UNIVERSIDADE DE SÃO PAULO

ESCOLA DE ENGENHARIA DE SÃO CARLOS

DEPARTAMENTO DE ENGENHARIA DE ESTRUTURAS
SET 5816 – ANÁLISE EXPERIMENTAL DE ESTRUTURAS

ALEX MACÊDO LEITE

Análise crítica do artigo “The Non-Destructive Test of Steel Corrosion in Reinforced

Concrete Bridges Using a Micro-Magnetic Sensor”

São Carlos
2016
1 INTRODUÇÃO

Ensaio Não Destrutivos (END) correspondem a métodos utilizados para examinar

materiais sem danificá-los, podendo ser executados nas fases de produção, construção,
montagem e manutenção. Os END são bastante utilizados no controle da qualidade de
materiais e produtos, abrangendo técnicas capazes de fornecer informações relacionadas
aos defeitos de uma certa ferramenta, características mecânicas de um material, ou até do
controle da degradação em atividades de equipamentos e estruturas (ABENDI, 2016).
Alguns ensaios não-destrutivos tradicionalmente utilizados para avaliar a corrosão
em estruturas de concreto armado são: ensaios eletroquímicos, teste de ultrassom linear,
ensaio de correntes induzidas, termografia infravermelha, difração de Raio-X. Em
materiais ferromagnéticos, como o aço, os domínios magnéticos e as macro propriedades
são alteradas quando há algum defeito na superfície, como fissuração, corrosão ou
concentração de tensões. Este fenômeno torna possível a utilização de ensaios não
destrutivos envolvendo o magnetismo na avaliação da situação do aço na estrutura
(ZHANG et al., 2016).
Os END magnéticos são divididos em ativos e passivos, dependendo da
necessidade de um intenso campo magnético externo para magnetizar a amostra. A
execução do END magnético ativo requer um forte campo magnético externo para que
sejam detectadas falhas nas amostras. Exemplos comuns deste tipo de ensaio são o Ruído
Magnético de Barkhausen (RMB) e o Fuga do Campo Magnético (FCM). O END
magnético passivo surgiu posteriormente e é caracterizado por utilizar somente o campo
magnético terrestre, dispensando o uso de qualquer outro campo magnético externo
artificial. O Memória Magnética de Metal (MMM) é um dos tipos de ensaio magnético
passivo, o qual consiste na medida da fuga do campo auto magnético (FCAM) do material
analisado. A técnica do MMM tem sido utilizada para se obter resultados qualitativos a
respeito de regiões da estrutura em estado crítico (ZHANG et al.,2016).
Neste trabalho é analisado um artigo recente envolvendo um novo método de END
para detectar a corrosão do aço, com base na técnica MMM. Inicialmente, é apresentada
uma breve pesquisa bibliográfica contendo algumas técnicas convencionais de END
envolvendo magnetização (FCM, RMB e MMM). Em seguida, são apresentados o
procedimento experimental e a análise numérica contidos no artigo e seus principais
resultados e conclusões. Por fim, é realizada uma análise crítica, contendo alguns
comentários a respeito do artigo abordado.
2 FUGA DO CAMPO MAGNÉTICO (FCM)

O princípio do FCM é que qualquer descontinuidade geométrica existente em uma

amostra de material ferromagnético, submetida a um forte campo magnético externo,
ocasionará fuga do campo, conforme ilustrado na Figura 1. O vazamento do fluxo pode
ser monitorado por um sensor de campo magnético e utilizado para avaliar a extensão do
defeito. Apesar da fuga do campo magnético ser de fácil compreensão, sua análise
abrange complexas interações entre a excitação do campo, a fuga do fluxo e o defeito do
material, o que dificulta obter conclusões a respeito de uma variável com base em outra
(WANG, Z.W.; GU; WANG, Y.S., 2012).

Figura 1 – Representação da fuga do fluxo magnético na presença de uma

descontinuidade geométrica.

Fonte: Adaptado de Wang, Z.W; Gu e Wang, Y.S. (2012).

A fuga do campo magnético pode ser representada por dois modelos. O modelo
de Forster foi idealizado de acordo com uma visão macroscópica do fenômeno,
considerando a alteração de parâmetros magnéticos, como a permeabilidade, no local da
falha. Já o modelo de Zatsepin–Scherbinin apresenta uma visão microscópica do assunto,
assumindo dipolos magnéticos na superfície da falha. Estes dois modelos consideram o
comportamento magnético linear e podem ser utilizados para caracterizar a FCM em
defeitos regulares (WANG, Z.W.; GU; WANG, Y.S., 2012).
Um dos desafios da técnica FCM é de como identificar os defeitos da amostra a
partir dos sinais FCM obtidos do ensaio. Os diversos parâmetros que caracterizam a falha,
como sua largura, sua espessura e sua localização, possuem efeitos significativos nos
sinais da FCM, o que dificulta fazer uma ligação entre tais parâmetros e os sinais da FCM.
Além disso, o comportamento da zona elasto-plástica perto do defeito é geralmente
ignorado, sendo adotadas as propriedades idealizadas do material. Isto ocorre devido à
pouca compreensão a respeito da deformação plástica de materiais ferromagnéticos.

3 RUÍDO MAGNÉTICO DE BARKHAUSEN (RMB)

O efeito Barkhausen está relacionado à variação da magnetização de um material

ferromagnético quando este é exposto a um campo magnético variado gradativamente
(CHAVEZ, 2009). Ao aplicar um campo magnético variável em um material
ferromagnético, ocorre a movimentação das paredes de domínio magnético. Caso haja
algum defeito na estrutura do material, a movimentação destas paredes é interrompida e,
ao modificar o campo magnético aplicado, há um “salto” das paredes de domínio nestes
defeitos, o que caracteriza o Ruído Magnético de Barkhausen (RMB) (ANTONIO et al.,
2013).
Em 1919, Barkhausen realizou experimentos submetendo amostras de diferentes
materiais a variações de campo magnético. Para isto, foram utilizadas duas bobinas: uma
bobina de excitação magnética, responsável pelo fornecimento de variação do campo, e
outra bobina leitora de RMB, a qual fica em torno da amostra ensaiada. O esquema deste
ensaio é mostrado na Figura 2. Foi observado que as amostras ferromagnéticas emitiam
um ruído quando havia um aumento do campo magnético a que estavam expostas e tal
ruído era capturado pela bobina leitora em forma de pulsos de voltagem (WANG, Z.W.;
GU; WANG, Y.S., 2012).

Figura 2 – Esquema do ensaio do Ruído Magnético de Barkhausen.

Fonte: LADIN - Laboratório de Dinâmica e Instrumentação (2016).

 O RMB é consequência do movimento irreversível das paredes do domínio
magnético de 180º e ocorrem nas regiões de maior variação do campo magnético
(ANTONIO et al., 2013). Vários fatores afetam a amplitude do RMB, como:
microestrutura do material, tensões residuais e aplicadas, entre outros. Dependendo do
tipo de esforço aplicado no material, o movimento das paredes de domínio é modificado
e a amplitude do RMB é diferente. Estudos comprovaram que determinado aço submetido
a esforços de tração na mesma direção da magnetização apresentaram maior variação do
RMB do que a mesma amostra submetida a tensões de compressão. Isto ocorreu por causa
da maior mobilidade das paredes de domínio em 180º no ensaio de tração, enquanto que
no ensaio de compressão ocorreu redução desta mobilidade (WANG, Z.W.; GU; WANG,
Y.S., 2012).
É importante lembrar que os sinais RMB são bastante vulneráveis à microestrutura
do material. O contorno dos grãos constitui um obstáculo intransitável para a
movimentação das paredes de domínio. Grãos maiores contribuem para uma menor
emissão de RMB, devido à redução dos defeitos estruturais que dificultam o movimento
das paredes de domínio (ANTONIO et al., 2013).

4 MEMÓRIA MAGNÉTICA DE METAL (MMM)

O MMM (Memória Magnética de Metal) corresponde a um método de ensaio não-

destrutivo passivo. Diferente dos métodos magnéticos ativos, que consomem muito
tempo e são inviáveis em algumas estruturas irregulares, o MMM se apresenta como um
método de ensaio mais simples para atender às exigências da engenharia, como a detecção
de defeitos em materiais. Desta forma, esta técnica tem ganhado relevância nos últimos
anos, possuindo a vantagem de utilizar o campo magnético terrestre como excitação da
amostra ensaiada (WANG,Z.W.; GU; WANG, Y.S., 2011).
O ferromagnetismo está relacionado às forças de troca que preservam os átomos
em suas posições na malha metálica. Em alguns metais, como ferro e níquel, tanto a
energia de troca quanto as energias de magnetização ordenam os átomos em grupos
microscópicos denominados domínio. Em cada domínio, a magnetização é saturada, o
que significa que há o alinhamento das rotações dos elétrons internos dos átomos. O
domínio funciona magneticamente semelhante a um átomo, podendo girar, crescer ou
reduzir de tamanho (SUÁREZ-ÁNTOLA, R.; SUÁREZ-BAGNASCO, D., 2004).
 Antes de ser submetido ao processo de magnetização, o material apresenta
domínios individuais com direções distribuídas aleatoriamente. Após a aplicação de um
campo magnético externo, dois fenômenos independentes ocorrem no material. Primeiro,
há um crescimento dos domínios magnéticos orientados na direção do campo aplicado,
enquanto os demais domínios, orientados nas outras direções, são reduzidos. Já quando a
magnetização está próxima da saturação, ocorre a rotação da direção de magnetização dos
domínios na orientação do campo aplicado (SUÁREZ-ÁNTOLA, R.; SUÁREZ-
BAGNASCO, D., 2004).
Quando um material ferromagnético é exposto ao campo magnético terrestre e a
uma determinada carga mecânica, ocorre a produção de sinais de fuga do campo auto-
magnético (FCAM) em regiões com concentração de tensões. Isto acontece por causa da
orientação fixa de domínios magnéticos devido a deformação plástica na região com
maiores concentrações de tensão. Enquanto a componente tangencial Hp(x) da FCAM
adquire valor máximo nestas regiões, a componente normal Hp(y) da FCAM possui valor
nulo, conforme exibido na Figura 3. Com isto, torna-se possível identificar zonas de
concentração de tensões em materiais ferromagnéticos a partir da análise da FCAM
(WANG, Z.W.; GU; WANG, Y.S., 2012).

Figura 3 – Representação da distribuição da FCAM em uma zona de concentração de

tensão.

Fonte: Wang, Z.W; Gu e Wang, Y.S. (2012).

A FCAM ocorre principalmente em duas situações. A primeira é quando a

temperatura do material ferromagnético é superior à Temperatura de Curie e ocorre uma
queda de temperatura até a temperatura ambiente. Com isso, há uma mudança nos padrões
de domínio magnéticos devido à magnetização não-uniforme resultante da
heterogeneidade estrutural. Outra situação se refere às regiões que apresentam
deslocamentos devido à deformação plástica. Para que a FCAM apresente resultados
mensuráveis, é importante que a densidade da energia magnética devido às deformações
locais seja da mesma ordem de grandeza da densidade da energia magnética do material
ensaiado. Quando a intensidade do campo magnético externo é elevada, a densidade da
energia magnética da amostra se torna alta e a FCAM pode passar despercebida nas falhas
estruturais (SUÁREZ-ÁNTOLA, R.; SUÁREZ-BAGNASCO, D., 2004). A Figura 4
mostra uma parte de um equipamento utilizado para medir a FCAM em materiais
ferromagnéticos e algumas de suas divisões.

Figura 4 – 1, 2 e 3: Transdutores fluxgate que medem a FCAM mais o campo magnético

externo; 4: transdutor fluxgate que mede apenas o campo magnético externo; 5: rodas e
medidor de comprimento; Δlb: distância entre transdutores.

Fonte: Suárez-Ántola e Suárez-Bagnasco (2004).

Devido às diversas vantagens que apresenta, como fácil operação e rápida

execução, o método MMM tem sido utilizado em diversas aplicações relacionadas a
diagnósticos de gasodutos e oleodutos, vasos de pressão, trilhos, entre outros. Um
exemplo de aplicação da técnica MMM é mostrado na Figura 5, que representa o ensaio
realizado em três diferentes anéis. No anel à esquerda, foi observado uma menor variação
da componente normal Hp(y) da FCAM, indicando boas condições de operação. Já no
anel intermediário, a amplitude do Hp(y) foi mais considerável, apontando deterioração
do material. No anel à direita houve uma variação perceptível da amplitude do Hp(y), o
que sugere a existência de zonas com concentração de tensões (WANG, Z.W.; GU;
WANG, Y.S., 2012).

Figura 5 – Resultados da componente normal Hp(y) da FCAM do ensaio MMM realizado

em três anéis.

Fonte: Wang, Z.W; Gu e Wang, Y.S. (2012).

Por se tratar de um método de ensaio desenvolvido recentemente, o MMM possui

algumas limitações. Uma delas está relacionada ao fato de que os resultados obtidos a
partir desta técnica são apenas qualitativos, sendo utilizada para indicar regiões do
material que possuem algum defeito, como concentração de tensões. Na última década,
os estudos envolvendo os sinais da FCAM e as características do material têm sido
aprimorados devido às seguintes razões (WANG, Z.W.; GU; WANG, Y.S., 2012):
- Carência de modelos físicos que avaliem a relação entre a deformação plástica e os
sinais da FCAM. Apenas modelos elásticos-magnéticos foram desenvolvidos para
analisar danos iniciais, sendo estes inadequados no estudo de deformações plásticas
(WANG, Z.W.; GU; WANG, Y.S., 2012);
- Ausência de parâmetro quantitativo para estimar as dimensões do defeito a partir dos
sinais da FCAM, os quais são utilizados apenas para identificar zonas do material com
alguma falha (WANG, Z.W.; GU; WANG, Y.S., 2012).
5 PROCEDIMENTO EXPERIMENTAL

Os materiais utilizados no procedimento experimental foram:

- Barra de aço com comprimento de 150 cm;
- Solução de 5% de NaCl;
- Equipamento para medição do campo magnético, composto por um sistema de
deslocamento mecânico 3D e um sensor micro magnético de alta precisão do tipo HMR
2300 capaz de medir a intensidade do campo magnético nas três direções ortogonais,
conforme ilustrado na Figura 6;

Figura 6 – Dispositivo utilizado no ensaio para medição da FCAM 3D, com especificação
do micro sensor no canto superior direito da ilustração.

Fonte: Zhang et al. (2016).

O nível de corrosão foi vistoriado de acordo com o desgaste da corrente elétrica e

com o tempo de sua passagem, segundo a lei de Faraday. Já o comprimento da corrosão
foi monitorado pela extensão da região infiltrada pela solução de NaCl. A Figura 7 contém
um esquema do ensaio executado.
Foram avaliadas duas amostras de concreto armado submetidas a uma corrente
elétrica de intensidade 0,5A e com período de corrosão de 96h, 120h e 144h. Os
comprimentos de corrosão foram de 20cm para a amostra 1 e de 15cm para a amostra 2.
Foram registrados os campos magnéticos em 4 diferentes alturas, como exibido na Figura
8.

Figura 7 – Esquema do ensaio executado.

Fonte: Zhang et al. (2016).

Figura 8 – Localização e trajeto do micro sensor durante o ensaio.

Fonte: Zhang et al. (2016).

6 MODELO TEÓRICO – ANÁLISE NUMÉRICA

Para simular o comportamento da FCAM na região de corrosão, foi utilizado um

modelo linear de redistribuição da carga magnética. No caso de materiais
ferromagnéticos, pode-se considerar o campo magnético externo gerado a partir de uma
carga magnética, representada por ρ (Equação 1).
   M (1)
Onde:
ρ = densidade de carga magnética;
M = magnetização.

A magnetização é expressa por:

M  (r 1) H mL (2)

Onde:
μr = permeabilidade magnética relativa;
HmL = campo de Weiss (campo efetivo responsável pela auto magnetização do
material ferromagnético).

Considerando HmL fixo, a intensidade do campo magnético fica em função apenas

de permeabilidade magnética, cuja variação é capaz de alterar a magnetização. Quando
um elemento de aço é submetido a um esforço local, como uma tensão mecânica, a
permeabilidade magnética do material é alterada nesta região, modificando a carga
magnética e, consequentemente, o campo magnético local. Neste trabalho, foi
considerado um modelo simplificado, com uma distribuição de carga linear.
Nesta análise numérica foi assumido um defeito com abertura 2b e profundidade h
em relação à superfície da barra devido à corrosão, como ilustrado na Figura 9. Nas duas
extremidades da região de corrosão há a concentração da carga magnética deslocada,
indicando alta densidade de carga magnética. Já na região de corrosão inferior ocorre
forte pressão na barra de aço, o que gera concentração de tensões locais e redistribuição
da carga magnética. Desta forma, a distribuição da carga magnética na região de corrosão
é expressa pela Equação 3.
Figura 9 – Esquema de distribuição de carga magnética e pressão de corrosão em região
corroída de uma barra de aço envolvida no concreto.

Fonte: Zhang et al. (2016).


  max , ( x  b, h  y  0)

 ( x, y )    max , ( x  b, h  y  0) (3)
 x
  max , (b  x  b, y  h)
 b
Onde:
ρmax = densidade máxima de carga magnética.

A FCAM em determinado ponto do espaço, de coordenadas (x,y), gerada a partir

de elementos carregados magneticamente das três bordas da região de corrosão, conforme
exibido na Figura 10, é dada pelas Equações 4, 5 e 6.

Figura 10 – FCAM em determinado ponto (x,y) devido à carga magnética.

Fonte: Zhang et al. (2016).

  ( x ', y ')dy '
dH1 ( x, y)  r1 (4)
20 r12

  ( x ', y ')dy '

dH 2 ( x, y )  r2 (5)
20 r2 2

 ( x ', y ')dx '

dH 3 ( x, y )  r3 (6)
20 r32

Onde:
μ0 = permeabilidade magnética no ar;
r1, r2 e r3 = vetores espaciais do elemento carregado até um ponto do espaço.

