Analyse du cisaillement simple en rupture
Analyse du cisaillement simple en rupture
T~ = 0 T~ = 0
0 pointe
1
front
Situation bidimensionnelle
Situation tridimensionnelle
Figure 1:
Figure 2: [1]
Figure 3: [2]
P
debut de la propagation
P
debut de la decharge
δ
P δ
Figure 4:
P debut de la propagation
debut de la decharge
debut de la
plasticite
Figure 7:
Figure 8: [2]
Élasticité plane:
Il existe deux cas : contraintes planes et déformations planes.
Généralement, les pièces de faible épaisseur sont en état de
contraintes planes, alors que les pièces de grande épaisseur sont en
état de déformation plane.
∆∆φ = 0
1 1 1
σrr = φ,θθ + φ,r , σθθ = φ,rr , σr θ = − φ,θ (1)
r2 r r ,r
u1 = u2 = 0 et u3 = u3 (x1 , x2 )
1 1
ε13 = u3,1 et ε23 = u3,2
2 2
σ31 − i σ32 = f 0 (z )
Singularité des contraintes en front de ssure
x1 x1
Mode I Mode II
σrθ σrr
σθθ
σrr P
σrθ σθθ
r
θ
x
k 4 − [α2 + (α + 2)2 ]k 2 + α2 (α + 2) = 0,
dont les racines sont k = ±α, k = ±(α + 2).
ψ(θ) est donc somme de e i αθ , e −i αθ , e i (α+2)θ , e −i (α+2)θ , ce qui donne
(en prenant la partie réelle):
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
ψ(θ) = Aα sin(αθ)+Bα cos(αθ)+Cα sin((α+2)θ)+Dα cos((α+2)θ), (5)
Bα + Dα = 0 Cα = −Aα α+α 2
( (
⇔ (8)
Aα α + Cα (α + 2) = 0 Dα = −Bα
Si α est demi-entier cos(απ) = 0 les équations (7) se réduisent à:
Aα + Cα = 0 Cα = −Aα
( (
⇔ (9)
Bα (α) + Dα (α + 2) = 0 Dα = −Bα α+α 2
1
α > −1 ⇔ α ≥ −
2
√KI 3θ KII 3θ
θ θ
σrr = 4 2πr (5 cos 2 − cos 2 ) + 4 2πr (−5 sin 2 + 3 sin 2 )
√
σθθ = 4√K2Iπr (3 cos 2θ + cos 32θ ) + 4√K2IIπr (−3 sin 2θ − 3 sin 32θ )
σ = √KI (sin θ + sin 3θ ) + √KII (cos θ + 3 cos 3θ )
rθ 4 2π r 2 2 4 2π r 2 2
(11)
Remarque1
u1 = 21E 2rπ (1 + ν) KI (2κ − 1) cos θ2 − cos 32θ + KII (2κ + 3) sin θ2 + sin 32θ
q h i h i
i ,
u2 = 21E 2rπ (1 + ν) KI (2κ + 1) sin θ2 − sin 32θ − KII (2κ − 3) cos θ2 + cos 32θ
q h i h
où :
3−ν
(
κ= 1+ν , contraintes planes
[[u1 ]] = 0 ⇔ KII = 0
[[u2 ]] 6= 0 ⇔ KI 6= 0
En mode II:
[[u1 ]] 6= 0 ⇔ KII 6= 0
[[u2 ]] = 0 ⇔ KI = 0
Remarque
KI toujours ≥ 0, jamais négatif !
KII de signe quelconque
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
Singularité des contraintes en mode III
x1
x3
x2
σ23
σ13 σ13
r σ23
θ 1
Aα sin(απ) = 0
(
σ32 (±π) = 0 ⇔ Aα sin(±απ) + Bα cos(±απ) = 0 (α 6= −1) ⇔
Bα cos(απ) = 0
1
Si α = n+ , n ∈ Z, cos(απ) = 0 donc Aα = 0, Bα quelconque.
