06_031-037_SH_engl.

qxd

15.08.2007

10:42 Uhr

Seite 31

Topics
DOI: 10.1002/best.200710103

Norbert Randl

Load Bearing Behaviour of Cast-in Shear Dowels

Dowel bars are often used to transfer shear forces across expansion joints and also to create shear key connections, for example
in precast concrete buildings, fixed on both sides into the adjacent reinforced concrete elements. This article gives an overview
of the current state of the technology.
Recent shear tests performed with cast-in rebars, away
from the edge, are presented and analysed. A design formula is
derived to determine the transverse shear capacity of dowels
cast into concrete, based on a modified support modulus method
and taking into consideration the concrete yielding, and this is
compared with existing theoretical approaches and experimental
results.

bar: With increasing shear load the concrete bearing stress

rises strongly and it can happen, like with an anchor
loaded in shear [2], that shell-shaped chips break out from
the concrete near the surface. This spalling of the concrete
further increases the inner lever arm of the load and the
bar is increasingly loaded in bending. This failure process
is considered and analysed in more detail in the following
sections. With a cast-in length of less than about 5 ds, failure can occur by the concrete breaking out behind the
dowel. Concrete failure can also happen if the distance to
the edge in the direction of the load is small [3]. In these
cases, approaches from fastening technology can be used
for design [3].

1 Problem statement

2 State of the technology

2.1 Investigations into the load bearing behaviour
of shear dowels

The way that a dowel cast-in on both sides and under

shear load functions was first clearly described by Paulay.
In [1], he differentiates between the effects of bending,
shear resistance and inclined tension (kinking effect) of
dowels (Fig. 1). For design purposes, the bending resistance is of most importance; the inclined tension effect
first builds up after considerable displacement. While the
maximum bending moment occurs in each case slightly
above or below the joint, the maximum shear under approximately symmetrical conditions results at the shear
surface, the bending moment at this section being M = 0.
Pure steel shear failure is, however, in general not significant, because the concrete cannot withstand the maximum bearing stress without correspondingly yielding.
Failure finally occurs, at sufficiently large distance
from the edge (min. 810 dowel diameters ds), and with a
cast-in length of at least 56 ds (see also section 2.2),
through steel rupture mainly caused by bending of the

Fig. 1. Dowel action according to Paulay et al. [1]

To investigate the load bearing behaviour of cast-in shear

dowels, the bars should either be cast-in on both sides of
the shear plane or, if they are cast-in only on one side, the
shear force has to be introduced without excentricity ideally at the concrete surface. The precondition for the first
case is that the adhesion of the two halves of the concrete
specimen has been effectively broken by a slip joint. Tests
where plastic sheets have been used to minimise the coefficient of friction are especially suitable for the study of the
load bearing behaviour.
Paulay et al. [1] paid special attention to the shear
capacity of cast-in bars during their investigations into
shear transfer at joints in 19721974: flat trowelled surfaces were subsequently painted with wax for this purpose,
and this reduced any friction effects to an absolute minimum. The results of the investigation are shown in the
force-displacement diagram in Fig. 2. The increase of
shear capacity proportional to steel area is noticeable. Using strain measurements, it was possible to confirm that
the bars (diameter 6.3/9.5/12.7 mm) started yielding at
displacements of about 2.5 mm. With further displacements up to more than 12 mm, further load increases of in
some cases up to 88 % were recorded on account of the
kinking effect.
Also worth mentioning are the tests by Bennett and
Banerjee [4] into the transfer of forces in the connections
between precast columns and beams. Their tests included
the investigation of the shear resistance of steel reinforcing bars, in which they arranged a joint with polyethylene

2007 Ernst & Sohn Verlag fr Architektur und technische Wissenschaften GmbH & Co. KG, Berlin Beton- und Stahlbetonbau 102 (2007), Special Edition

31

06_031-037_SH_engl.qxd

15.08.2007

10:42 Uhr

Seite 32

N. Randl Load Bearing Behaviour of Cast-in Shear Dowels

Fig. 2. Results of the tests from Paulay [1]

foil at the inner bearing edge of a single-field test beam.

