1 0, Ri ≥ β
COVER STORY
封面文章
断图像。 的偏斜效应,进而影响导波相控阵 D 作为一个关键参数,一般须根据结
波束成型方式可分为 3 种,包括: 的声束转向性能。选择具有准各向 构类型和关键缺陷位置来仔细选择。
阵元移动或阵列位置改变的物理波 同性性质及其邻域的导波模态可抑 另外,由于 SH0 具有非频散特性,是
束成型 [64]、模拟电路延迟的电子波 制各向异性影响,后续试验中 Yan 近年来研究热点,Huang 等 [86] 采用
束成型 [29] 以及数字信号处理的虚拟 等 [69] 使 用 相 控 阵 分 别 在 [(0/90) s ]2 全向 SH 波压电元件实现了 SH0 相
波束成型 [44,65–71]。第 1 种物理波束 和 [(0/45/90/–45) s ]2 两种铺层形式的 控阵高分辨率成像,成功识别了直径
成型技术的灵活性较差,后两种是 碳纤维环氧复合材料板上实现了损 2mm 的通孔损伤。
常用的波束成型技术。电子波束成 伤 诊 断 成 像。Vishnuvardhan [67] 和 2 空间 – 波数滤波器
型在各阵元上针对每个偏转方向独 Rajagopalan [68] 等发展了频域相位叠 波数是在传播方向上单位长度
立地延迟,再被同步激励,并遍历所 加的相控阵算法,采用一种单发多收 内的波长数量,即波动的空间频率,
有偏转方向,因此聚焦效果最强,信 (Single transmitter multi–receiver)的 包含了波动的传播速度、方向和波长
噪比最高。而在虚拟波束成形中,通 环形阵列,针对准各向同性石墨 – 环 等物理信息。采用空间 – 波数滤波
常是一个阵元激发,所有阵元接收, 氧树脂复合材料板的撞击分层损伤 器,当滤波器波数谱的方向与目标声
重复以上操作,完成遍历所有阵元 和复合材料层合板的钻孔损伤均取 源波数谱的方向一致时,声源的阵列
的全矩阵采集(Full matrix capture, 得良好检测效果。然而上述方法均 空间采样信号可以通过,否则无法
FMC)。虚拟波束成形避免了昂贵 仅限于探测无严重偏斜角的方向。 通过,此处考虑声源处于远场(图 3
且复杂的硬件系统,如电子移相器或 Tian[75] 和 Yu 等 [76] 考虑了偏斜效应, (b)[45])。针对近场盲区,刘彬 [87] 优
脉冲移位器和多路复用器等,且每个 允许在各向异性复合材料中形成相 化空间 – 波数滤波器的权重,在近场
阵元每次仅需激励一次,大大减少了 控阵波束,并通过非接触式的 SLDV 损伤识别情形下取得良好效果。
扫查时间。 系统实现了 CFRP 复材板中多损伤 通过空间 – 波数滤波器可以对
采 用 虚 拟 波 束 成 型 方 法, 的相控阵诊断成像,使用 SLDV 可以 结构上的空间波场提取特定波数信
Giurgiutiu 等 [44] 开 发 了 一 种 线 看作是对物理波束成型方式的发展。 息 [35,88–90]。Yu 等 [89] 使用 SLDV,结
型 相 控 阵,嵌 入 式 超 声 结 构 雷 达 为保证损伤诊断准确性,提高波 合空间 – 频率 – 波数联合分析方法,
(Embedded-ultrasonics structural 束成型质量,工作集中于优化波束指 获取由损伤引起的反射波、透射波以
radar,ESR),可以实现 0°~180° 的大 向性 [73],研究内容包括: 及模态转换波。由于空间 – 波数滤
面积损伤探测。对于各向同性结构, (1)最小化主波束(主瓣)宽度, 波器是在波数域中解构信号,因此具
若阵列元件表面激励均一且所有元 主瓣方向的宽度会影响最终损伤图 备分离特定波数模态、提取微弱散射
件频率响应一致,远场处由全向声 像的分辨率,主瓣越锐利,意味着能 信号的能力,还能够有效抑制复材
源阵列产生的位移场可认为是平面 量更精确地集中于偏转方向,周向分 结构由各向异性引起的模态混叠现
波阵面 [73]。而对于各向异性结构, 辨率就越高; 象 [70,90–92]。在复合材料结构上,Sohn
尤其是多层复合板,不能再将声源 (2)消除栅瓣,栅瓣是除偏转方 等 [90] 用 SLDV 获得空间波场后,从
视为全向型,但解析计算分析和有 向外其他方向上出现的强次级信号, 波场中用空间 – 波数滤波器分离出
限元模拟结果表明,点状声源传播 会导致图像出现重影,应完全消除; 损伤引起的驻波成分,成功识别分层
[74]
在远场中仍为平面波阵面 。基于 (3)抑制旁瓣,出现旁瓣表明超 和脱粘损伤。Michaels 等 [91] 使用空
上述,Leleux 等 [71] 使用相控阵探头 声能量向除偏转方向以外的方向发 间 – 波数滤波器从全波场信息中去
在带加筋肋的碳环氧复材板上检测 生了泄漏,应尽可能抑制旁瓣。 除入射波场,并分别提取 A0 和 S0 模
损伤,激励 0.5MHz(12 周期)S0 导 通过调整以下参数可以优化波 态的损伤散射波场。然而 SLDV 不
波模态,控制波束探查方向,成功识 束指向性 [77–78] :阵列形状 [65,79–81]、阵 适用于实施现场的在线监测,为此,
别出板的结构特征(边缘、加筋肋、 元直径 D(由阵元个数 N 与阵元间 研究人员发展了基于压电阵列的空
制孔)及撞击损伤并定位,最大定位 距 d 决定,通常与波长 λ 一同考虑, 间 – 波数滤波损伤成像方法 [70,93–95]。
误差为 ±10mm。各向异性复合材料 用 D/λ 表 示)[72]、幅 值 权 重 w n [82–85] 王 瑜 等 [93] 改 进 了 空 间 – 波 数 滤 波
的慢度(速度的倒数)与角度相关, 等。通常,增加阵元直径 D(线阵时 器,使用线型压电阵列获得损伤的角
Yan 和 Rose [69] 认为非圆形的慢度图 称为孔径),会提高远场分辨率,进而 度 – 时间诊断图像。在对声源成像
(Slowness map)导致波矢方向与能 改善成像效果。然而增加直径 D 将 时,采用空间 – 波数滤波器仅能获得
量流方向不一致,可能会引起严重 导致近场区域扩大。因此,阵元直径 声源的角度或方位信息,与相控阵类
2020年第63卷第18期·航空制造技术 29
封面文章 COVER STORY
30 航空制造技术·2020年第63卷第18期
∑ i ∑∑ i j ∑ ∑∑ i j k ∏ i
i =1 i =1 i < j i =1 i < j j < k i =1 N r − rn
1 N P = ∏ pi ∆tn =
P=
N
∑ pi i =1
c
i =1
N
P = ∏ pi N
r
COVER Sn
STORY
封面文章
1
i =1
∑ Sn =0 P = ∑ pi
致。由于复材结构中的分层损伤容 延迟后的信号叠加实现聚焦, N i之后 =1 播到时域激励起点, 如图 4(b)。对
r ≈ rn o
[49]
易引入非线性,因此 TR 技术成为无 1 N Michaels 等 将其进一步发展为延 于图 4(a) 其他非损伤处,如 p(x, y),
P = ∑ pi N −1
基准复合材料 NDT/SHM 的有效手 迟叠加(Delay and sum)成像方法。 τ p o,
N i =1 = ∑ wn f n (t − ∆tn ) ∑ Sn =0
由式(14) φ =求得波包时域平移量
r r
[55,113,115–118]
z (t )[117]
段 。在 TR 过程中影响重 Liu 等 n提出一种使用传递函数的 =0 反向传播 τ p o 后,波包将滞后于时域
包括
构信号准确度的因素有多种, 新 VTR 方法,结合基于概率的损伤 激励起点,ps(x, φ y)
⋅ Sn则相反。满足精
∑ S n =0 N − 1 ∆tn =
激励信号类型、激励模态、结构边界 检测重构算法,对 CFRP
∑ wn f n (t − ∆确反向传播的损伤轨迹是一个椭圆,
z (t ) =复材板中的 tn ) c
影响等内容 [55,112,119–120]
。为提高 TR 分层损伤实现了诊断成像。 A j (ωt − kn ⋅rn ) n =0 激励器和接收器是椭圆的两个焦点,
f n (t ) = e
信 号 的 聚 焦 能 力,PoddarN等 −1 [121]
