Sunteți pe pagina 1din 13

steel-concrete bridges - 2004

8. Calculation methods cracked concrete

8.1 Girder stiffness


Crack formation in concrete decks of steel-concrete bridges is caused both by direct and indirect loads. Indirect
loads originate from differences in thermal and reological behaviour between concrete and steel. Because the
concrete deck is restricted in its movement by the steel, indirect loads can cause considerable stresses already by
the end of the concrete hardening process. With bridges with an ongoing girder over various supports, a large
tensile force can develop in the concrete deck as a consequence of e.g. traffic loading. Another possibility of
tensile forces in the concrete roadway is e.g. caused by the structural systems of the bridge like a tied arch
bridge.
Crack formation in concrete decks can be dealt with in two different ways:
prevention of cracking
allow for cracking but limit crack width up to acceptable values

If cracking is to be prevented, concrete prestressing will have to be applied. However, application of prestress is
expensive and entails also practical discomforts. Moreover, the prestresssing of the cross section is often not
necessary regarding to ultimate limit state, because the construction steel and the reinforcement are sufficient to
warrant the resistance to exterior load. This aspect renders the steel concrete-bridge interesting, because it makes
the execution easier en reduces cost.

Fig. 71. Construction of the railway bridge crossing the A-12 at Nootdorp.

Nevertheless the influence of cracking on the behaviour of a steel-concrete bridge is an important design aspect.
Related to this are:
the influence of cracking on the durability of the deck
the influence of cracking on the mechanical behaviour of the steel-concrete bridge

Within this framework, we will go further into the consequence of cracking on the mechanical behaviour of the
composite bridge. NVN-ENV 1994-2: Eurocode 4: Design and calculation of steel-concrete structures Part 2:
Steel-concrete bridges, gives a number of methods by which cracking at the intermediate supports in the concrete
deck in case of a steel-concrete girder is taken into account. These methods are described in art. 4.5.3.4. In case

Dr. A. Romeijn 41
steel-concrete bridges - 2004

of ongoing girders, when the influence of concrete cracking on the results of a global analysis is small, two
simplified methods may be applied.

The first method prescribes that the appearing negative moments (hogging moments) at the intermediate supports
(the concrete roadway is supposed to be uncracked) are reduced by max. 10%. Here, the internal force division
after redistribution must be in balance with the exterior load.
To appraise the fatigue strength according categories A, B and C, see art. 4.4.0.3 of ENV 1992-2:1996, such a
way of working is not allowed.

With the second method, not in force for girders with a prestressed concrete flange or with imposed deformations
(e.g. jacking of supports), over an area as big as 15% of the span of an ongoing girder a both sides of an
intermediate support, calculated is with fully cracked concrete. The ratio of the lengths of two adjacent spans
should be at least 0:6. For the other parts, the bending stiffness is based on fully uncracked concrete.
By this method, not in force for girders with a prestressed concrete flange or with support deformations, over an
area as big as 15% of the span of an ongoing girder a both sides of an intermediate support, calculated is with
cracked concrete. The ratio of the lengths of two adjacent spans should be at least 0,6. For the other parts the
bending stiffness (flexural stiffness) is based on fully uncracked concrete.

Fig. 72. Stiffness reduction due to concrete cracking


at in between support.

EI1 = bending stiffness of the effective equivalent


cross section for an uncracked concrete deck.
EI2 = bending stiffness of the effective equivalent
cross section (incl. reinforcement) for a
cracked concrete deck.

The third method takes the effect of tension-stiffening into account. Over what areas of the girder this is
included, depends on the size of tensile stress in the ultimate concrete fibre. This method is described in Annex L
of the Eurocode and will be further explained.

8.2 Term tension stiffening


(reinforced concrete deck loaded by tension)
The characteristic strength-deformation relation of a tension member (like bowstring arches and upper and lower
chord member of a truss) can be divided into three regions, as indicated in fig. 73. The first region, the uncracked
section, runs from its origin to the point that is determined by the force N = Ncr and the strain ecr. For N< Ncr the
tension member reacts almost linearly-elastically. When reaching the ecr the first crack appears. This will be at
the spot where the concrete has the lowest tensile strength. By the coming into existence of the first crack, the
normal stiffness (axial stiffness) EcA c is reduced.
Because of the existence of
controlled deformation (instead
of a load displacement) due to
first crack initiation the
member force N decreases
slightly, so N < Ncr. By the
coming into existence of the
first crack, the axial stiffness
EcA c is reduced.

