plastic design of laterally patch loaded plates for ships
TRANSCRIPT
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
1/19
Marine Structures 20 (2007) 124142
Plastic design of laterally patch loaded
plates for ships
Lin Hong, Jrgen Amdahl
Department of Marine Technology, Norwegian University of Science and Technology, N-7491 Trondheim, Norway
Received 12 March 2007; received in revised form 29 May 2007; accepted 30 May 2007
Abstract
Laterally loaded rectangular plates are used extensively in various marine structures, and they are
often subjected to patch loading during ice action or accidental actions, such as collision and
grounding. Therefore, focus is placed on investigating the resistance of laterally patch loaded plates.
Plastic yield line theory has been adopted in this paper, since considerable plastic behavior is
exhibited. The beneficial influence of the membrane effect during finite deformations is taken intoaccount. The derivation of the roof-top-type patch loading mechanism using work energy
principles is described in some detail. An alternative collapse model, as named double-diamond
pattern herein, is proposed which could reduce the resistance and agrees better with the results from
nonlinear finite element analysis (NLFEA) in plastic bending phase compared to the conventional
roof-top model. Moreover, a plate length restriction factor is introduced to enhance the
applicability of the present formulation when free formation of the collapse mechanism is restricted
by the finite length of the plate. The developed formulae show reasonable agreement with the results
from NLFEA of the plate resistancedeformation relationships. The resistance according to the
proposed formulation is also compared with the recently developed International Association of
Classification Societies (IACS) unified requirements for plating design for polar ships.
r 2007 Elsevier Ltd. All rights reserved.
Keywords: Plastic design; IACS; Patch loading mechanism; NLFEA; Accidental loads; Yield line theory;
Load-carrying capacity; Membrane effect; Ship collision
ARTICLE IN PRESS
www.elsevier.com/locate/marstruc
0951-8339/$ - see front matterr 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.marstruc.2007.05.003
Corresponding author. Tel.: +47 73595301; fax: +47 73595697.E-mail address: [email protected] (L. Hong).
http://www.elsevier.com/locate/marstruchttp://localhost/var/www/apps/conversion/tmp/scratch_1/dx.doi.org/10.1016/j.marstruc.2007.05.003mailto:[email protected]:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_1/dx.doi.org/10.1016/j.marstruc.2007.05.003http://www.elsevier.com/locate/marstruc -
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
2/19
1. Introduction
The stiffened plate is one of the basic structural components for ships and other marine
structures. Usually, fairly uniform lateral load, in the form of hydrostatic or dynamic
pressure dominates in many parts of these structures, so that the performance of the plateunder such a load condition is of significant importance. The lateral load from accidental
actions, such as collision, grounding, is likely to take place over a limited length of the
plate. This is similar to the so-called patch loading, which has been widely studied in ice
design. For rare loads like ship collision and abnormal ice action, significant permanent
deformations should be considered acceptable, provided that no actual collapse takes place
or no fracture occurs. Consequently, the beneficial membrane stretching effect may be
included in the assessment of the resistance of the plate beyond the pure plastic bending
response. For this purpose it is natural to use the plastic design method, as described and
summarized, e.g. by Jones [1,2], which has gained its popularity due to its simplicity for
design purpose. The analysis is simplified considerably when elastic effects are disregarded
and the material is assumed to be rigid-perfectly plastic. The collapse load is determined by
postulating a kinematically admissible plastic mechanism on the basis of yield line theory
and then equating internal and external rate of virtual work.
Recently, the International Association of Classification Societies (IACS) has developed
unified requirements for shell plating for polar ships subjected to ice patch loads adopting the
conventional roof-top-type mechanism model, refer to Daley et al. [3] and IACS [4]. Signi-
ficant simplifications have been made through the development of unified requirements, which
are somewhat conservative. It is the object of this work to reassess the plating resistance by
removing the conservative simplifications in the plastic bending mechanism, and further toinclude the membrane effect at large deformations. Besides, an alternative patch loading mecha-
nism, named as double-diamond model, which shows better agreement with the results from
nonlinear finite element analysis (NLFEA) than roof-top-type mechanism in plastic bending,
is proposed. A correction factor for roof-top-type model is thus employed to adjust the plastic
bending capacity of patch loaded plates. Further, a plate length restriction factor is introduced
when the free formation of the entire collapse mechanism is restricted by the finite length of the
plate. Results from NLFEA are used to verify the usefulness of the proposed formulae.