A partir da integração das Equações 4, 5 e 6 são obtidas as componentes x e y da

FCAM em determinado ponto do espaço (x,y), conforme disposto nas Equações 7 e 8 .

max   yh  y   max   yh  y 

H x ( x, y )   arctan    arctan     arctan    arctan  
20   xb   x  b   20   x b   x b 
(7)
    bx   b  x   max  x (b  x)2  ( y  h)2 
 max  2b  ( y  h)  arctan    arctan      ln 2 
20b    y  h   y  h   2 0 b  2 (b  x)  ( y  h) 
2

max ( x  b) 2  ( y  h) 2 max ( x  b) 2  ( y  h) 2
H y ( x, y )  ln  ln
40 ( x  b) 2  y 2 40 ( x  b) 2  y 2
(8)
 max  ( y  h) (b  x) 2  ( y  h) 2   max   bx b  x  
  ln `  x  arctan  arctan 
20b  2 (b  x) 2  ( y  h) 2  20b   yh y  h 

7 ANÁLISE DOS RESULTADOS

As distribuições das componentes tangenciais do campo auto magnético (Hx)

obtidas nos ensaios das duas amostras são exibidas nas Figuras 11 e 12. Para as amostras
não corroídas, as curvas de distribuição do campo obtidas nas diferentes posições do
sensor são quase paralelas entre si (Figura 11a e 12a), não havendo cruzamento entre as
curvas em nenhum ponto. As Figuras 11b-d e 12b-d indicam as curvas de distribuição do
Hx da amostra 1 e da amostra 2, respectivamente, de acordo com a posição do sensor, em
diferentes tempos de corrosão. Para determinado período de corrosão, as curvas do Hx,
obtidas em diferentes alturas, se cruzam em dois pontos. Para períodos de corrosão
maiores, o valor da distância entre os pontos de cruzamento das curvas fica mais próximo
do comprimento de corrosão real da amostra. Desta forma, no maior tempo de corrosão
analisado (144 h), a distância entre os pontos de cruzamento das curvas foi de 18,5 cm
para a amostra 1 e de 14,5 para a amostra 2, valores próximos aos comprimentos de
corrosão de 20 cm e 15 cm para as respectivas amostras.

Figura 11 – Distribuição da componente tangencial Hx da FCAM na amostra 1 em

diferentes posições do micro sensor: (a) amostra não corroída, (b) corroída por 96 h, (c)
corroída por 120 e (d) corroída por 144 h.

Fonte: Zhang et al. (2016).

Figura 12 – Distribuição da componente tangencial Hx da FCAM na amostra 2 em
diferentes posições do micro sensor: (a) amostra não corroída, (b) corroída por 96 h, (c)
corroída por 120 e (d) corroída por 144 h.

Fonte: Zhang et al. (2016).

Após a realização dos ensaios, a camada de concreto foi removida e a barra de aço
foi exposta para avaliação da região de corrosão real, conforme exibido na Figura 13. Foi
verificado que não houve a remoção completa das nervuras da barra de aço na região
corroída, o que indica a pressão da corrosão como principal fator nos sinais da fuga do
campo auto magnético e não a remoção do material.
Figura 13 – Barras de aço no final do ensaio, após a remoção da camada de concreto: (a)
amostra 1 e (b) amostra 2.

Fonte: Zhang et al. (2016).

A distribuição da componente tangencial Hx da FCAM, calculada a partir da

Equações 7 e 8, para diferente alturas (y= 2 cm, 4 cm, 6 cm, 8 cm, 10 cm), indica um
comportamento semelhante ao obtido experimentalmente. Nos três casos analisados, com
comprimentos de corrosão de 10 cm, 20 cm e 40 cm, as curvas determinadas em diferentes
posições se cruzam em dois pontos e a distância entre estes pontos se aproxima do valor
do comprimento da região corroída, conforme exibido na Figura 14. Desta forma,
comprimentos de corrosão maiores implicam em pontos de cruzamento mais afastados.
O valor máximo absoluto do Hx ocorre no centro da zona de corrosão.
Figura 14 – Curvas da componente Hx da FCAM obtidas a partir da análise numérica,
considerando comprimento de corrosão de (a) 10 cm, (b) 20 cm e (c) 40 cm.

Fonte: Zhang et al. (2016).

8 CONCLUSÕES

Nesta pesquisa, foram obtidos experimentalmente os sinais da componente

tangencial Hx da FCAM em uma amostra composta por uma barra de aço imersa em
concreto submetida ao processo de corrosão. Para isto, foi utilizado um sensor micro
magnético. Foi observado que as curvas x-Hx, determinadas a partir de diferentes alturas
do sensor, se cruzaram em dois pontos e a distância entre tais pontos se aproximou do
comprimento de corrosão à medida que o tempo de exposição ao processo corrosivo foi
maior. Adotando o modelo de carga magnética linear, foi realizada uma análise numérica,
obtendo também a interseção das curvas x-Hx, conforme observado nos resultados
experimentais. Estes resultados indicam que a corrosão favorece a resistividade
magnética do material, promovendo concentração de carga magnética na região
danificada, o que ocasiona variação do fluxo magnético. Desta forma, o presente artigo
mostrou um novo método de ensaio não-destrutivo, considerado simples e eficiente, para
identificar a corrosão no interior de estruturas de concreto armado através de sensor micro
magnético de alta resolução.

9 ANÁLISE CRÍTICA E COMENTÁRIOS

 Os autores não apresentam justificativas para a utilização dos materiais. Por

exemplo, não indicam o porquê da utilização de solução de 5% de NaCl ou da
intensidade da corrente elétrica utilizada;
 Os autores não especificam alguns materiais empregados no ensaio, como o
concreto utilizado, qual o tipo de aço compõe a barra experimentada e o seu
diâmetro, impossibilitando outros pesquisadores de repetir o experimento com
condições semelhantes;
 Não são citados alguns cuidados que foram tomados ou, pelo menos, deveriam ter
sido tomados durante o experimento, como: o sensor deve estar próximo da região
da amostra com variação da FCAM, de modo a evitar a influência do ruído de
fundo na leitura da FCAM. Ao longo do artigo não é mencionada a profundidade
da posição do micro sensor em relação à amostra, o que deve refletir nos sinais da
FCAM medidos;
 Sabe-se que o nível de excitação e a homogeneidade do campo auto magnético
irão refletir na sua variação ao longo de um defeito. No artigo analisado, não é
exposto o nível de excitação do campo auto magnético da amostra;
 Observação: Considerando que o nível de corrosão de uma barra é dado pela
distância entre os pontos de cruzamento das curvas x-Hx em relação ao
comprimento de corrosão da amostra, a taxa de corrosão inicial em amostras com
maiores comprimentos de corrosão foi maior do que em amostras com
comprimentos de corrosão inferiores. Para a amostra 1 e tempo de corrosão de 96
h, a distância entre os pontos de cruzamento das curvas x-Hx obtidas
experimentalmente foi de 160 mm, valor correspondente a 80% do comprimento
de corrosão de 200 mm da amostra. Já para a amostra 2 e tempo de corrosão
similar, a distância entre tais pontos foi de apenas 100 mm, equivalente a 66,7%
do comprimento de corrosão de 150 mm da amostra. No tempo de corrosão final
de 144 h, ocorreu o oposto. Enquanto a distância entre os pontos de cruzamento
das curvas da amostra 1 foi de 185 mm, 92,5% de 200 mm, na amostra 2 o valor
 desta distância foi de 145 mm, 96,7% de 150 mm. A Figura 15 mostra um gráfico
da taxa de corrosão da armadura em função do tempo para as duas amostras
ensaiadas;

Figura 15 – Taxa de corrosão em função do tempo de corrosão para as amostras 1 e 2.

% de corrosão X Tempo de corrosão

120,00
100,00
% de corrosão

80,00
60,00 Amostra 1
Amostra 2
40,00
20,00
0,00
0 50 100 150 200
Tempo de corrosão (h)

Fonte: Elaborada pelo autor (2016)

 Enquanto nos resultados experimentais as curvas x-Hx em diferentes posições são

quase paralelas entre si nas regiões não corroídas, na análise numérica estas curvas
se sobrepuseram nestas regiões, conforme destacado na
Figura 16. Vale ressaltar que o experimento é uma atividade que envolve a posição
em 3D do micro sensor, já a análise numérica foi realizada considerando um
plano, ou seja, uma situação 2D. Além disso, os resultados experimentais podem
ter sido influenciados pelo ruído de fundo, conforme mencionado anteriormente;

Figura 16 – Divergência entre os comportamentos das curvas x-Hx nas regiões não
corroídas no resultado experimento e no resultado numérico.

Fonte: Elaborada pelo autor (2016).

 Para o comprimento de corrosão de 20 cm, o resultado experimental indicou uma
distância entre os pontos de interseção das curvas x-Hx de 18,5 cm no tempo de
exposição à corrosão de 144 h. Já o resultado da análise numérica apontou uma
distância entre os referidos pontos de 19,6 cm, o que sugere que os resultados
desta análise são referentes a um tempo de corrosão da amostra superior ao
máximo de 144 h considerado no ensaio experimental.
 O método utilizado no artigo deixa de ser interessante quando se possui um feixe
de barras de aço no concreto, como acontece geralmente nas estruturas de concreto
armado. Nestes casos, é possível que haja uma interferência da FCAM devido ao
processo corrosivo das diversas barras de aço, dificultando a determinação do
comprimento de corrosão de cada delas.
 REFERÊNCIAS BIBLIOGRÁFICAS

ANTONIO, P. P. C.; DE CAMPOS, M. F.; DIAS, F. M. S.; CAPO-SANCHEZ, J.;

PADOVESE, L R. Ruído Magnético Barkhausen em aços alta resistência baixa liga, IF
ao titânio e C-Mn levemente deformados plasticamente. In: 7° COBEF - Congresso
Brasileiro de Engenharia de Fabricação, 2013, Penedo, Itatiaia, RJ. Anais do 7°
COBEF - Congresso Brasileiro de Engenharia de Fabricação, 2013. p. 1-7.
ASSOCIAÇÃO BRASILEIRA DE ENSAIOS NÃO DESTRUTIVOS E INSPEÇÃO.
Ensaios não destrutivos e inspeção. Disponível em: <
http://www.abendi.org.br/abendi/default.aspx?mn=630&c=17&s=>. Acesso em: 04
nov. 2016.
LABORATÓRIO DE DINÂMICA E INSTRUMENTAÇÃO. Ensaios não destrutivos.
Ruído Magnético de Barkhausen. Disponível em:
<http://www.ladin.usp.br/END_BARKHAUSEN.html> . Acesso em: 08 nov. 2016.
SUÁREZ-ÁNTOLA, R.; SUÁREZ-BAGNASCO, D. The physical foundations of the
Magnetic Memory Method of NDT. In: Segundas Jornadas Técnicas de la Asociación
Uruguaya de Ensayos No Destructivos, 2004, Montevideo, Uruguay.
WANG, Z. D.; GU, Y.; WANG, Y. S. A review of three magnetic NDT technologies.
Journal of Magnetism and Magnetic Materials, Amsterdã, v. 324, n. 4, p. 382-388,
fev./2012.
ZHANG, H.; LIAO, L.; ZHAO, R.; ZHOU, J.; YANG, M.; XIA, R. The Non-
Destructive Test of Steel Corrosion in Reinforced Concrete Bridges Using a Micro-
Magnetic Sensor. Sensors, Basileia, Suiça, v. 16, n. 1439, p. 1-9, set./2016.
 sensors
Article
The Non-Destructive Test of Steel Corrosion
in Reinforced Concrete Bridges Using a
Micro-Magnetic Sensor
Hong Zhang 1 , Leng Liao 2 , Ruiqiang Zhao 2 , Jianting Zhou 1, *, Mao Yang 3 and Runchuan Xia 1
1 College of Civil Engineering, Chongqing Jiaotong University, Chongqing 400074, China;
superww058@163.com (H.Z.); Runchuan727@163.com (R.X.)
2 School of Materials Science and Engineering, Chongqing Jiaotong University, Chongqing 400074, China;
lengliao@cqjtu.edu.cn (L.L.); rqzhao@cqjtu.edu.cn (R.Z.)
3 College of Civil Engineering, Chongqing Three Gorges University, Chongqing 404100, China;
yangmao199104@163.com
* Correspondence: jtzhou@cqjtu.edu.cn or jt-zhou@163.com; Tel./Fax: +86-23-6265-2702

Academic Editor: Kenneth J. Loh

Received: 28 July 2016; Accepted: 2 September 2016; Published: 6 September 2016

Abstract: This paper presents a non-destructive test method for steel corrosion in reinforced
concrete bridges by using a 3-dimensional digital micro-magnetic sensor to detect and analyze
the self-magnetic field leakage from corroded reinforced concrete. The setup of the magnetic scanning
device and the measurement mode of the micro-magnetic sensor are introduced. The numerical
analysis model is also built based on the linear magnetic charge theory. Compared to the self-magnetic
field leakage data obtained from magnetic sensor-based measurement and numerical calculation,
it is shown that the curves of tangential magnetic field at different lift-off height all intersect near the
edge of the steel corrosion zone. The result indicates that the intersection of magnetic field curves
can be used to detect and evaluate the range of the inner steel corrosion in engineering structures.
The findings of this work propose a new and effective non-destructive test method for steel corrosion,
and therefore enlarge the application of the micro-magnetic sensor.

Keywords: steel corrosion; non-destructive test; micro-magnetic sensor; self-magnetic flux leakage;
numerical simulation

1. Introduction
Reinforced concrete is the most popular construction material in the world and widely used to
bridge construction due to the advantage of strong bearing, low-cost, and easy construction. However,
a major issue for reinforced concrete structures is the corrosion of the reinforcement steel bars exposed
to aggressive environmental conditions, such as a humid, saline-alkaline climate. The steel corrosion
makes a significant contribution to the failure of an engineering structure. Approximately 40% of
damages of engineering structure result from the steel corrosion and there is a loss of ~14 billion
dollars per year just in the United States. Therefore, the evaluation of corrosion in reinforced concrete
is very important for the management and maintenance of engineering structures. The traditional
nondestructive test techniques, such as electrochemical method [ 1], linear ultrasonic testing (UT) [2],
eddy current testing (ECT) [3], infrared thermography (IRT) [4], and X-ray diffraction (XRD) [5], are all
time-consuming technology or need expensive equipment for determining corrosion in the depth of
reinforced concrete.
Reinforcing steel bars are made from a typical ferromagnetic material, which is a very important
component of the reinforcing concrete structure. A defect (e.g., cracking, fatigue failure, stress
concentration and corrosion) in ferromagnetic materials can change the structure of magnetic domains

Sensors 2016, 16, 1439; doi:10.3390/s16091439 www.mdpi.com/journal/sensors

Sensors 2016, 16, 1439 2 of 9

and their macro-properties are consequently changed, such as magnetoconductivity, coercive force,
and hysteresis. This effect can be used for different applications in material science, especially in
the field of nondestructive testing by using magnetic measurement to evaluate the status of the steel
material in engineering structures, in particular to detect the stress-concentration, corrosion, and
failure-originated zones. A number of nondestructive magnetic techniques have been developed, such
as magnetic Barkhausen emission (MBE) [6], magnetoacoustic emission (MAE) [7], stress-induced
magnetic anisotropy (SMA) [8], and magnetic field leakage (MFL) [ 9–13]. The physical mechanics of
these techniques requires a strong magnetic field to magnetize the specimen for testing, and then detect
induced-related phenomena to evaluate the health status of a specimen, so these techniques could be
called active magnetic test methods. On the other hand, the shape and density of stimulated magnetic
fields differ from the characteristic of tested objects. These techniques are time-consuming and even
difficult to operate for some irregular structures [14]. For meeting the requirement of developing a
more simple and effective magnetic technique in engineering, a positive magnetic method called metal
magnetic memory (MMM) technique was proposed by A. A. Doubov in 1997 [ 15]. The advantage of
MMM technique is that the Earth’s magnetic field instead of an artificial strong field is used as the
stimulus source. Under the effect of the earth field and mechanical loads/defects, the self-magnetic
field leakage (SMFL) signals are generated from corroded or stress-concentration regions where the
tangential component reaches a maximum and the normal component transfers its polarity and has a
zero value. The MMM technique is suitable for many engineering practices [16–18], but up to now has
only been used as a qualitative test technique to determine the possible dangerous positions without
quantitative results. The more accurate and quantitative criteria are deficient and required for the
MMM technique.
In the present work, we introduce a passive magnetic test method to detect and evaluate the
corrosion of reinforcement steel bars in the reinforced concrete structure by detecting the SMFL from
the corrosion region. When a local corrosion happens, it brings a strong local corrosion pressure for
expanding of corrosion products, which breaks the magnetic continuity and improves the magnetic
resistivity of corrosion parts for the material loss and the magnetic-stress coupling [16,19]. The magnetic
permeability is accordingly changed because of the stress-magnetic effect of iron as a ferromagnetic
material. As a consequence, the SMFL is generated from the corrosion zone. By detecting and analyzing
the SMFL signal produced by the corrosion, the position and status of corrosion in the reinforced
concrete structure can be determined. This method does not need the equipment to actively excite a
magnetic field, it is an effective, time-saving, and easy-operation method to non-destructively test the
corrosion in the reinforced concrete structure.
The paper is organized as follows. In Section 2, the experimental setup and theoretical model
based on the micro-magnetic sensor are introduced. The experimental results and calculated data of
the SMFL signals are shown in Section 3, and the quantitative relationship between leaked magnetic
field and steel corrosion is given. In Section 4, we summarize all these results.