和 5 延迟叠加成像 rn 如图 4(a)。融合 DAiN+条路径,DiS 图像幅
[122]
z (t ) = ∑ wn f n (t − ∆tn ) τ p ( x, y ) =
Agrahari 研究了诸如频率、 n =0脉冲频 延 迟 叠 加 (Delay–and–sum, 值
P(x, y) 由各路径散射信号在飞 c g
A
带宽度、传感器尺寸、粘接层厚度、板 DAS)是一种经典有效且应用广泛 f n (t ) = e j (ωt − kn ⋅行时间处的幅值叠加得到
rn )
2020年第63卷第18期·航空制造技术 33
N
∑ i
yi =M∑Nx×jMD{x(t ), Ti , p j , RNi } + ei
i =1
y ∈ RN D∈R e∈R
yi = ∑j =x1 j D{x(t ), Ti , p j , Ri } + ei
2 j =1
∑ Sn =0
封面文章
min x-Gx 2 + λ y 1
x 2
COVER STORY min xy-∈
GxR N2 + λ y 1D ∈ R N ×M e ∈ RN
x N
y ∈R D ∈ R N ×M ∈R
e法 N
HARBUT [191–192] N −1 ,或开发结合压缩感 x 1 = x1 + x2 + + xM
i 条传感路径对应的非负路径加权函 [197–198]
。RAPID 在成像过程中无
z (t ) = ∑ wn f n (t −[193] ∆tn )
知的层析成像方法 ,均可以补偿 数, x代表第 = x1 i+条传感路径受损伤影响
x2 + + xM 2 需导波在结构中传播的先验知识,同
N x-Gx 2 + λ y 1
n =0 1 min
N
Wi [R
i i [+
欠采样引起的混叠,从而提高诊断图 y ) = ∑ pi ( x, y ) min
P( x, 的范围。此处将 =∑ x i ]W 2 R ( x, y ) ]
定义为随成 时避免了对真实结构中复杂时间历
N NxDI-Gx iλ y
2 1
= ∑ y)与第 pi ( x, y ) =i ∑
x
DI iWi [ Ri ( x, y ) ] 程的解释,计算效率高并且适合自动
像的鲁棒性和准确度。 i i
P( x像点
, y ) (x,
= 1 =条传感路径的相
1
DAi + DiS
Ri ( x, y ) =
D
COVER STORY
封面文章
度要求。但过多的传感路径又会增 权值差异,尽可能抑制“伪损伤”对 技术的飞速发展,为超声导波诊断成
加数据传输、采集和处理的负担,同 成像结果的影响,加筋肋复合材料板 像技术带来了新的契机。超声导波
时与在线测量系统轻量化小型化的 上的损伤诊断试验结果表明该方法 诊断成像技术在今后可能面临的挑
目标相背离。为此,Wang 等 [53] 提出 能有效提高定位精度。 战及其对应的发展方向概述如下:
一种虚拟路径(Virtual sensing paths) 9 成像技术对比 (1)应用对象方面。在飞行器
方法来增加路径数量,旨在提高定位 上述 8 种导波损伤诊断成像技 结构中,除板、管外,桁条 [210,214] 及起
精 度。 同 时,Wang 等 [53,206] 还 观 察 术的特点及优缺点比较见表 1。通 落架 [215] 等也可作为波导结构。针
到在多路径交叉点或传感器位置处 常对于不同的结构形式,应在综合考 对此类复杂波导结构问题,还需从理
容易出现“伪损伤”。Sharif–Khodaei 虑传感器可布置区域、待测损伤与导 论方法、数值模拟、试验研究等方面
等 [143] 在使用 RAPID 算法对一块复 波作用方式及各成像方法优缺点等 求解导波传播问题,为高质量成像奠
材板上进行缺陷诊断时也遇到了类 方面的基础上,选择合适的成像方法 定基础。
似情况。关于这个困扰 RAPID 技术 进行研究和应用。 (2)成像算法实现方面。真实
的“伪 损 伤”问 题,Liu 等 [12] 指 出, 飞行器结构不仅几何形式复杂,且待
在数据融合过程中,多路径交叉点的 主要挑战和发展方向 测区域面积大,现阶段单一的导波诊
权值会显著高于待测区域其他成像 基于超声导波的 NDT/SHM 技 断成像方法往往难以兼顾计算效率
点的权值,因此同样的损伤指数变 术在近年来取得了长足的进步,超声 和成像精度。需融合多种成像方法,
化下,诊断成像时映射到该点的几率 导波诊断成像技术作为其中一项重 在效率和精度二者之间寻找平衡点,
更大。为此,Liu 等 [12] 引入了同时 要研究热点,涉及到超声导波技术多 开发能够快速完成检测且提供高质
考量传感路径与路径加权函数影响 方面内容,包括导波传播理论和数值 量图像的混合诊断策略。
的单位权值分布函数,并提出一种路 求解 [210]、传感器激励问题 [211–212]、模 在飞行器结构的真实服役环境
径加权函数的优化选择方法,有助于 态选择与频散补偿方法 [133] 和先进 下,力热振噪等因素对导波信号和诊
降低选取参数时对人工经验的依赖。 导波信号处理技术 [213] 等,是超声导 断成像效果的影响不可忽视。发展
在此基础上,Liu 等 [196] 提出一种权 波技术的应用集大成者,但要在真实 无基准成像方法和带环境补偿的成
值补偿的加权分布诊断成像方法,在 飞行器结构上得到更广泛应用,仍面 像方法是两种可行的解决途径,部分
不增加传感器和传感路径数量前提 临不少问题。近年来,波动理论与建 相关研究已在实验室环境中取得良
下,补偿路径交叉点与其他成像点的 模方法、计算机信息科学、自动化等 好效果 [118,137],但在工程真实环境下
表1 基于超声导波的不同损伤诊断成像技术
Table 1 Damage diagnostic imaging approaches based on ultrasonic guided waves
适用散射信号
技术 传感器布置形式 适用范围 优点 缺点
特征形式
前处理硬件复杂,
相控阵 密集阵列 反射 简单结构 具备聚焦能力,高精度
需保证时延剖面精确度
空间 – 波数
密集阵列 反射 简单结构 波数域处理 计算复杂,涉及多域转换
滤波器
消除频散效应, 需结构先验知识
逆时偏移成像 密集阵列 反射 简单和复杂结构
适用多损伤 及保证波场采集完整性
密集阵列 反射,透射, 具备聚焦能力,
时间反转成像 简单和复杂结构 依赖时反算子稳定性
或稀疏网络 折射,模态转换 适用无基准方法
算法简单易实现,
延迟叠加成像 稀疏网络 折射 大面积结构 依赖波速,受环境影响大
无需提取 TOF
适用多损伤, 依赖于模型建立合理性,
基于模型成像 稀疏网络 折射 简单和复杂结构
受外部环境因素小 涉及反演求解时计算效率低
工程应用有局限,
层析成像 密集网络 透射,折射,衍射 结构热点区域 高精度
反演求解计算用时长
复杂和 算法简单,无需提取 TOF 和 成像效果与参数选择相关,
基于概率成像 较密集网络 折射,模态转换
大面积结构 结构先验知识 成像精度受传感器网络密度影响
2020年第63卷第18期·航空制造技术 35
封面文章 COVER STORY
36 航空制造技术·2020年第63卷第18期
COVER STORY
封面文章
components in aerospace structures[D]. Dalian: sensing[J]. Smart Materials and Structures, situ optimized PWAS phased arrays for Lamb
Dalian University of Technology, 2016. 2016, 25(8): 085042–085052. wave structural health monitoring[J]. Journal of
[16] WU Z J, XU L T, WANG Y S, et [26] QING X, BEARD S, KUMAR A, Mechanics of Materials and Structures, 2007,
al. Impact energy identification on a composite et al. Advances in the development of built– 2(3): 459–487.
