Beruflich Dokumente
Kultur Dokumente
net/publication/26850765
CITATIONS READS
4 838
4 authors, including:
Kottner Radek
University of West Bohemia
32 PUBLICATIONS 30 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Robert Zemčík on 28 May 2014.
D Delamination of composites
Mode B This type of failure is driven by the influence of interlaminar
shear stress or normal stress between adjacent laminae. The
mechanism of failure is similar to the problems of fracture
mechanics. The interlaminar fracture toughness is usually
Mode C
considered to be the critical value of the strain energy release
rate, denoted as Gc, or critical value of the stress intensity
factor Kc. Suitable experimental measurements must be carried
out to assess these values and for subsequent numerical
simulations. The interlaminar toughness GIc (corresponding to
Figure 1. Mechanisms causing matrix failure. Mode I loading) can be assessed according to the standard
YC S
R= , p ⊥( + ) = p ⊥( − ) = p ⊥⊥
(−)
,
(−)
S C = S 1 + 2 p ⊥⊥
2(1 + p ⊥⊥
(−)
) R
BULLETIN OF APPLIED MECHANICS 4(14), 81–87 (2008)
ASTM D5528-01 [1] using DCB (Double Cantilever Beam) specimen (beam) is not clamped perfectly (there can occur
specimens. certain rotation in the cross-section at crack tip), the ASTM
The specimens are laminated rectangular strips with suggests using a correction with additional crack length ∆,
included non-adhesive aluminum foil which serves as the hence the GIc is actually calculated as
initial crack (see Figure 3). The specimen is placed in testing 3F δ c . (4)
G Ic =
machine using piano hinges and thus loaded by tensile force F. 2b (a + ∆ )
The dependence of force F on the grip opening δ is recorded,
E Progressive failure analysis
together with the crack length. The critical value GIc can then
be readily evaluated from this dependency. The strain energy Several failure criteria were mentioned in the preceding
release rate is mathematically defined, in general, as section. These criteria can be used the find the location and
1 dU , moment of failure but whether and how the failure is going to
G=− (1)
b da propagate cannot be decided. One of the methods that can be
where dU is the increase in potential energy of internal forces, used for such simulation is a combination of so-called
b is specimen width and da is increase in crack length. In case progressive failure analysis (PFA) with, for instance, the
of Mode I loading the formula can be transformed using the material degradation strategy. During the numerical simulation
solution of Euler-Bernoulli beam theory to performed usually by FEA (finite element analysis), the
12F 2 a 2 degradation is applied to the elasticity constants of the material
GI = 2 2 , (2) (characterized by stress-strain matrix C), thus reducing the
b h E
local stiffness. Three main types of failure can occur in general
where E is the Young’s modulus in the axial direction and h is – for matrix failure transverse to lamina plane C reduces to
half-thickness of the specimen. According to ASTM [1] the CM, for fiber failure to CF and for simultaneous failure (or
formula which is used for the evaluation of interlaminar subsequent failure of both matrix and fibers) to CM/F.
toughness is in the form The constitutive relation for unidirectional fiber composite
3F δ
GIc = c c , (3) (modeled as transversely isotropic material) in the case of
2b a undamaged element can be written as
where Fc and δc are the critical values of load und σ = C ε, (5)
displacement at which the crack propagation occurs. Since the
where σ and ε are stress and strain vectors and
Table II. Calculation of LaRC04 failure criterion failure indices.
Fiber failure conditions
LaRC #3 σL ≥ 0 σL
FI F =
XT
LaRC #6 σ L < 0 , σ 2m 2m ≥ 0 σ 2m 2m σ 2m 2m
2
σ 1m 2 m
2
FI M / F = (1 − g ) T + T + L
Y Y S
LaRC #1 σT ≥ 0 σ σ τ
2 2
FI M = (1 − g ) TT + g TT + LTL
Y Y S
2 2
LaRC #5 σ T < 0 , σ L < −Y C τ Tm τ Lm
FI M = T
T m
+
L
L m
S −η σ n S −η σ n
2 2
LaRC #2 σ T < 0 , σ L ≥ −Y C τT τL
FI M = T T
+
L L
S −η σ n S −η σ n
BULLETIN OF APPLIED MECHANICS 4(14), 81–87 (2008)
Figure 3. Progressive failure analysis for quasi-static and non- In numerical simulation using FEA and Puck’s criterion, the
stationary problems (dashed). specimens were loaded in tension in displacement control, thus
BULLETIN OF APPLIED MECHANICS 4(14), 81–87 (2008)
MSC.Marc. Different finite element mesh was prepared for Table IV. Measured and calculated parameters on pre-
each calculated case with certain crack length a [16]. The cracked DCB specimen.
equality between values of energy release rate and J-integral in
Initial crack length [mm] a 60.37
the case of elastic analysis was used in the simulations. The
code calculates the J-integral using the DeLorenzi method [2].