Fig. 73. Load-displacement relation of a reinforced tension member (displacement controlled) .

If the elongation further increases (displacement controlled), also the force increases, until the second crack
comes into existence. This second crack will arise in the second weakest cross section. With a further increase of
the deformation, this deformation will concentrate in an increasing number of cracks. Every further crack will
arise at a little bigger force than at the preceding crack. At a certain moment, a situation has arisen that
aggravation of the deformation does not generate new cracks anymore. The crack pattern is then completed. The

Dr. A. Romeijn 42
steel-concrete bridges - 2004

average strain e = ?L/L, when the so-called completed crack pattern has been reached, is indicated by efdc (index
fdc stands for fully developed crack pattern). For simplicitys sake it is assumed that up to the point of reaching
the completed crack pattern the force N in the member remains constant.

The second, horizontal region of the load-displacement diagram is the determined by the points (Ncr, cr) and
Ncr, fdc). In the cracks, the force N = Ncr is taken up by the reinforcement resulting into a steel stress s,cr.

The force in the reinforcement is lead through by bonding between the concrete and the reinforcement into the
concrete. The length over which this takes place is called lst. Because in the second region, also called the stage of
single cracking, the force N = Ncr, is assumed to be constant, the stress in the reinforcement at the crack position
will be constant and thus also the crack width. In practice it appears that for most situations the second region
should be considered. The average crack width wmo at an incompleted crack pattern is equal to:
0 ,4 c r 2
0,85

wmo = 2 ( ) (1 + n )
ccm s
f E
In this, cr is the concrete strength at the moment of cracking and (= A s /A c ) the ratio of reinforcement steel.
Further:
fccm = average cubic compression strength of concrete
= diameter reinforcement
Es = modulus of elasticity of reinforcement
n=Es /Ec = modular ratio

Because the equations are valid in the area of the incompleted crack pattern, it has to be checked whether the
strainp opt is smaller than the fdc for which the completed crack pattern is reached. When the completed crack
pattern arises in a short time, for fdc approximately the following is in force:

fdc = (60 + 2,4 s,cr ) 10 6

When the completed crack pattern arises as a result of a long term loading, e.g. imposed deformation, for the
strain at completed crack pattern the following is in force:
fdc , = fdc
in which is a long term factor. This factor can reach the value 1.3 maximally.

In the third stage, the stage of stabilized cracking, the number of cracks is assumed to be constant. This means
that also the average crack distance lm remains the same. At further increasing of the imposed deformation, the
steel stress in the existing cracks will keep increasing until the yield stress is reached. The moment the completed
crack pattern is reached, the crack distances will vary between lst and 2ls . Then, there are no more cross sections
where the stress in the concrete is equal to the ultimate concrete tensile strength. The average crack distance lm
remains constant, also when the member is loaded, and is equal to 1,5 lst .

The average crack width lm in a completed crack pattern is calculated by:


E
lm = 1,8 wmo s
s ,cr
The average crack width w mv for the stage of stabilized cracking can be calculated by:
l
wm v = m ( s 0,5 s ,c r )
Es
with:
lm = the average crack distance
s = the steel stress at crack location (at SLS-state, if N>Ncr)
s,cr = the steel stress direct afterwards cracking (if N=Ncr); at crack lcation

The normal stiffness of the reinforcement bar for which the stage of stabilized crakcing is reached, will always
exceed that of the reinforcement alone. The contribution of the concrete between the cracks to the stiffness of the
member is known as tension-stiffening ts. For the relations between normal force and mean strain, tension-
stiffening can be found back as the difference in strain of the member when reaching the completed crack pattern
fdc and the strain of the bear reinforcement s = Ncr / (Es A s ):

Dr. A. Romeijn 43
steel-concrete bridges - 2004

ts = s fdc
with:
s = strain of bear reinforcement = N cr / (Es As )
fdc = strain of the member at reaching the stage of stabilized craking

It is assumed that the tension-stiffening ts has a constant value over the entire section of the
completed crack pattern. The phenomenon tension-stiffening is also expressed as a quotient, the
tension-stiffening factor ts .