plate using basis vectors[J]. Smart Materials and in diagnostic system for filament wound [37] QIU L, YUAN S F, ZHANG X Y,
Structures, 2015, 24(9): 095007. composite structures[J]. Composites Science et al. A time reversal focusing based impact
[17] QIU L, LIU B, YUAN S F, et al. and Technology, 2006, 66(11–12): 1694–1702. imaging method and its evaluation on complex
Impact imaging of aircraft composite structure [27] YANG B, XUAN F Z, CHEN S J, composite structures[J]. Smart Materials and
based on a model–independent spatial– et al. Damage localization and identification Structures, 2011, 20(10): 105014.
wavenumber filter[J]. Ultrasonics, 2016, 64: in WGF/epoxy composite laminates by using [38] PA I P F, D E N G H G ,
10–24. Lamb waves: Experiment and simulation[J]. S U N D A R E S A N M J . Ti m e – f r e q u e n c y
[18] REN Y Q, QIU L, YUAN S F, et Composite Structures, 2017, 165: 138–147. characterization of Lamb waves for material
al. A diagnostic imaging approach for online [28] GORGIN R, WU Z J, GAO D Y, et evaluation and damage inspection of plates[J].
characterization of multi–impact in aircraft al. Damage size characterization algorithm for Mechanical Systems and Signal Processing,
composite structures based on a scanning active structural health monitoring using the A0 2015, 62/63: 183–206.
spatial–wavenumber filter of guided wave[J]. mode of Lamb waves[J]. Smart Materials and [39] YU L Y, TIAN Z H. Lamb wave
Mechanical Systems and Signal Processing, Structures, 2014, 23(3): 035015. structural health monitoring using a hybrid
2017, 90: 44–63. [29] KUDELA P, RADZIENSKI M, PZT–laser vibrometer approach[J]. Structural
[19] ZHENG Y B, LIU K H, WU Z OSTACHOWICZ W, et al. Structural health Health Monitoring: an International Journal,
J, et al. Lamb waves and electro–mechanical monitoring system based on a concept of Lamb 2013, 12(5–6): 469–483.
impedance based damage detection using a wave focusing by the piezoelectric array[J]. [40] HONG J C, SUN K H, KIM Y Y.
mobile PZT transducer set[J]. Ultrasonics, 2019, Mechanical Systems and Signal Processing, Waveguide damage detection by the matching
92: 13–20. 2018, 108: 21–32. pursuit approach employing the dispersion–
[20] WU Z J, QING X P, CHANG F [30] IHN J B, CHANG F K. Pitch– based chirp functions[J]. IEEE Transactions
K. Damage detection for composite laminate catch active sensing methods in structural health on Ultrasonics, Ferroelectrics and Frequency
plates with a distributed hybrid PZT/FBG sensor monitoring for aircraft structures[J]. Structural Control, 2006, 53(3): 592–605.
network[J]. Journal of Intelligent Material Health Monitoring: an International Journal, [41] RAGHAVAN A, CESNIK C E S.
Systems and Structures, 2009, 20(9): 1069– 2008, 7(1): 5–19. Guided–wave signal processing using chirplet
1077. [31] ZHOU K, ZHENG Y B, ZHANG matching pursuits and mode correlation for
[21] REN B Y, LISSENDEN C J. PVDF J Q, et al. A reconstruction–based mode structural health monitoring[J]. Smart Materials
multielement lamb wave sensor for structural separation method of Lamb wave for damage and Structures, 2007, 16(2): 355–366.
health monitoring[J]. IEEE Transactions on detection in plate structures[J]. Smart Materials [42] SCHMIDT R. A signal subspace
Ultrasonics, Ferroelectrics, and Frequency and Structures, 2019, 28(3): 035033. approach to multiple emitter location and
Control, 2016, 63(1): 178–185. [32] STASZEWSKI W J. Intelligent spectral estimation[D]. Stanford: Stanford
[22] KWUN H, KIM S Y, LIGHT G M. signal processing for damage detection in University, 1982.
The magnetostrictive sensor technology for long composite materials[J]. Composites Science and [43] HE J Z, YUAN F G. Lamb wave–
range guided wave testing and monitoring of Technology, 2002, 62(7–8): 941–950. based sub wave length damage imaging using
structures[J]. Materials Evaluation, 2003, 61(1): [33] CROXFORD A J, WILCOX P the DORT–MUSIC technique in metallic
80–84. D, DRINKWATER B W, et al. Strategies for plates[J]. Structural Health Monitoring: an
[23] HARB M S, YUAN F G. Barely guided–wave structural health monitoring[J]. International Journal, 2016, 15(1): 65–80.
visible impact damage imaging using non– Proceedings of the Royal Society A: [44] G I U R G I U T I U V, B A O J .
contact air–coupled transducer/laser Doppler Mathematical, Physical and Engineering Embedded–ultrasonics structural radar for in
vibrometer system[J]. Structural Health Sciences, 2007, 463: 2961–2981. situ structural health monitoring of thin–wall
Monitoring: an International Journal, 2017, [34] OSTACHOWICZ W, KUDELA P, structures[J]. Structural Health Monitoring: an
16(6): 663–673. MALINOWSKI P, et al. Damage localization in International Journal, 2004, 3(2): 121–140.
[24] PARK B, SOHN H, MALINOWSKI plate–like structures based on PZT sensors[J]. [45] PUREKAR A S, PINES D J,
P, et al. Delamination localization in wind Mechanical Systems and Signal Processing, SUNDARARAMAN S, et al. Directional
turbine blades based on adaptive time–of–flight 2009, 23(6): 1805–1829. piezoelectric phased array filters for detecting
analysis of noncontact laser ultrasonic signals[J]. [35] PAI P F, SUNDARESAN M J. damage in isotropic plates[J]. Smart Materials
Nondestructive Testing and Evaluation, 2017, Space–wavenumber and time–frequency and Structures, 2004, 13(4): 838–850.
32(1): 1–20. analysis for damage inspection of thin–walled [46] ING R K, FINK M. Time–
[25] TIAN Z, YU L, LECKEY C. structures[J]. Structural Health Monitoring: an reversed Lamb waves[J]. IEEE Transactions
Rapid guided wave delamination detection and International Journal, 2012, 11(4): 452–471. on Ultrasonics, Ferroelectrics and Frequency
quantification in composites using global–local [36] Y U L Y, G I U R G I U T I U V. I n – Control, 1998, 45(4): 1032–1043.
2020年第63卷第18期·航空制造技术 37
封面文章 COVER STORY
[47] LIN X, YUAN F G. Damage FAN Shixuan, et al. Application of image C V, et al. Structural health monitoring of
detection of a plate using migration technique[J]. compounding in ultrasonic Lamb wave anisotropic plates using ultrasonic guided wave
Journal of Intelligent Material Systems and tomography[J]. Journal of Shanghai University STMR array patches[J]. NDT & E International,
Structures, 2001, 12(7): 469–482. (Natural Science), 2007(4): 430–433. 2009, 42(3): 193–198.