Crack length correction [mm] |∆| 7.41
foil
IV. CONCLUSION
The failure analysis of three different composite specimens
is performed in this work. The problems were both static and
dynamic. It can be concluded that the choice of failure criteria
plays an important role. Another important aspect is the
piano hinge reliability of the computational model, namely the material
properties used therein. The manufacturer of the composite
provides certain set of data which must be considered as
referential only. One option is to calculate the properties from
the properties of constituents but in order to obtain reliable
data it is necessary to carry out series of experiments and
foil corresponding numerical simulations [14, 15]. Currently, new
methodologies for faster and more precise identification of
Figure 9. DCB specimen (Mode I) experimental setup. material properties are investigated, as it shows that the failure
analysis has been of constantly growing importance by the
The values of GIc (and its limits GIcmin and GIcmax as shown in practice.
Table 4) were calculated from experimental data using (4) and
then used in the numerical model. As opposed to the theory, REFERENCES
the critical values change during the test. The reconstructed [1] ASTM D5528-01, Standard Test Method for Mode I Interlaminar Fracture
Toughness of Unidirectional Fiber-Reinforced Polymer Matrix Composites.
load-displacement curves are compared in Figure 10 Annual Book of ASTM Standard, pp. 249-260, 2002.
corresponding curve obtained experimentally. It is obvious [2] MSC.Marc Volume A: Theory and user information, Version 2005.
that for the two limit values the numerical simulation yields MSC.Software Corporation, 2004.
[3] Berthelot J. M.: Composite materials: Mechanical behaviour and
two different dependencies one of which approaches the
structural analysis. Springer – Verlag, New York, 1999.
experimental data at the beginning of delamination, while the [4] Knight N. F. Jr., Rankin C. C., Brogan F. A.: STAGS computational
latter at the moment of final rupture. Only in the case when the procedure for progressive failure analysis of laminated composite structures.
approximated values of GIc as a function of crack length a are International Journal of Non-Linear Mechanics, 37, pp. 833-849, Pergamon,
2002.
used it is possible to obtain good agreement with experimental [5] Kroupa T.: Poškození kompozitů vlivem rázu. Doktorská práce,
F vs. δ dependency. Západočeská univerzita v Plzni, Plzeň, 2006.
[6] Kroupa T., Zemčík, R., Laš V.: Progressive failure analysis of orthotropic
plate loaded by transverse low-velocity impact. In: 5th International Congress
of Croatian Society of Mechanics, Trogir/Split, 2006.
[7] Laš V., Zemčík R.: Progressive damage of unidirectional composite
panels, Journal of Composite Materials, Vol. 42, No. 1, pp. 25-44, 2008.
[8] Laš, V., Zemčík, R., Kroupa, T.: Numerické simulace porušení kompozitů
a jejich experimentální ověření. Transfer. 2007, roč.2, č.3, s.42-67, ISSN
1801-9315.
[9] Laš V., Zemčík R., Měšťánek P.: Numerical simulation of composite
delamination with the support of experiment. Acta Mechanica Slovaka, 1,
EAN 2006, TU Košice, 2006.
[10] Puck A., Schürmann H.: Failure analysis of FRP laminates by means of
physically based phenomenological models. Composites Science and
Technology, 58, pp. 1045-1067, Elsevier, 1998.
[11] Silvestre, T., P., Dávila, G., C., Camanho, P., P., Iannucci, L., Robinson,
O.: Failure models and criteria for FRP under in-plane or threedimensional
stress state including shear non-linearity. Research report, NASA/TM-2005-
213530, NASA Langley Research Center VA 23681 USA, 2005.
Figure 10. Comparison of load-displacement curves on pre- [12] Sleigh D. W.: Progressive failure analysis methodology for laminated
cracked DCB specimen. composite structures. NASA Langley Research Center, Hampton, 1999.
BULLETIN OF APPLIED MECHANICS 4(14), 81–87 (2008)