2. Review of plastic analysis for plates
The prediction of plastic capacity for plates subjected to lateral pressure based on yield
line theory was originally developed for the case of uniformly distributed load. Yield line
theory was first introduced by Wood [5] in the design of concrete slabs and plates for the
case of fully clamped condition. The plastic load-carrying capacity in bending for fully
clamped plates with length, L, and width, b, is predicted by
pc 48Mp
L2z20, (1)
in which, plate aspect parameter z0 is defined as
z0 b
L
ffiffiffiffiffiffiffiffiffiffiffiffiffiffi3
b2
L2
s
b
L
0@
1A, (2)
ARTICLE IN PRESSL. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 125
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
3/19
Mp represents the plastic bending moment capacity of a plate strip, with unit width and
height t, i.e. plate thickness
Mp syt
2
4
, (3)
sy is the yield strength of the material.
Sawczuk [6] included the membrane effect for large deflections. This was further
developed by Jones [7] using an approximate method to estimate the response of beams
and plates under finite permanent deflections, with elastic effects disregarded. The energy
dissipation per unit length of a hinge line for a clamped plate, which satisfies the normality
criterion of plasticity, is established as
D Mp 1 3w2
t2
_ym when
w
tp1, (4)
D 4Mpw
t_ym when
w
tX1, (5)
_ym represents the relative angular rotation rate across a hinge line; w is the transverse
deflection of the plate. Integrating the above expressions over all yield lines, the load-
carrying capacity for uniformly loaded, fully clamped plates is expressed by Jones and
Walters [8] in the following form:
p
pc 1
w2
3t2z0 3 2z0
2
3 z0
!when
w
tp1, (6)
p
pc
2w
t1
z02 z0
3 z0
t2
3w2 1
!when
w
tX1. (7)
Recently, lateral load pattern with finite length or finite height is of great concern and
the prediction of load-carrying capacity has been carried out by a number of authors in a
variety of ways: analytically, semi-analytically or empirically. Based on the plastic bending
behavior of the plates, Daley et al. [3] developed the plate thickness requirement into
practical use by IACS [4], which will be discussed further. Nyseth and Holtsmark [9]
proposed an alternative model for patch loading with a pattern of three parallel hinge lines.
Resistance expressions for loads with finite height and finite length are developed.Hayward [10] investigated the capacity of ship plating subject to loads of finite height
taking into account the membrane effect beyond the plastic bending phase by converting
the loads of finite height or length into equivalent uniform loads, as proposed by
Johansson [11].
3. Roof-top patch loading mechanism
The yield line theory and corresponding collapse mechanism have been proved to be
practical and useful for design of ship plating by Kmiecik [12]. It is generally used for
rectangular plates subjected to uniformly distributed lateral load, but the yield line modelcan also be used for plates subjected to lateral patch load as done by IACS.
The patch loading mechanism model with the shape of roof-top is demonstrated in
Fig. 1 based on the yield line concept consisting of two triangular regions and two
ARTICLE IN PRESSL. Hong, J. Amdahl / Marine Structures 20 (2007) 124142126
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
4/19
trapezoidal regions. The plate is assumed to be loaded over the entire width, but only a
part of the length, with the center of the patch in the middle of the plate. Dashed lines
represent yield lines where series of plastic hinges form, and the shaded area represents the
load patch acting laterally on the plate. The short yield lines are located outside the patch
area.
As illustrated, the length of the plate, L, the width of the plate, b, and the span of theload patch, s, are known parameters. The span of the patch may be smaller or larger than
the plate width. The hinge location parameter a and the yield line angle j are unknown, as
is the collapse pressure. The challenge for this mechanism is to identify where the plastic
hinge will form in order to give the least work solution, i.e. to find the hinge location
parameter a and the yield line angle j. It is supposed through the overall derivations that
the length of the plate is sufficient to support developing the yield line mechanism Lba.
Daley et al. [3] introduced considerable assumptions in order to simplify the problem.
First, the length of the mechanism, a, is assumed equal to the span of the load patch s. This
is essentially the same as assuming that the horizontal stiffeners are located at the patch
load boundary which is non-conservative. Later, this non-conservatism is removed byintroducing another simplification into the expression for the collapse resistance. The effect
of the two simplifications counteracts to some extent; the accuracy should thus be
investigated. In the following derivations, these simplifications will be removed.