2. Experimental Setup and Theoretical Model Based on Micro-Magnetic Sensor

2.1. Experiment Details Based on Micro-Magnetic Sensors

The corroded specimens of reinforced concrete are prepared by electrochemical method as shown
in Figure 1. A steel bar is enwrapped by concrete, whose thickness T = 3 cm and the steel bar length
L = 150 cm. The corrosion level can be controlled by the corroded current and time according to the
Faraday law and the corrosion length can be controlled by the size of region infiltrated by electrolyte
(5% NaCl solution).
Figure 2 shows the self-designed 3-dimensional (3D) scanning device for magnetic field
measurement based on the 3D mechanical displacement system and the high-precision micro-magnetic
sensor. The mechanical displacement system consists of 3D aluminum track and bracket system,
stepping motor driver, and a hollow bar for supporting the sensor. The displacement accuracy of
 Sensors 2016, 16, 1439 3 of 10

Sensors 2016, 16, 1439 3 of 9

the stepping motor driver is as high as 0.1 mm. To improve the stability of the supporting bar when
moving, a cable-stayed-like structure was used in the scanning device. The Honeywell HMR 2300
magnetometer is employed as the micro-magnetic sensor. It is a three-axis smart digital magnetometer,
and the three axes oriented in orthogonal directions of HMR 2300 can measure the X, Y and Z vector
components of a magnetic field. The output range of this micro-magnetic sensor is ±2 Gs with a
resolutionSensors
to less
2016, than
16, 143970 µGs. 3 of 10

Figure 1. The schematic of electrochemical corrosion.

Figure 2 shows the self-designed 3-dimensional (3D) scanning device for magnetic field
measurement based on the 3D mechanical displacement system and the high-precision
micro-magnetic sensor. The mechanical displacement system consists of 3D aluminum track and
bracket system, stepping motor driver, and a hollow bar for supporting the sensor. The
displacement accuracy of the stepping motor driver is as high as 0.1 mm. To improve the stability of
the supporting bar when moving, a cable-stayed-like structure was used in the scanning device.
The Honeywell HMR 2300 magnetometer is employed as the micro-magnetic sensor. It is a
three-axis smart digital magnetometer, and the three axes oriented in orthogonal directions of HMR
2300 can measure the X, Y and Z vector components of a magnetic field. The output range of this
micro-magnetic sensor Figure
is ±2
Figure Gs 1. The
1. The
with schematic of
schematic
a resolutionoftoelectrochemical
electrochemical corrosion.
less than 70 μGs. corrosion.
Figure 2 shows the self-designed 3-dimensional (3D) scanning device for magnetic field
measurement based on the 3D mechanical displacement system and the high-precision
micro-magnetic sensor. The mechanical displacement system consists of 3D aluminum track and
bracket system, stepping motor driver, and a hollow bar for supporting the sensor. The
displacement accuracy of the stepping motor driver is as high as 0.1 mm. To improve the stability of
the supporting bar when moving, a cable-stayed-like structure was used in the scanning device.
The Honeywell HMR 2300 magnetometer is employed as the micro-magnetic sensor. It is a
three-axis smart digital magnetometer, and the three axes oriented in orthogonal directions of HMR
2300 can measure the X, Y and Z vector components of a magnetic field. The output range of this
micro-magnetic sensor is ±2 Gs with a resolution to less than 70 μGs.

Figure 2. The 3D scanning device for magnetic field measurement based on the micro-magnetic
Figure 2. The 3D scanning device for magnetic field measurement based on the micro-magnetic sensor.
sensor. The inset shows the model of the micro-magnetic sensor.
The inset shows the model of the micro-magnetic sensor.
Two reinforced concrete samples are corroded under a current intensity of 0.5 A and the
Two corrosion
reinforced timeconcrete
is 96 h, 120 h, and 144
samples areh,corroded
respectively; whileathe
under corrosion
current length isof
intensity ~20 cmAand
0.5 and ~15 cmcorrosion
the
for the sample #1 and #2, respectively. The location and scanning paths of magnetic sensor
time is 96 h, 120 h, and 144 h, respectively; while the corrosion length is ~20 cm and ~15 cm for the
imbedding on the 3D scanning device are shown in Figure 3, where the Hx component of magnetic
sample #1 and
field is #2, respectively.
defined Thefield.
as a tangential location
Figure and scanning
3 shows pathspaths
four scanning of magnetic sensor
at different lift-off imbedding
height on
the 3D scanning device are shown in Figure 3, where the
(LFH) to scan the magnetic field around the reinforced concrete H component
x sample. of magnetic field is defined
as a tangential field. Figure 3 shows four scanning paths at different lift-off height (LFH) to scan the
magneticSensors
field2016, 16, 1439
around the reinforced concrete sample. 4 of 10
Figure 2. The 3D scanning device for magnetic field measurement based on the micro-magnetic
sensor. The inset shows the model of the micro-magnetic sensor.

Two reinforced concrete samples are corroded under a current intensity of 0.5 A and the
corrosion time is 96 h, 120 h, and 144 h, respectively; while the corrosion length is ~20 cm and ~15 cm
for the sample #1 and #2, respectively. The location and scanning paths of magnetic sensor
imbedding on the 3D scanning device are shown in Figure 3, where the Hx component of magnetic
field is defined as a tangential field. Figure 3 shows four scanning paths at different lift-off height
(LFH) to scan the magnetic field around the reinforced concrete sample.

3. The
FigureFigure location
3. The and
location scanning
and pathsofofthethe
scanning paths micro-magnetic
micro-magnetic sensor.sensor.

2.2. Theoretical Model for Measured Data

In order to establish a physical model for the SMFL signals from a corrosion region, a linear
model considering the redistribution of magnetic charge in the corrosion zone is presented.
Magnetic charge is an equivalent model, which has been developed to simulate the magnetic
field leakage [20–23]. For a ferromagnet, its exterior magnetic field would be considered to
Sensors 2016, 16, 1439 4 of 9

2.2. Theoretical Model for Measured Data

In order to establish a physical model for the SMFL signals from a corrosion region, a linear model
considering the redistribution of magnetic charge in the corrosion zone is presented. Magnetic charge
is an equivalent model, which has been developed to simulate the magnetic field leakage [20–23]. For a
ferromagnet, its exterior magnetic field would be considered to originate from the magnetic charge:
ρ = −∇·M. M is the magnetization satisfying M = (µr − 1) HmL , where µr is the relative magnetic
permeability and H mL is the Weiss field which is the effective field producing self-magnetization in the
ferromagnet. It is obvious that the quantity of charge is decided by the magnetization M, and then is
only decided by µr when the Weiss field H mL is fixed. In other words, the magnetic charge is changed
with the magnetic permeability of the material. For the steel bar under a local mechanical stress, the
magnetic permeability would change due to the stress-magnetic effect. As a result, magnetic charge
changes accordingly in these zones. For simplicity, here a phenomenological model is built in which a
linear charge distribution is adopted.
As shown in Figure 4, we assume a defect with width 2b and depth h at the bar surface when
the steel bar is partially lost due to the corrosion. Under the earth magnetic field, the dislocated
magnetic charge is concentrated at both ends of the corrosion zone. Moreover, the corrosion induces
a strong pressure stress on the steel bar and there exists a local stress concentration in the corrosion
zone. For this reason, a redistribution of magnetic charge occurs at the bottom of the corrosion zone.
The magnetic charge density in the corrosion zone can be expressed as

 ρmax , ( x = −b, −h < y < 0)

ρ( x, y) = −ρmax , ( x = b, −h < y < 0) (1)
 −ρ x
maxb , (− b < x < b, y = − h )


where ρmax denotes the maximum charge density, and the charge density is considered as the uniform
Sensors 2016, 16, 1439 5 of 10
distribution at the edges of the corroded region for simplicity.

Figure 4. The
Figure schematic
4. The of the
schematic magnetic
of the magneticcharge
charge distribution andcorrosion
distribution and corrosion pressure
pressure at corrosion
at the the corrosion
region on locally corroded steel bar enwrapped by concrete.
region on locally corroded steel bar enwrapped by concrete.

As shown in Figure 5, the self-magnetic leakage field at a space point (x, y) generated by the
As shown in Figure 5, the self-magnetic leakage field at a space point (x, y) generated by the
charge elements of the three regions can be expressed by Equation (2), respectively.
charge elements of the three regions can be expressed by Equation (2), respectively.
 ( x, y)dy
dH1 ( x, y ) 
ρ( x 0 ,y0 )dy20
r
dH1 ( x, y) = 2  r r 1
2πµ00r112 1
−ρ((xx 0 ,y0 )dy0
dH2 ( x, y) =  2πµ , y2 )dy
r2  (2)
dH 2 ( x, y )  0 r2 2 r2 (2)
dH3 ( x, y) = 2πµ r 02r2 r3
ρ( x20
,y0 )dx 0

0 3
 ( x, y)dx
dH 3 ( x, y )  r3
where µ0 denotes the magnetic permeability 21 ,0 r322, and r3 denote the space vectors from the
in the air,r
charge element to the space point, respectively.
where μ0 denotes the magnetic permeability in the air, r1, r2, and r3 denote the space vectors from
the charge element to the space point, respectively.
 20 r2
 ( x, y)dx
dH 3 ( x, y )  r3
20 r32

where μ0 denotes the magnetic permeability in the air,

Sensors 2016, 16, 1439
r1, r2, and r3 denote the space vectors from 5 of 9
the charge element to the space point, respectively.

Figure 5. The
Figure self-magnetic
5. The leakage
self-magnetic field
leakage generated
field from
generated thethe
from magnetic charge
magnetic at at
charge thethe
corrosion region.
corrosion
region.
Thus, the x and y components of total leaked magnetic field at space point (x, y) can be obtained
Thus, the x and y components of total leaked magnetic field at space point (x, y) can be
by the integral of Equation (2). It can be expressed as
obtained by the integral of Equation (2). It can be expressed as
 
Hx ( x, y) =
ρmax
2πµ0 arctan ( yx+ h y
+b ) − arctan( x +b)

ρmax y+h y
− 2πµ 0 
arctan x−b − arctan x−b
 
ρmax b− x −b− x
+ 2πµ 0b 
2b − ( y + h ) arctan − arctan
 y+h y+h
2 2
ρmax x (b− x ) +(y+h)
+ 2πµ 0b 2 ln (b+ x )2 +(y+h)2
ρmax ( x +b)2 +(y+h)2 (3)
Hy ( x, y) = 4πµ0 ln ( x +b)2 +y2
ρmax ( x −b)2 +(y+h)2
− 4πµ ln ( 2 )
0
 ( x − b ) + y2 
ρmax (y+h) (b− x ) +(y+h)2
2
− 2πµ ln
0b 2 (b+ x )2 +(y+h)2
  
− 2πµ0 b x arctan yb−
ρmax x
+h − arctan y+h
−b− x

3. Results and Discussion for Steel Corrosion Based on Micro-Magnetic Sensor

First, we scanned the magnetic field around the two samples and obtained the tangential field
(Hx ) distributions as shown in Figures 6 and 7, respectively. For the existence of demagnetization
field of steel bar and the earth magnetic field, the Hx is nonzero, but the x-Hx curves at different LFH
are almost parallel to each other for no damaged condition. There is no curve intersection, as shown
in Figures 6a and 7a, which is a typical Hx distribution for a healthy steel bar. Figure 6b–d plot the
Hx dependent on the sensors position of sample #1 for the corrosion time of 96 h, 120 h, and 144 h,
respectively. It is obvious that all the curves with different LFH intersect to two points when the
samples are corroded. The distances between two intersecting points are 16.0 cm, 17.5 cm, and 18.5 cm
for the different corrosion time, respectively. The longer the corrosion time is, the closer the distance is
to the length of corrosion region. A similar result is found from the measurement for sample #2 for
which curves are shown in Figure 7b–d. The result indicates that the position and regions of corrosion
can be obtained by analyzing the SMFL signals outside of reinforced concrete.
For testing the real corroded area, the partial concrete has been removed to expose the corroded
zone of the steel bar after magnetic detection. The morphology of the corrosion zone is shown in
Figure 8. It can be found that the corrosion is actually not serious because the average corrosion
depth is ~2 mm. Even the screw threads of the steel bar are not completely removed in the corroded
zone. This means that the material removal is not the leading factor for the SMFL, but the corrosion
pressure is.
For better understanding the experimental results, the tangential componentHx of self-magnetic
leakage field outside of concrete can be determined according to Equation (3). Figure 9 shows the Hx
Sensors 2016, 16, 1439 6 of 9

distributions affected by the corrosion length (the value of 2b in the Figure 4). Clearly, the phenomena
observed in the experiment have been captured in this model. The peaks ofx-Hx exhibit the center of
the corrosion region. Furthermore, it can be found that the x-Hx curves at different lift-off height (the
value of y in Figure 5) all intersect at the same two points observed in the experiments. The distance
∆x between the intersecting points is enlarged by increasing corrosion length and ∆x almost equal to
Sensors 2016,
Sensors 2016, 16,
16, 1439
1439 77 of
of 10
10
the corrosion length.

Figure 6.
Figure 6. The
The self-magnetic
self-magnetic field around the
field around the sample
sample #1
#1 depends
#1 depends on on the
on the location
the location of
location of sensor
of sensor at
sensor at different
at different
different
lift-off height
height (LFH):
(LFH): (a)
(a) non-corroded;
non-corroded; (b)
(b) corroded
corroded for
for 9696
h; h;
(c)(c) corroded
corroded forfor
120120
h; h;
(d) (d) corroded
corroded
lift-off height (LFH): (a) non-corroded; (b) corroded for 96 h; (c) corroded for 120 h; (d) corroded for for
for 144
144 h.
144 h. h.

Figure 7.
Figure The self-magnetic
7. The self-magnetic field
field around
around the
the sample
sample #2#2 depends
depends onon the
the location
location of
of sensor
sensor at
at different
different
Figure 7. The self-magnetic field around the sample #2 depends on the location of sensor at different
LFH: (a)
LFH: (a) non-corroded;
(a) non-corroded; (b)
non-corroded; (b) corroded
(b) corroded for
corroded for 96
for 96 h;
96 h; (c)
h; (c) corroded
(c) corroded for
corroded for 120
for 120 h;
120 h; (d)
h; (d) corroded
(d) corroded for
corroded for 144
for 144 h.
144 h.
h.
LFH:
Sensors 2016, 16, 1439 7 of 9
Sensors 2016, 16, 1439 8 of 10

Figure 8. The topography of corroded steel bar after removal of the cover concrete: (a) sample #1;
(b) sample #2.

For better understanding the experimental results, the tangential component Hx of

self-magnetic leakage field outside of concrete can be determined according to Equation (3).
Figure 9 shows the Hx distributions affected by the corrosion length (the value of 2b in the Figure 4).
Clearly, the phenomena observed in the experiment have been captured in this model. The peaks of
x-Hx exhibit the center of the corrosion region. Furthermore, it can be found that the x-Hx curves at
different lift-off
Figure 8. Theheight (the value
topography of y insteel
of corroded Figure
bar 5) allremoval
after intersectofat
thethe same
cover two points
concrete: observed
(a) sample #1; in
Figure 8. The topography of corroded steel bar after removal of the cover concrete: (a) sample #1;
the experiments.
(b) sample #2. The distance Δx between the intersecting points is enlarged by increasing
(b) sample #2.
corrosion length and Δx almost equal to the corrosion length.
For better understanding the experimental results, the tangential component Hx of
self-magnetic leakage field outside of concrete can be determined according to Equation (3).
Figure 9 shows the Hx distributions affected by the corrosion length (the value of 2b in the Figure 4).
Clearly, the phenomena observed in the experiment have been captured in this model. The peaks of
x-Hx exhibit the center of the corrosion region. Furthermore, it can be found that the x-Hx curves at
different lift-off height (the value of y in Figure 5) all intersect at the same two points observed in
the experiments. The distance Δx between the intersecting points is enlarged by increasing
corrosion length and Δx almost equal to the corrosion length.

Figure 9.
Figure The calculated
9. The calculated tangential Hxx of
tangential H of self-magnetic
self-magnetic leakage
leakage field
field at
at the
the different
different lift-off
lift-off height
height
(y = 2 cm, 4 cm, 6 cm) for the corrosion length 2b = 10 cm, 20 cm, 40
(y = 2 cm, 4 cm, 6 cm) for the corrosion length 2b = 10 cm, 20 cm, 40 cm. cm.

4. Conclusions
In this work, by detecting and analyzing the SMFL signals outside of corroded reinforced concrete
using a micro-magnetic sensor, we find that the x-Hx curves obtained at different lift-off heights all
intersect at the same points and the distances between intersecting points are basically equal to the
length of the corrosion region. By a linear magnetic charge model, the distribution of Hx component is
Figure 9. The calculated tangential Hx of self-magnetic leakage field at the different lift-off height
(y = 2 cm, 4 cm, 6 cm) for the corrosion length 2b = 10 cm, 20 cm, 40 cm.
Sensors 2016, 16, 1439 8 of 9

simulated and the intersection of curves is also found in the calculated results. This means that the
corrosion improves the magnetic resistivity of the steel bar and produces magnetic charge concentration
in the corrosion region, which generates this abnormal magnetic field distribution. The results propose
a new magnetic NDT technique to detect and evaluate the inner corrosion in engineering structures
using high-resolution micro-magnetic sensors. This method has many advantages over traditional
techniques. It is a simple, inexpensive and efficient method to non-destructively test the corrosion in
the reinforced concrete structure.

