See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327282194

FLEXURAL RESISTANCE AND BEARING STRENGTH OF SOLDIER BEAMS USED

IN TEMPORARY WORKS

Conference Paper · July 2018

CITATIONS READS

0 73

3 authors:

Marina Bock Marios Theofanous

University of Wolverhampton University of Birmingham
14 PUBLICATIONS   86 CITATIONS    20 PUBLICATIONS   39 CITATIONS   

SEE PROFILE SEE PROFILE

Samir Dirar
University of Birmingham
41 PUBLICATIONS   187 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

frp strengthening View project

Alternative Raw Materials with Low Impact (ARLI) View project

All content following this page was uploaded by Marina Bock on 29 August 2018.

The user has requested enhancement of the downloaded file.

 Eighth International Conference on
THIN-WALLED STRUCTURES  ICTWS 2018
Lisbon, Portugal, July 24-27, 2018

FLEXURAL RESISTANCE AND BEARING STRENGTH OF SOLDIER

BEAMS USED IN TEMPORARY WORKS

Marina Bock*, Marios Theofanous** and Samir Dirar**

* School of Architecture and Built Environment, University of Wolverhampton, Wolverhampton, UK

e-mail: marina.bock@wlv.ac.uk
** Department of Civil Engineering, University of Birmingham, Birmingham UK
e-mail: m.theofanous@bham.ac.uk, S.M.O.H.Dirar@bham.ac.uk

Keywords: Temporary structure; Demountable structure; Experimental behaviour; Cold-formed;

Local buckling; Web crippling.

Abstract. Structures used for temporary works are lightweight so that they are easy to transport, erect
and dismantle. Design of such structures shall account for the various construction stages and hence
various actions that may be applied to ensure that the structure remains stable and safe. Particular
care should be taken in their design as local instabilities could arise due to their thin-walled nature.
This investigation examines the behaviour of cold-formed C shaped sections connected back to back
with spacers. The members, also referred to as soldiers, were loaded as simply supported beams to
determine their flexural resistance while their performance when subject to bearing load was
investigated based on transverse loading applied on both flanges simultaneously. A total of five
flexural tests and 13 load bearing tests were undertaken. In light of no specific rules to design such
members, the experimental results were used to assess the suitability of the design provisions for cold-
formed steel members. Focus is placed upon the provisions for flexural and web crippling design.
Experimental and predicted resistances are compared and recommendations are provided.

1 INTRODUCTION
Soldier beams are often used in forming and construction as temporary structures that were
developed to replace former timber based products that could not cope with the demands of
frequent reuse. They are economic backing members suitable for many propping applications
such as horizontal and vertical systems supporting façades or bridge components, braces,
trusses spanning openings, secondary beams, heavy duty towers and struts, wailing frames
and many others [1]. Soldiers are lightweight cold-formed members made of high grade steel
that comprise two C shaped members (lipped channel sections) connected by internal spacers
at regular intervals. Due to their susceptibility to local instabilities, soldiers are often stiffened
with transversal steel plates along their length and are provided with end plates perpendicular
to the C shaped members. Upon assembly of all components, the soldier becomes a straight
beam with high load capacity. The webs of the C shaped members have a pattern of holes that
makes them attachable to other components and/or accessories almost anywhere along the
member length. In addition, soldier beams can be bolted together end-to-end to create
continuous structural members. Their high reusability and recyclability, in combination with
their strength and lightweight properties, leads to decreases in CO2 emissions and energy use
hence substantial society-wide benefits.
As a solution suitable for temporary structures, allowing for all construction stages and
loads over the working life of the structure is of paramount importance to achieve a safe
design. In the UK, the standard that deals with temporary structures is BS 5975 [2]. The code
is based on the old Allowable Stress Design (ASD) and mainly addresses the management
 M. Bock et al.