Regarding the boundary conditions as been discussed by Amdahl [13], shell plating
generally deforms between the supporting members such as frames, longitudinals or
stringers. Provided that these supporting members are capable of supporting the plate
boundary, it is considered that the clamped boundary condition is more relevant than
simply supported, because significant plastic rotations will take place, especially during
finite deformations. The edges are assumed fixed against in-plane inward motion. For
symmetry reasons this is considered reasonable when adjacent panels are subjected tosimilar loading. If this is not the case, the fixity may be slightly overestimated, because
adjacent panels will deform in plane to some extent in order to mobilize the required
membrane stresses at the edges.
ARTICLE IN PRESS
L
Patch loading
s
b
a
p
w
p
w
Fig. 1. Roof-top-type mechanism for laterally patch loaded plates.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 127
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
5/19
Normally, shell plating can deform extensively before rupture takes place. Local finite
deformation is acceptable provided that this does not compromise the overall strength or
watertight integrity of ships. This is evident when ships suffer from collision and grounding
where large deformations occur without the structure suffering from rupture [14].
Permanent deformations of one times plate thickness are typically allowed in the design ofbow shell plating subjected to slamming loads [15]. This is believed to give enough margins
against possible rupture of shell plating. However, for the moment, there are no universal
acceptance criteria for permanent deformation for shell plating. Wang et al. [16] have made
some efforts to summarize the state-of-art criteria of allowable deformation for shell
plating. Frame spacing, plate thickness or fabrication tolerance may be chosen as the
governing parameters for allowable deformation. Notwithstanding this, the plastic
membrane behavior will be employed herein to evaluate the structural response of plates
under patch loading, especially under the situation of extreme or abnormal ice action and
accidental actions such as ship collision.
4. Structural response of plates under patch loading
The plastic design and yield line theory are based on rigid-perfectly plastic material
model; the elastic part of the plate deformation is disregarded. The yield lines are formed
by a series of plastic hinges; all plastic deformations are assumed to be concentrated at
plastic hinge locations.
First, the internal and external rates of virtual work are calculated. Considering pure
plastic bending, the rate of internal virtual work of the mechanism, dwi, is calculated as
dwi 4Mpatgj bdy, (8)
where dy is the virtual rotation angle of the short yield lines. The rate of external virtual
work, dwe, is
dwe
ZF
p dwx;y, (9)
where w(x,y) is the plate deflection in a point with coordinates (x,y) and F is the patch
loading area. The expression for the rate of external virtual work is obtained by integrating
equation (9):
dwe p dy
123ab b2 tan jb tan j a s3 cot j 3a s2b
. (10)
Equating the internal and external rate of virtual work, dwi dwe, there comes out the
following expression for the load-carrying capacity of patch loaded plates:
p p0 Kp, (11)
p0 is the plastic resistance in bending for a plate strip with unit width, p0 16MP=b2. The
coefficient Kp represents the patch loading factor and is derived as
Kp 31 a=b tan j
3a=b tan j tan j a s=b3 cot j 3a s=b2 . (12)
In case of uniform pressure, all quantities are predefined except the oblique yield line
angle j. In case of patch loading, both the yield line angle j and the length of the
ARTICLE IN PRESSL. Hong, J. Amdahl / Marine Structures 20 (2007) 124142128
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
6/19
mechanism, a, are unknown. It is natural to determine j and a such that the critical
pressure is minimized:
dp
da
0;dp
dj
0. (13)
Analytical minimization of Eq. (11) yields tremendously complicated expressions for j
and a. To circumvent this problem, Daley et al. [3] first assumed the length of the failure
mechanism a is equal to the length of the patch load s. By restoring a safety factor later, the
resistance of patch loaded plate is obtained:
p 16Mp
b21 0:5
b
s
2. (14)
And the plate thickness requirement which has been put into practice by IACS [4] is
obtained as
t b
2
ffiffiffiffiffip
sy
r1
1 0:5b=s. (15)
Removing the simplifications used in IACS, numerical optimization is adopted to find
the relationships between the unknown and known parameters.
From the optimization of Eq. (13), the relationships for a and tan(j) with respect to
patch loading aspect ratio, s/b, are shown in Fig. 2.
It is observed from Fig. 2 that the length of the plastic mechanism is approximately
equal to two times the length of the patch load for small patches (s/bo0.5) andasymptotically approaches the length of the patch load for very large patches (s/b46). The
hinge location parameter a and the yield line angle j, in the moderate range, can be
approximated as
a
s 1
1
2s=b, (16)
ARTICLE IN PRESS
Fig. 2. Non-dimensional relationships for a and tan(j) with respect to patch aspect ratio s/b.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 129
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
7/19
tan j 3
2
1
4s=b. (17)
As shown in Fig. 3, Eqs. (16) and (17) give quite accurate results when they are
substituted into Eq. (12). Introducing Eqs. (16) and (17) into Eq. (12), we obtained:
Kp 6b=s2 18b=s 60 72s=b
b=s3 12b=s2 60b=s 208 432s=b. (18)
Further approximation could be made as
Kp 1:0 1:3b
s 0:18
b
s
2. (19)
The expressions for Kp in Eqs. (18) and (19) are plotted in Fig. 3 under different patch
loading aspect ratio, s/b, together with the numerical optimization results for Eq. (12).