Acknowledgments: This work was supported by the National Science Fund for Distinguished Young Scholars
(No. 51425801), the National Natural Science Foundation of China(Nos. 51278512, 51508058, 11404045),
the Social livelihood science and technology innovation special of Chongqing (No. cstc2015shmszx30012 and
No. cstc2015shmszx30029), the Science and Technology Project of Guizhou Provincial Transportation Department
(No. 2016-123-006, 2016-123-039, 2016-123-040), the Science and Technology Planning Project of Yunnan Province
(No. 2014 (A) 27), and the Communications Science and Technology Project of Guangxi Province of China
(No. 20144805).
Author Contributions: J. Zhou arranged all the work in this project and gave a keen insight in this manuscript;
H. Zhang and M. Yang performed the experiments; R. Zhao and R. Xia analyzed the measured data; L. Liao and
R. Zhao carried out the simulation for the experimental results; H. Zhang and L. Liao wrote the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Lorenz, W.J.; Mansfeld, F. Determination of Corrosion Rates by Electrochemical DC and AC Methods.
Corros. Sci. 1981, 21, 647–672. [CrossRef]
2. Yeih, W.; Huang, R. Detection of the Corrosion Damage in Reinforced Concrete Members by Ultrasonic
Testing. Cem. Concr. Res. 1998, 28, 1071–1083. [CrossRef]
3. Tian, G.Y.; Rangarajan, S.; Sophian, A. Stress Measurement using Novel Eddy Current Probe. In Proceedings
of the BINDT Annual Conference, Torquay, UK, 14–16 September 2004.
4. Kobayashi, K.; Banthia, N. Corrosion Detection in Reinforced Concrete using Induction Heating and Infrared
Thermography. J. Civil Struct. Health Monitor. 2011, 1, 25–35. [CrossRef]
5. Suzuki, S.; Suzuki, S.; Matsubara, E.; Waseda, Y. In-situ X-ray Diffraction of Corrosion Products Formed on
Iron Surfaces. Mater. Trans. 2005, 46, 637–662.
6. Sakamoto, H.; Okada, M.; Homma, M. Theoretical Analysis of Barkhausen Noise in Carbon Steels.
IEEE Trans. Magn. 1987, 23, 2236–2238. [CrossRef]
7. Piotrowski, L.; Augustyniak, B.; Chmielewski, M.; Kowalewski, Z. Possibility of Application of
Magnetoacoustic Emission for the Assessment of Plastic Deformation Level in Ferrous Materials.
IEEE Trans. Magn. 2011, 47, 2087–2092. [CrossRef]
8. Langman, R.A.; Mutton, P.J. Estimation of Residual Stresses in Railway Wheels by means of Stress-induced
Magnetic Anisotropy. NDT E Inter. 1993, 26, 195–205. [CrossRef]
9. Fitzgerald, C.S. Inspection for Rail Defects by Magnetic Induction. In Proceedings of the SPIE 2458
Nondestructive Evaluation of Aging Railroads, Oakland, CA, USA, 6 June 1995.
10. Hall, D.R. Electromagnetic Inspection of Wire Ropes and strand-supported structures. Insight 2003, 45,
471–475. [CrossRef]
11. Bergamini, A.; Christen, R. A Simple Approach to the Localization of Flaws in Large Diameter Steel Cables.
In Proceedings of the SPIE 5047 Smart Nondestructive Evaluation and Health Monitoring of Structural and
Biological Systems II, San Diego, CA, USA, 2 March 2003.
12. Krieger, J.; Krause, H.J.; Gampe, U.; Sawade, G. Magnetic Field Measurements on Bridges and Development
of a Mobile SQUID-System. In Proceedings of the SPIE 3587 Nondestructive Evaluation of Bridges and
Highways III, Newport Beach, CA, USA, 3 March 1999.
13. Xiong, E.G.; Wang, S.L.; Zhang, J.F.; Li, B. Magnetic Flux Leakage Testing Mechanism and Its Application.
In Proceedings of the International Symposium on Innovation & Sustainability of Structures in Civil
Engineering, Guangzhou, China, 29–30 November 2009.
14. Wang, Z.D.; Gu, Y.; Wang, Y.S. A Review of Three Magnetic NDT Technologies. J. Magn. Magn. Mater. 2012,
324, 382–388. [CrossRef]
Sensors 2016, 16, 1439 9 of 9

15. Dubov, A.A. A Study of Metal Properties Using the Method of Magnetic Memory.Met. Sci. Heat Treat. 1997,
39, 401–405. [CrossRef]
16. Shi, Y.; Zhang, C.; Li, R.; Cai, M.; Jia, G. Theory and Application of Magnetic Flux Leakage Pipeline Detection.
Sensors 2015, 15, 31036–31055. [CrossRef] [PubMed]
17. Han, W.H.; Shen, X.H.; Xu, J.; Wang, P.; Tian, G.Y.; Wu, Z.Y. Fast Estimation of Defect Profiles from the
Magnetic Flux Leakage Signal Based on a Multi-Power Affine Projection Algorithm. Sensors 2014, 14,
16454–16466. [CrossRef] [PubMed]
18. Yao, K.; Deng, B.; Wang, Z.D. Numerical Studies to Signal Characteristics with the Metal Magnetic
Memory-effect in Plastically Deformed Samples. NDT E Inter. 2012, 47, 7–17. [CrossRef]
19. Morón, C.; Cabrera, C.; Morón, A.; García, A.; González, M. Magnetic Sensors Based on Amorphous
Ferromagnetic Materials: A Review. Sensors 2015, 15, 28340–28366. [CrossRef] [PubMed]
20. Förster, F. Nondestructive Inspection by the Method of Magnetic Leakage Fields: Theoretical and
Experimental Foundations of the Detection of Surface Cracks of Finite and Infinite Depth. Sov. J. Nondestr. Test
1982, 8, 841–859.
21. Zatsepin, N.N.; Shcherbinin, V.E. Calculation of the Magnetostatic Field of Surface Defects I. Field to
Pography of Defect Models. Sov. J. Nondestr. Test 1966, 5, 385–393.
22. Shur, M.L.; Zagidulin, R.V.; Shcherbinin, V.E. Theoretical Problems of the Field Formation from a Surface
Defect. Defektoskopiya 1988, 3, 14–25. (In Russian)
23. Zagidulin, R.V. Calculation of the Remnant Magnetic Field of a Discontinuity Defect in a Ferromagnetic
Sample. II. Remnant Magnetic Field of a Defect in the Air. Defektoskopiya 1998, 10, 33–39. (In Russian)

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
 Journal of Magnetism and Magnetic Materials 324 (2012) 382–388

Contents lists available at SciVerse ScienceDirect

Journal of Magnetism and Magnetic Materials

journal homepage: www.elsevier.com/locate/jmmm

Review

A review of three magnetic NDT technologies

Z.D. Wang n, Y. Gu, Y.S. Wang
Department of Mechanics, School of Civil Engineering, Beijing Jiaotong University, Beijing 100044, China

a r t i c l e i n f o a b s t r a c t

Article history: Magnetic techniques are most important NDT technologies to characterize the mechanical features of
Received 27 May 2011 ferromagnetic materials based on the physical principle of magnetic-stress coupling. A review is presented
Received in revised form in this paper about the development of the magnetic NDT technologies. After a brief outline of the
26 August 2011
theoretical studies of the magnetic-stress coupling effect, the three popular magnetic NDT technologies are
Available online 7 September 2011
reviewed, which are magnetic flux leakage (MFL), magnetic Barkhausen noise (MBN) and recently
Keywords: developed metal magnetic memory (MMM). The first two are ascribed to the active magnetic method,
Non-destructive technique and the last one is the passive method. Based on an extensive literature survey in this field, this paper
Magnetic-stress coupling effect focuses on the discussion of the physical mechanism and some important experimental results relevant to
Magnetic flux leakage
the three NDT technologies. The challenges for each technique in this field are also summarized.
Barkhausen noise
& 2011 Elsevier B.V. All rights reserved.
Metal magnetic memory

1. Introduction 2. Magnetic-stress coupling theory

A fundamental feature of ferromagnetic materials, which consist
of numerous small magnetic domains in the microstructure, is the
coupling between stress and magnetic field [1–6]; that is, the
magnetization may result in variations in the dimension of
ferromagnetic materials, namely magnetostriction [7,8], and on
the other hand the stress may also change the magnetization of
ferromagnetic materials, the so-called piezomagnetic effect [9,10].
These macro-phenomena are related to magnetic-moments rota-
tion and domain-wall displacement in the micro-structure when
ferromagnetic materials are subjected to an applied magnetic field
or mechanical stress [11,12].
Compared with the magnetostriction, the piezomagnetic effect
has received more attention in non-destructive test (NDT) field,
because it is the physical basis to evaluate the stress status of
ferromagnetic structures and components by magnetic measuring
methods. As a result, a great number of non-destructive magnetic
techniques have been developed over the last decades such as
magnetic flux leakage (MFL), magnetic Barkhausen noise (MBN),
magnetoacoustic emission (MAE), stress-induced magnetic aniso-
tropy (SMA) and recently developed metal magnetic memory
(MMM) [13,14].
In the remainder of this paper, the concept and theory of
magnetic-stress coupling models in ferromagnetic materials will
be presented in Section 2, the development of three typical NDT
magnetic techniques are described in Section 3 and the conclu-
sions are presented in Section 4.
n
Corresponding author.
E-mail address: zhdwang@bjtu.edu.cn (Z.D. Wang).
It is well-known that magnetic properties of ferromagnetic
materials are stress-dependent. For example, the coercivity Hc and
permeability m of ferromagnetic materials may be changed in the
order of 100% by the stress within the elastic limit [15]. Fig. 1
presents a schematic description of a cyclic stress affecting the
magnetization of ferromagnetic materials. The total magnetization
M includes the reversible component, Mre, and irreversible compo-
nent, Mirr. In every loading cycle, an irreversible magnetization
difference DMirr is obtained when the domains pass through pinning
sites under the action of the mechanical stress [16].
From the microstructural point of view, a ferromagnetic sample
is composed of numerous magnetic domains. Every domain is
10  8–10  12 m3 in volume and includes approximately 1012–1015
atoms. The magnetic moments distribute randomly in the initial
state, and no magnetization appears in the macro-scale. When an
external load is applied to a ferromagnetic sample, a tension tends
to orient the domains in the direction of the applied load for a
positive magnetorestrictive material and a compression orients the
domains perpendicularly to the loading direction due to the piezo-
magnetic effect [17]. Fig. 2a displays a symmetric structure in which
the applied field H¼0. Only in the case of Ha0 (see Fig. 2b), the
symmetric structure is destroyed and the ferromagnetic sample is
magnetized. Thus, the presence of an external field (e.g. an artificial
field or Earth’s magnetic field) is a necessary condition for all
magnetic NDT techniques. This may also be concluded from the
following experiment: the spontaneous stray field signal is detected
when plastically deforming a ferromagnetic sample under the action
of the Earth’s magnetic field, but not obtained when the sample was
plastically deformed in a magnetic-prohibited environment.
Theoretical studies of the magnetic-stress coupling effect provide
a physical understanding of various magnetic NDT technologies.

0304-8853/$ - see front matter & 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmmm.2011.08.048
 Z.D. Wang et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 382–388
Fig. 1. Schematic diagram of the stress affecting the magnetization of a ferromagnetic
sample. The irreversible component of the residual magnetization ðDMirr s Þ remains in
every loading cycle.
Fig. 2. Magnetic field and stress affecting the movement of magnetic domains in
the micro-structure. Tensile stress orients the domains in the loading direction
and compressive stress causes the domains perpendicularly to the loading
direction. The magnetic field destroys the systematic domain structure and lead
to magnetization (a) without defect and (b) corrupt wall.
In fact, the conceptual description of the interaction between the
stress and magnetism may date back to more than half a century
ago. Brown [18] first presented a theoretical analysis of the
magneto-mechanical effect in ferromagnetic materials by replacing
the applied stress with an equivalent field in 1949. Later on, Cullity
[19] focused on such phenomena in terms of Le Chatelier’s principle.
Further, Sablik et al. [20] considered the variations in hysteresis of
ferromagnetic materials under a constant stress. Additionally, the
studies of the magneto-mechanical effect based on the concepts of
the ‘‘effective field theory’’ and ‘‘law of approach’’ were performed
by Jiles [21–24].
The classical magneto-elastic constitution for a ferromagnetic
material is generally expressed as [25]
( l11, l12, l21 and l22 magnetostriction coefficients; ep plastic
e ¼ es þ eH ¼ Sr þ Ds H,
ð1Þ
B ¼ Bs þBH ¼ DH r þ lH,
where S, l Ds and DH are tensors of the stiffness matrices,
effective magnetic permeability, magnetic–elastic coefficient
and elastic–magnetic coefficient, respectively; r and H are ten-
sors of the stress and external magnetic field, respectively; the
functions of e and B denote strain tensor and effective magnetic
field tensor, respectively; es and eH denote the strain components
caused by the stress and magnetic field, respectively; Bs and BH
denote the magnetization components caused by the stress and
magnetic field, respectively.
For an isotropic ferromagnetic material, a number of representa-
tive macro-models have been presented in order to characterize the
383
magneto-mechanical effect (see e.g. [19–21]), as follows:
   
dl dB
¼ , ð2Þ
dH s ds H
dM 1 dMan
¼ 2 sð1cÞðMan Mirr Þ þ , ð3Þ
ds k ds
3
Es ¼  ls cos2 y: ð4Þ
2
In Eq. (2), (dl/dH)s is the changing rate of the magnetostriction
with the applied field at a constant stress. (dB/ds)H is the change of
the magnetic induction with the stress at a constant field. Eq. (2)
quantifies the relationship between the magnetostrictive effect
and magneto-elastic effect. Eq. (3) describes the dependence of
the material magnetization on the stress as well as anhysterestic
magnetization—Man (the ideal or lossless magnetization of a
material) and Mirr (the irreversible component of magnetization);
k and c are material constants. Eq. (4) gives the dependence of the
stress energy (Es) on the stress (s), bulk magnetostriction (l) and
the angle between the applied stress and field (y).
Under consideration of the different physical mechanisms of the
magnetization of ferromagnetic materials at the different stages,
nonlinear magnetic–elastic models, involving the magneto-mechan-
ical effect, have been proposed in [26–28], which however do not
differentiate elastic and plastic deformation. It is well-known that
elastic and plastic deformation represents different deformation
modes in the microstructure. The former involves an increase or
decrease in the atomic spacing while the latter may result in the
generation and accumulation of various defects such as dislocation,
twinning and shear band. Clearly, the different deformation modes
have different impacts on the magnetic behavior of ferromagnetic
materials. In order to characterize such a difference in quantity,
Wang et al. [29] recently proposed a magnetic–elastic–plastic model
by considering the different mechanisms of the elastic and plastic
deformation:
   
H m a
Magnetization strength : M ¼ Ms coth total  0 ,
m0 a Htotal
e p
Effective field : Htotal ¼ HH þ Hs þ Hs ,
Magnetic-induced effective field : HH ¼ H þ aM,
3seq cos2 b
Elastic-induced effective field : Hse ¼
m0
 