and control aspects of these structures rather than their structural design and it is not part of
the family of the Eurocodes. Falsework and scaffolding is covered by BS EN 12812 [3] and
BS EN 12813 [4], respectively, and even though approaches to design falsework are given,
the provisions are skewed towards either too general structural checks or limited applicability.
Moreover, the codes are not aligned with the design philosophy of the Eurocodes as they are
not structural codes and do not follow the limit states design philosophy.
In America, the standard that concerns the design of temporary structures is ASCE/SEI 37
[5], but again the structural requirements are missing with focus given to specifying minimum
design loads and load combinations. In both European and American codes [3, 5] reduced
load return periods are assumed for meteorological actions which leads to smaller design
values for such actions compared to those adopted for permanent structures. The
appropriateness of using small return periods in the design of temporary structures, especially
for the wind load, has been demonstrated to be unsafe [6]. Other specific design codes for
falsework system are available such as the AASHTO standard GSBTW-1-M [7] and the ACI
standard 347-04 [8] which cover the design of bridge falsework systems and formwork for
concrete construction, respectively.
Soldier beams are often proprietary products which are used in construction using
manufacturer’s technical sheets reporting the load capacity recommended based on in-house
research and testing. Even though the products fully comply with the codes, these technical
sheets often overlook structural aspects that play a key role in the overall safety of the
temporary development such as design tolerances, imperfections, special loading and
foundation conditions among others. Other than soldier beams for falsework applications,
scaffolds, telescopic props and bridge construction equipment are also categorized as
temporary structures. The design approach used is different for all of them and so is their
main role [9]. The significant level of uncertainty associated with temporary structures is not
well captured in existing codes arguably due to the lack of understanding of how such
structures actually behave. Research has been conducted on the causes of collapse in
temporary structures and methods have been proposed to minimise that risk [9]. It was
concluded that temporary structure failures occur mainly due to design and operation issues as
well as events associated with underestimating the applied loading [9] and emphasis is placed
upon the importance of understanding the limitations of the current existing design methods.
In a numerical study by Peng et al. [10], a finite element model was developed to analyse
the buckling behaviour of shoring falsework subjected to high axial loads which was later on
validated against test data [11,12]. The same authors proposed analysis and design methods
that allow for the different loading conditions during the construction phase in modular
falsework [13]. Different modular falsework systems were tested and modelled in [14] that
were used by subsequent investigations as benchmark tests. Research conducted in Australia
examined the material and structural response of the individual components of falsework and
assemblies that were utilized to generate data and propose reliable design models [15-17].
Event tough research conducted over the last years has facilitate a better understanding of
how temporary structures behave there is yet a significant gap to cover concerning not only
different cross-sectional shapes and loading arrangements but also verifying the suitability of
current structural codes to cover temporary structures. The research presented in this study
addresses structural testing of soldier beams subjected to bending and bearing loads (i.e local
transverse loading) and assesses the suitability of existing European and American structural
design codes for cold-formed steel for application to soldier beams. To achieve this, the
design provisions for local buckling and web crippling given in EN 1993-1-3 [18] and the
North American Specification (NAS) [19, 20] have been considered.

2
 M. Bock et al.

2 EXPERIMENTAL TESTS
The behavior of soldier beams commonly used for temporary structures has been
experimentally investigated and is reported herein. The scope of the experimental program
covered material tests, flexural tests in the form of 3-point bend test as well as bearing tests
which were undertaken at the Structures laboratory of the University of Birmingham.
Five beam soldiers of the “Slim-Lite” type were supplied by Leada Acrow for the purpose
of this investigation. All soldier beams were 3.6m long with the cross-sectional shape shown
in Figure 1. The sections had a nominal overall width of 152 mm, a nominal overall height H
of 170 mm, a lip length of c 25 mm and a thickness t of 32 mm. The measured dimensions of
the five beam soldiers are given in Table 1 where the relevant symbols are defined in Figure
1.
The C shaped side members are connected to one another via four internal spacers in the
form of U-profiles at distances shown in Figure 2, where half of a typical soldier length is
shown. The U-profiles adopted as internal spacers are 3.9 mm thick, have a flange width of 25
mm flange and a web height of 42 mm. Additionally, each C shape side has five plate
stiffeners every 600 mm along the length and comprises a pattern of large and small holes
with different diameters located at 300 mm intervals along their webs. The diameter ϕ of the
large holes is 62 mm whereas the diameter of the small holes, which are located on either side
of the large holes at 75 mm distance, is 17 mm. The arrangement of internal spacers, plate
stiffeners and holes is illustrated in Figure 2.2. Note that this figure shows half of the total
length of the beam soldier. Each soldier member was cut in half along the length; one half
was tested as a beam to obtain the flexural response and the other half was used for the
bearing strength tests and to extract the coupons used in the material testing described in the
next section.

c
ri

s
t
w
Figure 1: Cross-section of the beam soldier.

600mm 600mm 600mm

150mm 300mm ϕ=62mm ϕ=17mm

End Plate Internal Plate Internal Plate

plate Stiffener spacer Stiffener spacer Stiffener
75mm
900mm 600mm
Figure 2: Arrangement of holes, plate stiffeners and internal spacers.

3
 M. Bock et al.

Table 1: Measure geometric properties of the soldiers.