[48] CAI J, SHI L H, YUAN S F, et al. [58] SU Z Q, WANG X M, CHENG [68] R A J A G O PA L A N J,
High spatial resolution imaging for structural L, et al. On selection of data fusion schemes BALASUBRAMANIAM K, KRISHNAMURTHY
health monitoring based on virtual time for structural damage evaluation[J]. Structural C V. A phase reconstruction algorithm for Lamb
reversal[J]. Smart Materials and Structures, Health Monitoring, 2009, 8(3): 223–241. wave based structural health monitoring of
2011, 20(5): 055018. [59] C L A R K E T , C A W L E Y P. anisotropic multilayered composite plates[J].
[49] MICHAELS J E, MICHAELS T E. Enhancing the defect localization capability of a The Journal of the Acoustical Society of
Guided wave signal processing and image fusion guided wave SHM system applied to a complex America, 2006, 119(2): 872.
for in situ damage localization in plates[J]. Wave structure[J]. Structural Health Monitoring: an [69] YAN F, ROSE J L. Guided wave
Motion, 2007, 44(6): 482–492. International Journal, 2011, 10(3): 247–259. phased array beam steering in composite
[50] LEVINE R M, MICHAELS J E. [60] LU G T, LI Y R, SONG G B. A plates[C]//Proceedings of Health Monitoring of
Model–based imaging of damage with Lamb delay–and–Boolean–ADD imaging algorithm Structural and Biological Systems, International
waves via sparse reconstruction[J]. The Journal for damage detection with a small number of Society for Optics and Photonics. San Diego,
of the Acoustical Society of America, 2013, piezoceramic transducers[J]. Smart Materials 2007.
133(3): 1525–1534. and Structures, 2016, 25(9): 095030. [70] PUREKAR A S, PINES D J.
[51] MALYARENKO E V, HINDERS M [61] LEONARD K R, HINDERS Damage detection in thin composite laminates
K. Fan beam and double cross hole Lamb wave M K. Lamb wave tomography of pipe–like using piezoelectric phased sensor arrays and
tomography for mapping flaws in aging aircraft structures[J]. Ultrasonics, 2005, 43(7): 574–583. guided lamb wave interrogation[J]. Journal of
structures[J]. The Journal of the Acoustical [62] K U N D U T. A c o u s t i c s o u r c e Intelligent Material Systems and Structures,
Society of America, 2000, 108(4): 1631–1639. localization[J]. Ultrasonics, 2014, 54(1): 25–38. 2010, 21(10): 995–1010.
[52] PRASAD S M, BALASUBRAMANIAM [63] 詹湘琳 , 韩红斌 , 周德新 . 民用 [71] L E L E U X A , M I C H E A U P,
K, KRISHNAMURTHY C V. Structural health 飞机复合材料结构件超声相控阵无损检测 CASTAINGS M. Long range detection of
monitoring of composite structures using Lamb 技 术 进 展[J]. 航 空 制 造 技 术 , 2014, 57(15): defects in composite plates using lamb waves
wave tomography[J]. Smart Materials and 124–127. generated and detected by ultrasonic phased
Structures, 2004, 13(5): N73–N79. ZHAN Xianglin, HAN Hongbin, ZHOU array probes[J]. Journal of Nondestructive
[53] WA N G D , Y E L , L U Y, e t a l . Dexin. Progress of ultrasonic phased array Evaluation, 2013, 32(2): 200–214.
Probability of the presence of damage estimated nondestructive testing technique for civil [72] GIURGIUTIU V. Structural health
from an active sensor network in a composite aviation composite structures[J]. Aeronautical monitoring with piezoelectric wafer active
panel of multiple stiffeners[J]. Composites Manufacturing Technology, 2014, 57(15): 124– sensors[M]. Amsterdam: Elsevier Academic
Science and Technology, 2009, 69(13): 2054– 127. Press, 2007.
2063. [64] SICARD R, CHAHBAZ A, [73] YU L Y, GIURGIUTIU V. In situ 2–
[54] MIAO X T, WANG D, YE L, et al. GOYETTE J. Guided Lamb waves and L–SAFT D piezoelectric wafer active sensors arrays for
Identification of dual notches based on time– processing technique for enhanced detection and guided wave damage detection[J]. Ultrasonics,
reversal lamb waves and a damage diagnostic imaging of corrosion defects in plates with small 2008, 48(2): 117–134.
imaging algorithm[J]. Journal of Intelligent depth–to wavelength ratio[J]. IEEE Transactions [74] VELICHKO A, WILCOX P D.
Material Systems and Structures, 2011, 22(17): on Ultrasonics, Ferroelectrics and Frequency Modeling the excitation of guided waves in
1983–1992. Control, 2004, 51(10): 1287–1297. generally anisotropic multilayered media[J]. The
[55] PARK H W, SOHN H, LAW K H, [65] WILCOX P D. Omni–directional Journal of the Acoustical Society of America,
et al. Time reversal active sensing for health guided wave transducer arrays for the rapid 2007, 121(1): 60–69.
monitoring of a composite plate[J]. Journal of inspection of large areas of plate structures[J]. [75] TIAN Z, LECKEY C A C, YU
Sound and Vibration, 2007, 302(1–2): 50–66. IEEE Transactions on Ultrasonics, Ferroelectrics L. Phased array beam forming and imaging
[56] WANG J Z, SHEN Y F. An enhanced and Frequency Control, 2003, 50(6): 699–709. in composite laminates using guided
Lamb wave virtual time reversal technique [66] R A J A G O PA L A N J, waves[C]//Proceedings of Health Monitoring
for damage detection with transducer transfer BALASUBRAMANIAM K, KRISHNAMURTHY of Structural and Biological Systems 2016.
function compensation[J]. Smart Materials and C V. A single transmitter multi–receiver (STMR) Colorado: International Society for Optics and
Structures, 2019, 28(8): 085017. PZT array for guided ultrasonic wave based Photonics,2016.
[57] 张海燕 , 刘倩玮 , 范仕轩 , 等 . 图 structural health monitoring of large isotropic [76] YU L Y, TIAN Z H. Guided wave
像合成在超声兰姆波层析成像中的应用 [J]. plate structures[J]. Smart Materials and phased array beam forming and imaging in
上海大学学报(自然科学版), 2007(4): 430– Structures, 2006, 15(5): 1190–1196. composite plates[J]. Ultrasonics, 2016, 68:
433. [67] V I S H N U VA R D H A N J , 43–53.
ZHANG Haiyan, LIU Qianwei, MURALIDHARAN A, KRISHNAMURTHY [77] DRINKWATER B W, WILCOX
38 航空制造技术·2020年第63卷第18期
COVER STORY
封面文章
P D. Ultrasonic arrays for non–destructive LIU Bin. Research on spatial–wavenumber 085024.
evaluation: A review[J]. NDT & E International, filter based structural health monitoring [97] KWON H S, KIM J Y. An analytical
2006, 39(7): 525–541. and imaging method[D]. Nanjing: Nanjing filter design method for guided wave phased
[78] 崔媛智. 基于 Lamb 波频率调谐和 University of Aeronautics and Astronautics, arrays[J]. Mechanical Systems and Signal
相控阵的金属板损伤检测研究[D]. 镇江 : 江 2015. Processing, 2016, 81: 433–446.
苏大学 , 2016. [88] RUZZENE M. Frequency– [98] B AY S A L E , K O S L O F F D D ,
CUI Yuanzhi. Research of metal plate wavenumber domain filtering for improved SHERWOOD J W C. Reverse time migration[J].
damage detection based on Lamb wave damage visualization[J]. Smart Materials and Geophysics, 1983, 48(11): 1514–1524.
frequency tuning and phased array[D]. Structures, 2007, 16(6): 2116–2129. [99] LIN X, YUAN F G. Detection of
Zhenjiang: Jiangsu University, 2016. [89] YU L Y, LECKEY C A C, TIAN multiple damages by prestack reverse–time
[79] MALINOWSKI P, WANDOWSKI Z H. Study on crack scattering in aluminum migration[J]. AIAA Journal, 2001, 39: 2206–
T, TRENDAFILOVA I, et al. A phased array– plates with Lamb wave frequency–wavenumber 2215.
based method for damage detection and analysis[J]. Smart Materials and Structures, [100] LIN X, YUAN F G. Experimental
localization in thin plates[J]. Structural Health 2013, 22(6): 065019. study applying a migration technique in
Monitoring: an International Journal, 2009, 8(1): [90] SOHN H, DUTTA D, YANG J Y, structural health monitoring[J]. Structural Health
5–15. et al. Automated detection of delamination and Monitoring: an International Journal, 2005, 4(4):
[80] WANDOWSKI T, MALINOWSKI disbond from wave field images obtained using 341–353.