 ðl11 þ l12 sÞ þ 2ðl21 þ l22 sÞM 2 M,
1 b/ep S
Plastic-induced effective field : Hsp ¼ k9ep 9, where k ¼ ,
m0 2Ms
ð5Þ
where Ms is the saturation magnetization; m0 the permeability
of vacuum; a mean field parameter representing the interdomain
coupling; seq the equivalent stress for a complex stress state; b the
angle between the magnetization and equivalent-stress directions;
deformation; /epS average pinning energy of the site for 1801
wall; b material’s constant.
This model leads to some results verified by experimental
observations: elastic tensile stress in the field direction accelerates
the magnetization for positive magneto-restrictive materials, but
the compressive stress opposes the magnetization; different from
the elastic deformation, both tensile and compressive plastic
deformation may decrease the magnetization of ferromagnetic
materials. However, similar to most proposed magnetic-stress
coupling models, the actions of the applied field and stress in the
magnetization are decoupled in Eq. (5). Clearly, the decoupling
in the magnetic-stress coupling models results in a confusing
conclusion; that is, a ferromagnetic sample may be magnetized
by the stress even if no magnetic field exists, which is in conflict
384 Z.D. Wang et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 382–388
with the reported experimental observations. Therefore more
convincible models should be proposed in order to have a thorough
understanding of the magnetic-stress coupling effect of ferromag-
netic materials.
3. Magnetic NDT technology
Magnetic NDT technologies have been extensively adopted in
engineering to ensure the operating safety of ferromagnetic
structures and components [30]. In this section, the studies of
three typical magnetic NDT technologies (MFL, MBN and MMM)
will be summarized. Among them, MFL and MBN techniques may
be ascribed to active magnetic test methods in which a strong
magnetic field is applied. However, the MMM technique is a
weak-field test method in which the Earth’s magnetic field
instead of an artificial field is used as the stimulus.
3.1. Magnetic flux leakage
Magnetic flux leakage (MFL) is one of the traditional electro-
magnetic NDT techniques, originated from magnetic particle
technique. Hoke first discovered the MFL phenomenon in 1918.
However, due to the lack of magnetization techniques in the
early time, the first application of the MFL technique was
performed by Watts in 1933 in assessing the quality of the
welded joints. Since 1960s, this technique has been extensively
used as an inspection technique in the petrochemical engineering
and transportation, energy and metal industries [31–37]. One
successful application of the MFL technique is the device, called
‘‘pipeline-pig’’, which is developed to detect the corrosion and
metal loss in oil and gas in-service pipelines [38–40]. Fig. 3 shows
a scheme of the structure and operating principle of the ‘‘pipeline-
pig’’. A strong permanent magnet in the ‘‘pipeline-pig’’ nearly
saturates the pipe wall when it is propelled by the oil/gas
pressure or driving equipments. No flux is leaked out if the pipe
wall is perfect (Fig. 3a). However, the flux ‘‘leaks’’ out of the wall
at the location of a metal loss defect (Fig. 3b). The ‘‘leakage flux’’ is
detected by an array of circumferentially distributed sensor
assembly. The MFL data are sampled and stored using an on-
board data acquisition system, and subsequently analyzed offline
by trained data analysts.
As a classical NDT method, the principle of the MFL technique is
relatively simple. That is, when a strong magnetic field is applied to
a ferromagnetic material, any geometrical discontinuity in the test
object will cause the field to leak out of the object into the air (see
Fig. 4). The flux leakage can be monitored by a magnetic field
sensor and used to estimate the dimensions of the defect. Although
the MFL phenomenon is easily understood, the design and analysis
of MFL systems involve complicated interactions between the
excitation field, leakage flux and the defects in the material. There
are several important aspects to be considered. Firstly, the level of
Fig. 3. Structure and principle of the ‘pipeline-pig’. (a) No magnetic flux leakage for perfect pipe wall; (b) magnetic flux leakage in defect position.
the excitation magnetic flux should be large and homogenous to
allow the magnetic flux variation to occur at the location of a
defect; secondly, the sensors should be located close to the position
at which the changes in the magnetic field density, originating
from the defect, are distinct from the background noise; addition-
ally, developing an effective inversion method to identify the
defective characters by the recorded MFL signals is difficult since
the defect is irregular.
A great number of efforts have been devoted to develop a simple
analytical model to explain the formation of the MFL signal [41–49].
Those developed models may be classified into two types. One was
originated from Förster model [41] by considering the change of the
magnetic parameters (e.g. permeability and coercive field) in the
local defect region, and the other was developed from Zatsepin–
Scherbinin’s model [42] where magnetic dipoles were assumed in
the surface of the defect. Clearly, Förster model described the MFL
phenomenon in the macroscopic view but Zatsepin–Scherbinin’s
model in the microscopic view. In order to consider the non-linear
magnetic behavior of ferromagnetic materials, more complicated
models have been proposed in [43–49].
The reported models can effectively describe the fundamental
features and the topography of the MFL field for regular defects
(e.g. rectangular slot), but not for irregular defects. In order to
analyze the MFL signal generated by irregular defects, some
integral equations in describing the defect-induced magnetic
charges were given based on the linear approximation of ferro-
magnetic materials, which can be solved numerically using the
iteration method [50–53].
Compared with the theoretical analysis, the magnetic finite
element method (FEM) is a powerful tool for the investigation of
the MFL signal due to its flexibility in the simulations of varied
irregular geometrical defects. 2D magnetic FEM methods [54,55]
provide sufficient information for sharp-shaped defect characteriza-
tion, but do not accurately quantify the natural defects, e.g. stress
corrosion cracks. Therefore, 3D magnetic FEM has received more
attention in recent years. In this aspect, the study in [56] presented
comparisons between 2D and 3D models. The studies in [57–60]
Fig. 4. Schematic representation of the flux leakage in the presence of a geometric
discontinuity.
 Z.D. Wang et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 382–388
investigated the characteristics of MFL signals due to corrosion
defects. Additionally, the effect of the local dent-induced stresses on
the MFL signal was studied using both magnetic FEM modeling and
experimental tests in [61]. A 3D simulation, aimed at studying the
influence of the defect geometry and lift-off value, was performed in
[62]. Further, the study in [63] examined the effects of both dent
geometry and localized residual stresses on the MFL signal using a
3D magnetic FEM method.
It may be pointed out that the MFL technique is one of the most
popular magnetic NDT techniques and extensively used in various
engineering fields. In the MFL test, the key is how to inversely
determine the defects of the investigated object using the recorded
MFL signals. Various new algorithms (e.g. wavelets, neutral networks
and genetic algorithm) were applied in [64,65]. However, there are
two primary obstacles in the defect inversion. Firstly, the defect in
the reality is commonly complicated and usually characterized by
many parameters such as the width, thickness, location and edge
condition of the defect, which in turn significantly impact on the
measured MFL signals. Clearly, it is generally difficult to characterize
every parameter based on the measured MFL signals. Secondly, it is
still a challenge to deal with the elastic–plastic zone near the cracks,
and idealized material properties are commonly assumed. This is in
part due to the fact that the effects of the plastic deformation on the
magnetic characteristics of ferromagnetic materials are not thor-
oughly understood.
3.2. Magnetic Barkhausen noise
Different from the MFL technique developed for inspecting the
macro-defects, the magnetic Barkhausen noise (MBN) technique
is usually used to determine the magnetic easy axis [66], the
residual and applied stresses [67] and the grain size [68] of
ferromagnetic materials.
In 1919 Barkhausen [69] found that when a ferromagnetic
material was magnetized by an increasing field, the noise in the
form of voltage pulses was generated in a coil placed near the
material. Fig. 5a shows a schematic of the Barkhausen noise
during the magnetization process. The test system is presented
in Fig. 5b. From the microstructural point of view, the develop-
ment of Barkhausen noise is due to the abrupt motion of 1801
domains across local pinning sites [70–72].
The parameter commonly used in the analysis of the detected
MBN signal voltage V is the MBNenergy defined as
X Z
MBNenergy ¼ A V 2 dt: ð6Þ
events
Fig. 5. Barkhausen noise and test system. The left figure shows the development of Barkhausen noise during the magnetization of specimen, and the right figure shows of
MBN test system.
385
Generally, MBNenergy may be displayed in a polar scale and
described by [73]
MBNenergy ¼ a cos2 ðyjÞ þ b, ð7Þ
where y is the angle of the applied magnetic field with respect to
the reference loading direction; a the angular dependence of the
variation of the MBN signal; b the angular independent signal
(isotropic background); j the direction of the magnetic easy axis
with respect to the reference direction.
The MBN signal is generated due to the irreversible movement
of magnetic domain walls. The amplitude depends on residual
and applied stresses. Thus, it is a useful NDT technique to detect
the presence of the residual and applied stresses in ferromagnetic
samples. In steels with a positive magnetostrictive constant,
experiments [74–81] confirmed that the MBN amplitude was
increased due to the elastic tensile stress in the direction of the
magnetization but decreased due to the compressive stress. It is
due to the different effects of tensile and compressive stresses on
the movement of domain walls [82–84]: tensile stress increases
the number of mobile 1801 domain walls, reorients domains or
modifies the pinning barriers, but compressive stress makes the
flux closure domain formation by decreasing the number of 1801
domain walls.
Plastic deformation results in a significant increase in the
dislocation density and changes the domain wall energy at some
pinning sites. It should be noted that the effect of plastic
deformation on the MBN signals is more complicated than that
of elastic deformation [85–87]. Such complexity is especially
reflected by some contradictory claims reported regarding the
dependence of MBN signals on plastic deformation. For example,
most studies [88–93] reported a continuous decrease of the
Barkhausen noise as the plastic strain increases, while in the
other studies [73,94–97] an increase at the low plastic strain
followed by a decrease at the high plastic strain was observed.
The above-mentioned results were obtained by the quasi-
static experiments. Further work has also been performed on
fatigue loads affecting the MBN signal. Kettunen and Ruuskanen
[98] measured the MBN signals of two low-carbon structural
steels under tension–compression fatigue cycles. It was found
that the MBN signals increased in the beginning and then kept an
approximate constant value without large oscillations until the
final failure of the specimens. Karjalainen and Moilanen [99]
further tested the MBN signals of the mild steel under the
constant deflection bending fatigue. It was demonstrated that
the overall signal amplitude decreased during the fatigue test. In
some cases, there existed a sudden increase in the signal level
386 Z.D. Wang et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 382–388
prior to the failure. Furuya et al. [100] tested the MBN signal in
the low cycle fatigue experiments in mild steels. They pointed out
that the MBN amplitude decreased continuously during the cyclic
loading and no particular changes in the amplitude were
observed before the failure. Lindgren and Lepistö [101] used a
continuous monitoring method to test the MBN signal during the
whole fatigue loading process. It was demonstrated that the
maximum MBN signal increased with the increase of the stress
amplitude but the minimum value was approximately constant.
The changes of the MBN amplitude as a function of fatigue
cycles were found to be different between the mild steel and
high strength steel samples. Other interesting results have been
reported by Refs. [102–106].
Clearly, the above reports are somewhat contradictory. This is
because the MBN signals are very sensitive to the microstructure
of ferromagnetic materials [107,108] and measuring systems
[109,110]. In general, developing an efficient model to quantify
the interaction between microstructure features and Barkhausen
noise is still a challenge in the field.
3.3. Metal magnetic memory
As above-mentioned, both MFL and MBN are active magnetic
test techniques where a strong applied field is used to magnetize
the tested object. Thus, the two techniques are usually time-
consuming and even impractical for some irregular structures. For
the purpose of developing a more simple and effective NDT
magnetic technique to meet the requirements in engineering,
the passive magnetic technique has received a great interest in
Fig. 6. Schematic representation of SMFL distributions in the stress-concentration
zone. Hp(x) reaches the maximum and the normal component Hp(y) is zero in
maximum stress-concentration zone.
Fig. 7. Testing results of normal component Hp(y) of three rings. The results show that the left ring is well and the right is seriously damaged.
the recent years. In the late 1990s, a passive magnetic NDT
technique, namely metal magnetic memory (MMM) technique,
was first proposed by a Russian group [111,112]. The advantage of
the MMM technique is that the Earth’s magnetic field instead of
an artificial strong field plays as the role of the stimulus source. In
addition, the MMM technique displays other attractive advan-
tages, as follows:
 it is effective in assessing the early damage and developed
defects;
 it has a detecting depth of up to a few millimeters in
comparison with macrons in X-ray diffraction technique;
 it is easy-operation and time-saving, and has the maximum
measuring velocity of up to several meters per second.
The physical mechanism of the MMM technique is described
in [113–116]. Under the effect of the earth field and mechanical
load, self-magnetic-flux-leakage (SMFL) signals are generated in
the stress-concentration zones where the tangential SMFL com-
ponent Hp(x) reaches the maximum and the normal component
Hp(y) changes polarity and has a zero value (see Fig. 6).
The magnetic state is still retained even if the load is removed.
Therefore, the stress concentration zones can be detected by
measuring the SMFL signals on the surface of the structure.
In the view of micro-structures, it is due to the irreversible
orientation of magnetic domains caused by plastic deformation
in the maximum stress-concentration zone.
The MMM technique has received extensive attention in
engineering due to its advantages of easy-operation, time-saving
and simple criteria. Various applications have been reported for
diagnosing gas and oil pipelines, rails, turbine wheels, pressure
vessel and others in [113]. Fig. 7 shows the testing results of
normal component Hp(y) of three rings. The much lower ampli-
tude of Hp(y) about the left ring means that it is under well-
operation condition. However, the much higher amplitude of
Hp(y) about the right one indicates that this one is seriously
deteriorated. Moreover, the results also show the possible loca-
tions of the stress-concentration (SC) zone as labeled in the right
picture of Fig. 7.
The MMM technique is suitable for many engineering prac-
tices. However, as a comparatively new test method, it still has a
large room to be improved, as shown in the following section. One
of the critical points is that more accurate and quantitative
criteria are required. Up to now this technique is only used as a
qualitative test method to determine the possible dangerous
positions without quantitative results (e.g. stress-concentration
intensity and residual fatigue life). In order to understand the
physical mechanism and provide more information about the
 Z.D. Wang et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 382–388
relations between the SMFL signal and structure characters,
further studies have been recently performed.
Dong et al. [117] measured the SMFL signal for a series of tensile
18CrNiWA steels. The magnetic curve on the measured line shows
a good linearity after loading. It rotates in the counterclockwise
direction with the increase of the tensile stress during the elastic
deformation, while gathers together in the plastic deformation stage.
Wang et al. [118,119] performed numerical simulations of the SMFL
signals using a revised magnetic-charge model where critical factors
(e.g. the size, location and density of the local plastic zone and the
lift-off values of the sensor) affecting the SMFL signal are presented.
Yao et al. [120] experimentally confirmed that the SMFL signal and
its gradient were significantly different during which Q235-steel
specimens are deformed from elastic to plastic deformation under
the tension, but no detectable change could be found during the
whole compressive loading. They presented an explanation by
considering the different movement modes of the domain structures
subjected to tensile and compressive loads. Shi et al. [121] measured
the SMFL signal and its gradient during which 18CrNi4A steel
specimens were subjected to tension–tension fatigue loading where
the effect of the local stress concentration factor on magnetic test
results was especially considered. Leng et al. [122] tested the SMFL
signal of 18CrNi4A steel specimens induced by cyclic bending
stresses. The experimental results were qualitatively explained by
the ‘‘effective field model’’ developed by Jiles [24]. Wilson et al.
[123] introduced a novel three-axis magnetic sensor to confirm that
the parallel component Bx is more related to the applied stress than
the perpendicular component Bz. Roskosz and Gawrilenko [124]
further presented experimental and numerical analyses of the SMFL
distribution in loaded notched samples.
The MMM technique, although has received considerable
attention over the last decade, still needs further studies, based
on the following reasons:
(1) Lack of physical models in quantifying the relations between the
SMFL signal and plastic deformation. As mentioned above, the
MMM technique is developed for inspecting the initial
damage or plastic deformation induced by the local stress
concentration. Naturally, traditional magneto-elastic models
[19–21] cannot describe such phenomenon. Although the
newly developed magneto-plastic model [29] differentiated
the actions of the elastic and plastic deformation on the
magnetization of ferromagnetic materials, no direct relation-
ship between the plastic strain and SMFL signal has been
presented.
(2) Lack of quantitative criteria in capturing the features of defects.
There are only two qualitative criteria for which the tangen-
tial component Hp(x) reaches the maximum value, and the
normal component Hp(y) changes positive–negative symbols
and thus has a zero in the location of maximum stress. Due to
the lack of quantitative defect criteria, the MMM technique
can only be used as an assistant tool to diagnose the possible
dangerous zone presently. More accurate results must be
determined by other techniques. In fact, there must be some
quantitative relations between SMFL signals and defect char-
acters in MMM tests. In this aspect, systematic experiments
and numerical simulations should be performed.
4. Conclusions
The magnetic-stress coupling effect is one of the basic characters
of ferromagnetic materials. Based on such an effect, various
magnetic NDT techniques in assessing the stress status of ferro-
magnetic structures have been developed, some of which have been
successfully used in engineering. However, understanding of the
387
magnetic-stress coupling effect is incomplete due to the complexity
of the magnetic-stress action and high-sensitivity to the material’s
microstructure. Thus, the applications of the NDT techniques in
industry still face considerable challenges. Further studies of the
NDT techniques are desirable.
In this paper, the current physical models for magnetic-stress
coupling are introduced, and three representative magnetic NDT
techniques (MFL, MNB and MMM) are reviewed. The fundamental
features, advantages, disadvantages and future work for these
NDT techniques are summarized based on a literature review of
more than one hundred references.
Acknowledgments
This work was funded by Natural Science Foundations of China
(no. 11072027), and the Fundamental Research Funds for the
Central Universities and Ministry of Education of the People’s
Republic of China (NECT).
References
[1] R. Becker, W. Döring, Ferromagnetismus, Springer, Berlin, 1939.
[2] R.M. Bozorth, H.J. Williams, Effect of small stresses on magnetic properties,
Review of Modern Physics 17 (1945) 1772.
[3] R.M. Bozorth, Ferromagnetism, Van Nostrand, New York, 1951.
[4] S. Chikazumi, Physics of Magnetism, Wiley, New York, 1964.
[5] H. Träuble, Magnetism and Metallurgy, Academic Press, New York, 1969.
[6] D.P. Bulte, R.A. Langman, Journal of Magnetism and Magnetic Materials 251
(2002) 229.
[7] E.W. Lee, Reports on Progress in Physics 18 (1955) 184.
[8] R. Birss, IEEE Transactions on Magnetics 7 (1) (1971) 113.
[9] R. Birss, C.A. Faunce, E.D. Isaac, Journal of Applied Physics 10 (1971) 1040.
[10] P.J. Banks, IEEE Transactions on Magnetics Mag. 13 (3) (1977) 1000.
[11] J. Degauque, Solid State Phenomena 35/36 (1994) 335.
[12] D.C. Jiles, Introduction to Magnetism and Magnetic Materials, 2nd edn.,
Chapman and Hall Press, London, 1998.
[13] D.C. Jiles, NDT & E International 23 (2) (1990) 83.
[14] J. Blitz, Electrical and Magnetic Methods of Nondestructive Testing, Adam
Hilger IOP Publishing, Ltd, Bristol, 1991.
[15] D.C. Jiles, Journal of Applied Physics 21 (1988) 1196.
[16] J.J. Zheng, S.Y. Cao, H.L. Wang, Sensors and Actuators A 143 (2008) 204.
[17] M. Lindgren, T. Lepistö, NDT & E International 34 (2001) 337.
[18] W.F. Brown, Physical Review 75 (1949) 147.
[19] B.D. Cullity, Introduction to Magnetic Materials, Addison-Wesley Publishing
Company, New York, 1972.
[20] M.J. Sablik, G.L. Burkhardt, H. Kwun, D.C. Jiles, Journal of Applied Physics 63
(1988) 3930.
[21] D.C. Jiles, D.L. Atherton, Journal of Applied Physics 17 (1984) 1265.
[22] D.C. Jiles, D.L. Atherton, Journal of Magnetism and Magnetic Materials 61
(1986) 48.
[23] D.C. Jiles, M.K., Journal of Magnetism and Magnetic Materials 140–144
(1995) 1881.
[24] D.C. Jiles, Journal of Applied Physics 28 (1995) 1537.
[25] Z.D. Wang, K. Yao, K.Q. Ding, NDT & E International 43 (6) (2010) 513.
[26] G.P. Carman, M. Mitrovic, Journal of Intelligent Material Systems and
Structure 6 (1996) 673.
[27] Y.P. Wang, D.N. Fang, K.C. Hwang, International Journal of Non-Linear
Mechanics 38 (2003) 1053.
[28] H.M. Zhou, Y.H. Zhou, X.J. Zheng, Q. Ye, J. Wei, Journal of Magnetism and
Magnetic Materials 321 (2009) 281.
[29] Z.D. Wang, B. Deng, K. Yao, Journal of Applied Physics 109 (2011) 083928.
[30] Handbook on the Magnetic Examination of Welds, ISBN:0-85300195-2, The
Welding Institute, Cambridge, UK, 1988.
[31] A. Khalid, Insight 41 (1999) 232.
[32] M.D. Pandey, NDT & E International 31 (1998) 349.
[33] R.W.F. Shannon, I.C. Braithwaite, L.L. Morgan, Oil and Gas Journal 86 (32)
(1988) 47.
[34] K.S. Ryu, D.L. Atherton, L. Clapham, Insight 44 (5) (2002) 285.
[35] K. Hwang, S. Mandayam, S.S. Udpa, L. Udpa, W. Lord, M. Atzal, NDT & E
International 33 (2000) 531.
[36] R. Pohl, A. Erhard, H.J. Montag, H.M. Thomas, H. Wüstenberg, NDT & E
International 37 (2) (2004) 89.
[37] R. Christen, A. Bergamini, M. Motavalli, NDT & E International 42 (2009) 22.
[38] H.J.M. Jansen, M.M. Festen, Insight 37 (6) (1995) 421.
[39] M. Afzal, S. Udpa, NDT & E International 35 (7) (2002) 449.
[40] V. Babbar, L. Clapham, J. NDT, Evaluation 22 (4) (2003) 117.
[41] F. Förster, I. Parts II, Material Evaluation 43 (1995) 1154 (see also 1398).
388 Z.D. Wang et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 382–388
[42] N.N. Zatsepin, V.E. Shcherbinin, Soviet Journal of Nondestructive Testing 5
(1966) 385.
[43] M.L. Shur, R.V. Zagidulin, V.E. Shcherbinin, Defektoskopiya 3 (1988) 14–25
(in Russian).
[44] R.V. Zagidulin, Defektoskopiya 10 (1998) 33 (in Russian).
[45] W. Lard, J.M. Bridges, W. Yen, R. Palanisamy, Material Evaluation 37 (1978) 47.
[46] K. Mandal, D.L. Atherton, A study of magnetic flux leakage signals, Journal of
Applied Physics 31 (1998) 3211.
[47] C. Mandache, L. Clapham, Journal of Applied Physics 36 (2003) 2427.
[48] I. Uetake, T. Saito, NDT & E International 30 (1997) 371.
[49] G. Dobmann, P. Höller, Physical analysis methods of magnetic flux leakage,
in: Sharpe R.S. (Ed.), Research Techniques in NDT London, Academic Press,
1980.
[50] G. Dobmann, Magnetic leakage techniques in NDT. a state-of-the-art survey
of the capabilities for defect detection and sizing, in: Lord W. (Ed.),
Electromagnetic Methods of NDT, Gordon & Breach, London, 1985.
[51] R.B. Dingle, Asymptotic Expansions: their Derivation and Interpretation,
Academic Press, London, 1973.
[52] F.W.J. Olver, Asymptotics and Special Functions, Academic Press, New York,
1974.
[53] S. Lukyanets, A. Snarskii, M. Shamonin, V. Bakaev, NDT & E International 36
(2003) 51.
[54] M. Katoh, N. Masumoto, K. Nishio, T. Yamaguchi, NDT & E International 36
(7) (2003) 479.
[55] Y. Li, G.Y. Tian, S. Ward, Insight 48 (2) (2006) 103.
[56] F.I. Al-Naemi, J.P. Hall, A.J. Moses, Journal of Magnetism and Magnetic
Materials 304 (2006) e790.
[57] N. Ida, W. Lord, IEEE Transactions on Magnetics 19 (5) (1983) 2260.
[58] D.L. Atherton, British Journal of Non-destructive Testing 30 (3) (1988) 159.
[59] P.A. Ivanov, Z. Zhang, C.H. Yeoh, L. Udpa, Y. Sun, S.S. Udpa, IEEE Transactions
on Magnetics 34 (5) (1998) 3020.
[60] W. Mao, C. Mandache, L. Clapham, D.L. Atherton, Insight 43 (10) (2001) 688.
[61] V. Babbar, B. Shiari, L. Clapham, IEEE Transactions on Magnetics 40 (1)
(2004) 43.
[62] Z.Y. Huang, P.W. Que, L. Chen, NDT & E International 39 (2006) 61.
[63] B. Vijay, B. James, C. Lynann, NDT & E International 38 (2005) 471.
[64] A. Ovanesova, L.E. Suarez, Engineering Structures 26 (1) (2004) 39.
[65] A. Joshi, L. Udpa, S. Udpa, A. Tamburrino, IEEE Transactions on Magnetics 42
(10) (2006) 3168.
[66] T.W. Krause, L. Clapham, D.L. Atherton, Journal of Applied Physics 75 (12)
(1994) 7983.
[67] K. Mandal, D. Dufour, R. Sabet-Sharghi, B. Sijgers, D. Micke, T.W. Krause,
Journal of Applied Physics 80 (1996) 6391.
[68] R. Ranjan, D.C. Jiles, P.K. Rastogi, IEEE Transactions on Magnetics 23 (1987)
1869.
[69] H. Barkhausen, Physics Z 29 (1919) 401.
[70] S. Chikazumi, S.H. Charap, Physics of Magnetism, Krieger Publishing
Company, Florida, 1964.
[71] T.W. Krause, J.M. Makar, D.L. Atherton, Journal of Magnetism and Magnetic
Materials 137 (1–2) (1994) 25.
[72] L. Clapham, S. White, J. Lee, D.L. Atherton, Journal of Applied Physics 88 (4)
(2000) 2163.
[73] C.G. Stefanita, D.L. Atherton, L. Clapham, Acta Materials 48 (2000) 3545.
[74] C. Jagadish, L. Clapham, D.L. Atherton, IEEE Transactions on Magnetics 26
(3) (1990) 1160.
[75] D.C. Jiles, IEEE Transactions on Magnetics 25 (5) (1989) 3455.
[76] M. Lindgren, T. Lepistö, NDT & E International 34 (2001) 337–344.
[77] T.W. Krause, L. Clapham, A. Pattantyus, D.L. Atherton, Journal of Applied
Physics 79 (8) (1996) 4242.
[78] A. Dhar, D.L. Atherton, Non-destructive Testing Evaluation 10 (1993) 287.
[79] R. Pasley, Materials Evaluation (1970) 157.
[80] R. Rautioaho, P. Karjalainen, Journal of Magnetism and Magnetic Materials
73 (1988) 96.
[81] K. Titto, Non-Destructive Testing-Australia 26 (1989) 36.
[82] J.M. Makar, D.L. Atherton, IEEE Transactions on Magnetics 30 (4) (1994) 1380.
[83] L. Clapham, C. Jagadish, D.L. Atherton, Acta Metallurgica et Materialia 39 (7)
(1991) 1555.
[84] T.W. Krause, N. Pulfer, P. Weymann, D.L. Atherton, IEEE Transactions on
Magnetics 32 (5) (1996) 4764.
[85] A.J. Birkett, W.D. Corner, B.K. Tanner, S.M. Journal of Applied Physics 22
(1989) 1240.
[86] X. Kleber, A. Vincent, NDT & E International 37 (2004) 439.
[87] A. Dhar, L. Clapham, D.L. Atherton, NDT & E International 34 (2001) 507.
[88] R.F. Krause, B.D. Cullity, Journal of Applied Physics 39 (1968) 5532.
[89] A. Birkett, W.D. Corner, B.K. Tanner, S.M. Thompson, Journal of Applied
Physics 22 (1989) 1240.
[90] U. Lieneweg, IEEE Transactions on Magnetics 10 (2) (1974) 118.
[91] S. Vaidyanathan, V. Moorthy, P. Kalyanasundaram, T. Jayakumar, R. Baldev,
Metallurgical and Materials Transactions A 30 (1999) 2067.
[92] D.J. Buttle, C.B. Scuby, J.P. Jakubovics, G.A.D. Briggs, Philosophical Magazine
55 (6) (1987) 717.
[93] A.J. Birkett, W.D. Corner, B.K. Tanner, S.M. Thompson, Journal of Applied
Physics 22 (1989) 1240.
[94] L.J. Swartzendruber, G.E. Hicho, H.D. Chopra, S.D. Leigh, G. Adam, E. Tsory,
Journal of Applied Physics 81 (8) (1997) 4263.
[95] D.G. Hwang, H.C. Kim, Journal of Applied Physics 21 (1988) 1807.
[96] C.G. Stefanita, D.L. Atherton, L. Clapham, Acta Materialia 48 (13) (2000)
2545.
[97] A. Dhar, L. Clapham, D.L. Atherton, NDT & E International 34 (2001) 507.
[98] P. Kettunen, P. Ruuskanen, Scandinavian Journal of Metallurgy 8 (1979) 112.
[99] L.P. Karjalainen, M. Moilanen, NDT & E International 13 (3) (1979) 51.
[100] Y. Furuya, H. Shimada, K. Yamada, T. Suzuki, Journal of the Japanese Society
for Non-Destructive Inspection 41 (4) (1991) 215.
[101] M. Lindgren, T. Lepistö, NDT & E International 33 (2000) 423.
[102] Z.J. Chen, A. Strom, D.C. Jiles, IEEE Transactions on Magnetics 29 (6) (1993)
3031.
[103] L.P. Karjalainen, M. Moilanen, IEEE Transactions on Magnetics 16 (3) (1980)
514.
[104] K. Tiitto, Nondestructive Testing Evaluations 5 (1989) 27.
[105] M.R. Govindaraju, A. Strom, D.C. Jiles, S.B. Biner, Z.J. Chen, Journal of Applied
Physics 10 (1993) 6165.
[106] V. Moorthy, B.K. Choudhary, S. Vaidyanathan, T. Jayakumar, K.B.S. Rao,
R. Baldev, International Journal of Fatigue 21 (1999) 263.
[107] O. Saquet, J. Chicois, A. Vincent, Materials Science Engineering A269 (1999) 73.
[108] C. Gatelier-Rothea, J. Chicois, R. Fougers, P. Fleischmann, Acta Materialia 46
(14) (1998) 4873.
[109] D.K. Bhattacharya, S. Vaidyanathan, Journal of Magnetism and Magnetic
Materials 166 (1997) 111.
[110] A. Dhar, D.L. Atherton, IEEE Transactions on Magnetics 28 (6) (1992) 3363.
[111] A.A. Doubov, Welding in World 41 (1998) 196.
[112] A.A. Doubov, Diagnostics of Metal Items and Equipment by Means of Metal
Magnetic Memory, in: Proceedings of CHSNDT 7th Conference on NDT and
International Research Symposium, Shantou China: Non-Destructive Test-
ing Institution, CEMS, 1999,p. 181.
[113] A.A. Doubov, G.V. Vstovsky, Physical Base of the Method of Metal Magnetic
Memory, Moscow, 2000.
[114] J.L. Ren, K. Song, G.H. Wu, J.M. Lin. Mechanism Study of Metal Magnetic
Memory Testing, in: Proceedings of the 10th Asia-Pacific Conference on
Non-Destructive Testing Brisbane, Australia, 17–21 September, 2001.
[115] A.A. Doubov, Inspection Diagnostics 6 (2001) 19.
[116] A.A. Doubov, Principal Features of Metal Magnetic Memory Method and
Inspection Tools as Compared to known Magnetic NDT Methods, in:
Proceedings of the 16th Annual World Conference on Non-Destructive
Testing, Montreal, Canada, 2004.
[117] L.H. Dong, B.S. Xu, S.Y. Dong, L. Song, Q.Z. Chen, D. Wang, NDT & E
International 42 (2009) 323.
[118] Z.D. Wang, K. Yao, K.Q. Ding, NDT & E International 43 (4) (2010) 354.
[119] Z.D. Wang, K. Yao, K.Q. Ding, NDT & E International 43 (6) (2010) 513.
[120] K. Yao, Z.D. Wang, B. Deng, K. Shen, Experimental Mechanics, doi:10.1007/
s11340-011-9490-3.
[121] C.L. Shi, S.Y. Dong, B.S. Xu, P. He, NDT & E International 43 (1) (2010) 8.
[122] J.C. Leng, M.Q. Xu, M.X. Xu, J.Z. Zhang, NDT & E International 42 (2009) 410.
[123] J.W. Wilson, G.Y. Tian, S. Barrans, Sensors and Actuators A 135 (2007) 381.
[124] M. Roskosz, P. Gawrilenko, NDT & E International 41 (2008) 570.
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/216606747