2
(13)
(
KI < KIc ⇒ pas de propagation
KI ≥ KIc ⇒ propagation
B
P
B0 Epaisseur
D=4m
50
MPa
√
Acier Limite élastique (MPa) Tenacité (MPa m)
A 1250 90
B 900 120
C 650 190
1
0
0
1
0
1
0
1
0
1
0
1
0
1
p 0
1
0
1
0
1
0
1
0
1
R 0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1 e
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
σzz σzz
pR
p π R 2 = σzz 2π Re ⇒ σzz =
2e
σθθ
pR
p 2R = σθθ 2e ⇒ σθθ =
e
σrr est négligeable devant σzz et σθθ (épaisseur faible). On a donc:
0 0 0
pR
[σ] = 0 1 0
e
0 0 1/2
−1/2 0 0
pR
⇒ [s ] = 0 1/2 0
e
0 0 0
√ √
pR π a pR π a
KIc = 1, 12 0 ⇒ emin
0
= 1, 12
emin KIc
On obtient pour les aciers A,B,C : 15,6 cm; 11,7 cm; 7,4 cm.
Finalement, l'épaisseur minimale (compte tenu des deux
mécanismes de ruine) imposée par les aciers A,B et C est : 15,6
cm; 11,7 cm; 13,3 cm. Il convient donc de choisir l'acier B avec
l'épaisseur 11,7 cm.
√
K = Y σ π a, Y est un facteur adimensionnel, appelé facteur de forme
x2
σ∞
−a x1
a
σ∞
√
KI = σ∞ π a
σ∞
√
KI = 1, 12σ∞ π a
Remarque
K = KA + KB + KC
Exemple
Poutre ssurée de section rectangulaire d'épaisseur B (a/w=0.2).
a
P W P
M M
KI = KIexion + KItraction
Pour la traction :
0.857+ 0.265a/w
Y (a/w ) = 0.265(1 − a/w )4 +
(1 − a/w )3/2
P √
KItraction = 1, 38 πa
BW
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
d'après Tableau :
M √
6
KIexion = 1.055 πa
BW 2
Forces imposees
da
a
Deplacements imposes
− d Π − d Γ = dK (17)
Remarques
avec :
A : surface de la ssure
−d Π − γs dA = dK
1 ∂Π
− d Π − 2γs Bda = Bda − − 2γs = dK (19)
B ∂a
On pose :
1 ∂Π
G =− , : taux de restitution de l'énergie potentielle J/ m2
B ∂a
(G − Gc ) Bda = dK , (20)
G ≥ Gc , Critère de Grith
∂G
(
∂a > 0, propagation instable
∂G
∂a < 0, propagation stable
Plaque de verre
2a
T 2 aπ
Plaque mince innie ⇒ G= , E: module de Young.
E
Gc E
r
Rupture ⇒ G = Gc ⇒ σr =
πa
(MPa)
5
4
σr
3
2
10 15 20 ( √1m )
√1
a
v h
a x
Bh3
I= moment d'inertie de la section.
12
D'aprés le théorème de Clapeyron : = W Fv = 2UE
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
F 2 a3 1 ∂Π F 2 a2
Π = UE − W = −UE = − ⇒G =− =
3EI B ∂a BEI
On constate que G(a) est une fonction croissante. La propagation
de la ssure est donc instable.
Déplacement contrôlé
coin rigide
111111
000000
000000
111111
000000
111111
000000
111111
000000
111111 v a h x
000000
111111
000000
111111
000000
111111
000000
111111
000000
111111
3EIv 2 ∂Π 9EIv
2
Π = UE = G =− =
4a 3 B ∂a B 4a4
On constate que G(a) est une fonction décroissante, la propagation
est donc stable.