The results from Eq. (19) agree quite favorably with the results from Eq. (18) and the
numerical optimization solutions. As Kp is only a function of patch loading aspect ratio
s/b, the comparison will be valid for all cases.
Plates subjected to large deflections will produce large in-plane membrane stresses which
increase the resistance significantly compared to pure bending. As a matter of fact, pure
bending mechanism is impossible provided that the long edges of the plate are not free to
deform, such as in a continuous plate field, e.g. the shell plating of ships side structure.
This is easily recognized by inspection of the roof-top-type collapse mechanism adopted.
In the middle of the plate, large deformations produce a tendency of pull-in of the
boundary. However, towards the short edges the pull-in effect is much smaller andvanishes at the edges.
By introducing the coefficient Km which represents the membrane effect, into Eq. (11),
the expression for the load-carrying capacity of patch loaded plates can be written as
p p0 Kp Km. (20)
ARTICLE IN PRESS
Fig. 3. Comparison of different expressions for Kp under different patch loading aspect ratios.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142130
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
8/19
The original expressions for the membrane effect for uniformly loaded plates are:
Km 1 1
3z2
3a=b tan j 2tan2 j 1
a=b tan j 1where zp1, (21)
Km 2z 1 tan2 j 1
2a=b tan j 2
1
3z2 1
& 'when zX1, (22)
in which z is the plate deflection normalized with respect to plate thickness, i.e. z w/t.
These expressions will remain the same for the case of patch loading because the factor for
patch loading appears merely in calculating the external virtual work.
The challenge is still to determine the hinge location parameter a and the yield line angle
j. As the yield line mechanism has been formed in the stage of plastic bending, it is
assumed that the mechanism will remain the same until ultimate collapse, i.e. a and j will
not change in the membrane stretching phase. Thus, substituting Eqs. (16) and (17) intoEqs. (21) and (22), the membrane coefficients are expressed as
Km 1 1
3z2b=s2 9b=s 16 36s=b
b=s 12 12s=bwhen zp1, (23)
Km 2z 1 b=s2 12b=s 52
4b=s 48 48s=b
1
3z2 1
!when zX1. (24)
Approximate formulation for Km which will be valid over the entire deflection range
may be obtained by an elliptic fit to the asymptotic solutions:
Km 1; when z ! 0, (25)
Km 2zb=s2 8b=s 4 48s=b
4b=s 48 48s=b
!when z !1. (26)
Then there is obtained:
Km
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1 2z
b=s2 8b=s 4 48s=b
4b=s 48 48s=b
2s. (27)
The expressions for the membrane coefficient for two patch loading aspect ratios arecompared in Fig. 4. It is observed that the approximate formulation Eq. (27) agrees
reasonably well with Eqs. (23) and (24) (within a few %).
From the derivations above, the required thickness for shell plating subjected to patch
load could thus be formulated as
t b
2
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffip
sy
1
Kp
1
Km
s. (28)
5. Proposed double-diamond patch loading mechanism
As will be seen in Figs. 7 and 8 in Section 6, the resistance of patch loaded plate is over-
predicted in the plastic bending stage based on the conventional roof-top-type collapse
mechanism adopted, especially for the case of small patch aspect ratios, e.g. s/b 1. From
ARTICLE IN PRESSL. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 131
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
9/19
the observation of structural behavior of plates by finite element simulation, an alternative
kinematically admissible patch loading mechanism is proposed in Fig. 5 which is based on
a similar yield line model as the conventional roof-top-type model, named as double-
diamond collapse mechanism. The triangular hinge formations of short edges are
replaced by diamond hinge formations. The model increases the rate of external work, andreduces the rate of internal work, resulting in a somewhat lower collapse load in bending.
When the width of the patch, s, approaches plate length, L, the second yield line angle,
j2, is restricted and will diminish gradually. Consequently, for a uniformly loaded plate
the proposed double-diamond mechanism condenses into the conventional roof-top
pattern.