P, OSTACHOWICZ W M. Damage detection a scanning laser vibrometer[J]. Smart Materials [101] HE J Z, YUAN F G. Damage
with concentrated configurations of piezoelectric and Structures, 2011, 20(4): 045017. identification for composite structures using
transducers[J]. Smart Materials and Structures, [91] MICHAELS T E, MICHAELS J E, a cross–correlation reverse–time migration
2011, 20(2): 025002. RUZZENE M. Frequency–wavenumber domain technique[J]. Structural Health Monitoring: an
[81] LIU Z H, SUN K M, SONG G R, analysis of guided wavefields[J]. Ultrasonics, International Journal, 2015, 14(6): 558–570.
et al. Damage localization in aluminum plate 2011, 51(4): 452–466. [102] 严刚 , 周丽 . 基于频率 – 波数域
with compact rectangular phased piezoelectric [92] TIAN Z H, YU L Y, LECKEY C, 偏移的损伤被动成像识别研究 [J]. 固体力学
transducer array[J]. Mechanical Systems and et al. Guided wave imaging for detection and 学报 , 2011, 32(S1): 273–279.
Signal Processing, 2016, 70–71: 625–636. evaluation of impact–induced delamination in YAN Gang, ZHOU Li. Passive imaging
[82] STERNINI S, PAU A, LANZA composites[J]. Smart Materials and Structures, and identification of damages using migration
D I S C A L E A F. M i n i m u m – v a r i a n c e 2015, 24(10): 105019. technique in frequency–wavenumber domain[J].
imaging in plates using guided–wave–mode [93] 王瑜 , 袁慎芳 , 邱雷 . 基于改进空 Acta Mechnica Solida Sinica, 2011, 32(S1):
b e a m f o r m i n g [ J ] . I E E E Tr a n s a c t i o n s o n 间滤波器的复合材料结构损伤成像方法[J]. 273–279.
Ultrasonics, Ferroelectrics, and Frequency 复合材料学报 , 2011, 28(1): 186–193. [103] 严刚 , 周丽 . 基于 Mindlin 板理论
Control, 2019, 66(12): 1906–1919. WANG Yu, YUAN Shenfang, QIU Lei. 的偏移损伤成像数值仿真研究 [J]. 力学学报 ,
[83] ENGHOLM M, STEPINSKI T. Improved spatial filter based damage imaging 2010, 42(3): 499–505.
Adaptive beam forming for array imaging of method on composite structures[J]. Acta YAN Gang, ZHOU Li. A simulation study
plate structures using Lamb waves[J]. IEEE Materiae Compositae Sinica, 2011, 28(1): 186– on migration technique for damage imaging
Transactions on Ultrasonics, Ferroelectrics, and 193. based on mindlin plate theory[J]. Chinese
Frequency Control, 2010, 57(12): 2712–2724. [94] 刘彬 , 邱雷 , 袁慎芳 , 等 . 基于多 Journal of Theoretical and Applied Mechanics,
[84] KWON H S, LEE S S, KIM J Y. 维阵列和空间滤波器的损伤无波速成像定位 2010, 42(3): 499–505.
Beam pattern improvement by compensating 方法 [J]. 复合材料学报 , 2014, 31(3): 835–844. [104] WANG L, YUAN F G. Damage
array nonuniformities in a guided wave phased LIU Bin, QIU Lei, YUAN Shenfang, et identification in a composite plate using prestack
array[J]. Smart Materials and Structures, 2013, al. Damage imaging and localization method reverse–time migration technique[J]. Structural
22(8): 085002. based on multi–dimension arrays and spatial Health Monitoring: an International Journal,
[85] HUA J D, LIN J, ZENG L, et al. filter without wave velocity[J]. Acta Materiae 2005, 4(3): 195–211.
Minimum variance imaging based on correlation Compositae Sinica, 2014, 31(3): 835–844. [105] ZHOU L, YUAN F G, MENG W
analysis of Lamb wave signals[J]. Ultrasonics, [95] QIU L, LIU B, YUAN S F, et al. A J. A pre–stack migration method for damage
2016, 70: 107–122. scanning spatial–wavenumber filter and PZT identification in composite structures[J]. Smart
[86] HUANG Q, CHEN M T, SU Z 2–D cruciform array based on–line damage Structures and Systems, 2007, 3(4): 439–454.
Q, et al. A high–resolution structural health imaging method of composite structure[J]. [106] HE J Z, YUAN F G. A quantitative
monitoring system based on SH wave Sensors and Actuators A: Physical, 2016, 248: damage imaging technique based on enhanced
piezoelectric transducers phased array[J]. 62–72. CCRTM for composite plates using 2D scan[J].
Ultrasonics, 2019, 97: 29–37. [96] ENGHOLM M, STEPINSKI T, Smart Materials and Structures, 2016, 25(10):
[87] 刘彬 . 基于空间 – 波数滤波器的 OLOFSSON T. Imaging and suppression of 105022.
结构健康监测成像方法研究[D]. 南京 : 南京 Lamb modes using adaptive beam forming[J]. [107] HE J Z, ROCHA D C, LESER P
航空航天大学 , 2015. Smart Materials and Structures, 2011, 20(8): E, et al. Least–squares reverse time migration
2020年第63卷第18期·航空制造技术 39
封面文章 COVER STORY
(LSRTM) for damage imaging using Lamb Journal of Aerospace Engineering, 2007, 20(3): and Structures, 2004, 13(2): 415–423.
waves[J]. Smart Materials and Structures, 2019, 141–151. [128] MICHAELS J E. Enhanced
28(6): 065010. [117] LIU Z H, YU H T, FAN J W, et differential methods for guided wave phased
[108] F I N K M . Ti m e r e v e r s a l o f al. Baseline–free delamination inspection in array imaging using spatially distributed
ultrasonic fields. I. Basic principles[J]. IEEE composite plates by synthesizing non–contact piezoelectric transducers[C]//Proceedings of
Transactions on Ultrasonics, Ferroelectrics and air–coupled Lamb wave scan method and virtual AIP Conference, American Institute of Physics.
Frequency Control, 1992, 39(5): 555–566. time reversal algorithm[J]. Smart Materials and Brunswick, 2006.
[109] ING R K, FINK M. Time Structures, 2015, 24(4): 045014. [129] MICHAELS J E, CROXFORD A J,
recompression of dispersive Lamb waves using [118] MUSTAPHA S, YE L. Propagation WILCOX P D. Imaging algorithms for locating
a time reversal mirror – Application to flaw behaviour of guided waves in tapered sandwich damage via in situ ultrasonic sensors[C]//
detection in thin plates[C]//Proceedings of 1996 structures and debonding identification using Proceedings of 2008 IEEE Sensors Applications
IEEE Ultrasonics Symposium. San Antonio, time reversal[J]. Wave Motion, 2015, 57: 154– Symposium. Atlanta, 2008: 63–67.
1996. 170. [130] MICHAELS J E. Detection,
[110] PRADA C, FINK M. Separation [119] PARK H W, KIM S B, SOHN H. localization and characterization of damage
of interfering acoustic scattered signals using Understanding a time reversal process in Lamb in plates with an in situ array of spatially
the invariants of the time–reversal operator. wave propagation[J]. Wave Motion, 2009, 46(7): distributed ultrasonic sensors[J]. Smart Materials
Application to Lamb waves characterization[J]. 451–467. and Structures, 2008, 17(3): 035035.