8.3 Tension stiffening for composite beams

The characteristic M-K relation


of a composite beam is
comparable to the F- relation
of a tension member. Also here,
three stages can be distinguished
(see fig. 74).

Mcr = cracking moment


Mcr,ts = crakcing moment
incl. tension stiffening

Fig 74.
M-K relation of a reinforced
composite beam..

When a composite cross section is loaded on bending and the concrete deck is loaded by a tensile force (e.g.
above intermediate supports), three different stages in the crack formation process can be distinguished. In the
first phase, the uncracked condition, the girder will behave totally elastically. The moment that at the topside of
the concrete deck the tensile strength (fcrm) is reached, and so will cause the first crack, the second stage begins,
the stage of single cracking. The crack pattern in this phase is also called the uncompleted crack pattern. This
stage of single cracking ends when the load is further increased and no new cracks arise. The third phase, the
stage of stabilized cracking, is characterised by an increasing deformation, but with a number of cracks that
remains equal, and so also an equal crack distance, a so-called completed crack pattern.

As explained earlier, the average crack distance lm depends on among other things the reinforcement steel stress
right after cracking s,cr. For a single member loaded by tension only, because of equilibrium just before and
after cracking, the balance equation reads:

Ac cr (1 + n ) = As s, cr

and for the reinforcement

1 + n 1
s,cr = cr = cr ( + n)

with
= As /Ac ; the ratio on area of reinforcement within the effective width
s cr = tensile strength of concrete of moment of first cracking (also depending on type of loading: long term
loading results into a lower value compared to short term loading)
s,cr = stress in the reinforcement steel directly after first cracking (if N=Ncr); at crack position

The concrete tensile strength s cr can be considered a function of the average tensile strength fctm which in its turn
is linked to the characteristic cubic compression strength fcck . E.g. in case of long term loading s cr =0,60 fctm .

Dr. A. Romeijn 44
steel-concrete bridges - 2004

Eurocode 4: Determination of Mcr en Mcr,ts


Ns,cr is the internal normal force in the concrete as a result of shrinkage and the crack moment M cr.
The crack moment is defined by the Eurocode as the bending moment that in the ultimate (concrete) fibre of the
steel-concrete cross section makes the steel-concrete cross section exceed the average concrete tensile strength
(fctm). The load distribution (like bending moments, shear forces and normal forces) and internal stresses (like
bending stress, shear stress and axial stress amongst the cross section) must be determined with the help of the n-
value (modular ratio) for short term loading, indicated by n 0 .

Fig. 75. Determination of the cracking moment Mcr,


region (a), taking into account shrinkage.

In other words, the cracking moment M cr causes a force in the concrete cross section (Nc) and in the
reinforcement cross section (Ns ). Shrinkage causes a force Nc,e in the concrete.

Ns,cr = Nc + Ns + Nc,e (the normal force in the concrete deck due to cracking moment and shrinkage)

M cr zo M cr zo Ea
Nc = Ac Ns = As
Ea Ea
I1 I1 E c
Ec Ec

with

As Ea
s = and n0 = gives
Ac Ec

N s,cr = Ac
Mcr z0
n0 I 1
b g
1 + n0 s + N c,

FG z + h IJ
H 2K c
M cr 0
bb =
I1 n0

I1 k c 0
Mcr = n0 bb
z0

The maximum stress capacity of the uncracked concrete cross section is

c
bb = f ctm c , h
1
kc 0 =
h
1+ c
2 z0

Therefore,

Dr. A. Romeijn 45
steel-concrete bridges - 2004

c h
N s,cr = Ac f ctm c ,
1
hc
b
1 + n0 s g (L.9 of ENV 1994 - 1:1997)
1+
2 z0
or

I1 k c0
c h
n0 f ctm c , z0
N s,cr = Ac
n0 I1
z0
b
1 + n0 s g
with:
Ac = effective concrete cross section
z0 = distance between centroidal axes of the uncracked composite section and concrete deck
hc = thickness of the concrete deck
s = reinforcement ratio; ratio of reinforcement area and area effective concrete deck
n0 = n-value for short term loading (n = n 0)

After the stress in the ultimate fibre has exceeded the average concrete tensile strength, the second stage starts. In
fig. 76 the internal force distribution, including shrinkage, is given.