Specimen B (mm) H (mm) c (mm) ri (mm) t (mm)
S1 55.9 169.6 17.2 1.98 3.95
S2 55.3 169.3 17.6 1.47 2.94
S3 56.3 168.8 19.1 1.37 2.93
S4 55.2 169.3 17.8 1.86 3.75
S5 55.2 169.3 17.8 1.86 3.75

2.1 Material testing

A total of 7 coupon tests at a static strain rate were conducted in accordance with the
provisions of BS EN ISO 6892-1:2016 [21]. The coupons were cut and milled from the
flanges of the C shape members in the longitudinal (rolling) direction. Upon extracting the
tensile coupons from the flat parts of the flanges, a small curvature of the coupons was
observed due to the release of bending residual stresses. The residual stresses were, however,
reintroduced when placing the coupon in the testing machine and are hence implicitly
reflected in the obtained stress-strain curve [22]. Prior to testing, the protective paint coating
and in some cases a layer of rust were removed from the external surfaces of the coupons. The
nominal dimensions of the tensile coupons were 290 mm × 30 mm with the parallel length
region having a nominal width of 20 mm. A Vernier digital calliper was used to measure the
width and thickness at 3 different locations upon the sample, to the nearest 0.01 mm. A
summary of the measured coupon dimensions is presented in Table 2, in which the reported
cross-sectional area Ac is based on average values of the measured width and thickness of the
coupons, Lc is the parallel length and Lt is the overall length of the coupon. A Zwick 1484
Universal test machine (UTM) was used to load the coupons in tension as shown in Figure 3.
An extensometer was attached to measure the strain and control the loading rate at 2 s
intervals along with a data acquisition system to log the data recorded during the test.
The steel used for manufacturing the soldier beams was designated as Grade HA350 to AS
1534 [1] with a minimum yield strength of 350 MPa and a minimum tensile strength of 430
MPa. The results of the tension coupon tests are graphically shown in a stress-strain diagram
in Figure 4. The obtained values for the Young’s modulus E, yield strength σ0.2 (0.2 proof
stress for coupons not exhibiting yield plateau), ultimate tensile stress σu and corresponding
strain εu and strain at fracture εf presented in Table 2.2. Due to the cold-forming nature of the
product, some coupons showed a non-linear material response with minimal yield plateau and
even though the material was specified to a single steel grade, two distinct steel grades with
yield strengths approximately equal to 350 MPa and 450 MPa can be identified, hence the
coupon designation adopted in Table 2.

Table 2: Tensile coupon specimen geometry and test results.

Extracted Ac Lc Lt
Coupon
from (mm2) (mm) (mm) E σ0.2 σu εu εf
designation
specimen (GPa) (MPa) (MPa) (%) (%)
350-1 S2 75.2 101.59 289 221 360 441 18.2 30.9
350-2 S2 70.9 99.53 290 -* 375 473 17.6 25.6
350-3 S3 77.0 101.57 286 216 343 420 14.5 22.5
450-1 S1 73.3 99.17 289 184 434 495 12.6 19.3
450-2 S4 70.9 99.88 288 211 450 504 12.0 18.3
450-3 S4 67.7 99.44 286 218 442 539 12.1 17.9
450-4 S5 67.2 99.48 286 193 392 503 8.3 13.1
*Young’s modulus not logged due to a faulty grip at the beginning of the test

4
 M. Bock et al.

Figure 3: Apparatus used for the tensile coupon tests.

600
450-3
500

400
Stress σ (MPa)

450-4
300 450-2
450-1 350-2
350-1
200 350-3

100
450 family
350 family
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Strain (%)
Figure 4: Stress-strain curves for all coupon tensile tests.

2.2 Bending tests

A total of five bending tests in the 3-point-bending configuration was carried out to failure
to obtain the flexural response. The experimental set up is shown in Figure 5 where the
employed instrumentation is also depicted. The load was applied through a 500 kN capacity
hydraulic jack at midspan coinciding with the location of the first internal spacer that connects
both C cross-sectional shaped members. A 50 mm wide steel plate was placed between the
specimens and the roller underneath the hydraulic jack. The locations at which the applied
load or the support reaction were acting coincided with the location of either an end plate or
an internal u-shaped stiffener connecting the C sections comprising B1, B2 and B3. For
specimens B4 and B5 where the C sections comprising each specimen were connected to one
another only at one end via an end plate, steel plates were wedged between the C sections at
the other end over the support to eliminate any torsion due to the support reaction. The
deflection and rotation of the beams during the test were monitored and recorded by using
five LVDTs. One LVDT was monitoring the vertical deflection of the tension flange at
midspan to measure the vertical deflection of the soldier members while two LVDTs were
placed at the ends of the members spaced 100 mm apart to measure the end rotation. All the
instrumentation was connected to a data acquisition system recording data at 2 s intervals. A
loading rate of 1.6 mm/min was applied throughout testing with pauses at 20 kN intervals to
check the obtained data.

5
 M. Bock et al.

P LVDTs

Span

(a) Setup and instrumentation

(b) Specimen B5 at failure

Figure 5: Bending test set up.