The Journal of the Acoustical Society of [120] WATKINS R, JHA R. A modified [131] LEVINE R M, MICHAELS J E.
America, 1998, 104(2): 801–807. time reversal method for Lamb wave based Block–sparse reconstruction and imaging for
[111] 王强 , 严夏君 , 陈小惠 , 等 . 无基 diagnostics of composite structures[J]. Lamb wave structural health monitoring[J].
准 Lamb 波时间反转损伤概率成像监测方法 Mechanical Systems and Signal Processing, IEEE Transactions on Ultrasonics,
[J]. 仪器仪表学报 , 2013, 34(7): 1588–1594. 2012, 31: 345–354. Ferroelectrics, and Frequency Control, 2014,
WA N G Q i a n g , YA N X i a j u n , C H E N [121] PODDAR B, KUMAR A, MITRA 61(6): 1006–1015.
Xiaohui, et al. Baseline free Lamb wave damage M, et al. Time reversibility of a Lamb wave for [132] KIM H W, YUAN F G. Enhanced
probability imaging monitoring method based damage detection in a metallic plate[J]. Smart damage imaging of a metallic plate using
on time reversal theory[J]. Chinese Journal of Materials and Structures, 2011, 20(2): 025001. matching pursuit algorithm with multiple wave
Scientific Instrument, 2013, 34(7): 1588–1594. [122] AGRAHARI J K, KAPURIA S. paths[J]. Ultrasonics, 2018, 89: 84–101.
[112] XU B L, GIURGIUTIU V. Single Effects of adhesive, host plate, transducer and [133] KIM H, YUAN F G. Adaptive
mode tuning effects on lamb wave time reversal excitation parameters on time reversibility of signal decomposition and dispersion removal
with piezoelectric wafer active sensors for ultrasonic Lamb waves[J]. Ultrasonics, 2016, based on the matching pursuit algorithm using
structural health monitoring[J]. Journal of 70: 147–157. dispersion–based dictionary for enhancing
Nondestructive Evaluation, 2007, 26(2–4): [123] WANG Q, YUAN S F. Baseline– damage imaging[J]. Ultrasonics, 2020, 103:
123–134. free imaging method based on new PZT sensor 106087.
[113] SOHN H, PARK H W, LAW K H, arrangements[J]. Journal of Intelligent Material [134] CROXFORD A J, WILCOX P D,
et al. Combination of a time reversal process Systems and Structures, 2009, 20(14): 1663– DRINKWATER B W. Guided wave SHM with
and a consecutive outlier analysis for baseline– 1673. a distributed sensor network[C]//Proceedings of
free damage diagnosis[J]. Journal of Intelligent [124] C I A M PA F, M E O M . I m p a c t Health Monitoring of Structural and Biological
Material Systems and Structures, 2007, 18(4): detection in anisotropic materials using a Systems 2008. San Diego: International Society
335–346. time reversal approach[J]. Structural Health for Optics and Photonics, 2008.
[114] 蔡笑风 , 刘继方 , 李永峰 , 等 . 复 Monitoring, 2012, 11(1): 43–49. [135] LU G T, LI Y R, WANG T, et al.
合材料 Lamb 波时间反转检测方法研究现状 [125] ZENG L, LIN J, HUANG L P. A A multi–delay–and–sum imaging algorithm
[J]. 无损检测 , 2018, 40(11): 64–69. modified lamb wave time–reversal method for for damage detection using piezoceramic
CAI Xiaofeng, LIU Jifang, LI Yongfeng, health monitoring of composite structures[J]. transducers[J]. Journal of Intelligent Material
et al. Research status of Lamb wave time– Sensors, 2017, 17(5): 955. Systems and Structures, 2017, 28(9): 1150–
reversal method for composite materials[J]. [126] 王强 , 袁慎芳 . 复合材料板脱层 1159.
Nondestructive Testing, 2018, 40(11): 64–69. 损伤的时间反转成像监测 [J]. 复合材料学报 , [136] F E N D Z I C , R É B I L L AT M ,
[115] GANGADHARAN R, MURTHY 2009, 26(3): 99–104. MECHBAL N, et al. A data–driven temperature
C , G O PA L A K R I S H N A N S , e t a l . Ti m e WANG Qiang, YUAN Shenfang. Time compensation approach for Structural Health
reversal health monitoring of composite plates reversal imaging method for composite Monitoring using Lamb waves[J]. Structural
using Lamb waves[J]. International Journal of delamination monitoring[J]. Acta Materiae Health Monitoring: an International Journal,
Aerospace Innovations, 2011, 3(3): 131–142. Compositae Sinica, 2009, 26(3): 99–104. 2016, 15(5): 525–540.
[116] SOHN H, PARK H W, LAW K [127] WANG C H, ROSE J T, CHANG F [137] CHEN X, MICHAELS J E, LEE S
H, et al. Damage detection in composite plates K. A synthetic time–reversal imaging method for J, et al. Load–differential imaging for detection
by using an enhanced time reversal method[J]. structural health monitoring[J]. Smart Materials and localization of fatigue cracks using Lamb
40 航空制造技术·2020年第63卷第18期
COVER STORY
封面文章
waves[J]. NDT & E International, 2012, 51: vessel by guided waves based on a real–time method for composite laminates[J]. Composite
142–149. monitoring system[J]. International Journal of Structures, 2019, 209: 233–241.
[138] YU L Y, LECKEY C A. Lamb Hydrogen Energy, 2019, 44(40): 22740–22751. [159] HARLEY J B, MOURA J M
wave–based quantitative crack detection [149] 徐冠基 , 许才彬 , 杨志勃 , 等 . 碳 F. Data–driven matched field processing for
using a focusing array algorithm[J]. Journal of 纤维层合板 Lamb 波损伤检测的加权块稀疏 Lamb wave structural health monitoring[J]. The
Intelligent Material Systems and Structures, 成像法 [J]. 西安交通大学学报 , 2019, 53(6): Journal of the Acoustical Society of America,
2013, 24(9): 1138–1152. 176–182. 2014, 135(3): 1231–1244.
[139] HALL J S, MICHAELS J E. XU Guanji, XU Caibin, YANG Zhibo, [160] MESNIL O, RUZZENE M. Sparse
Minimum variance ultrasonic imaging applied et al. Weighted group–sparse imaging method wave field reconstruction and source detection
to an in situ sparse guided wave array[J]. IEEE for lamb wave damage detection of CFRP using Compressed Sensing[J]. Ultrasonics,
Transactions on Ultrasonics, Ferroelectrics and laminates[J]. Journal of Xi’an Jiaotong 2016, 67: 94–104.
Frequency Control, 2010, 57(10): 2311–2323. University, 2019, 53(6): 176–182. [161] XU C B, YANG Z B, CHEN X F,
[140] H A L L J S , M C K E O N P, [150] EBRAHIMKHANLOU A, et al. A guided wave dispersion compensation
SATYANARAYAN L, et al. Minimum variance DUBUC B, SALAMONE S. Damage method based on compressed sensing[J].
guided wave imaging in a quasi–isotropic localization in metallic plate structures using Mechanical Systems and Signal Processing,
composite plate[J]. Smart Materials and edge–reflected lamb waves[J]. Smart Materials 2018, 103: 89–104.
Structures, 2011, 20(2): 025013. and Structures, 2016, 25(8): 085035. [162] G O L ATO A , S A N T H A N A M
[141] HALL J S, MICHAELS J E. [151] OSTIGUY P C, QUAEGEBEUR S, AHMAD F, et al. Multifrequency and
Adaptive dispersion compensation for guided N, MASSON P. Comparison of model–based multimodal sparse reconstruction in Lamb wave
wave imaging[C]// Proceedings of AIP damage imaging techniques for transversely based structural health monitoring[C]//Structural
Conference. Puertollano, 2012. isotropic composites[J]. Structural Health Health Monitoring and Inspection of Advanced
[142] NG C T, VEIDT M. A Lamb– Monitoring: an International Journal, 2017, Materials, Aerospace, and Civil Infrastructure
wave–based technique for damage detection in 16(4): 428–443. 2015. San Diego: International Society for
composite laminates[J]. Smart Materials and [152] HALL J S, MICHAELS J E. Optics and Photonics, 2015.