0000000000000000000000
1111111111111111111111
1111111111111111111111
0000000000000000000000
0000000000000000000000
1111111111111111111111
0000000000000000000000
1111111111111111111111
1
Théorème de Clapeyron : UE = Pu (a)
2
1
W = Pu (a), Π = − Pu (a)
2
∂Π P du
1
G =− =
B ∂a 2B da
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
Sachant que u (a) = C (a)P , on obtient :
P 2 dC
G= formule de la complaisance
2B da
P du P dQ
u (a ) = ⇒ =− 2
Q (a ) da Q da
P 2 dQ1 u 2 dQ
G =− = − formule de la raideur
2B Q 2 da 2B da
Détermination expérimentale de Gc
Les formules de la raideur et de la complaisance possèdent une
interprétation graphique trés simple sur la courbe
eort-déplacement, d'où découle une méthode de mesure de G
(donc de Gc , puisque G = Gc lorsqu'il y a propagation):
B
a
a+ ∆a
H
u
O
K
Aire OAB w Aire OAH (on néglige ABH car ∆a est trés petit)
1 1 1 dQ
Aire OAB w u 2 Q (a) − u 2 Q (a + ∆a) = − u 2 ∆a = Gc ∆aB
2 2 2 da
Realtion entre G et K (Formule d'IRWIN)
On va montrer que les FIC (notion locale) et le taux de restitution
d'énergie potentielle G (notion globale) sont reliés par la formule
suivante :
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
K
2
1−ν 2
G =
E (KI2 + KII2 ) + 2III
µ en déformation plane
1 2 KIII
2
G =
E (KI + KII2 ) + 2µ en contrainte plane
Démonstration-Intégrale de fermeture
G ∆a représente l'énergie potentielle élastique restituée par la
structure lors d'un accroissement de ssure de ∆a. Il est aussi égal
au travail eectué pour refermer la ssure de ∆a .
Soit une ssure d'une certaine longueur (Etat I). Supposons que la
ssure se referme d'une petite longueur ∆a (Etat II). Calculons le
travail eectué lors de cette fermeture. Nous supposons que la
ssure est en mode I et en déformation plane :
∆a
x1 x1
Z ∆a
1 II [[u ]]I dx
W = G ∆a = σ22 2 1
2 0
Puisque ∆a est petit, on prend les expressions asymptotiques pour
les contraintes et les déplacements :
KI ∆a − x1
r
4(1 − ν)
σ II
22 =√ , [[u2 ]]I = KI
2π 1 x µ 2π
R ∆ a q ∆ a − x1 ∆a π
sachant que :
0 x1 dx1 = 2 , on obtient :
KI2 (1 − ν 2 )
G=
E
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
Intégrale de RICE
Considérons un contour quelconque entourant la pointe de la
ssure.
x2
x1
Γ
w = 12 σij εij
(
w : densité d'énergie de déformation
Démonstration de J=G
On se met en mode III (élasticité anti-plane):
u1 = u2 = 0 u3 = u (x1 , x2 )
On peut montrer que :
µ
w= 5 u · 5u 5u : gradient de u
2
∂u
σij ui ,1 nj = µ u ,1
∂n
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
Donc, l'intégrale de Rice en élasticité anti-plane devient :
∂u
Z
µ
J= ( 5 u · 5un1 − µ u,1 )ds
Γ 2 ∂n
On choisit comme contour Γ un cercle de rayon r trés petit autour
de la pointe:
x2
eθ n=er
r
e2
θ
x1
e1
∂u 1 ∂u
5u = e~r + e~
∂r r ∂θ θ
FERDJANI Hicheme, Blida, h_ferdjani@[Link] Mécanique de la rupture fragile
Puisque r est petit, on prend le premier terme (asymptotique) de u:
KIII r
r
2 θ
u= sin
µ 2π 2
K θ
u,1 = 5u · e~1 = − √III sin
µ 2π r 2
∂u ∂u KIII θ
= = √ sin
∂n ∂r µ 2π r 2
2
KIII
5u · 5u =
µ2 2 π r
En remplaçant dans J, obtient après simplication :
2
KIII
Z π
1 θ
J= ( cos θ + sin2 )d θ
µ2π −π 2 2
2
KIII
J= =G
2µ
δ : deplacement
h
x1
h Γ
λ −ν1 δ2
w= (tr ε)2 + µtr (ε2 ) = E
2 2(1 + ν)(1 − 2ν) h 2
d'où :
E (1 − ν)δ 2
J=G =
(1 + ν)(1 − 2ν)h
EG Eδ
r
KI = =
1 − ν2 (1 + ν) (1 − 2ν)h
p
K
2
1−ν 2
G =
E (KI2 + KII2 ) + 2III
µ en déformation plane
1 2 KIII
2
G =
E (KI + KII2 ) + 2µ en contrainte plane
KI KII
σθθ (r , θ) = √ cos3 (θ/2) + √ [−3 sin(θ/2) cos2 (θ/2)] (21)
2π r 2π r
K 3 KII 3
! " # !