The Physical Foundations of Magnetic Memory

Method for NDT

Conference Paper · January 2004

DOI: 10.13140/2.1.3752.4484

CITATIONS READS

2 606

2 authors, including:

Roberto Suárez-Ántola
Ministry of Industry, Energy and Mining
205 PUBLICATIONS 298 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Roberto Suárez-Ántola
letting you access and read them immediately. Retrieved on: 22 October 2016
 The physical foundations of the Magnetic Memory Method of NDT

Roberto Suárez-Ántola
Diego Suárez-Bagnasco

Department of Electrical Engineering, School of Engineering and Technologies,

Catholic University of Uruguay

Contact address: dsuarezb@adinet.com.uy

Summary

The magnetic memory method (MMM) of non-destructive inspection is based on the analysis of the distribution of
scattering magnetic fields (SMF) which result from strains and structural changes produced during fabrication.
The method was developed by Professor A. Dubov in Moscow and has been used for several decades in Russia,
the former East Germany, Poland, Ukraine, Bulgaria, Yugoslavia, China, India, and other countries. In this paper,
the authors review some aspects of the physical foundations of the method and its relation with several well
known phenomena that link the magnetic field with stress and strain fields in metals (mainly ferromagnetic). A
brief discussion of technical interest about the possibility of an indirect measure of stresses in metal bodies using
measured values of scattering magnetic fields is given.

Keywords: scattering magnetic field, linear and non-linear magneto-elastic effect, dislocations, stress,
strain.

Introduction
The purpose of this paper is to review and discuss some of the foundations of a new magnetic NDT
method, known as the Magnetic Memory Method.
The concept of Metal Magnetic Memory was first introduced by Anatoli Dubov in 1994 (Dubov, 2004).
Apparently, before that time it seems that it was not used in the technical literature, in the sense in
which it is now used in NDT.
The uniqueness of the metal magnetic memory method is that it is based on use of the scattering
magnetic field (SMF). This SMF appears broadly under two circumstances. In first place, it appears
when the temperature of a ferromagnetic metallic body is above Curie’s temperature and it cools
down below the abovementioned temperature to ambient temperature. The resulting magnetization
presents non-uniformities related to structural heterogeneities that modify the pattern of magnetic
domains produced due to the Earth magnetic fields. These non-uniformities are very significant, even
in the case in which the ferromagnetic body cools down in weak magnetic fields as the Earth’s
magnetic one. In second place, but equally important, a non-uniform magnetization appears in
regions with steady dislocation patterns, produced during plastic sliding by local stresses induced by
working loads action. In both cases the variations of magnetization from one point to another in the
material is correlated with a measurable variation in the scattering magnetic field in the surface of the
body. It seems that it is very difficult, if not impossible, to obtain information from SMF with artificial
magnetization in working constructions. Such information is available and can be obtained only in a
small external field, as the Earth magnetic field. This is so because a correlation exists between the
density of elastic or elastic-plastic strain energy and the density of induced magnetic field energy. The
measurable effect over the SMF is higher when the density of magnetic energy produced by local
strains in dislocation patterns is at least of the same order of magnitude as the density of magnetic
energy of the bulk of the material. In a strong external field this last energy density may be so high
that in the SMF the perturbation due to the structural defect may pass unnoticed. It is shown in
practical works that MMM can be used both during equipment operation and after working loads relief
during the repairs. Magnetic texture, formed under the action of working loads, becomes frozen after
unloading by virtue of the same mechanisms of magnetic dislocation hysteresis that explain
hysteresis loops. Thus, there appears a unique possibility to evaluate the actual stress-strained state
of the equipment and to reveal at an early stage maximal damage zones in metal by measuring SMF
using suitable equipment. Moreover, also by measuring SMF, it is possible to asses the quality of new
machine-building products directly after their manufacturing.

In Figure 1 we can see one piece of equipment used to measure SMF on the surface of ferromagnetic
bodies. In this case it is being used to inspect butt welded joints of a pipe.

Figure 1 References 1, 2, 3: flux-gate transducers of the scanner that measure the SMF plus the
external magnetic field. 4: flux gate transducer that measure the external magnetic field only. 5:
wheels and length meter. ∆lb: distance between flux-gate transducers.

Special scanning equipment including not only the known flux-gate sensors but also the length meter,
analogue to digital converter (ADC), the processor and other devices was first developed to register
local micron SMF areas. The MMM characteristics thus obtained correlated well with stress
concentration zones thus detected. Such scanning devices have never been used in other magnetic
NDT methods before the MMM birth (there are no analogues in the world). Scanning devices and the
control method are patented in Russia, Germany and Poland.

Description of the method as applied to NDT welding inspection: basic definitions

according to an ISO standard draft

Metal Magnetic Memory method (MMM)

Method of non-destructive inspection based on the analysis of the distribution of scattering magnetic
fields (SMF) on product surfaces which result from strains and structural changes formed during
welding. The SMF represents the residual magnetization of welded joints after their manufacture and
cooling in the Earth’s magnetic field.

Scattering magnetic field (SMF)

Is the scattering magnetic field measurable on a surface of products and welded joints.

Stress concentration zone (SCZ)

Contour on a product surface limited by a line of change of the sign of SMF normal component (or
zero value of SMF). The SCZ is characterized by an abrupt local change of magnetization value ∆M
relative to a matrix magnetization M0 that appears as an abrupt “splash” of SMF. It is formed in areas
of defect concentrations of metal and impurities when a welded joint is cooled below the Curie point in
the Earth’s magnetic field.
Magneto-mechanical hysteresis
Hysteresis stipulated by domain boundaries fixing on dislocation pile-ups and other structural
heterogeneities during natural magnetization of welded joint metal as it cools below the Curie point in
the Earth’s magnetic field.

Intensity of scattering magnetic field (Hp [A/m])

The numerical characteristic (parameter) of magnetic field intensity measured on the surface of a
welded joint.

Coefficient of intensity of scattering magnetic field change (Kin[A/m2])

The gradient of the measurable Hp field, the ratio of the module of a magnetic field Hp difference to
the distance between the corresponding two points on the surface of the tested body

Magnetic parameter of deformability of metal (m)

max med
The ratio of a maximum field gradient ( К in ) to its medium value ( К in )
The parameter m can be obtained from laboratory and industrial investigations.

Channel of Hp field measurements (H1, H2 etc.)

The magnetic field intensity measured by one flux-gate transducer. Such designation of a measurable
field is used in the multi-channel condition of measurement (two and more flux-gate transducers).

Discreteness of Hp field recording ( ∆l к )

Step of scanning by flux-gate transducer or distance between two adjacent points of an inspection for
each measurements channel

Basic distance ( ∆l b )
Distance between two channels (flux-gate transducers) of magnetic field Hp measurements.