Table 3 reports the distance between supports of the tested specimens, the observed failure
modes, the ultimate load Pu,test and the corresponding ultimate bending moment Mu,test. Note
that the name of the specimen has changed from S1, S2, etc. to B1, B2 to identify the
specimen name with the test type. The recorded load-mid-span deflection response is shown
in Figure 6, where marked differences in initial elastic stiffness and ultimate loads is
displayed by nominally identical specimens due to variations in measured thickness, varying
material properties as discussed previously and slight differences in tested length. An
additional reason relates to the end conditions of some specimens, which did not include a
direct connection of the two C sections via a welded spacer of end plate.

Table 3: Basic tensile coupon specimen geometry and measured tensile coupon test results.
Span Pu,test Mu,test Failure
Specimen
(mm) (KN) (KNm) mode
B1 1700 126.4 53.72 LB
B2 1660 101.8 42.25 LB
B3 1640 99.4 40.75 LB
B4 1660 92 38.18 LTBa) b)
B5 1575 111.6 43.94 LB+LTBa)
a)End plates provided at one end only b)Significant initial imperfections

Specimens B1, B2 and B3 failed by local buckling (denoted LB in Table 3). Figure 7 (a)
displays the observed failure mode of specimen B1, which is characterized by local buckling
of the compressed flanges at mid-span. Specimens B2 and B3 exhibited similar failure modes.
Specimen B4 failed prematurely by lateral torsional buckling (LTB) without visible signs of
local buckling of the compressed flanges. This is attributed to the inefficiency of the
employed end detail (i.e. wedged plates between C sections) to eliminate the torsion induced
at the support. Moreover, specimen B4 had a markedly large initial bow imperfection, which
is believed to have further exacerbated its response. The employed end detail was improved in

6
 M. Bock et al.

the subsequent test and specimen B5 failed by local buckling (LB) with some slight evidence
of lateral torsional buckling (LTB) as shown in Figure 7(b), which however was not
noticeable prior to the attainment of the maximum load and is not believed to have
significantly affected the response.

140
B1
120 B5
B3
100
Load (kN)

80
B4
B2
60

40

20

0
0 5 10 15 20
Mid-span deflection (mm)
Figure 6: Load deflection response of tested beams.

(a) local buckling (B1) (b) local and lateral torsional buckling (B5)
Figure 7: Observed failure modes.

2.3 Bearing tests

The bearing strength of the beam soldiers was investigated by applying a localized
transverse load on the top flange of the soldiers directly over a support location.Both the
applied load and support reaction were applied via 75 mm long bearing plates. The bearing
strength was determined at four locations along the soldier as shown in Figure 8; at the first
hole located 150 mm from the edge, at the second hole located 450 mm from the edge (or 300
mm from the first hole), at the plate stiffener, and at the internal spacer. Figure 9 depicts 3
tests in progress, where the different loading locations can be clearly seen.
The bearing resistance of the beams was tested at the above mentioned four different
locations for soldiers S1, S2 and S3, thus comprising a total of 12 bearing strength tests. Due
to the loading applied on both flanges, the specimens are termed TFL1, TFL2 and TFL3. The
ultimate bearing resistance attained at the first hole location Pu,h1,test, at the second hole
location Pu,h2,test, at the plate stiffener Pu,ps,test and at the internal spacer Pu,is,test are reported in
Table 4. Shortening of the overall height of the beam was being monitored and recorded with
two LVDTs, one on either side of the bearing plate loading the top flange. The load –

7
 M. Bock et al.

shortening response of the beams is shown in Figure 10. Tests were stopped either at
pronounced deformation beyond ultimate load or upon fracture of the weld between the
internal spacer and the C shape members. Local yielding failure of the web, a phenomenon
also refer to as web crippling or crushing, was observed in all 12 specimens. It can be clearly
seen from the reported response, that the existence of a stiffener or u-shaped spacer at the
loading location significantly enhances the bearing strength of the specimens as web-crippling
does not occur until either the stiffener buckles or the welds between the u-shaped internal
spacer and the C-sections fail, whereupon a sudden drop of the recorded load occurs. The
response of the specimens when loaded over the 1st or 2nd hole is identical, and is
characterized by gradual loss of stiffness and a smooth post-ultimate response.

P P

75mm
150mm 300mm 900mm bearing plate

1st hole test Internal spacer test

P P

450mm 600mm

2nd hole test Plate stiffener location test

Figure 8: Location of the bearing loads.

1st hole test Plate stiffener location test 2nd hole location test
Figure 9: Test set up for three locations.