The shear force at the bearing had to be transferred entirely by the reinforcing bars crossing the joint. The most
important result was the proportionality between shear resistance and steel cross-sectional area. The cause of failure in the end was the formation of a plastic hinge in the
connecting reinforcement.
In connection with their investigations into aggregate
interlock across cracks, Millard and Johnson at the University of Warwick carried out a series of tests in order to
establish the effect of the reinforcement [5]. Half of the
test samples were cast in steel formwork and covered with
two thin polyethylene sheets to minimise the friction effect. The other halves of the test specimen were then cast
against the plastic sheet. Strain gauges were glued to the
reinforcing bars in grooves (about 1 mm deep). Only for
the 8 mm-bars, they were glued directly to the surface of
the bar without grooves (which incidentally did not disturb the bearing behaviour). The measured strain values
showed a large scatter overall and allowed no clear conclusions. Because of the short distance to the edge in the
loading direction of mostly not more than 6ds, splitting
failures occurred in some places in the concrete specimen,
so that the measured ultimate loads lie on average somewhat under those of other researchers.
At the end of the 1980s, tests were carried out at the
Polytechnic University of Milan to record the shear resistance of steels under transverse loading [6]. These deserve special attention, in connection with the research
into the load bearing capacity of cast-in shear studs, because a special experimental setup was used to make sure
that the shear load acted exactly in the plane of the concrete surface without lever arm. A section of the test specimen (the load introduction zone) is shown in Fig. 3.
Bars of diameters 14, 18 and 24 mm cast-in on one side

32

Beton- und Stahlbetonbau 102 (2007), Special Edition

into a concrete cube were tested in a total of 27 shear

tests. Above these ribbed steel bars, a total of up to six
small thin tubes were cast in at 13 mm spacing, so that
displacement transducers could be installed later and
thus the deformation of the bar during the loading
process recorded. The first shell-shaped spalling of concrete ocurred at about 80 % of the peak load, as also observed by other researchers, and the final cause of failure
was yielding of the bars. The development of the curvature could be established from the measured deformations. The calculations showed that the first yielding of
the steel occurred at about 40 % of peak load; with further loading, the yielding extended rapidly, starting from a
point about one bar diameter below the concrete surface.
According to the details in [6], the yield zones finally extended over a length of about 1 to 2ds.

2.2 Influence of the embedment depth

There are many investigations and parameter studies into
the influence of the relationship of embedment depth to
bar diameter (lb/ds) on the shear bearing capacity of castin shear studs, which are all based on bars cast-in on one
side. However, because of the comparable load transfer into the concrete, the observations regarding the influence
of embedment depth can be made use of.
Starting with the support modulus method, Friberg
[7] suggested as far back as 1938, in connection with investigations into the joints in pavement slabs in the USA,
the selection of a length of at least 75 % of the distance
from the concrete surface up to the second zero crossing
in the concrete stresses of an endlessly long bar. This corresponds to an embedment length of about 6ds. From deformation measurements of the bar at the concrete surface, Friberg calculated figures for the modulus of dowel

06_031-037_SH_engl.qxd

15.08.2007

10:42 Uhr

Seite 33

N. Randl Load Bearing Behaviour of Cast-in Shear Dowels

Fig. 3. Testspecimen at the Polytechnic University of Milan [6]

support from 270 to 400 N/mm3. The maximum calculated concrete pressures reached about four times the cylinder strength.
Extensive investigations into this question were published in issue 386 of the DAfStb [8]. Based on tests and
FE calculations, the conclusion is made that an embedment length of about 5ds is sufficient to reach the full
shear bearing capacity. This statement agrees with the investigations carried out by Fuchs in connection with his
dissertation [2].

2.3 Models for calculating the load bearing capacity

For the first estimation of shear bearing capacity, the literature often refers to the failure hypothesis of Huber/von
Mises. This is generally and frequently used in steel design, because it delivers good agreement with test results
and takes into account that the shear yield limit is lower
than the tension yield limit. Theoretically, one would then
obtain an ultimate shear load for a cast-in bar under pure
shear loading of 0.58 As fy. In practice this bearing capacity is actually not reached on account of the yielding of
the concrete and, in consequence, the bending of the bar.
Refined methods of calculation mostly assume the
failure at the section of maximum bending moment, although only bars cast-in on one side are considered. This
is a simplification, which is justified in the case where the
concrete strengths above and below the joint are similar. A
full collection of the most important knowledge and
methods is to be found in [9]. In the following sections
only those approaches are mentioned, which were innovative and are still relevant today.
Using the theory of a beam on an elastic foundation,
Friberg first developed equations in 1938. The basis for
these was the mathematical formulation of the problem
based on the support modulus theory, where an elastic bar
of endless length, embedded into an elastic continuum, is
considered (Fig. 4). Friberg chose as failure criterion the
reaching of the maximum stress in the concrete directly
under the steel bar close to the concrete surface. In reality,
this peak of stress does not occur, rather the concrete