Structures, 2009, 18(7): 074006. Multipath ultrasonic guided wave imaging [163] G O L AT O A , A H M A D F ,
[143] SHARIF–KHODAEI Z, in complex structures[J]. Structural Health SANTHANAM S, et al. Multipath exploitation
ALIABADI M H. Assessment of delay– Monitoring: an International Journal, 2015, for enhanced defect imaging using Lamb
and–sum algorithms for damage detection in 14(4): 345–358. waves[J]. NDT & E International, 2017, 92: 1–9.
aluminium and composite plates[J]. Smart [153] LEVINE R M. Ultrasonic guided [164] GAO F, HUA J D, ZENG L, et al.
Materials and Structures, 2014, 23(7): 075007. wave imaging via sparse reconstruction[D]. Amplitude modified sparse imaging for damage
[144] RÉBILLAT M, MECHBAL N. Atlanta: Georgia Institute of Technology, 2014. detection in quasi–isotropic composite laminates
Damage localization in geometrically complex [154] 许成 , 张海燕 . 稀疏重建的兰姆 using non–contact laser induced Lamb waves[J].
aeronautic structures using canonical polyadic 波结构损伤定位 [J]. 声学技术 , 2016, 35(6): Ultrasonics, 2019, 93: 122–129.
decomposition of Lamb wave difference signal 537–541. [165] HAMIDI S, SHAHBAZPANAHI
tensors[J]. Structural Health Monitoring: an XU Cheng, ZHANG Haiyan. Damage S. Sparse signal recovery based imaging in the
International Journal, 2020, 19(1): 305–321. localization with Lamb waves via sparse presence of mode conversion with application
[145] DAFYDD I, SHARIF KHODAEI reconstruction[J]. Technical Acoustics, 2016, to non–destructive testing[J]. IEEE Transactions
Z. Analysis of barely visible impact damage 35(6): 537–541. on Signal Processing, 2016, 64(5): 1352–1364.
severity with ultrasonic guided Lamb waves[J]. [155] DONOHO D L. Compressed [166] HUA J D, ZENG L, GAO F, et
Structural Health Monitoring: an International sensing[J]. IEEE Transactions on Information al. Dictionary design for Lamb wave sparse
Journal, 2019,19(4):1104–1122. Theory, 2006, 52(4): 1289–1306. decomposition[J]. NDT & E International, 2019,
[146] QIU L, LIU M L, QING X L, et al. [156] FIGUEIREDO M A T, NOWAK R 103: 98–110.
A quantitative multidamage monitoring method D, WRIGHT S J. Gradient projection for sparse [167] NOKHBATOLFOGHAHAI A,
for large–scale complex composite[J]. Structural reconstruction: Application to compressed NAVAZI H M, GROVES R M. Use of delay
Health Monitoring: an International Journal, sensing and other inverse problems[J]. IEEE and sum for sparse reconstruction improvement
2013, 12(3): 183–196. Journal of Selected Topics in Signal Processing, for structural health monitoring[J]. Journal of
[147] SHAN S B, QIU J H, ZHANG 2007, 1(4): 586–597. Intelligent Material Systems and Structures,
C, et al. Multi–damage localization on large [157] HUA J D, GAO F, ZENG L, et al. 2019, 30(18–19): 2919–2931.
complex structures through an extended delay– Modified sparse reconstruction imaging of lamb [168] WILLIAMS R A, BECK M S.
and–sum based method[J]. Structural Health waves for damage quantitative evaluation[J]. Process tomography: principles, techniques,
Monitoring: an International Journal, 2016, NDT & E International, 2019, 107: 102143. and applications[M]. Oxford: Butterworth–
15(1): 50–64. [158] XU C B, YANG Z B, ZHAI Z, Heinemann, 1995.
[148] YANG B, XIANG Y X, XUAN F Z, et al. A weighted sparse reconstruction–based [169] RADON J. On the determination of
et al. Damage localization in hydrogen storage ultrasonic guided wave anomaly imaging functions from their integral values along certain
2020年第63卷第18期·航空制造技术 41
封面文章 COVER STORY
manifolds[J]. IEEE Transactions on Medical the limits of resolution imposed by scattering Ultrasonics, 2020, 101: 105988.
Imaging, 1986, 5(4): 170–176. in ray tomography[J]. Geophysics, 1991, 56(2): [194] HAY T R, ROYER R L, GAO H,
[170] HOUNSFIELD G N. Computerized 202–207. et al. A comparison of embedded sensor Lamb
transverse axial scanning (tomography): Part 1. [182] MCKEON J C P, HINDERS M K. wave ultrasonic tomography approaches for
Description of system[J]. The British Journal of Parallel projection and cross hole Lamb wave material loss detection[J]. Smart Materials and
Radiology, 1973, 46(552): 1016–1022. contact scanning tomography[J]. The Journal Structures, 2006, 15(4): 946–951.
[171] D I N E S K A , LY T L E R J . of the Acoustical Society of America, 1999, [195] ZHAO X L, GAO H D, ZHANG
Computerized geophysical tomography[J]. 106(5): 2568–2577. G F, et al. Active health monitoring of an
Proceedings of the IEEE, 1979, 67(7): 1065– [183] ROSE L R F, WANG C H. Mindlin aircraft wing with embedded piezoelectric
1073. plate theory for damage detection: Source sensor/actuator network: I. Defect detection,
[172] 王增勇 , 汤光平 , 李建文 , 等 . 工 solutions[J]. The Journal of the Acoustical localization and growth monitoring[J]. Smart
业 CT 技术进展及应用 [J]. 无损检测 , 2010, Society of America, 2004, 116(1): 154–171. Materials and Structures, 2007, 16(4): 1208–
32(7): 504–508. [184] WA N G C H , C H A N G F K . 1217.
WANG Zengyong, TANG Guangping, LI Scattering of plate waves by a cylindrical [196] LIU K, MA S, WU Z, et al. A
Jianwen, et al. Development and application of inhomogeneity[J]. Journal of Sound and novel probability–based diagnostic imaging with
industrial CT[J]. Nondestructive Testing, 2010, Vibration, 2005, 282(1–2): 429–451. weight compensation for damage localization
32(7): 504–508. [185] ROHDE A H, VEIDT M, ROSE using guided waves[J]. Structural Health
[173] JANSEN D P, HUTCHINS D A. L R F, et al. Imaging flexural in homogeneities Monitoring: an International Journal, 2016,
Lamb wave tomography[C]//Proceedings of using plate— wave diffraction tomography[C]// 15(2): 162–173.
IEEE Symposium on Ultrasonics. Honolulu, Proceedings of AIP Conference. Vancouver, 2007. [197] MOUSTAFA A, SALAMONE
1990: 1017–1020. [186] ROHDE A H, VEIDT M, ROSE S. Fractal dimension–based Lamb wave
[174] 张海燕 , 周全 , 吕东辉 , 等 . 各向 L R F, et al. A computer simulation study of tomography algorithm for damage detection in
同性薄板中横穿孔缺陷的超声兰姆波层析成 imaging flexural inhomogeneities using plate– plate–like structures[J]. Journal of Intelligent
像 [J]. 声学学报 , 2007, 32(1): 83–90. wave diffraction tomography[J]. Ultrasonics, Material Systems and Structures, 2012, 23(11):
ZHANG Haiyan, ZHOU Quan, LÜ 2008, 48(1): 6–15. 1269–1276.
Donghui, et al. Ultrasonic Lamb wave [187] CHAN E, ROSE L R F, WANG C [198] ZENG L, LIN J, HUA J, et al.
tomography of through hole flaws in isotropic H. A comparison and extensions of algorithms Interference resisting design for guided wave
thin plates[J]. Acta Acustica, 2007, 32(1): 83– for quantitative imaging of laminar damage in tomography[J]. Smart Materials and Structures,
90. plates. II. Non–monopole scattering and noise 2013, 22(5): 055017–055028.