Table 4: Bearing resistance attained for all the beams and locations.
Specimen Pu,h1,test (kN) Pu, h2,test (kN) Pu,ps,test (kN) Pu,is,test (kN)
TFL1 150.1 147 308 240
TFL2 126 127 328 255
TFL3 107 122 230 255

8
 M. Bock et al.

350

300

250

Load (kN)
200

150

100
Plate stiffener
50 Internal spacer
1st hole
2nd hole
0
0 2 4 6 8 10 12 14
Shortening (mm)
a) B1
350

300

250
Load (kN)

200

150

100
Plate stiffener
50 Internal spacer
1st hole
2nd hole
0
0 2 4 6 8 10 12 14
Shortening (mm)
b) B2
350

300

250
Load (kN)

200

150

100
Plate stiffener
50 Internal spacer
1st hole
2nd hole
0
0 2 4 6 8 10 12 14
Shortening (mm)
c) B3
Figure 10: Load shortening response for tested beams.

3 ANALYSIS OF RESULTS AND RECOMMENDATIONS

The experimental data generated as part of this investigation has been used to assess the
suitability of existing design standards for cold-formed steel. The flexural and web crippling

9
 M. Bock et al.

resistances predicted by EN 1993-1-3 [18] and the North American Specification (NAS) [19,
20] have been compared against the experimental values and design recommendations are
presented in this section. In the assessment of the design methods all safety factors were set to
unity and the measured geometric and material properties were used. It should be noted that
provisions for determining ultimate resistances in soldier beams are not explicitly given in
those standards and that in practice the soldiers are to be reused; hence the applied stresses
and corresponding deformations are to be limited below those given for members in
permanent structures.

3.1 Flexural resistance

This section compares the experimental bending moment resistance Mu,test against the
bending moment resistance predicted by EN 1993-1-3 [18] Mu,EC3 and the Direct Strength
Method (DSM) given in [20] Mu,DSM. Following the EN1993-1-3 [18] design procedures, all
cross-sections were classified as class 3 whilst, with respect to the provisions for distorsional
buckling given in clause 5.5 of EN 1993-1-3, a reduction to the effective thickness (from 5 to
7%) was required for the compression flange/lip. Therefore, the design moment resistance of
all sections was lower than their elastic moment resistance Mel,R. However, as previously
discussed, no distortional buckling was evident in the observed failure modes. This is
primarily attributed to the effect of the regularly spaced stiffeners and spacer plates, which
effectively restrain the distortional buckling halfwave length and hence increase the resistance
of the specimens to distortional buckling. Additionally, the existence of a moment gradient
due to the employed configuration leads to overall smaller compressive stresses applied over
the distortional buckling halfwave length and is also believed to diminish the susceptibility to
distortional buckling.
The assessment of the design methods considered is presented in Table 5 on a test to
predicted ratio basis where it is observed that EN 1993-1-3 displays a mean ratio value of
1.17, lower than the 1.20 value observed for the DSM, and coefficient of variation (COV) of
0.221, higher than the 0.148 value shown by the DSM. The mean values and COVs between
brackets were determined without considering specimen B4 which failed prematurely by
lateral torsional buckling, a failure mode addressed in other provisions of the design standards
considered for this assessment.

Table 5: Assessment of provisions for the flexural resistance.

Mel Mu,EC3 Mu,DSM Mu,test/ Mu,test/
Specimen
(KNm) (KNm) (KNm) Mu,EC3 Mu,DSM
B1 57.66 46.50 42.03 1.16 1.28
B2 36.71 30.10 31.84 1.40 1.33
B3 34.46 28.70 30.18 1.42 1.35
B4 a) 54.51 44.30 39.7 0.86 a) 0.96
B5 54.51 44.30 39.7 0.99 1.11
Mean 1.17 (1.20)b) 1.20 (1.27) b)
COV 0.221 (0.123) b) 0.148 (0.075) b)
a)Lateral Torsional Buckling affected response
b) Number between brackets means mean and COV values without considering Specimen B4

Overall, all code predictions with the exception of the ones for specimen B4 can be
considered safe (i.e. test/predicted >1) but EN 1993-1-3 is more efficient as the mean ratio is
closer to the unity. However, the DSM shows lower scatter (i.e. lower COV) compared to EN
1993-1-3. This is because EN 1993-1-3 becomes more conservative with increasing cross-
sectional slenderness as observed for the thinner specimens B2 and B3 while the DSM better
predicts the flexural strength over the full slenderness range considered. On this basis, it can

10
 M. Bock et al.

be concluded that both design standards are adequate to determine the flexural capacity of
soldier beams, with DSM providing more consistent predictions. The effect of the spacers and
stiffeners, which was not reflected in any of the two design procedures as well as the effect of
moment gradient warrant further research.