Fig. 4. Shear dowel embedded in an elastic mass [7]

yields from the surface down to a depth of more than

1.0 ds [2]. Also not considered are the non-linear behaviour of the concrete and the contribution of the concrete
to the side of the dowel. Nonetheless, the support modulus theory delivers useable values for estimating the load
bearing capacity of dowels; the calculated ultimate load
reacts relatively unaffected by a variation of the support
modulus.
The semi-empirical method of Rasmussen is still used
today. Rasmussen makes no assumption in his model [10]
for the distribution of the stresses under the dowel, more
he uses an average value for the concrete pressures under
the dowel. From the conditions, that at the section of
maximum moment in the dowel the shear force becomes
zero and that max M = 1/6 fy ds2, the ultimate load can
be calculated as follows:
Fu = k ds2 fc,cyl fy

(1)

The value for the constant k has been experimentally determined at 1.3. Using this empirical constant, Rasmussen
calculates maximum concrete pressures for his tests
reaching about 4 to 5 times the cylinder strength.
The formula developed by Vintzileou and Tassios
[11] for the shear bearing capacity in failure mode I
(concrete spalling and formation of a plastic hinge) is
based on the same basic premises (Fig. 5). Assuming that

Beton- und Stahlbetonbau 102 (2007), Special Edition

33

06_031-037_SH_engl.qxd

15.08.2007

10:42 Uhr

Seite 34

N. Randl Load Bearing Behaviour of Cast-in Shear Dowels

Table 1. Survey of the experimental program (concrete compression strength measured with 200 mm cubes)
Baseplate

Roughness

Number
of tests

3b

very smooth
(release oil)

very smooth
(plastic sheet,
steel plate)

2
6
2
1
1
1

Bar diafcw,bottom fcw,top

meter [mm] [N/mm2] [N/mm2]
6
12
20
6
12
20

54
54
54
23
23
23

44
44
44
20
20
20

Fig. 5. Model of Vintzeleou and Tassios [11]

the concrete between surface and plastic hinge is fully

plasticising and the pressures reach 5 times the cylinder
strength fc,cyl, the following condition for the ultimate load
can be derived:
Fu2 + (10 fc,cyl e ds) Fu 1, 7 ds4 fc,cyl fy = 0
( e = lever arm)

(2)

For e = 0, the equation corresponds to the ultimate load

derived according to Rasmussen.

3 More recent investigations with reinforcement bars

In the years 19951997, the load bearing behaviour of
shear joints with a subsequently added layer of new concrete for various joint roughnesses was investigated in detail in a project carried out at the University of Innsbruck
in cooperation with the Hilti Corporation [9], [12], [13].
In the course of this investigation, the shear resistance of
reinforcing bars crossing the joint and cast-in on both
sides was also analysed, in order to filter out their contribution to the total shear resistance. To this purpose, concrete slabs were cast, then turned over with the formed
side upward and reinforcing bars of BSt 500 with various
diameters (6 mm, 12 mm, 20 mm) subsequently mortared
in (test programme in Table 1). The friction effect along
the surface of the joints was minimised by applying oil as
normally used for formwork release agent and additionally plastic sheets and thin steel plates were inserted. Finally, concrete cuboids with the dimensions 0.20 0.30
0.40 m were cast on top and, after hardening, sheared off.
A special test-setup ensured that there was no eccentricity
of the shear load to the joint (Fig. 6).

Typical force-displacement curves for various dowel

sizes are shown in Fig. 7. With respect to serviceability
limit state only the displacement until reaching the first
plateau in the force-displacement curve is considered.
The load reached was proportional to the steel cross-section area and increased with about the square root of the
concrete strength. The reaching of this yield plateau can
be explained by the formation of plastic hinges in the reinforcement bar above or below the shear surface. A renewed slow increase of the load was observed with increasing deformation of up to 20 mm and more. The reason for this is the increasing kinking effect of the joint reinforcement (Fig. 1). The final failure thus occurred after
large deformations, either through steel breakage or sometimes by the deformed bar being pulled out (Fig. 8), accompanied by concrete spalling around the bar.