√ I
θ̂ θ̂ θ̂ θ̂θ̂ θ̂
− cos2 sin +√ 3 sin cos sin − cos3 =0
2π r 2 2 2 2π r 2 2 2 2 2
" #
3 θ̂ θ̂ θ̂ θ̂ 1 θ̂
⇒ KI cos2 sin = 3KII cos sin2 − cos2
2 2 2 2 2 2 2
θ̂ 1 θ̂ 1
sin2 − cos2 = (1 − 3 cos θ̂)
2 2 2 4
θ̂ θ̂ 1
KI cos sin = KII (1 − 3 cos θ̂)
2 2 2
KIc
σc = σθθ (rc , 0) = √ (23)
2π rc
KI KII KIc
σθθ (rc , θ̂) = √ cos3 (θ̂/2) + √ [−3 sin(θ̂/2) cos2 (θ̂/2)] = √
2π rc 2 π rc 2π rc
(24)
Léquation (24) donne le critère de propagation en mode mixte:
" #
r 2
1 KI KI
KII > 0
− +8
4 KII KII
θ̂
tan = " r 2
#
2
1 KI KI
KII < 0
+ +8
4 KII KII
Critère de propagation :
Y y
X
2σ 2a φ x
2σ
√
On donne : KIc = 45MPa m, 2a=0.02 m, φ = 30◦ . On demande :
1 Angle de bifurcation
2 contrainte critique σc
Remarque
KIc
σc = √
π a (2 sin 2
φ + cos2 φ) cos3 (θ̂/2) + 3 sin φ cos φ sin(θ̂/2) cos2 (θ̂/2)
=298.64 MPa
III
II
I
KImax = KIc
log∆KI
log∆K0
da
= C (∆KI )m
dN
avec C et m des constantes matérielles. C'est le domaine de la
fatigue à grand nombre de cycles (stade II).
da
= C (∆KI )3 C = 0.42 × 10−11 (cycle .MPa3 .m1/2 )−1
dN
KIII √
σ23 (θ = 0) = √ ⇒ KIII = τ π a
2π r
1 1 00
σ11 = Re [φ − 2 z̄ φ − 2 ψ ]
0 00
σ11 = Re φ − x2 Imφ
0 00
σ22 = Re φ0 + x2 Imφ00
σ12 = −x2 Re φ00
σ22 = x2 Imφ00
σ12 = −Imφ0 − x2 Re φ00
x2
p2(t)
p1 (t) x1
−a p1(t) a
p2 (t)
x2
σ∞ σ∞
x2 x2
σ∞
=
x1 +
−a a x1
x1 σ∞
σ∞
Probleme 2
Probleme I
σ11 = Re φ0 −x2 Imφ00 , σ22 = Re φ0 +x2 Imφ00 +σ∞ , σ12 = −x2 Re φ00
KI √
σθθ (θ = 0) = √ ⇒ KI = σ∞ π a
2π r
Z a
a+t
r
1
KII = √ p1 (t ) dt .
πa −a a−t
x2
σ∞ σ∞
0
1 x2
x2 0
1 σ∞
111
000 0
1
0
1
0000
1111 0000 x1
1111 0
1
00000
11111
1111
0000 1111
0000 0
1
0000
1111 0000 000
1111 111 0
1 1111
0000 1111
0000
−c −a a c 0
1
0
1 0000
1111 0000x1
1111
x1 σ∞
σ∞
Probleme 2
Probleme I
c +t t t
Z r
dt = c arcsin − cos arcsin ,
c −t c c
(25) donne :
a a
−2σc arccos + σ∞ π = 0 ⇔ c = .
c cos πσ
2σc
∞
!
1
c −a = −1 a
cos 2πσ ∞
σ+c
Boundary conditions significantly influence the solutions to the Beltrami equation, which governs potential functions characterizing stress fields in crack mechanics. In cases of a crack with conditions like σ_θθ(r, ±π) = 0, these define the behavior of the potential functions along the crack surfaces, necessitating solutions for α that satisfy both the differential equation and the boundary limits. This results in unique formulations for stress components where both integer and semi-integer values of α yield non-trivial solutions, reflecting the crack's stress distribution continuity and symmetry .