Three unknown parameters, hinge location parameter a, yield line angle j1 and j2, are
employed in this model, which will make the derivation more complicated than before.
Adopted the same analytical procedure as for the roof-top-type collapse model, the
expression for the pressure of the proposed double-diamond model is derived as
p p0
3a
b
tan2 j2 1
tan j1 tan j2
2a s
btan2 j1 a s=b
2
tan j1a s=b= tan j2tan j1 a s=b
2
tan j1 tan j2 tan j1
:
(29)
The minimization of the pressure requires:
dp
da 0;
dp
dj1 0;
dp
dj2 0. (30)
Apparently, it is even more difficult to get analytical solution for the double-diamondcollapse model. Numerical results show that the bending capacity reduces significantly
compared with roof-top-type model. For instance, for a plate with width 0.6 m, length
3.0 m and thickness 35 mm, the collapse load of plastic bending for different patches is
ARTICLE IN PRESS
Fig. 4. Comparison of different expressions for Km under different patch loading aspect ratios.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142132
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
10/19
calculated in Table 1. The material is assumed rigid-perfectly plastic, with yield strength
sy 300 MPa.
On the basis of the values in Table 1, an approximate correction factor fB is introducedinto Eq. (11), which adjusts the resistance in pure bending for the roof-top mechanism to
the corresponding resistance for the double-diamond mechanism:
fB 1 0:075s
b
0:5. (31)
It is a very challenging task to determine analytically the effect of membrane stresses on
the resistance in the finite deformation range. Comparison with NLFEA shows (refer
Figs. 7 and 8 in Section 6), however, that the expression ofKm, Eqs. (23) and (24), is a good
predictor also for the proposed double-diamond mechanism. Consequently, this factor
is continuously used as an approximation of the membrane effect. The load-carryingcapacity from the double-diamond mechanism becomes:
p p0 Kp Km fB. (32)
ARTICLE IN PRESS
Table 1
The bending resistance calculated from roof-top model and double-diamond model for different patch aspect
ratios
s/b p
Roof-top Double-diamond
1 10.13 9.36
1.5 7.95 7.42
2 6.92 6.49
2.5 6.32 5.95
Unit: MPa.
L
1
Patch loading
s
b
a
p
p
2
w
w
Fig. 5. Double-diamond collapse mechanism for laterally patch loaded plates.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 133
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
11/19
As a result, the plate thickness requirement can be written in the form:
t b
2 ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffip
sy
1
Kp
1
Km
1
fBs . (33)
6. Finite element verification
NLFEAs have been conducted using the progressive collapse analysis software USFOS
[17] to validate the proposed formulae for the load-carrying capacity of rectangular plates
subjected to lateral patch load.
In the following examples, two plates, denoted plates A and B, are studied for generality.
The different geometry particulars are shown in Table 2.
In Fig. 6, the collapse pattern of patch loaded plate obtained from NLFEA is shown.
The dashed lines represent the patch load boundary, and the stapled lines represent yield
lines. It is obvious that the proposed double-diamond mechanism is appropriate as the
roof-top mechanism when the plate is partially loaded.
For plate A, patch loadings with lengths of 0.40 and 0.60 m are studied. Results from
NLFEA are compared with the resistance predicted by the proposed formula Eq. (20)
from roof-top model in Fig. 7. Comparatively, the prediction by using Eq. (32) adjusted
from double-diamond model is also plotted in Fig. 7.
Three patches with different lengths are investigated for plate B, namely 0.60, 0.90 and
1.50 m, respectively. The resistances predicted using formula in Eqs. (20) and (32) are
compared with the results of NLFEA in Fig. 8.It is seen from Figs. 7 and 8 that the resistances predicted by Eq. (20) overestimate the
resistances in the plastic bending phase, notably for small patch aspect ratios, e.g. s/b 1.
When entering the membrane stage, the resistances agree reasonably well with the results
from NLFEAs. The predicted resistance agrees better and satisfactorily with the results
from NLFEA when it is adjusted by the proposed correction factor, i.e. Eq. (32), notably
for small patch aspect ratios. However, adopting Eq. (32) the resistance becomes
somewhat conservative for plate of large patch aspect ratios, e.g. s/b 2.5, when finite
deformation is allowed.
The principal limitation of the plastic yield line method is that it ignores the elasto-
plastic behavior that foregoes edge hinge formation. This becomes particularly importantfor small patch aspect ratios. Hence, the proposed formulation should be used with care, if
small or no permanent deflections are acceptable. On the other hand, the formula adjusted
by the correction factor from double-diamond mechanism model shows great
applicability, and it can be applied for patch loaded plates in most cases.