4 Model for the determination of the failure load

The shear resistance of cast-in steel studs can be derived, if
the arrangement is symmetrical, from the consideration of
a bar cast-in on one side. A realistic estimation of the
shear capacity can be gained by regarding it as an elastically supported beam and additionally taking into account
that yielding of the concrete occurs near the surface. The
limit state is reached when the steel bar forms a plastic
hinge at the section of maximum moment.
The starting point for the following calculations is an
approach from Ackermann/Burkhardt [14], which is refined and extended. Near the surface, yielding of the concrete is assumed, so that there are two zones for the calculation: below the concrete surface there is a plastic zone,
and the distribution of the pressure for the failure condition is approximated here by a rectangle. Connecting to
this, from a depth to be determined, simplified ideal-elastic behaviour of the concrete is assumed and the bar in

Fig. 6. Test specimen and load application [9]

34

Beton- und Stahlbetonbau 102 (2007), Special Edition

06_031-037_SH_engl.qxd

15.08.2007

10:42 Uhr

Seite 35

N. Randl Load Bearing Behaviour of Cast-in Shear Dowels

Fig. 8. Test specimen after failure

Fig. 7. Load-displacement curves due to shear loading

Fig. 9. Modelling the cast-in dowel under shear loading

this lower zone is treated like an elastically supported

beam (Fig. 9).
Starting with the basic formula for the average concrete pressure under the bar b = c y and assuming a
constant support modulus, the differential equation for
the elastically supported beam can be derived:
d4y
EI
=cy
dz4

(3)

In the solution of this differential equation, the following

expression occurs as parameter:
L=4

4 EI
ds c

(4)

It has the dimension of a length and is normally called

elastic or characteristic length. Under the assumption
that the embedment length lb 3 L, one can solve the differential equation by using the solution for an endlessly
long bar (the error is less than 5 %).
The boundary conditions for moment and shear at
the bar end are then M = 0 and V = 0. Thus the mathematical solution for a bar loaded by a horizontal load and
a moment simplifies to (for plus/minus definition, see
Fig. 9):
y1 =

2 V0
2M
z
n ) and := 1
e cos + 2 0 e (cos sin
Ldsc
L
L dsc
(5a)

y1 =

2 V0
L2d

sc

e (cos + sin )

4 M0
L3dsc

e cos

(5b)

M1 = + V0 L e sin + M0 e (cos + sin )

V1 = + V0 e (cos sin )

2 M0
e sin
L

(5c)
(5d)

The boundary conditions at depth h are for the elastically

supported bar (plus/minus, again as per Fig. 9):
V0 = F pmax ds h
M0 = F h pmax ds

(6a)
d
h2
= h (F pmax s h)
2
2

b (z1 = 0) = c y1 (z1 = 0) = pmax

(6b)
(6c)

Substituting equation (5a) in (6c) gives:

c y1(z1 = 0) =

2 V0
2 M0
+
= pmax
Lds
L2ds

(7)

For V0 and M0, the boundary conditions from (6a, b) are

now substituted into (7):
ds
2 (F pmax ds h) 2 h (F pmax 2 h)
+
= pmax (8)
L ds
L2ds
which gives after reworking:
2
2

h 2L
2FL
2 F h L

=0

d + d p
2
ds ds pmax ds 3
s s
max d s

(9)

Beton- und Stahlbetonbau 102 (2007), Special Edition

35

06_031-037_SH_engl.qxd

15.08.2007

10:42 Uhr

Seite 36

N. Randl Load Bearing Behaviour of Cast-in Shear Dowels

From this equation, the plastic depth h related to the

diameter ds can be determined:

h
L
2F
=

ds pmax ds 2 ds

(10)

At the section of maximum moment, the shear force (derived from equation (5d)) V1() = 0 must be:
V1() = +V0 e (cos sin )

2 M0
e sin = 0 (11)
L

Assuming that the location of the plastic hinge is close to

where the plastic zone starts, the following approximations can be made: cos 1, sin .
Thus follows:
V0 L
=
V0 L + 2 M0

(12)

From equation (10), the shear load F can be expressed

explicitly:
h
ds2
L p
F=
+ max
2
ds ds

(13)