Fig. 76. Internal forces of the


individual cross-sections for
single cracking, region (b),
taking into account shrinkage.

In the stage of single cracking it may be assumed that the internal normal force of the concrete deck, Ns,cr, due to
shrinkage and a bending moment M greater than M cr is constant.
This is further explained in fig. 77. From the left graph it appears that the value of Ns, = Nc, during the stage of
single cracking reduces linearly to zero and that this normal force is taken over by the reinforcement, so that Ns,cr
remains equal. In the graph to the right, where the building up of the internal forces during the various phases is
made clear, the same reduction of Nc, is shown. This way, the Eurocode takes the reduction of the isostatic
effects of shrinkage by crack formation into account.

Fig. 77.
Distinction of the three main
stages: normal forces and
external moments.

Ns = normal force in the concrete deck


Ma = bending moment in the steel girder

Dr. A. Romeijn 46
steel-concrete bridges - 2004

When at further loading no new cracks arise anymore, stage three begins. The internal force distribution in this
phase is elucidated in fig. 78. In determining the internal force distribution (Nso , Mso , Nao en M ao) the concrete is
neglected.

Nso = normal force in the reinforcement at


stage c.

Fig. 78.
Internal force distribution composite beam, at
stage c.

The additional internal forces are calculated by:

N s = N so + N s,ts N a ,ts = N s, ts M a ,ts = N s, ts a

The additional normal force Ns,ts of the concrete deck due to tension-stiffening without considering prestressing
is given by:
f ctm As
N s,ts = 0,4
s st
with:
fctm = the mean tensile strength of concrete
As = the area of reinforcement steel within the effective width

A I2
and: st =
Aa I a

with:
A = total area
As = area of the reinforcement steel within the effective width
Aa = area of the steel girder
I2 = the second moment of area of the effective composite section neglecting concrete
Ia = second moment of area of the steel girder

Next, with the known values of Ns.cr and Ns.ts , M cr.ts can be determined with the following equation:
I2
M cr , ts = ( N s, cr N s, ts )
As z 2
with:
z2 = the distance between the controidal axes of the cracked composite section with the second
moment of where I2 and the centre of area of the reinforcement

In case of the use of prestressing steel:

f ctm ( As + Ap )
N s,ts = 0,4
eff p st
with:
fctm = mean tensile strength of concrete
As = area of reinforcement steel within the effective width
Ap = area of prestressing steel within the effective width

Dr. A. Romeijn 47
steel-concrete bridges - 2004

8.4 Eurocode 4: Effect of concrete cracking on the flexural stiffness


To determine the areas where the concrete is assumed being cracked, the Eurocode prescribes the following rules
regarding the flexural stiffness to be applied:
In areas where the tensile stress in the ultimate concrete fibre exceeds 1,3 fctk, 0.95 , stiffness must be reduced
to EaI2
In areas where the tensile stress in the ultimate concrete fibre exceeds fctk, 0.95 , the stiffness must be reduced
to EaI2, ts.

In fig. 74, the flexural stiffness is set as a function of the bending moment for the various phases of the crack
forming process. The stiffness EaI1 , Ea I2 and EaI2,ts (using the relevant load dependent modular ratio n) have been
defined as follows:

Ea = modulus of elasticity of steel


I1 = second moment of area of the composite section without neglecting concrete (within effective width)
I2 = second moment of area of the effective equivalent steel section (structural steel, reinforcement
and prestressing steel) calculated neglecting concrete in tension but including reiforcement within
the effective width
I2,ts = the second moment of area of the effective equivalent steel section as for I2 , except that the effect
of tension stiffening of concrete between cracks is included.