[175] ZHAO X, ROYER R L, OWENS tolerance[J]. Wave Motion, 2016, 66: 220–237. [199] WANG D, YE L, LU Y, et al. A
S E, et al. Ultrasonic Lamb wave tomography in [188] SIMONETTI F, HUANG L. From damage diagnostic imaging algorithm based
structural health monitoring[J]. Smart Materials beam forming to diffraction tomography[J]. on the quantitative comparison of Lamb wave
and Structures, 2011, 20(10): 105002. Journal of Applied Physics, 2008, 103(10): 103110. signals[J]. Smart Materials and Structures, 2010,
[176] KAK A C, SLANEY M, WANG [189] B E L A N G E R P, C AW L E Y P, 19(6): 065008.
G. Principles of computerized tomographic SIMONETTI F. Guided wave diffraction [200] MEMMOLO V, MONACO E,
imaging[J]. Medical Physics, 2002, 29(1): 107– tomography within the born approximation[J]. BOFFA N D, et al. Guided wave propagation
108. IEEE Transactions on Ultrasonics, and scattering for structural health monitoring of
[177] HUTCHINS D A, JANSEN D P, Ferroelectrics, and Frequency Control, 2010, stiffened composites[J]. Composite Structures,
EDWARDS C. Lamb–wave tomography using 57(6): 1405–1418. 2018, 184: 568–580.
non–contact transduction[J]. Ultrasonics, 1993, [190] HUTHWAITE P, SIMONETTI F. [201] MOIX–BONET M, WIERACH P,
31(2): 97–103. High–resolution guided wave tomography[J]. LOENDERSLOOT R, et al. Damage assessment
[178] LEVENT DEGERTEKIN F, PEI Wave Motion, 2013, 50(5): 979–993. in composite structures based on acousto-
J, KHURI–YAKUB B T, et al. Insituacoustic [191] SHI F, HUTHWAITE P. Ultrasonic ultrasonics— evaluation of performance[M]//
temperature tomography of semiconductor wave–speed diffraction tomography with under Smart Intelligent Aircraft Structures (SARISTU).
wafers[J]. Applied Physics Letters, 1994, 64(11): sampled data using virtual transducers[J]. IEEE Cham: Springer International Publishing, 2016:
1338–1340. Transactions on Ultrasonics, Ferroelectrics, and 617–629.
[179] JANSEN D P, HUTCHINS D Frequency Control, 2018, 65(7): 1226–1238. [202] VA N V E L S O R J K , G A O
A, MOTTRAM J T. Lamb wave tomography [192] D R U E T T, TA S T E T J L , H, ROSE J L. Guided–wave tomographic
of advanced composite laminates containing CHAPUIS B, et al. Autocalibration method for imaging of defects in pipe using a probabilistic
damage[J]. Ultrasonics, 1994, 32(2): 83–90. guided wave tomography with under sampled reconstruction algorithm[J]. Insight— Non-
[180] MALYARENKO E V, HINDERS data[J]. Wave Motion, 2019, 89: 265–283. Destructive Testing and Condition Monitoring,
M K. Ultrasonic Lamb wave diffraction [193] CHANG M, YUAN S F, GUO F Y. 2007, 49(9): 532–537.
tomography[J]. Ultrasonics, 2001, 39(4): 269–281. Corrosion monitoring using a new compressed [203] QING X P, BEARD S, SHEN S B,
[181] WILLIAMSON P R. A guide to sensing–based tomographic method[J]. et al. Development of a real–time active pipeline
42 航空制造技术·2020年第63卷第18期
COVER STORY
封面文章
integrity detection system[J]. Smart Materials structural health monitoring of rocket engines[J]. methods and their engineering applications:
and Structures, 2009, 18(11): 115010. Journal of Intelligent Material Systems and A review of recent advances[J]. Mechanical
[204] LIU K H, WU Z J, JIANG Y Q, Structures, 2006, 17(7): 619–628. Systems and Signal Processing, 2019, 119:
et al. Guided waves based diagnostic imaging [209] MICHAELS J E. Effectiveness of 182–221.
of circumferential cracks in small–diameter in situ damage localization methods using sparse [214] ZHANG J Q, WU Z J, YANG Z Y,
pipe[J]. Ultrasonics, 2016, 65: 34–42. ultrasonic sensor arrays[C]//Proceedings of et al. Multimode guided waves–based structural
[205] GAO D Y, WU Z J, YANG L, et al. Health Monitoring of Structural and Biological defect localization longitudinally and cross–
Guide waves–based multi–damage identification Systems 2008. San Diego: International Society sectionally in T–bars[J]. Journal of Aerospace
using a local probability–based diagnostic for Optics and Photonics, 2008. Engineering, 2020, 33(4): 04020017.
imaging method[J]. Smart Materials and [210] YANG Z Y, WU Z J, ZHANG J Q, [215] 叶婷 . 飞机起落架 300M 钢材料
Structures, 2016, 25(4): 045009. et al. Acoustoelastic guided wave propagation 典型结构的健康监测研究[D]. 南京 : 南京航
[206] WANG D, YE L, SU Z Q, et al. in axial stressed arbitrary cross–section[J]. Smart 空航天大学 , 2011.
Probabilistic damage identification based on Materials and Structures, 2019, 28(4): 045013. YE Ting. Research on health monitoring
correlation analysis using guided wave signals [211] ZHOU K, XU X, ZHAO Z, et al. system for typical structure of 300M steel
in aluminum plates[J]. Structural Health Excitation of plane Lamb wave in plate–like materials of aircraft’s landing–gear[D].
Monitoring: an International Journal, 2010, 9(2): structures under applied surface loading[J]. Nanjing: Nanjing University of Aeronautics and
133–144. Smart Materials and Structures, 2018, 27(2): Astronautics, 2011.
[207] S H E E N B , C H O Y. A s t u d y 025011–025020.
on quantitative lamb wave tomogram via [212] ZHOU K, ZHOU Z H, YANG Z 通讯作者 : 武湛君,教授、博士,科技部重点研
modified RAPID algorithm with shape factor Y, et al. Axisymmetric and non–axisymmetric 发计划项目首席科学家,国家奖励办自然科
optimization[J]. International Journal of Lamb wave excitation using rectangular 学奖评审专家,研究方向为耐极端环境复合材
Precision Engineering and Manufacturing, 2012, actuators[J]. Smart Materials and Structures, 料设计与性能表征;复合材料低温贮箱;结构
13(5): 671–677. 2019, 28(11) : 115024. 健康监测理论、方法与技术;飞行器关键结构
[208] QING X P, CHAN H L, BEARD [213] YANG Y, PENG Z K, ZHANG W 设计与分析;纳米 / 智能材料与结构,E–mail :
S J, et al. An active diagnostic system for M, et al. Parameterised time–frequency analysis wuzhj@dlut.edu.cn。
[ABSTRACT] Damage diagnostic imaging approach based on ultrasonic guided waves (UGW) is a research hotspot
in the fields of non-destructive testing and structural health monitoring, which can directly reflect the damage location
and severity on aerospace structures. This paper summarizes the general process of UGW diagnostic imaging technology,
introduces the research status of eight representative imaging algorithms, including phased array imaging, space-wave
number filter imaging, inverse-time migration imaging, time reversal imaging, delay-and-sum imaging, model-based
imaging, tomographic imaging and probabilistic-based imaging, and compares the application scope, advantages and
disadvantages of these imaging algorithms. Then, the challenge and development trend of UGW diagnostic imaging
technology are discussed according to the actual industrial application.
Keywords: Ultrasonic guided waves; Diagnostic imaging; Structural health monitoring; Non-destructive testing;
Aerospace structure; Composite structures
(责编 大漠)
2020年第63卷第18期·航空制造技术 43