Substituting this in (6a) and (6b) gives:

h L 1 h
V0 = pmax ds 2 + =
ds ds 2 ds

L h 1
pmaax ds 2
d
s ds 2
h L h 1 h 2
M0 = pmax ds 2 +
=
ds ds 2 2 ds
Lh
pmax ds 2
2 ds

z1 L h
=
L L+h

(14a)

(14b)

(15)

The ultimate moment is, with the formation of a plastic

hinge, the fully plastic moment Mpl. The plastic moment
of resistance Wpl for a circular section is around 1.7 times
larger than Wel:
Wel =

ds3
d3
, Wpl = s
32
6

(16)

At the limit state, the moment according to equation (5c)

reaches the following maximum value:
max M1 = V0

L e

sin + M0

d3
(cos + sin ) = fy s
6
(17)

Equation (17) can be further developed by substituting the

formulae (6a) and (6b) and with the approximations cos

36

Beton- und Stahlbetonbau 102 (2007), Special Edition

F (2h L2 + 2h 2L) pmax ds h (2hL2) =

d3
fy s (L + h)2
6

(18)

Substituting equation (10) for h, after some reworking, the

following condition can be defind:

12 L 2
8L
2 fy ds
2
F3
+

+ F p
2
2
max ds 3 pmax 2
pmax ds

+ F ( 8 L 3 ) + 2 L4 pmax ds = 0

(19)

This cubic equation can be reworked and finally be

solved, with the aid of the Cardano formula and intermediate steps [9]. For the range of interest in construction
applications (1 fy / pmax 10 and 1,0 L/d 2,5), the resulting complex solution functions can be replaced by a
more simple form as a good approximation, so the simplified condition for the ultimate limit load Fu becomes finally:
Fu = pmax ds 2
2
3

fy
ds ds fy
L

0
,
005
0, 46
+ 0,187

ds
pmax L L pmax

(20)

and further through substituting into (12) and reworking:

(V1 = 0) =

1, sin and e 1 (the plastic hinge is near the

start of the plastic zone), finally resulting in:

The maximum possible concrete pressures pmax under the

dowel, uniformely distributed over the width of the bar diameter, are about 34 times the cube strength of the concrete. There are various suggestions given in the literature
for the support modulus c; it is mostly assumed to be between 100 and 1000 N/mm3. The calculated load bearing
capacity according to equation (20) reacts insensitively to
c, and therefore the result is scarcely influenced by the
support modulus. Following [11] and [2], pmax is set to 3.5
fcwm, and a constant value of 500 N/mm3 is taken for c.
The dependency of the ultimate limit load on the individual parameters cannot be directly concluded from
equation (20). Variation of the various parameters in (20)
shows that for bar diameters ds 10 mm und fy 400
N/mm2 the ultimate limit load Fu increases approximately
proportionately with the steel cross-sectional area and the
root of the concrete compression strength. The load bearing capacity can then for this limited range of validity be
approximated by the following equation, which corresponds to the approach of Rasmussen (equation (1)):
Fu = 1.5 A s fy fcwm = 1.5

ds2
fy fcwm (21)
4

The assumption of the increase with the root of the steel

yield limit is, for fy 400 N/mm2 in connection with the
constant factor in equation (21), on the safe side. Actually
the exact solution of equation (19) would result in a more
considerable increase of the load bearing capacity with fy
and thus give for the approximation formula (21) a rather
higher exponent for fy of about 0.70.8. However such an

06_031-037_SH_engl.qxd

15.08.2007

10:42 Uhr

Seite 37

N. Randl Load Bearing Behaviour of Cast-in Shear Dowels

Fig. 10. Comparison of calculated and experimental results

approach would have to be verified through tests with

higher strength steel in combination with various concrete
strengths, taking into account the possibility of early concrete spalling.

5 Comparison with test results and conclusion

The shear resistance values according to equation (20)
have been compared with test results from literature [1],
[4], [5], [6], [13] (Fig. 10). Thereby a value of 500 N/mm3
was chosen for the support modulus c, and for pmax three
times the cube compression strength of the concrete. The
data for concrete strength fcwm were calculated as average
values from the test specimen halves above and below the
shear planes.
The values predicted by calculation are shown in Fig.
10, related to the relevant experimental failure loads. The
mean values of the calculated loads agree with the test results. The low scatter of less than 15 % shows the good
agreement between experimentally determined and calculated loads and confirms the realistic description of the
failure load using equation (20). If the failure loads are determined using equation (21), then the scatter of the related values rises slightly to 16 %. Taking into account the restricted range of validity described in section 4, the simplified approach (21) does represent an adequate approximation to determine the load bearing capacity of cast-in
shear studs.