The stress intensity factor (SIF), denoted as K, characterizes the stress state near the tip of a crack under loading conditions and is essential in describing the singular stress fields in fracture mechanics. It provides a measure of the severity of the crack tip stress field and delineates conditions for crack propagation. Stress asymptotic expressions illustrate the dominance of terms proportional to r^{-1/2} as r approaches 0, demonstrating the singular nature of stresses near the crack tip. The SIF, therefore, appears in these asymptotic expressions as a scaling factor, directly relating to the energy release rate G, further linking local stress states to global fracture energy criteria .
The dominance of the r^{-1/2} term in asymptotic stress expressions near a crack tip reflects the singular nature of stress fields in the vicinity of the crack. This term emerges from the Fourier-Laplace transformation of stress solutions characterized by the mode I, II, or III stress intensity factors and aligns with the geometric singularity created at the crack tip. As r approaches zero, the r^{-1/2} term becomes significant, overpowering other stress components, which dictates the asymptotic stress behavior crucial for fracture mechanics analysis and prediction of crack growth behavior under varying loading conditions .
The Rice integral, denoted J, represents the energy release rate for a crack growth process, quantifying the change in potential energy of a system per unit increment of crack extension. This integral is defined as J = ∫_Γ (w·n₁ - σij·ui,1·nj) ds, where it calculates the balance of mechanical work and deformation energy around the crack tip. Being path-independent, it offers a robust measure of energy flow and can be used for detecting crack initiation and progression, independent of the specific contour path chosen, which makes it crucial for assessing fracture toughness and instability .
Integer and semi-integer values of α lead to distinct solution forms for the fourth-order differential equation derived from the Beltrami equation in crack mechanics. Integer values of α resolve in the equation's terms simplifying under boundary conditions to yield valid non-trivial solutions, as terms become multiply zero when sin(απ) = 0. Similarly, semi-integer values, where cos(απ) = 0, offer solutions under different boundary conditions. These conditions define whether coefficients like Aα, Bα, Cα, and Dα become zero or are interrelated non-trivially, dictating resultant stress distributions around crack tips .
The crack opening displacement in mode III is proportional to the mode III stress intensity factor (KIII). Specifically, the displacement is expressed as u3 = 2KIII/μ√(r/2π) sin(θ/2), showing that the opening displacement increases with the intensity of the applied stress at the crack tip proportionally to KIII. This relationship underscores how local stress fields in shear mode loading affect crack face displacements, linking deformation directly to the stress intensity factor .
Compliance, defined as the change in displacement per unit load, is used in experimental fracture mechanics to determine the critical strain energy release rate, Gc, by observing changes in compliance with crack length during loading. This relationship is expressed as Gc = P²/2B(dC/da), where P is the applied load, B the specimen thickness, and dC/da the rate of change of compliance with crack length. By measuring these variations experimentally, one can estimate the energy released per unit crack extension, providing a direct method for determining material toughness .
The concept of compliance, when related to crack propagation, involves measuring how easily a structure deforms under applied load. By plotting load-displacement curves and determining change in compliance with respect to crack length, one derives the energy release rate, G. This measure is done by G = -P²/2B·dQ/da, where changes in stiffness and flexibility (du/da and dC/da) quantify the energy required to extend a crack. This method offers an experimental approach to determine fracture toughness characteristics of materials, essential for understanding failure mechanisms .
Griffith's criterion relates to stable and unstable crack propagation through the balance between available energy and energy required to create new crack surfaces. When the energy release rate G exceeds the critical energy release rate Gc, the crack propagates unstably (∂G/∂a > 0), leading to rapid fracture. However, if G is less than Gc, the crack does not propagate spontaneously, indicating potential stability unless external conditions such as increased loading change this balance. The criterion, therefore, describes the transition between stable and unstable states as a function of material properties and loading conditions .
The biharmonic nature of the Airy function, expressed by the equation ΔΔΦ = 0, is fundamental in plane elasticity as it ensures the compatibility of strains by satisfying the equilibrium equations devoid of body forces. It allows for the description of stress fields in a plane by deriving it from a potential function, offering a path to solutions using complex variable methods. This is crucial in characterizing stress states and deformation patterns in elasticity problems, enabling the application of boundary conditions to find specific problem solutions .