ARTICLE IN PRESS
Table 2
The geometry particulars of two plates
Plate A (m) Plate B (m)
Width 0.4 0.6Length 2.0 3.0
Thickness 0.02 0.035
sy 300MPa, E 2.1 105, n 0.3.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142134
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
12/19
The formulation predicts the resistance quite well in the range of finite deformations,
where a favorable increase of the resistance is experienced. Examples where finite
deformations could be accepted are cases when the plate is subject to extreme/abnormal iceactions or accidental actions such as ship collision.
Fig. 9 shows the resistances of plate B obtained from proposed Eq. (32) and IACS rule,
Eq. (14). Only small discrepancy is observed between the collapse bending resistance.
ARTICLE IN PRESS
Fig. 6. Collapse pattern for laterally patch loaded plates from NLFEA.
Fig. 7. Resistance predictions for plate A according to proposed Eq. (20), Eq. (32) and NLFEA.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 135
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
13/19
Finite deformations and the beneficial effect of membrane forces are not taken into
account in IACS plating design formulation.
It is also of interest to investigate the required plate thickness by using proposed formulaEq. (33) and IACS I2 structural requirement [4], Eq. (15). The design pressure is set to
5 MPa. The results are shown in Table 3 for a plate with width 0.6 m for various patch
aspect ratios and different allowable permanent deformations.
ARTICLE IN PRESS
Fig. 8. Resistance predictions for plate B according to proposed Eq. (20), Eq. (32) and NLFEA.
Fig. 9. Resistance predictions for plate B according to Eq. (32) and IACS rule.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142136
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
14/19
It is observed that when no permanent deformation is allowed, the required plate
thickness attained from Eq. (33) is virtually identical to the IACS requirement. The
maximum deviation is 1.4%. This shows that the IACS requirementin spite of the two
simplifications introducedis surprisingly good when the length of the plate is sufficient to
support the mechanism.
If permanent deformations are allowed, a noticeable thickness reduction is experienced.
For example, when the allowable permanent deformation is set equal to half of plate
thickness, the thickness reduction from adjusted formulae is up to 8.4%. If the allowable
permanent deformation is one times plate thickness, the required plate thickness may be
reduced up to 23%. The reduction becomes more significant if more deformations areacceptable.
Permanent deformations may often be accepted for abnormal ice loads or accidental
loads like ship collision, but acceptable criterion should be set appropriately due to
considerations of strain levels and proper safety margins with respect to fracture.
It should also be noticed that the present formulation and the IACS requirement are
appropriate as long as the plate length is much longer than the patch length, i.e. stiffeners
on short edges do not play any role. If the patch length approaches the plate length, these
stiffeners will impose restrictions on developing the mechanism and the resistance increases
correspondingly. This effect is particularly significant for plates with small aspect ratios.
Fig. 10 shows the plate resistance in bending versus patch aspect ratio s/b using variousformulations. The resistance is normalized versus that of a plate strip, p0, corresponding to
an infinitely long plate. In all these formulations, it is implicitly assumed that the plate
length, L, is sufficiently larger than the patch length, s, so as to contain the mechanism
yielding the minimum resistance. In the same diagram, the resistance for the plate with a
stiffener at the patch short edge, i.e. L s implying uniform loading, is plotted for
comparative purpose. It appears that the increase in resistance is significant for small plate
aspect ratios, but negligible for large aspect ratios. The IACS formulation fails to give
credit to this increased resistance. The present formulation may take this into account by
imposing restrictions on the mechanism length, a, and the second yield line angle j2 when
the plate length is limited for developing the entire mechanism. For this purpose, anotherfactor containing the plate aspect ratio must be introduced to enhance the applicability of
the proposed formulae. The restriction influence of plates with finite length will be
evaluated in Section 7.
ARTICLE IN PRESS
Table 3
Required plate thickness according to proposed formula Eq. (33) and IACS rule Eq. (15) for various patch aspect
ratios and different allowable permanent deformations
t s/b
1 1.5 2 3
tIACS 25.8 29.0 31.0 33.2
tEq. (33) (z 0) 25.6 28.6 30.5 32.8
tEq. (33) (z 0.5) 24.4 27.0 28.6 30.4
tEq. (33) (z 1) 21.6 23.4 24.4 25.5
sy 300 MPa, unit: mm.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 137
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
15/19
Fig. 11 shows numerical analysis including unloading of plate B with a patch loading of
s 0.9 m. For the present geometry the unloading stiffness is large, and the totaldisplacement is fairly equal to the permanent (plastic) displacement. This substantiates
direct comparison with results from plastic analysis by neglecting the elastic displacement
component.