Uncracked section
In the stage of uncracked concrete, the flexural stiffness of the composite section is equal to:
EI = Ea I1

Stage of single cracking


As discussed before, the present normal force remains constant in the concrete deck, Ns.cr, , during the stage of
single cracking and consists, as appears from the graph (fig. 77) from two contributions. The contribution by
shrinkage is reduced from Ns,, to the value zero when the bending moment changes from M sd to M cr,ts . To
determine the flexural stiffness in this stage of single cracking, first the various contributions to the normal forces
in the reinforced concrete deck must be determined. The equation for Ns then reads as follows:
M cr , ts M sd
N s, ( M sd ) = N c ,
M cr , ts M cr
The second contribution to the normal force in the reinforced concrete deck, Ns , depends on the present moment
and can be described as follows:

N s ( M sd ) = N s, cr Ns , ( M sd )
Now, with Ns the flexural stiffness can be determined during the stage of single cracking with the aid of equation
L.8 in attachment L.4 of the eC4-2. This equation is given below and is, in principle, in force for both the stage
of single cracking and the stabilized crack forming phase.
Ea I a
Ea I 2, ts =
N a
1 s
M
with:
a = the distance between the neutral axes of the structural steel section and the reinforced uncracked
concrete section

Above equation follows from the equation


M Ma M N s a
= = =
E I 2 ts Ea I a Ea I a

Stage of stabilized cracking


As indicated before, for the determination of the flexural stiffness in this phase, the following equation is used:

Ea I a
Ea I 2. ts =
N a
1 s
M

Dr. A. Romeijn 48
steel-concrete bridges - 2004

It is remarked that the EC4-2 has made the calculation method for short term loading, but does not mention this
specifically. In case of a (railway) bridges, where variable loading has a predominant part in the total loads, this
would mean that the actual contribution of tension-stiffening is greatly overestimated.

8.5 Sensitivity results tension-stiffening for various design parameters


Case study: starting point HSL-bridge over the Hollandsch Diep

Railway bridge over the Hollandsch Diep: general description


This is the largest steel-concrete railway bridge in the Netherlands. The construction of the bridge is a great
logistic operation, because the construction parts are fabricated at various locations in The Netherlands. The
design consists of a river part, flanked by two concrete approaching bridges with a total length of about 2
kilometres. The river span of the bridge is a continuous steel-concrete box girder with 10 main spans of about
105 m each and two end spans of 70 m. One of the starting points in the design is that the piers (substructure) of
the new bridge are exactly in line with the piers of the existing bridge and may not be wider than the existing
piers in the direction of the span. The river span can be divided into 59 m. long field sections and 46 m long
hammer sections, see fig. 79.

23000 23000 59000 23000 23000 59000 23000 23000


105000 105000

Fig. 79. Layout of the main span of the railway bridge crossing Hollandsche Diep.

The field girder is a composite box girder, the cross section of which is formed by 1:10 oblique web panels that,
at the upper flanges, are at 6 m. distance from each other. With a c.t.c. distance of the railway tracks of 5.5 m, a
favourable load transfer is made possible. The steel structure looks like an open U-shape with a girder height of
4.5 m. The concrete deck, placed on the upper flanges, causes the box shape to come into existence. All webs of
the steel girders are located on the same level and under the same slope. In fig. 80 a characteristic cross section is
given.

Fig. 80. Cross section at midspan of railway bridge


crossing Hollandsch Diep.

The hammer parts consist of a composite girder, constructed from seven sub-sections:
two steel-concrete support girders with a steel height of about 2.5 m
two steel legs with a smooth changing pentagonal cross section
two steel connecting parts forming the joint between main girder and leg
a hammer part

Stiffness is mostly decisive in a design of railway bridges. In the case of structures on behalf of the high-speed
train, a complementary stiffness criterion, the so-called comfort demand is applicable. This demand has been
prescribed by the Dutch Railway Company and has been included in the high-speed train guidelines.

Dr. A. Romeijn 49
steel-concrete bridges - 2004

Requirements are put to the vertical acceleration to a traveller experiences when passing the bridge. The
allowable acceleration depends on the exposure time and the resonance sensitivity. Due to its great length of the
bridge, large number of equal spans (10), and the own frequency of the bridge (1 Hz) being close to the passing
time of the train of a span (1s), the acceptance levels for this bridge are relatively low.

A birds eye view of the execution method is as follows:

prefab girders located at the supports


fabricated incl. corrosion protected at an alternative location
without any restraint condition (with the aid of supports) placing the steel sections (without concrete deck)

prefab field girders


fabricated incl. corrosion protected at an alternative location
without any restraint condition (with the aid of temporary supports) placing the steel sections incl. concrete
deck

ongoing bridge girder


connecting both girder sections (field hammer)
pouring of concrete deck at support
concrete casting of the joints between field and support girder
.