3.2 Resistance to local transverse forces

The provisions for resistance to local transverse forces given in EN 1993-1-3 [18] and
NAS [19] are assessed based on the test results in this section. With regards to EN 1993-1-3
[18], the equations taken into consideration were those given in clause 6.1.7.2 for two
opposing local transverse forces closer together than 1.5hw, where hw is the web height
between the midlines of the flanges. Clause 6.1.7.2 of EN 1993-1-3 [18] specifies different
design equations depending on the type of loading (i.e. one flange or two flange loading),
whether the distance from the free end at which the bearing loads are applied is greater or
smaller than 1.5hw, hw being the depth of the flat part of the web, and on whether the rotation
of the web is prevented or not. The codified design predictions for each of the locations
subjected to two flange loading are termed Pu,h1,EC3, Pu,h2,EC3, Pu,ps,EC3 and Pu,is,EC3 for loads
applied at the 1st hole, 2nd hole, plate stiffener and internal u-shaped spacer. The resulting
numbers are presented in Table 6, while Figure 11 shows the test-to-predicted ratio Pu,test/Pu,EN
1993-1-3 plotted against the slenderness ratio hw/t of the web of the soldier beam.
For the NAS procedure [19], the clause covering web crippling design is C3.4.1. A unified
equation to determine the web crippling resistance is presented in combination with a set of
tables stating dimensionless coefficients for various cross-sectional geometries. For the
assessment conducted herein, the coefficients for built-up sections and single web channel and
C-sections given in tables C3.4.1-1 and C3.4.1-2 of the NAS [19], respectively, have been
used allowing for unfastened support conditions and stiffened flanges. Likewise in the
assessment of EN 1993-1-3, end loading and interior loading conditions have been considered
in the procedure based on the location of the two-flange opposing loads. The predictions by
clauses C3.4.1-1 and C3.4.1-2 for the first Pu,h1,NAS and second hole Pu,h2,NAS, the plate
stiffener Pu,ps,NAS and the internal spacer Pu,is,NAS are presented in Table 7 and 8 respectively.
According to clause C3.4.2 of the NAS, the web crippling resistance must be reduced in C-
section webs with holes. Even though the tested specimens did not meet the applicability
limits set out in the NAS, the predicted resistances by both clauses C3.4.1-1 and C3.4.1-2 at
the first and second hole have been reduced by approximately 12%. The values are reported
between brackets in Tables 7 and 8 respectively. The test-to-predicted ratios Pu,test/Pu,NAS, C3.4.4-
1 and Pu,test/Pu,NAS, C3.4.4-2 for built-up and single web channel and C-sections plotted against the
slenderness parameter hw/t is presented in Figures 12 and 13 respectively.

Table 6: Assessment of provisions for the web crippling resistance given in EN 1993-1-3.
Pu,h1,EC3 Pu,h2,EC3 Pu,ps,EC3 Pu,is,EC3 Pu,h1,test/ Pu, h2,test/ Pu,ps,test/ Pu,is,test/
Specimen
(KN) (KN) (KN) (KN) Pu,h1,EC3 Pu,h2,EC3 Pu,ps,EC3 Pu,is,EC3
B1 67.83 203.86 207.00 207.00 2.21 0.72 1.49 1.16
B2 36.47 98.24 114.85 114.85 3.45 1.29 2.86 2.22
B3 31.57 83.65 99.39 99.39 3.39 1.46 2.31 2.57
Mean 3.02 1.16 2.22 1.98
COV 0.189 0.273 0.253 0.302

11
 M. Bock et al.

Table 7: Assessment of provisions for the web crippling resistance given in NAS clause C3.4.4-1 for
built-up sections.
Pu,h1,NAS Pu,h2,NAS Pu,ps,NAS Pu,is,NAS Pu,h1,test/ Pu, h2,test/ Pu,ps,test/ Pu,is,test/
Specimen
(KN) (KN) (KN) (KN) Pu,h1,NAS Pu,h2,NAS Pu,ps,NAS Pu,is,NAS
196.00 438.02 0.77 0.34
B1 438.02 438.02 0.70 0.55
(172.92)* (386.12)* (0.87) *
(0.38)*
89.18 199.30 1.41 0.64
B2 199.30 199.30 1.65 1.28
(78.87)* (175.75)* (1.60)* (0.72)*
76.47 170.90 1.40 0.71
B3 170.90 170.90 1.35 1.49
(67.64)* (150.71)* (1.58)* (0.81)*
1.19 0.56
Mean 1.23 1.11
(1.35)* (0.64)*
0.253 0.290
COV 0.319 0.365
(0.252)* (0.290)*
* Reduced values to allow for hole under bearing load as set out in clause C3.4.2 of the NAS