In the present investigation, it is assumed that the boundaries are fixed against pull-in.
For illustration purpose, the resistance when the boundaries are free against pull-in, but
constrained to remain straight, is also illustrated in Fig. 11. It is interesting to see that there
is a significant increase of the resistance during finite deformations for straight boundaries,
although less pronounced than that for fixed boundaries. Actually, the roof-top
mechanism will imply development of internally balancing in-plane tensile and
compressive stresses as long as the boundaries remain straight. The pure bendingmechanism is therefore hypothetical when finite deformations are considered.
7. Patch loaded plates with finite length
It is explicitly assumed for all derivations that the plate is sufficiently long that no
restrictions apply for the development of the double-diamond mechanism. When the
plate is not much longer than the patch width, restrictions must be imposed on the
mechanism length, a, and the second yield line angle j2. This occurs when
Loa b tan j2, in which a and j2 are the optimal values for an infinitely long plate.
By applying restrictions to Eq. (29), the minimum resistance increases. The yield lineangle j2 vanishes gradually and the mechanism length a approaches the plate length L as L
approaches the span of the patch loading s. The relative increases in the bending resistance
for patch aspect ratios of s/b 1, 1.5 and 2 are shown in Fig. 12. It is observed that for a
ARTICLE IN PRESS
0
1
2
2.5
3
1 3 5
normalizedbendingresistance
Uniform loading
IACS
Roof-top model
Double-diamond model
3.5
1.5
0.5
2
patch aspect ratio s/b
4
Fig. 10. Normalized plate bending resistance versus patch aspect ratio s/b for an infinitely long plate using various
formulations.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142138
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
16/19
square patch (s/b 1), the stiffeners/frames have no influence as long as the plate length is
L4(1.7s 1.7b). The relative resistance increases exponentially up to a maximum value of
1.31 when L s, i.e. a square plate under uniform loading. For a longer patch (s/b 2),there is no influence of stiffeners for L4(1.25s 2.5b). The maximum relative increase of
the resistance is 1.1 for L 2s, i.e. a uniformly loaded plate. In this aspect, it is concluded
that stiffeners/frames play virtually no role when patch aspect ratios s/b42.
A reasonable approximation for plate length restriction factor is given by the following
expression:
fL 0:84 0:45L
s
L
b
!1X1. (34)
The length restriction factor should be no less than 1 in all cases, i.e. it will be employedinto Eq. (32) only when fL is larger than 1. The final expression for the load-carrying
capacity of laterally patch loaded plates is
p p0 Kp Km fBfL. (35)
Some NLFEAs have been carried out for plate B subjected to a square patch load to
validate fL when the plate length is insufficiently long to allow free formation of the
double-diamond mechanism. From Fig. 12, the plate length restriction will play roles
when L=so1:7 for a plate with a square patch loading. For illustration purpose, L/s 1.4and 1.2 are considered herein. In Fig. 13, the results from NLFEA are compared with that
from simplified calculations when fL is introduced. It is noted that the resistance increasessignificantly when the plate length is finite and insufficient to develop the entire
mechanism. The resistance compares reasonably well using Eq. (35) with respect to the
results from NLFEA, especially when finite deformation is allowed.
ARTICLE IN PRESS
Fig. 11. NLFEA predictions of resistance versus deflection of patch loaded plates with fully clamped and axially
free, but straight boundary conditions, together with a loadingunloading history curve (b 0.6m, s 0.9 m,
t 0.035 m, sy 300 MPa).
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 139
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
17/19
8. Conclusions
The plastic, large deflection resistance of laterally patch loaded plates has been analyzed,
adopting the approach developed by Jones and Walters [8] for uniformly loaded plates. It
presupposed that the patch loading extends over the entire plate width, but only a part of
ARTICLE IN PRESS
Fig. 13. Resistance predictions when the collapse mechanism is restricted by the finite length of the plate.
Fig. 12. Length restriction factor for plates with insufficient length under various patch aspect ratios.
L. Hong, J. Amdahl / Marine Structures 20 (2007) 124142140
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
18/19
the plate length. The approach resembles the one adopted by IACS in their unified
requirements for the design of plates against ice action, but the simplifications introduced
in the IACS formulations are not used and effect of membrane stresses are employed.