Fig. 81.
Pictures paper Engineering the
bridge over Hollandsch Diep,
IABSE 2003, W. tHart.

Bridge case study


With somewhat simplified geometry the high-speed railway bridge described above has been used as a starting
point. The data used are given in table 9 and fig. 82.

Field Support
Thickness concrete deck (at center), d c* 380 mm 380 mm
Thickness concrete deck (at edge) d c.zijk 200 mm 200 mm
Total width concrete deck, b c.tot 14290 mm 14290 mm
Width steel upper flange, b fl 700 mm 1000 mm
Thickness steel upper flange, d bf 20 mm 40 mm
Width steel lower flange, b of 5575 mm 5575 mm
Thickness steel lower flange, d of 50 mm 30 mm
Web height, h lijf 4430 mm 4430 mm
Web thickness, d lijf 22 mm 22 mm
Diameter reinforcement upper layer, Ds.b 12 mm 25 mm
C.t.c. distance reinforcement upper layer, hoh s.b 125 mm 125 mm
Diameter reinforcement lower layer, Ds.o 12 mm 25 mm
C.t.c. distance reinforcement lower layer, hohs.o 125 mm 125 mm
Tabel 9. Dimensions of the steel girder and concrete deck as used for the case-study (reference situation)

Dr. A. Romeijn 50
steel-concrete bridges - 2004

Fig. 82.
Length of the girder parts.

For the analysis a comparison has been opted for between on the one hand taking into account of tension-
stiffening (art. 4.5.3.4 (5), EC4-2) and on the other hand using the simplified method for cracked concrete (art.
4.5.3.4 (4), EC4-2). The effect of this simplified method is that on both sides of an intermediate support over a
length of 15% of the span, calculations are made with a flexural stiffness of the steel cross section only and for
the remaining parts of the girder with the composite flexural stiffness. Both methods are represented in fig. 13 for
further clarification. A comparison was opted for with this simplified method, because this method greatly
resembles the method used in practice today.

Fig. 83. Girder parts used for studying the consequency of tension-stiffening on load distribution.

The grouping in flexural stiffness is done based on the appearing tensile stresses in the concrete deck due to the
infrequent combinations of actions including long-term effects. The construction of the concrete deck at the
intermediate support is therefore particularly important for the appearing tensile stresses in the concrete deck.
The self-weight of the field girder (construction steel and concrete), for instance, will not be supported by the
composite supporting girder, but only by the construction steel themselve. The loads that are to be taken into
account when determining tensile stresses in the concrete deck are then:
Permanent loading
construction needed in perspective to avoiding derailment train
self-weight other parts (like lightning, etc...)
Variable loading
2 UIC train load models
shrinkage
250 kN 250 kN 250 kN 250 kN

80 kN/m 80 kN/m

onbeperkt 800 1600 1600 1600 800 onbeperkt

Fig. 84. Traffic loading vertically; UIC71.

The infrequent load in conformance with the Eurocode is to be read as short term loading.

Dr. A. Romeijn 51
steel-concrete bridges - 2004

According to the EC4-2, for the determination of the tensile stress use has to be made of a moment envelope,
relevant to the infrequent effects mentioned before. To obtain this moment envelope, the moment lines of all
possible positions of both UIC load models has to be obtained.
Obtained results for the various girder flexural stiffness are mentioned in tables 10 and 11.

girder part position flexural stiffness


[108 kNm2 ]
xbegin xeind
0 mm 5128 mm at support 4,738
5128 mm 7529 mm nearby support 5,938
7529 mm 12500 mm inbetween 8,684
12500 mm 35000 mm field 12,432
Table 10. Input data ESA-Prima-Win ( method incl. tension-stiffening).
girder part position Flexural stiffness
[108 kNm2 ]
xbegin xeind
0 mm 10500 mm at support 4,738
10500 mm 12500 mm nearby support 8,684
12500 mm 35000 mm field 12,432
Tabel 11. Input data ESA-Prima-Win (simplified method).