Table 8: Assessment of provisions for the web crippling resistance given in NAS clause C3.4.4-2 for
single web channel and C-sections.
Pu,h1,NAS Pu,h2,NAS Pu,ps,NAS Pu,is,NAS Pu,h1,test/ Pu, h2,test/ Pu,ps,test/ Pu,is,test/
Specimen
(KN) (KN) (KN) (KN) Pu,h1,NAS Pu,h2,NAS Pu,ps,NAS Pu,is,NAS
122.63 333.95 333.95 333.95 1.22 0.44
B1 0.92 0.72
(108.19)* (294.39)* (1.39)* (0.50)*
55.13 164.46 164.46 164.46 2.29 0.77
B2 1.99 1.55
(48.76)* (145.03)* (2.58)* (0.88)*
47.11 143.92 143.92 143.92 2.27 0.85
B3 1.60 1.77
(41.67)* (126.92)* (2.57)* (0.96)*
1.93 0.69
Mean 1.50 1.35
(2.18)* (0.78)*
0.258 0.258
COV 0.294 0.337
(0.257) (0.258)*
*
* Reduced values to allow for hole under bearing load as set out in clause C3.4.2 of the NAS

The results show that EN 1993-1-3 is the most conservative approach yet the most reliable
as most of the values are on the safe side (i.e. Pu,test/Pu,EN 1993-1-3 ratio >1). The predictions by
Clause C3.4.1-1 for built-up sections fall mostly on the unsafe side, 55% of the whole data
set, while that predicted by Clause C.3.4.1-2 show 44% of the data set on the unsafe side. The
three design provisions exhibit significant scatter throughout with COV values high ranging
from 0.189 (lowest) to 0.365 (highest); lowest value was predicted by EN 1993-1-3, and
highest was predicted by clause C3.4.1-1 for built-up sections. In addition, all three
approaches show increasing test-to-predicted ratios with increasing slenderness ratios hw/t.
The predictions by the NAS at the second hole, which is categorized as an interior loading
condition, are unsafe hence EN 1993-1-3 is more suitable for loading conditions where two
bearing loads are applied further than a distance of 1.5hw from a free end. However, the high
scatter and inconsistency observed in the results suggests that further investigation is needed
to better capture the overhang effects and define a more accurate boundary between Interior
and Exterior loading. Based on the predicted resistances by the NAS allowing for cross-
sections with holes, the main recommendation that can be drawn is that the application of a
reduction factor leads to higher test-to-ultimate ratios that are more suitable when dealing
with holes at interior loading conditions.
Overall, and even though its predictions are overly conservative, it is therefore
recommended to adopt the provisions for web crippling design given in EN 1993-1-3 when

12
 M. Bock et al.

dealing with soldier beams subjected to two opposite loads acting simultaneously at any
location along the member.

4.0
3.5
Pu,test/Pu,EN 1993-1-3 3.0
2.5
2.0
1.5
1.0
1st hole
0.5 2nd hole
Plate stiffener
Internal spacer
0.0
20 40 60 80
Slenderness hw/t
Figure 11: Test-to-predicted ratio by EN 1993-1-3.
3.0

2.5

2.0
Pu,test/Pu,NAS,C3.4.1-1

1.5

1.0 1st hole

2nd hole
Plate stiffener
0.5 1st hole reduced
Internal spacer
2nd hole reduced
0.0
20 40 60 80
Slenderness hw/t
Figure 12: Test-to-predicted ratio by NAS, C3.4.1-1 for built-up sections.
3.0

2.5

2.0
Pu,test/Pu,NAS,C3.4.1-2

1.5

1.0 1st hole

2nd hole
Plate stiffener
0.5 1st hole reduced
Internal spacer
2nd hole reduced
0.0
20 40 60 80
Slenderness hw/t
Figure 13: Test-to-predicted ratio by NAS, C3.4.1-2 for single web channel and C-sections.

13
 M. Bock et al.

4 CONCLUSION
A series of tests on beam soldiers was conducted to obtain their flexural response and load
bearing strength when subjected to two flange loading. Auxiliary material coupon tests were
also conducted to obtain the material response and are reported herein. Upon removing the
external layer of paint and rust, variations in thickness of the specimens were observed.
Furthermore, significant differences in the flexural response between nominally identical
specimens were evident because of corrosion and the accompanying loss of effectiveness of
the more severely corroded specimens.
Although not directly applicable to soldier beams, the resistance of which is traditionally
assessed based on allowable stresses, the suitability of international design standards for the
design of cold-formed steel structures has also been assessed. Both American [19, 20] and
European [18] standards were shown to significantly (approximately 20%) underestimate the
flexural resistance of the tested beams, primarily due to their inability to account for the
beneficial effect of both the plate stiffeners and u-shaped spacers to the stability and strength
of the compound members. Similarly, the design equations for bearing strength were also
shown to severely underestimate the observed resistance with a high scatter, with the
exception of the NAS design predictions over the 2nd hole which were unsafe. The high
conservatism and scatter of the predictive models for bearing resistance is not confined to
solider beams or C sections, as similar conclusions have been drawn by several researchers
for various cross-sectional shapes and materials [23-27].