The expressions for yield line angle j and patch mechanism length a are obtained in the
present study for the conventional roof-top-type collapse model. The resistancepredicted with the proposed formulations is compared with the results from NLFEAs.
The resistance is somewhat overestimated in the plastic bending phase, notably for small
patches, but becomes increasingly accurate when entering the stage of finite deformations.
Hence the formulation seems to be a versatile tool for predicting the plate resistance when
finite, permanent deformations are accepted. This may, for example, be the case for plates
subjected to extreme/abnormal ice action and accidental actions such as ship collision.
In addition, an alternative and more comprehensive yield line mechanism model for
patch loaded plate is proposed, named as double-diamond collapse model. This model
yields improved prediction of the pure bending resistance, but at the expense of very
complex analytical solutions. Instead, the new model is used to derive an approximate
correction factor, which is implemented into the proposed formulae.
The required plate thickness according to the IACS unified requirement for polar ships
is in very good agreement with that obtained from the proposed formulae when no
permanent deformation is allowed, and shows that the IACS requirement is reasonable for
plates with sufficient length supporting the mechanism.
In cases of plates with small plate aspect ratios and patch length approaching plate
length, a plate length restriction factor is introduced which takes into account of the
influence of the plate with finite length imposing significant restrictions on collapse
mechanism. The formulation of the length restriction factor is verified through NLFEAwhich shows great agreement.
Acknowledgments
The present work is carried out within the scope of the Strategic University Programme
(SUP) ScenaRisC&G in Norwegian University of Science and Technology (NTNU)
which is funded by the Research Council of Norway (NFR). The authors wish to thank the
Research Council of Norway for supporting this project.
References
[1] Jones N. Structural impact. Cambridge: Cambridge University Press; 1989.
[2] Jones N. Review of the plastic behavior of beams and plates. Int Shipbuild Progr 1972;19:31327.
[3] Daley CG, Kendrick A, Appolonov E. Plating and framing design in the unified requirements for polar class
ships. In: Proceedings of the 16th international conference on port and ocean engineering under arctic
conditions, vol, 3, Ottawa, Canada, 2001. p. 77991.
[4] International Association of Classification Societies. I2 Structural requirements for polar class ships, 2006.
[5] Wood RH. Plastic and elastic design of slabs and plates. New York: The Ronald Press; 1961.
[6] Sawczuk A. On initiation of the membrane action in rigid-plastic plates. J Mech 1964;3(1):1523.
[7] Jones N. A theoretical study of the dynamic plastic behavior of beams and plates with finite-deflections. Int
J Solids Struct 1971;7:100729.[8] Jones N, Walters RM. Large deflections of rectangular plates. J Ship Res 1971;15(2):16471.
[9] Nyseth H, Holtsmark G. Analytical plastic capacity formulation for plates subject to ice loads and similar
types of patch loadings. In: Proceedings of 25th international conference on offshore mechanics and arctic
engineering (OMAE2006), Hamburg, Germany, p. 12.
ARTICLE IN PRESSL. Hong, J. Amdahl / Marine Structures 20 (2007) 124142 141
-
7/29/2019 Plastic Design of Laterally Patch Loaded Plates for Ships
19/19
[10] Hayward RC. Plastic response of ship shell plating subject to loads of finite height. Master thesis, Faculty of
Engineering and Applied Science, Memorial University of Newfoundland, 2001.
[11] Johansson BM. On the ice-strengthening of ship hulls. Int Shipbuild Progr 1967;14:23145.
[12] Kmiecik M. Usefulness of the yield line theory in design of ship plating. Mar Struct 1995;8(1):6779.
[13] Amdahl J. Plate and stiffener ice action design consideration for the Shtokman platform. Technical notes,
Trondheim, 2006.
[14] Wang G. Some recent studies on plastic behavior of plates subject to large impact loads. J Offshore Mech
Arctic Eng 2002;124(3):12531.
[15] Wang G, Tang S, Shin Y. Direct calculation approach and design criteria for wave slamming of an FPSO
bow. Int J Offshore Polar Eng 2002;12(4):297304.
[16] Wang G, Basu R, Chavda D, Liu S. Rationalizing the design of ice strengthened side structures. In:
Proceedings of maritime transportation and exploitation of ocean and coastal resources, vol. 1. Lisbon,
Portugal, 2005. p. 54957.
[17] /www.usfos.comS.
ARTICLE IN PRESSL. Hong, J. Amdahl / Marine Structures 20 (2007) 124142142
http://www.usfos.com/http://www.usfos.com/http://www.usfos.com/http://www.usfos.com/http://www.usfos.com/