The same empirical parameters are used to describe the SMF for other applications.

Foundations of the method: basic physical principles

Domain structure: a qualitative approach

Ferromagnetism is associated with the exchange forces which hold atoms in place in the metal lattice.
In certain metals, such as iron, nickel, cobalt, gadolinium and the Heusler alloys, the exchange energy
and the different magnetization energies organize these atoms into microscopic groups called
domains. There are usually a large number of domaíns within each metal grain. Within each domain,
the magnetization is saturated, that is, the spins of the magnetically important internal electrons of all
the atoms are aligned. The domain acts magnetically as though it were a single atom. It can rotate.
Moreover, it can grow and shrink (Kittel, 1990).
Before magnetization, the directions of the individual domains are randomly distributed. When an
external magnetic field is applied, two independent processes account for the magnetic response of
the material. The first process is the growth, by boundary displacement, of domains which are
favourably oriented with respect to the field at the expense of domains not so favourably oriented. The
second process predominates at field strengths approaching saturation and is unimportant in weak
fields, is the rotation of the direction of magnetization of the domains into the direction of the applied
field. Predominating in different parts of the magnetization cycle, these processes account for the
non-linear shape of the familiar hysteresis curve of a material (Wert and Thomson, 1970).
Magneto-elastic effects: a thermodynamic approach

As early as 1864, E. Villari observed that mechanical loads alter the magnetic properties of
ferromagnetic materials. The magnetic effects of internal or applied stresses are known to be closely
associated with the well known phenomenon of magnetostriction, discovered by Joule in 1847. For a
historical approach to magnetostriction and magneto-elastic effects, see Hunt (1954).
Magnetostriction is the change of dimensions of ferromagnetic materials upon magnetization. It is not
unreasonable therefore that strain conversely should result in changed magnetic properties. Stress
effects then may be thought of as inverse magnetostriction. If changes are small enough to be
reversible, a relationship can be derived from thermodynamic principles, between tensile stress σ,
strain ε (here change in length per unit length), magnetic field strength H and magnetic induction B, as
follows (Mason, 1958; Vonsovskii, 1975; Herpin, 1968):

 ∂ε   ∂B 
  =  [1]
 ∂H  T , P ,σ  ∂σ  T , P , H

(both partial derivatives taken at constant temperature and pressure)

 ∂ε 
λS =  
 ∂H  T ,P ,σ

is known as the magnetostriction. This equation shows that when λS is positive, as it is, for example,
in iron and permalloy, B will increase with applied tensile stress for a fixed value of H. When λS is
negative, as it is in nickel, B will decrease. Equation [1] applies only over the comparatively small
range of reversible permeability. This is the range of applied fields and stresses within which domain
boundaries can still return to their original position when the field or stress is released.
Since stress is a source of potential energy, the origin of its influence on magnetic properties lies
necessarily in the fact that stress changes the energy condition of the domains. In addition, a number
of other factors contribute to the energy of the domain. These include exchange forces between
atoms in the lattice, the thermal motion of these atoms, magnetic energy stored in the field
surrounding each domain, crystal anisotropy and the presence of impurities and lattice imperfections.
Any magnetic response of the material to applied forces and fields will invariably be due to a re-
adjustment of the size and orientation of the domains so that their energy will be a minimum under the
applied conditions (the thermodynamic theory of magnetic domains was developed by Landau and
Lifchitz in 1935 (Landau and Lifchitz, 1969)).
If λS is positive and the applied stress tensile, then the domains will tend to line up with the stress
direction. However, if λS is negative, the domains will orient themselves more nearly perpendicular to
the stress. In the meantime, the applied magnetic field is trying to line up the domains in the field
direction. If this direction differs from that of the domain stress orientation there will be a competition
between them. A third effect operating on the domains is the crystal anisotropy, which tends to line up
the domains along certain preferred crystallographic axes. The thermal energy of the lattice tries to
destroy the magnetization by randomizing the spin directions of the individual atoms (it succeeds in
this above the Curie temperature). Suffering the effects of the abovementioned factors, the domains
orient themselves in a configuration of lowest energy. Because the price of lining up in any direction is
paid in terms of increased energy of the domains, complete alignment is approached only under
intense applied fields or stresses.
Now,

B=µ0H + M [2]

Then

 ∂B   ∂M 
  =  [3]
 ∂σ  T , P , H  ∂σ  T , P , H
Equation [3] jointly with equation [2] shows that local variation in stresses produce a local variation in
magnetization. However, magnetization is a measure of the relative ease with which the applied
magnetic field can line up the domains.

The scattering magnetic fields

From equations [1] and [3] it follows that λS measures the sensitivity of magnetization M to a variation
in local stress σ. In order to see how this variation can be used for NDT purpose, let us suppose that
the material is in a relatively weak external magnetic field, such as earth’s magnetic field. Let us
suppose also that the material is above its Curie’s temperature. Then it begins to cool down. When
the temperature is below Curie’s point, a macroscopic magnetization emerges, produced by the
external field and due to the abovementioned domain structure. When significant defects in the
material modify the stress distribution, producing internal stresses or modifying external ones, the
local macroscopic magnetization correspondingly varies. Then, accordingly to equation [2], magnetic
induction B varies. If such a variation is big enough, it can be measured and the measured induction
can be correlated with the presence of significant defects in the ferromagnetic material. The spatial
variations of magnetization due to applied or internal stresses produce corresponding spatial
variations in the so called scattering magnetic field. This field is the leakage magnetic field in the air
surrounding the magnetized body. Due to the effect of well known boundary conditions (Landau and
Lifchitz, 1969) both the magnetic field and the magnetic induction in air are almost normal to the air-
r
solid interface. However, inside the material, magnetic induction B is almost parallel to the interface.
Let us consider two points in the air-solid interface, near enough to each other. Due to the
r r
restriction ∇ ⋅ B = 0 , a variation of the tangential component of
B must be compensated by an equal
r
and opposite variation in its normal component. But the normal component of B in the material is
r
equal to the normal component of B in air. Therefore, if magnetization varies inside the body from
r 1 r
point to point, parallel to the interface, the normal magnetic field in air H = B suffers a
µ0
corresponding variation. This variation can be measured displacing a magnetometer over the
interface, and obtaining the corresponding scattering magnetic field gradient.
This is the physical foundation of Magnetic Memory Method for non-destructive testing of materials
(see also Vlasov, 2003).

Figure 2 shows the results of the calculus of the component of the magnetic field in air near the
surface of a cylindrical ferromagnetic pipe, normal to pipe’s surface.

Figure 2
The x axis is taken parallel to pipe’s axis and on the surface of the body. In ordinates we have the
normal component of the magnetic field, as a function of position along pipe’s surface and at a short
and constant distance from that surface. In order to obtain the field, we assume that the ferromagnetic
r r
pipe is uniformly magnetized with magnetization vector M 0 in the external field H 0 due to the Earth.
r
The magnetization vector M 0 is everywhere parallel to pipe axis. The small rectangles on the x axis
show the beginning and the end of the pipe. Figure 2 shows a perturbation in the normal component
of the magnetic field in air, over the right side of the pipe (near the middle of the pipe). This
r
perturbation corresponds to a localized magnetization ∆M . This localized magnetization produces a
field that is superposed with the field produced by the magnetization of the matrix of the ferromagnetic
material. If equipment like the one shown in Figure 1 is displaced along the pipe a measured
magnetic field H r (x ) similar to the pattern shown in Figure 2 could be obtained. Figure 2 shows that
the field produced by the uniformly magnetized matrix of the pipe is significant only near the ends of
the pipe. This is so because the field is measured very near to the surface of the matrix. As a
consequence, near the ends of the pipe, the SMF produced by the uniformly magnetized matrix
doesn’t allow the measurement of the SMF produced by patterns of dislocation. But in between the
SMF of the matrix is almost zero, so that the SMF of another origin can be measured. The origin of
r
∆M in the material could be as follows. Let us suppose that a scanning searching anomalous signals
of SMF is done after manufacturing the steel pipe. Let us suppose that during manufacturing the
material of the pipe was heated above its Curie’s temperature. Then it was cooled in the weak
external magnetic field of the Earth. It gains a degree of magnetization that can be reached at normal
temperature only in a high-intensity magnetic field. Due to the so called “Para process” (Livshits and
others, 1982) this magnetization is correlated with significant strains (the magnetostrictive constant of
the paraprocess is much greater that the magnetostrictive constant at ambient temperature, the so
called technical magnetostrictive constant). As Duvob (2004) put it: “natural magnetization at
manufacturing of machine-building products forms, as a rule, exactly at such conditions. The
mechanism of the product’s real magnetic texture formation (fusing, forging, thermal treatment,
welding) occurs directly after crystallization at cooling below the Curie point. Here the process of
cooling is, as a rule, non-uniform. The metal external layers cool faster than the internal ones.
Thermal stresses form across the product’s volume. They form with the lattice and the appropriate
magnetic texture. They appear mainly in areas of the greatest concentration of lattice defects (i.e.,
clusters of dislocations) and of structural non-uniformity”.
The domain boundaries (DB) get attached and a strong local magnetization is produced that
overcomes average matrix magnetization. The magnetic field thus produced emerges in the surface
of the material and gives a strong signal of SMF when measured as explained above. The points
where the SMF signal goes to zero in between two maxima of opposite sign, as shown in Figure 2,
correspond to the projection onto the surface of lines of residual stress concentration. It was
established in industrial researches that natural magnetization, formed in such a way, reflects the
product’s structural and technological heredity, and as mentioned above, the lines where the field
changes its sign, correspond to the lines of concentration of residual stresses.

A brief discussion and conclusions

The magnetostrictive effect changes its sign when the external magnetic fields are strong enough.
The same occurs with the converse effect that is used as a framework in MMM method. This non-
linear behaviour makes the interpretation of experimental results fairly complex. In principle it is
possible to estimate stresses produced by work loads in ferromagnetic bodies used for structural
purposes or as machine elements. But in this case if magneto-elastic effect is used for stress
calculation, not taking into account the non-linearity of the relation between magnetization and stress
can result in a significant underestimation of the stress.
The MMM method can be applied to scan, depending of the size, near the 100% of a piece. It can
detect stress and strain concentration zones that can not be detected by other NDT methods. It can
be used to follow the evolution of these zones and it can be used for predictive purposes. It can be
combined with more traditional NDT methods to study more in depth the troublesome zones detected
by measuring the scattering magnetic fields in the surface of materials (Suárez-Antola, Vedovatti and
Zirnhelt, 2004).
Acknowledgements

The authors give warm thanks to John Zirnhelt and Eduardo Vedovatti, for several interesting discussions about
the magnetic memory method.

Suggested references

A. Dubov, “Principal features of metal magnetic memory method and inspection tools as compared to
known magnetic NDT methods”, Montreal World Conference on Non-Destructive Testing, August
2004
A. Herpin, “Théorie du magnetisme”, Presses Universitaires de France, Paris, 1968
F. V. Hunt, “Electroacoustics”, Harvard University Press, Cambridge, Ma, 1954
Ch. Kittel, “Introduction to solid state physics”, Wiley, N.Y., 1990
L. Landau and E. Lifchitz, “Électrodynamique des milieux continus”, Mir, Moscow, 1969
B. G. Livshits, V.S. Kraposhin, Y.L. Linetski, “Propiedades físicas de metales y aleaciones”, Mir,
Moscú, 1982
W. P. Mason, “Physical acoustics and the properties of solids”, Van Nostrand, NY, 1958
R. Suárez Antola, E. Vedovatti, J. Zirnhelt “Magnetic Memory Method, a critical review”, Montreal
World Conference on Non-Destructive Testing, August 2004
V.T. Vlasov, “Physical base of the method of metal magnetic memory – own magnetic field of
dislocations” in the 3rd International Conference on Diagnostics of Equipment and Structures using the
Magnetic Memory Method”, Moscow, Russia, February 2003
S.V. Vonsovskii, “Magnetism”, Halstead Press, N.Y., 1975
Ch. Wert and R. Thomson, “Physics of solids”, McGraw-Hill, N.Y., 1970
 7º CONGRESSO BRASILEIRO DE ENGENHARIA DE FABRICAÇÃO
7th BRAZILIAN CONGRESS ON MANUFACTURING ENGINEERING
20 a 24 de maio de 2013 – Penedo, Itatiaia – RJ - Brasil
th th
May 20 to 24 , 2013 – Penedo, Itatiaia – RJ – Brazil

RUÍDO MAGNÉTICO DE BARKHAUSEN EM AÇOS ALTA RESISTÊNCIA

BAIXA LIGA, IF AO TITÂNIO E C-Mn LEVEMENTE DEFORMADOS
PLASTICAMENTE
Pedro Paulo de Carvalho Antônio, pedropaulo86@hotmail.com 1
Marcos Flávio de Campos, mcampos@metal.eeimvr.uff.br 1
Fábio Moreira da Silva Dias, fabio.dias@csn.com.br 2
Julio Capo-Sanchez, jcapo51@yahoo.com ³
Linilson Padovese, lrpadovese@usp.br ³
1
Universidade Federal Fluminense, Av. dos Trabalhadores 420, 27255-125,Volta Redonda, RJ, Brasil.
2
Companhia Siderúrgica Nacional,Centro de Pesquisas, Rua 17 A, Bela Vista, 27259-200, Volta Redonda, RJ, Brasil.
3
Universidade de São Paulo, Escola Politécnica, Av. Prof. Mello Moraes 2463, 05508-900, São Paulo, SP, Brasil.

Resumo: A movimentação das paredes de domínio pode ser interrompida por defeitos na estrutura no material,
também chamados de pontos de ancoragem. Ao variar o campo magnético as paredes de domínio realizam “saltos”
sobre os pontos de ancoragem. Esses saltos geram o Ruído Magnético de Barkhausen (RMB). A emissão de RMB sofre
variação por conta de fatores microestruturais que agem como pontos de ancoragem e impedem o movimento das
paredes de domínio. Baseado nisso o RMB se mostra uma alternativa tecnológica mais barata e eficaz na identificação
de materiais que sofreram deformações plásticas, mudanças na microestrutura ou estão sujeitos a tensões mecânicas
por exemplo. O objetivo desse trabalho é relacionar a emissão de RMB com a deformação plástica, levando em conta
outros fatores como composição química e microestrutura. Para isso avaliou-se a emissão de RMB de três tipos
distintos de aços na condição natural e com diferentes níveis de deformação.

Palavras-chave: Ruído Magnético Barkhausen, (RMB), Ensaio de Tração, Materiais.

1. INTRODUÇÃO

Materiais ferromagnéticos possuem momentos magnéticos espontâneos, ou seja, mesmo sem um campo
magnético aplicado os materiais se comportam como um pequeno imã permanente.
Quando se aplica um campo magnético variável em um material ferromagnético percebe-se um aumento do
campo externo, o aumento desse campo é devido à soma do campo magnético variável aplicado e o campo magnético
do corpo que foi magnetizado, esse aumento é chamado de indução magnética (Delgado Júnior, 2010; Grijalba, 2010).
A variação da indução magnética acontece decorrência da movimentação das paredes de domínios magnéticos
devido à presença do campo magnético variável, mas a indução magnética não é um processo contínuo, as paredes de
domínio têm sua movimentação interrompida na presença dos chamados pontos de ancoragem, e, ao variar o campo
magnético aplicado as paredes de domínio “saltam” sobre os pontos de ancoragem, onde esses saltos na indução
magnética são chamados de Ruído Magnético de Barkhausen (RMB) (Padovese et al, 2012; De Campos et al, 2013).
Como o RMB tem sua origem na movimentação das paredes de domínio magnético, fatores como intensidade
do campo magnético, freqüência de excitação, tensões mecânicas, deformações plásticas e microestrutura afetam a
emissão de RMB.
Em virtude dessa variação, a emissão de RMB vem se mostrando uma alternativa tecnológica mais barata e
eficaz para a identificação de materiais que sofreram deformações plásticas, variação na microestrutura ou estão
submetidos a tensões mecânicas por exemplo (Grijalba, 2010).
O objetivo desse trabalho é relacionar a emissão de RMB com a deformação plástica, levando em conta outros
fatores como composição química e microestrutura. Para isso avaliou-se a emissão de RMB de três tipos distintos de
aço na condição natural e com diferentes níveis de deformação.

©
Associação Brasileira de Engenharia e Ciências Mecânicas 2013
 7 º C O N G RE S S O B R A S I L E I R O DE E N G E N H A R I A D E F A B RI C A Ç Ã O 2 0 a 2 4 d e M a io d e 2 0 1 3 . P e n e d o, I t a t i a i a - R J

2. RUÍDO MAGNÉTICO DE BARKHAUSEN

O físico alemão Heinrich Barkhausen descobriu em 1919 que durante a magnetização de amostras ferrosas,
ruídos eram emitidos, sendo que esses ruídos poderiam ser captados como cliques audíveis por um alto-falante
(Delgado Júnior, 2010).
O experimento de Barkhausen era relativamente simples, consistia na aplicação de uma força magnetomotriz
variável no tempo, aproximando e afastando um imã no formato de U. Através desse experimento foi observado que
durante a magnetização de uma amostra de ferro, pulsos de tensão elétrica de curta duração são induzidos em uma
bobina enrolada em torno da barra (Delgado Júnior, 2010; Huallpa, 2011; Grijalba, 2010).
Mesmo quando o campo de magnetização foi variado lentamente as descontinuidades persistiram,
demonstrando que o fenômeno era um processo intrínsecamente discreto. Além disso, o ruído variava de acordo com o
tipo de amostra testada.

Figura 1: Esquema de um dispositivo para medição de RMB composto de um yoke e uma bobina leitora
(Delgado Júnior, 2010).