REFERENCES
[1] Acrow Slim-Lite Soldier System. General Technical and Application Manual. Formwork product
Technical Guide, Slough, 2018.
[2] BS 5975:1982 – Code of practice for falsework (1st ed.). London, UK: British Standards
Institution (BSI), 1982.
[3] BS EN 12812:2008. Falsework — Performance requirements and general design. London, UK:
British Standards Institution (BSI), 2008.
[4] BS EN 12813:2004. Temporary works equipment - Load bearing towers of prefabricated
components - Particular methods of structural design. London, UK: British Standards Institution
(BSI), 2004.
[5] ASCE/SEI 37–14 – Design loads on structures during construction. American Society of Civil
Engineers (ASCE), 2014.
[6] Sexsmith R.G., and Reid S.G, “Safety factors for bridge falsework by risk management”,
Structural Safety, 25(2), 227–243 (2003).
[7] AASHTO GSBTW–1–M – Guide design specifications for bridge temporary works, with 2008
interim revisions (1st ed.). Washington, DC: American Association of State Highway and
Transportation Officials (AASHTO), 2008.
[8] ACI 347–04 – Guide to formwork for concrete. Washington, DC: American Concrete Institute
(ACI), 2004.
[9] Beale R. and João A. Design Solutions and Innovations in Temporary Structures. IGI Global,
Hershey PA, USA, 2017.
[10] Chan S., Zhou Z., Chen W., Peng J. and Pan A, “Stability analysis of semirigid steel scaffolding”,
Engineering Structures, 17(8), 568–574, 1995.
[11] Peng J., Pan A., Rosowsky D., Chen W., Yen T. and Chan S., “High clearance scaffold systems
during construction - I. Structural modelling and modes of failure”, Engineering Structures, 18(3),
247–257, 1996.

14
 M. Bock et al.

[12] Peng J., Pan A., Rosowsky D., Chen W., Yen T. and Chan S., “High clearance scaffold systems
during construction - II. Structural analysis and development of design guidelines”, Engineering
Structures, 18(3), 258–267, 1996.
[13] Peng J., Pan A. and Chan S., “Simplified models for analysis and design of modular falsework”,
Journal of Constructional Steel Research, 48(2–3), 189–209, 1998.
[14] Weesner L. and Jones H., “Experimental and analytical capacity of frame scaffolding”,
Engineering Structures, 23(6), 592–599, 2001.
[15] Chandrangsu T. and Rasmussen K., “Investigation of geometric imperfections and joint stiffness
of support scaffold systems”, Journal of Constructional Steel Research, 67(4), 576–584, 2011.
[16] Chandrangsu T. and Rasmussen K., “Structural modelling of support scaffold systems”, Journal
of Constructional Steel Research, 67(5), 866–875, 2011.
[17] Zhang H., Chandrangsu T. and Rasmussen, K., “Probabilistic study of the strength of steel
scaffold systems”, Structural Safety, 32(6), 393–401, 2010.
[18] EN 1993-1-3. Eurocode 3 – Design of Steel Structures – Part 1-3: Genaral rules – Supplementary
rules for cold-formed members and sheeting. London, UK: British Standards Institution (BSI),
2006.
[19] NAS American Specification for the Design of Cold-Formed Steel Structural Members, (Draft
Edition), American Iron and Steel Institute, Washington, D.C, 2001.
[20] AISI. North American Specification. Appendix 1: Design of cold-formed steel structural
members using the Direct Strength Method. 2004 supplement to the north American specification
for the design of cold-formed steel structures, American Iron and Steel Institute, Washington
(DC), 2004.
[21] BS EN ISO 6892-1:2016. Metallic materials — Tensile testing, Part 1: Method of test at room
temperature, London: BSI, 2016.
[22] Cruise R. and Gardner L. “Residual stress analysis of structural stainless steel sections”, Journal
of constructional steel research, 64(3), 356-366, 2008.
[23] Young B, Hancock GJ. Web crippling behavior of cold-formed unlipped channels. In:
Proceedings of the 14th International specialty conference on cold-formed steel structures; 1998
October; Missouri, U.S.A; 1998.
[24] Beshara B, Schuster M. Web crippling data and calibrations of cold-formed steel members. AISI
Research Report RP00-2. 2000 American Iron and Steel Institute; Washington, DC.
[25] Zhou F, Young B. Experimental investigation of cold-formed high-strength stainless steel tubular
members subjected to combined bending and web crippling. J Struct Eng (ASCE)
2007;133:1027–34.
[26] Gardner L, Talja A, Baddoo NR. Structural design of high-strength austenitic stainless. Thin-
Walled Struct 2006;44:517–28.
[27] Bock M, Real E. “Strength curves for web crippling design of cold-formed stainless steel hat
sections”.Thin-Walled Structures 2014, 85, pp. 93-105.

15

View publication stats