[3] Eligehausen, R., Malle, R and Silva, J. F.: Anchorage in

Concrete Construction. Berlin: Ernst & Sohn 2006.
[4] Bennett, E. W. and Banerjee, S.: Strength of beam column
connections with dowel reinforcement; The Structural Engineer, Vol. 51, No. 4, 1976, S. 133139.
[5] Millard, S. G. and Johnson, R. P.: Shear Transfer across
Cracks in Reinforced Concrete due to Aggregate Interlock
and Dowel Action; Magazine of Concrete Research, Vol. 36,
No. 126, 1984, S. 921.
[6] Dei Poli, S., Di Prisco, M. and Gambarova, P. G.: Shear
Response, Deformations and Subgrade Stiffness of a Dowel
Bar Embedded in Concrete; Journal of the American Concrete Institute, ACI, Vol. 89, No. 6, 1992, S. 665675.
[7] Friberg, Bengt F.: Design of Dowels in Transverse Joints of
Concrete Pavements; Proceedings of the American Society of
Civil Engineers, Vol. 64, No. 9, 1938, S. 18091828.
[8] Utescher, G. und Herrmann, H.: Versuche zur Ermittlung
der Tragfhigkeit in Beton eingespannter Rundstahldollen
aus nichtrostendem austenitischem Stahl (Tests to determine
the bearing capacity of round bar stainless austenitic steel);
Deutscher Ausschuss fr Stahlbeton, H. 346, Verlag Wilhelm
Ernst & Sohn, Berlin, 1983.
[9] Randl, N.: Untersuchungen zur Kraftbertragung zwischen
Alt- und Neubeton bei unterschiedlichen Fugenrauhigkeiten
(Investigations into the force transfer between old and new
concrete with various joint roughnesses); Dissertation, Universitt Innsbruck 1997, 379 S.
[10] Rasmussen, B. H.: Strength of Transversely Loaded Bolts
and Dowels Embedded in Concrete; Laboratoriet for
Bygningastatik, Denmark Technical University, Meddelelse,
Vol. 34, No. 2, 1962, S. 3955 (in Danish).
[11] Vintzileou, E. N. and Tassios, T. P.: Mathematical Models
for Dowel Action under Monotonic Conditions; Magazine of
Concrete Research (Wexham Springs), Vol. 38, No. 134,
1986, S. 1322.
[12] Randl, N., und Wicke, M.: Schubbertragung zwischen
Alt- und Neubeton (Shear transfer between old and new concrete). Beton- und Stahlbeton 95 (2000), Heft 8, S. 461473.
[13] Hilti Konzernforschung: Concrete Concrete Load
Transfer; Versuchsbericht Nr. A-IF6-2/96, 1997, not published.
[14] Ackermann, G. und Burkhardt, M.: Tragverhalten von bewehrten Verbundfugen bei Fertigteilen und Ortbeton in den
Grenzzustnden der Tragfhigkeit und Gebrauchstauglichkeit (Load bearing behaviour of reinforced composite
joints in precast elements and in-situ concrete at the ultimate
and serviceability limit states); Beton- und Stahlbetonbau,
Vol. 87, 1992, H. 7, S. 165170, H. 8, S. 197200.

Literature
[1] Paulay, T., Park, R. and Phillips, M. H.: Horizontal Construction Joints in Cast-in Place Reinforced Concrete; ACI
Special Publication SP-42 Shear in Reinforced Concrete,
1974, Vol. II, S. 599616.
[2] Fuchs, W.: Tragverhalten von Befestigungen unter Querlast
im ungerissenen Beton (Load bearing behaviour of fastenings
under shear load in uncracked concrete). Dissertation, Universitt Stuttgart, 1990.

Dipl.-Ing. Dr.techn. Norbert Randl

Professorship for Concrete and Steel Construction
Fachhochschule Krnten
9800 Spittal an der Drau, Austria
n.randl@fh-kaernten.at

First publication (in German language):

Beton und Stahlbetonbau 100 (2005), issue 6, pp. 467474.

Beton- und Stahlbetonbau 102 (2007), Special Edition

37