When the static strength of a structure is decisive and the structure is optimal used (design-wise), the tensile
stresses in the concrete deck as a result of the decisive moment envelope will be relatively high. Due to that, the
lines relevant to M III_II en M II_I will move vertically to the x-axis (fig. 83). The result is that a very great part of
the girder gets the flexural stiffness of the steel cross section only and a small part gets the stiffness of the
uncracked composite cross section. The size of the so-called area II is very small and therefore plays a minor part
in the total analysis. For this reason, the influence of the adaptation of the calculation method regarding tension
stiffening is very small. The modification of the bending stiffness by 3% in such a small zone on that position of
the girder gives negligible differences regarding the stiffness behaviour and load distribution.
The obtained stiffness layout of the whole girder differs only little from the girder arrangement of the simplified
method and the mutual differences between both methods regarding stiffness and load distribution are therefore
small. Further it is to be noted that the flexural stiffness in area II varies between 60 and 75% of the uncracked
composite stiffness. When during varying of a design parameter both the supporting girder and the field girder
were adapted, this ratio remained as good as constant (about 70%).

A design in which another criterion than the static strength is decisive, like for instance the fatigue criterion or
plate stability, will give, when using the method including tension stiffening as a result of the decisive moment
envelope, lower tensile stresses in the concrete deck. The lines relevant to the moments will then move away
from the x-axis. The consequence being that a greater part of the girder gets the uncracked composite flexural
stiffness, because area II moves further in the direction of the intermediate support. The difference between both
methods regarding the stiffness and load distribution therefore increases. In the extreme case that for the entire
girder the uncracked composite stiffness may be granted, the difference between both methods can get as high as
15-20% regarding the maximum additional displacement.

As part of the analysis of the calculation method of EC4-2 the sensitivity of the calculation results to changes of
the size of design parameters and other entrance data are given. For this sensitivity analysis, the following
parameters and data were varied:
the reinforcement percentage
the web height of the plate girders
the thickness of the concrete deck
the width of the concrete deck
the factor 0.4 in equation L.6 in contribution L.4 of EC4-2
the tensile stresses in the concrete deck caused by shrinkage

In table 12, an overview is given of the extremes regarding the differences between the calculation results
according to the method including tension stiffening on the one hand and the simplified method on the other

Dr. A. Romeijn 52
steel-concrete bridges - 2004

hand. The table shows the differences regarding the maximum additional displacement and the relation between
bending moment at the intermediate support and the moment sum being 1/8 q l2 .

parameter range max. additional vertical ratio Mtstp and 1/8 q l 2


displacement
max min gem max min gem
[%] [%] [%] [%] [%] [%]
Concrete strength class C35/45 - C55/67 5,5 2,6 4,2 -1,1 -3,4 -2,3
Reinforcement 0,54 - 3,04% 5,4 4,9 5,1 -2,1 -2,4 -2,3
Construction height 4000 - 6000 mm 8,6 2,8 5,5 -1,1 -4,5 -2,6
Thickness concrete deck 220 - 500 mm 6,9 2,6 4,8 -0,9 -3,7 -2,2
Tensile stress caused by 0,00 - 3,00 N/mm2 8,1 1,7 4,9 -0,5 -4,9 -2,5
shrinkage
(c.b. = c.o. )

Table 12. Summary on differences regarding both analytical models on the deflection and load distribution.

The general conclusion of the sensitivity analysis us that ts and ts are very sensitive to variations of the design
parameters. Taking it roughly, ts varies for realistic choices of the design parameters between 1.5 and 1.6.
This means a 50 to 60% higher flexural stiffness of the cracked composite cross section incl. tension-stiffening
with respect to the stiffness of the steel cross section (cracked composite cross section excl. tension-stiffening.)
The value of ts is determined by M sd = M cr.ts and is there maximal. For larger bending moments the value of ts
reduces greatly.

For the bridge geometry under consideration, the following conclusions can be made from the foregoing:

Bending stiffness of cracked composite cross section incl. tension-stiffening plays a minor part in
determining rough stiffness behaviour according to the method including tension-stiffening for prismatic
girder.

If the tensile stresses in the concrete deck due to the static load are relatively low, the difference between the
method including tension-stiffening and the simplified method regarding stiffness behaviour of a composite
girder is the biggest (around 10% for the girder of the case-study); for high tensile stresses the difference
between both methods can be neglected.

Tensile stresses caused by shrinkage have a great negative influence on the contribution of tension-
stiffening.

Dr. A. Romeijn 53

S-ar putea să vă placă și