O conceito de domínios e paredes de domínios ainda não havia sido estabelecido, o que ocorreu por volta de
1950, depois de estabelecido o conceito de domínios e paredes ficou claro que os sinais ouvidos por Barkhausen eram
devidos aos movimentos e as rotações das paredes de domínio (Delgado Júnior, 2010).
A maioria dos saltos de Barkhausen são produzidos pelo movimento irreversível de paredes de domínio de
180○ e acontecem na região de maior inclinação da curva de magnetização (Grijalba, 2010; Padovese et al, 2012;
Huallpa, 2011).
No avanço do processo de magnetização outros fatores contribuem para a ocorrência de ruído de Barkhausen
como: o movimento das paredes de 90○, rotação de domínios, e aniquilaçao de paredes. Conforme o processo vai se
aproximando da saturação magnética, o movimento das paredes cessa, desaparecendo assim a emissão do RMB. Fatores
como precipitados, contornos de grão, discordâncias e concentradores de tensões atuam como barreia para a
movimentação das paredes de domínio, de modo que a emissão do RMB é sensível aos efeitos produzidos pelo material
(Delgado Júnior, 2010; Grijalba, 2010; Padovese et al, 2012; De Campos et al, 2013).
A seguir são apresentados alguns conceitos sobre como a emissão do RMB se comporta em relação aos
diferentes efeitos produzidos sobre os materiais. Sendo os principais fatores abordados nesse trabalho: microestrutura e
deformação plástica

2.1. Microestrutura

As paredes de domínio estão sempre confinadas dentro de um mesmo grão com mesma orientação
cristalográfica. Sendo assim, os contornos de grão acabam sendo uma barreira intransponível para a movimentação das
paredes de domínio (Grijalba, 2010; Delgado Júnior, 2010; Padovese et al, 2012).
Materiais com tamanho de grão maior a emissão de RMB diminui. Tal diminuição se deve ao fato de que ao
aumentar o tamanho do grão, ocorre uma redução do número de lugares para nucleação das paredes de domínio,
também há uma redução do número de pontos de ancoragem que servem parar o movimento das paredes de domínio, o
que reduz a emissão de RMB (Delgado Júnior, 2010; Franco et al, 2013). Ou seja, o número de paredes de domínio
ativas diminui quando o tamanho de grão aumenta, gerando assim menos sinal (ruído).
No caso de presença de perlita, a emissão de RMB apresente diferença em relação a ferrita. Na ferrita o
tamanho de grão tem grande influência devido ao fato de que os contornos de grão atuam como barreiras na
movimentação das paredes de domínio, já na perlita o formato de lamelas no qual ela se apresenta na microestrutura dos
materiais agem como pontos de ancoragem, fazendo com que as paredes de domínio necessitem de maior energia para
se movimentarem o que consequentemente aumenta a amplitude da emissão de RMB em relação à ferrita (Grijalba,
2010; Delgado Júnior, 2010; Huallpa, 2011; Padovese et al, 2012).

©
Associação Brasileira de Engenharia e Ciências Mecânicas 2013
 7 º C O N G RE S S O B R A S I L E I R O DE E N G E N H A R I A D E F A B RI C A Ç Ã O 2 0 a 2 4 d e M a io d e 2 0 1 3 . P e n e d o, I t a t i a i a - R J

2.2. Deformação Plástica

Quando os materiais são deformados plasticamente, ocorre um aumento da densidade das discordâncias, as
quais irão atuar como barreiras para a movimentação das paredes de domínio (De Campos et al, 2013). Quando as
paredes de domínio se movimentam, conseguindo ultrapassar essas barreiras, ocorre a emissão do RMB. Porém, na
medida em que a deformação aumenta, ocorre significativa fragmentação dos grãos deformados. Ou seja, os pontos de
ancoragem formados na deformação plástica são muito fortes, impedindo a movimentação das paredes de domínio,
diminuindo assim a emissão do RMB (Grijalba, 2010; Padovese et al, 2012).

3. MATERIAIS E METODOLOGIA

3.1. Materiais

Para a realização do estudo proposto nesse trabalho foram confeccionados 15 corpos de prova de três aços
distintos, sendo: 1 – Aço de alta resistência e baixa liga (ARBL) microligado ao Nióbio; 2 – Interstitial Free
estabilizado ao Titânio (IF-Ti); 3 – Carbono-Manganês (C-Mn) correspondente à especificação SAE 1006–1008, sendo
os dois primeiros utilizados no segmento automobilístico para a fabricação de diferentes componentes dos veículos e o
terceiro comumente encontrado no mercado. Os aços utilizados no presente estudo são galvanizados.
As faixas de composição química para cada aço, as espessuras nominais e o percentual de encruamento das
chapas dos aços utilizados estão apresentadas nas tabelas abaixo.

Tabela 1: Composição química dos aços utilizados no estudo.

Aço C Mn P S Ti Nb
ARBL <0,1200 <1,000 <0,030 <0,030 - <0,050
IF–Ti <0,0050 <0,180 <0,030 <0,030 <0,080 -
C-Mn <0,0800 <0,300 <0,030 <0,030 - -

Tabela 2: Espessuras Nominais e Percentuais de Laminação de Encruamento.

Percentual de Laminação de Encruamento

Aço Espessuras Nominais (mm)
(%)
ARBL 1,0 0,8
IF-Ti 1,0 1,2
C-Mn 0,95 1,4

3.2. Metodologia

Para os ensaios foram confeccionados 15 corpos de prova na base 50 mm, na direção de laminação, sendo
cinco corpos de prova de cada aço utilizado nesse trabalho.
O ensaio de tração uniaxial foi realizado para deformar plasticamente os materiais com as variações
percentuais de alongamento propostas nesse trabalho e obter suas propriedades mecânicas.
Do total de corpos de prova, seis deles, sendo dois de cada aço foram submetidos ao ensaio de tração até
atingirem os percentuais de alongamento de 0,5% e 4,0%, que foram propostos para a realização desse estudo.
Dois corpos de prova de cada aço foram submetidos ao ensaio de tração para avaliar suas propriedades
mecânicas e, ainda um corpo de prova de cada aço não foi submetido ao ensaio de tração, com a finalidade de
servir de parâmetro de comparação para a medição do Ruído Magnético de Barkhausen com os demais.
Amostras dos materiais usados foram preparadas para que fossem realizadas análises metalográficas com a
finalidade de caracterizar a microestrutura dos materiais e determinar o tamanho de grão ferrítico.
As medições da emissão de RMB foram realizadas com o sensor colocado perpendicularmente em pontos
na região central das amostras. Para cada ponto no corpo de prova foram realizadas 10 medições, como os materiais
apresentam comportamento diferente, as medições foram feitas em pontos em cima como embaixo dos corpos de
prova.
Os materiais foram magnetizados com uma freqüência de excitação de 10 Hz e uma corrente de 1A
suficiente para atingir a saturação magnética. Como os sinais de RMB possuem amplitude da ordem de microvolts
foi necessário o uso de um amplificador de 5000, que não alterou o comportamento do sinal, mantendo-o constante
durante o processo de medição.
A banda de frequência analisada foi de 150 kHz, isso é dizer que o espectro de frequência da emissão de
RMB terá frequências a partir de 2 kHz (uma vez que valores menores são produtos da magnetização do material e
não são sinais de RMB propriamente ditos) até 150 kHz. Para alcançar essa banda de freqüência de 2 kHz até 150

©
Associação Brasileira de Engenharia e Ciências Mecânicas 2013
 7 º C O N G RE S S O B R A S I L E I R O DE E N G E N H A R I A D E F A B RI C A Ç Ã O 2 0 a 2 4 d e M a io d e 2 0 1 3 . P e n e d o, I t a t i a i a - R J

kHz foi utilizada uma freqüência de exploração de 300 kHz, pois é necessário capturar os sinais com uma
exploração do dobro da freqüência máxima analisada.
Um filtro passa alta foi utilizado para eliminar as freqüências menores que 2 kHz.

4. RESULTADOS E DISCUSSÕES

4.1. Resultados

A partir dos ensaios de tração foi possível obter algumas das propriedades mecânicas das amostras de aço
utilizadas para a realização do estudo. Os valores obtidos estão apresentados na tabela 3 apresentada a seguir.

Tabela 3: Propriedades mecânicas

.
Propriedades Mecânicas
Aço
LEsup (MPa) LEinf (MPa) LE 0,2%(MPa) LR(MPa) AL(%) n
ARBL 377 371 - 448 31 0,17
IF – Ti - - 171 290 51 0,23
C – Mn - - 281 358 36 0,14

A caracterização microestrutural possibilitou identificar as fases presentes nos materiais, e também, determinar o
tamanho médio dos grãos. A figura 2 a seguir mostra os resultados obtidos na caracterização microestrutural das
amostras.

a) b) c)
Figura 2: Microestrutura dos aços: a) Aço ARBL, aumento 200x; b) Aço IF-Ti, aumento 500x;
c) Aço C-Mn, aumento 200x.

A Análise da amostra de aço ARBL apresenta uma microestrutura formada por grãos ferríticos poligonais
heterogêneos associados à cementita lamelar fina precipitada nos contornos de grão. Notou-se também a formação de
perlita.
A amostra do aço IF-Ti apresentou uma microestrutura formada por grãos ferríticos poligonais.
A análise do aço C-Mn mostrou uma microestrutura formada por grãos ferríticos poligonais heterogêneos
levemente alongados associados à cementita globular fina. Notou-se também a formação de perlita, porém, em menor
quantidade que a amostra de ARBL.

A tabela 4 a seguir mostra os tamanhos de grão médio das amostras.

Tabela 4: Tamanho de Grão.

Aço Tamanho de grão em Micrometros Tamanho de Grão Médio (G) - ASTM

ARBL 7,73 10,5
IF-Ti 12,86 9,0
C-Mn 8,45 10,5

O tamanho de grão em numeração ASTM foi obtido conforme a equação 1 que correlaciona o tamanho de grão em
micrometro com o respectivo valor em numeração ASTM:

G  10  6,64 log d (1)

©
Associação Brasileira de Engenharia e Ciências Mecânicas 2013
 7 º C O N G RE S S O B R A S I L E I R O DE E N G E N H A R I A D E F A B RI C A Ç Ã O 2 0 a 2 4 d e M a io d e 2 0 1 3 . P e n e d o, I t a t i a i a - R J

Onde:

G: tamanho de grão em numeração ASTM.

d: tamanho de grão em cm.

Para a análise apresentada nesse trabalho, foram considerados os valores do envelope do ruído RMB, utilizados
para relacionar a influência da microestrutura e das tensões na emissão de RMB. As figuras a seguir (figs. 2, 3 e 4)
ilustram a emissão de RMB em função do tempo para as amostras utilizadas. Essas figuras mostram que o pico diminui
com o aumento da deformação, e que também que o pico de desloca para a direita. O fato do pico se deslocar para a
direita significa que foi necessário um maior campo aplicado para que ocorra movimentação das paredes de domínios.

ARBL

Tempo (s)

Figura 2: Emissão de ruído Barkhausen das amostras de aço ARBL deformadas de 0% a 4%.

IF-Ti

Tempo (s)
Figura 3: Emissão de ruído Barkhausen das amostras do aço IF-Ti deformadas de 0% a 4%

©
Associação Brasileira de Engenharia e Ciências Mecânicas 2013
 7 º C O N G RE S S O B R A S I L E I R O DE E N G E N H A R I A D E F A B RI C A Ç Ã O 2 0 a 2 4 d e M a io d e 2 0 1 3 . P e n e d o, I t a t i a i a - R J

C-Mn

Tempo (s)
Figura 4: Emissão de ruído Barkhausen das amostras de aço C-Mn deformadas de 0% a 4%.

Os gráficos da emissão de RMB para os três aços apresentam resultados semelhantes, no sentido de que
conforme os cp’s foram sofrendo deformações, a emissão de RMB foi diminuindo, o que mostra a tendência de que
conforme o material sofre deformações a emissão de RMB vai diminuindo.
O efeito da laminação de encruamento aparentemente não influenciou de forma significativa a emissão de
RMB, pois os valores apresentaram pequenas variações se comparados um com o outro.
A diferença de valores entre os picos obtidos para os aços se deve a diversos fatores, como composição
química, tamanho de grão, entre outros.

4.2. Discussões

A emissão de RMB apresentou menores valores conforme os materiais foram submetidos a maiores níveis de
deformação plástica. Esse efeito pode ser explicado pelas tensões internas criadas na deformação plástica que dificultam
a movimentação das paredes de domínio, já que, o RMB mede essencialmente a movimentação das paredes de domínio,
na medida em que a deformação aumenta, ocorre fragmentação dos grãos deformados, impedindo a livre movimentação
das paredes de domínio.
Os resultados mostram que para materiais com maiores tamanhos de grão, como no aço IF-Ti a emissão é menor.
Para os aços com menor tamanho de grão como o ARBL e o C-Mn a emissão de RMB é maior devido ao aumento do
número de paredes de domínio ativas.
Relacionando os resultados obtidos para as propriedades mecânicas e a microestrutura dos materiais com os
resultados na emissão de RMB, observa-se que para valores de limite de escoamento, limite de ruptura, menores
tamanho de grão e menores porcentagens de encruamento a emissão de RMB foi maior.
Os picos da emissão de RMB do aço C-Mn apresentaram diferenças maiores entre 0% e 0,5% de deformação se
comparados aos outros materiais analisados. Esse efeito pode ser explicado pela diferença de espessura existente, pois,
como a emissão de RMB é originada pela movimentação das paredes de domínio, qualquer variação na dimensão do
material compromete a emissão de RMB.
Os resultados apresentados na tabela de propriedades mecânicas mostram que o corpo de prova com 0,5% de
deformação se encontra no início do regime plástico, então se espera que os valores da emissão de RMB para 0,5% seja
semelhante a 0%, o que foi observado. Tal resultado também foi observado para o aço ARBL.
O efeito da laminação de encruamento não influenciou a emissão de RMB, pois os valores apresentaram pequenas
variações se comparados um com o outro e o estudo apresentou resultados semelhantes.
Os resultados apresentados para todos os materiais analisados mostraram uma forte tendência em relação à
diminuição da emissão de RMB conforme a deformação plástica aumenta.

©
Associação Brasileira de Engenharia e Ciências Mecânicas 2013
 7 º C O N G RE S S O B R A S I L E I R O DE E N G E N H A R I A D E F A B RI C A Ç Ã O 2 0 a 2 4 d e M a io d e 2 0 1 3 . P e n e d o, I t a t i a i a - R J

5. AGRADECIMENTOS

Os autores agradecem à Companhia Siderúrgica Nacional e ao Laboratório de Dinâmica e Istrumentação (LADIN)

da Universidade de São Paulo pelo fornecimento das amostras utilizadas e por possibilitar a realização dos
experimentos, e também ao CNPq.

6. REFERÊNCIAS

De Campos, M. F., Da Silva, F., LINS, J., Monlevade, E., Alberteris Campos, M., Perez-Benitez, J., Goldenstein,
H., Padovese, L.R., 2013, “Comparison of the Magnetic Barkhausen Noise for low carbon steel in the deformed
and annealed conditions.”, IEEE Transactions on Magnetics.

Delgado Júnior, H. G., 2010, “Modelamento Matemático e Investigação Experimental da Evolução das Tensões de
Soldagem de Aços.”, Tese de doutorado, Universidade Federal Fluminense, Volta Redonda.

Franco, F.A., González, M.F.R., De Campos, M.F., Padovese, L. R., 2013, “Relation Between Magnetic
Barkhausen Noise and Hardness for Jominy Quench Tests in SAE 4140 and 6150 Steels.”, Journal of
Nondestructive Evaluation.

Grijalba, F. A. F., 2010, “Desenvolvimento de Ensaio Não Destrutivo Baseado no Ruído Magnético de Barkhausen
para Caracterização de Tensões Elásticas e Deformações Plásticas em Aços.”, Tese de Doutorado, Escola
Politécnica da Universidade de São Paulo, São Paulo.

Huallpa, E. A., 2011, “Caracaterização da Transformação Martensítica em Temperaturas Criogênicas.”, Tese de

Mestrado, Escola Politécnica da Universidade de São Paulo, São Paulo.

Padovese, L.R., Capó Sánchez, J., De Campos, M.F., 2012, “Magnetic Barkhausen emission in lightly deformed
AISI 1070 steel”. Journal of Magnetism and Magnetic Materials, vol. 324, pp. 11-14.

MAGNETIC BARKHAUSEN NOISE IN HIGH RESISTENCE LOW ALLOY,

TITANIUM-IF AND C-Mn STEELS LIGHTLY DEFORMED PLASTICALLY
Pedro Paulo de Carvalho Antônio, pedropaulo86@hotmail.com 1
Marcos Flávio de Campos, mcampos@metal.eeimvr.uff.br 1
Fábio Moreira da Silva Dias, fabio.dias@csn.com.br 2
Julio Capo-Sanchez, jcapo51@yahoo.com ³
Linilson Padovese, lrpadovese@usp.br ³
1
Universidade Federal Fluminense, Av. dos Trabalhadores 420, 27255-125,Volta Redonda, RJ, Brasil.
2
Companhia Siderúrgica Nacional,Centro de Pesquisas, Rua 17 A, Bela Vista, 27259-200, Volta Redonda, RJ, Brasil.
3
Universidade de São Paulo, Escola Politécnica, Av. Prof. Mello Moraes 2463, 05508-900, São Paulo, SP, Brasil.

Abstract: The movement of the domain walls can be interrupted by defects in the structure of the materials also called
pinning sites. With the change of the applied magnetic field, the domain walls jump between pinning sites. These jumps
produce the Magnetic Barkhausen Noise (MBN). The emission of MBN depends on microstructural features acting as
pinning sites, impeding the domain wall moving. With basis on these behaviors, the MBN is an cheap and efficient
technological alternative for identifying materials that were submitted to plastic deformations or microstructural
changes or are under mechanical stresses. The objective in this study is obtaining a relationship between MBN and
plastic deformation, taking into account chemical composition and microstructure. Thus, the MBN emission was
evaluated for three different steel types with different levels of deformation.

Keywords: Magnetic Barkhausen Noise, (MBN), Tensile Test, Materials.

The authors are the only responsible for the printed material included in this paper.

©
Associação Brasileira de Engenharia e Ciências Mecânicas 2013