Vol. 12, Issue 2/2016, 103-110, DOI: 10.

1515/cee-2016-0014

EXISTING STEEL RAILWAY BRIDGES EVALUATION

Josef VIČAN1,*, Jozef GOCÁL1, Jaroslav ODROBIŇÁK1, Peter KOTEŠ1
1
Department of Structures and Bridges, Faculty of Civil Engineering, University of Žilina, Univerzitná
8215/1, 010 26 Žilina, Slovakia.
*
corresponding author: vican@fstav.uniza.sk.

Abstract Keywords:
The article describes general principles and basis of evaluation of Evaluation;
existing railway bridges based on the concept of load-carrying Existing railway bridge;
capacity determination. Compared to the design of a new bridge, the Bridge load-carrying capacity;
modified reliability level for existing bridges evaluation should be Passage of service load;
considered due to implementation of the additional data related to Partial safety factors for loads
bridge condition and behaviour obtained from regular inspections. and resistances.
Based on those data respecting the bridge remaining lifetime, a
modification of partial safety factors for actions and materials could
be respected in the bridge evaluation process. A great attention is
also paid to the specific problems of determination of load-caring
capacity of steel railway bridges in service. Recommendation for
global analysis and methodology for existing steel bridge
superstructure load-carrying capacity determination are described
too.

1. Introduction
In the period of 2013 - 14 the collective of the Department of Structures and Bridges has worked
up the Guideline “Determination of load-carrying capacities of railway bridges” [1] for Slovak Railways.
Therefore, the paper describes general concept and basic assumptions for evaluations of existing
railway bridges and determining their load-carrying capacities based on the principles of Eurocodes.
Evaluation of existing railway bridges represents the key process of the Bridge Management
Systems relevant from the viewpoint of decision-making processes. Therefore, it should be based on
the reliability concept respecting the load-carrying capacity as the basic parameter of the existing
bridge reliability. The existing bridge evaluation should be considered when relevant deviations from
the project documentation are found, when significant deterioration from the viewpoint of bridge
reliability is observed, or when the bridge exceeds its recommended design life. Concurrently, the
bridge load-carrying capacity is the decision parameter for determining the passage of corresponding
railway service load over the bridge.

2. General concept of the railway bridge load-carrying capacity determination

2.1 General
The concept of the above mentioned existing bridge evaluation is fully compatible with the
principles of Eurocodes, so that the structure of the new Guideline is consistent with European
standards. The Guideline introduces general rules of determining the load-carrying capacity of the
permanent and temporary railway bridges on tracks with speed up to 200 km per hour with respect to
different level of its accuracy. It also presents provisions for assessment of the passage of the railway
service load classified into the corresponding Line Categories (LC). Concurrently, it should also be
used for estimation of the load-carrying capacity of new bridge with respect to reliability level adequate
to design of the new bridge without alleviations and simplifications specified only for existing bridge.
The Guideline introduces two types of the bridge load-carrying capacities. Normal load-carrying
capacity RFLM71 is specified as a dimensionless ratio of the marginal effect of variable vertical rail traffic
load satisfying relevant ultimate or serviceability limit state, to the effect caused by the LM71 in
accordance with the standard [2]. The exceptional load-carrying capacity should be estimated using
analysis of the existing bridge in accordance with approaches and principles presented in the
Unauthenticated
Download Date | 1/16/17 11:24 PM
Civil and Environmental Engineering Vol. 12, Issue 2/2016, 103-110

Guideline respecting additional alleviations in comparison to normal load-carrying capacity. According

to the precision of methodology used for the load-carrying capacity determination, the four categories
(A, B, C and D) of load-carrying capacities are defined. A great attention is given to the estimation of
the load-carrying capacity of the category C and D, which are calculated using analysis of existing
bridge based on its assessed actual condition. Considering a new bridge, its load-carrying capacity
should be determined on results of its design analysis. In the case of the categories A or B, the
determining the bridge load-carrying capacity is based on the estimation or comparative analysis.

2.2 Actions on railway bridges

All main principles in accordance with relevant Eurocodes for actions on bridges shall be
respected. The current state, location and values of loads shall be taken into account by in-situ
measurements compared to the available bridge documentation. Characteristic values of the
permanent actions on existing bridges shall be considered in accordance with standard [4] respecting
Annex D in standard [5], and standard [6] in cases, when the results of in-situ measurements on the
bridge are allowed for. Density of material should be stipulated indirectly on the basis of standard or
tabulated values or directly using specimens or samples obtained from bridge structural members.
For calculation of load-carrying capacity of existing bridge, the vertical rail traffic load
represented by the LM71 according to standard [2] shall be applied with α = 1.00. Nosing force,
centrifugal force, braking and acceleration forces multiplied also by α = 1.00, which shall be
considered to act concurrently with the vertical rail traffic load, shall be also taken in accordance with
the code [2].
Dynamic effects of the vertical rail traffic load may be considered using the dynamic factors Φ2
or Φ3 in accordance with standard [2]. Unless the relevant authority of the railways specifies, the
dynamic factor, Φ3 shall be applied for determining the existing bridge load-carrying capacity.
Disregarding the bridge deck type, redistribution of vertical and horizontal forces to the three rail
support close to the acted axle forces of LM71 or other concentrated wheel loads may be considered.
To determine the design values of the effects of permanent, variable (LM71, nosing force,
centrifugal force and braking and acceleration forces) and non-traffic actions on bridge (wind and
thermal loads), the partial factors of load effects are used. The following values of partial factor γG
should be considered for determination of the design values of permanent load effects on existing
bridges, unless more precise calculation is assumed:
- bridge structural members whose geometrical parameters were checked by measurements:
γG = 1.20,
- bridge structural members whose geometrical parameters were not checked:
γG = 1.30.
Partial factor γQ,LM71 for the vertical and horizontal variable rail traffic load effects should be
taken into account dependent on the age of bridge structural member and its planned remaining
lifetime using values as follows:
- bridge structural member younger than 30 years: γQ,LM71 = 1.40,
- bridge structural member older than 30 years: γQ,LM71 = 1.25.
The partial factors for non-traffic load effects (wind and thermal load) should be considered by
the following values dependent on the age of bridge structural member and its planned remaining
lifetime:
- bridge structural member younger than 30 years: γQ = 1.50,
- bridge structural member older than 30 years: γQ = 1.35.
Partial factors for the load effects concerning the serviceability limit states should be taken by
values of γF,ser = 1.0.
There is a possibility to determine values of partial factors for load effects using the more
precise approach according to articles [7], [8] given in Annex F of this Guideline in dependence on the
age of bridge structural member and bridge planned remaining lifetime. This procedure should be
stated by the relevant authority of the Slovak Railways. More precise approach was also presented in
paper [9].

Unauthenticated
Download Date | 1/16/17 11:24 PM
Stavebné a Environmentálne Inžinierstvo Vol. 12, Issue 2/2016, 103-110

3. Analysis of existing steel railway bridges

3.1 Material characteristics

Characteristics of material and their design values should be specified either by means of the
by-inspection verified documentation and the standards valid at the time of bridge design, or on the
results of the bridge diagnostics and material tests. The guaranteed yield strength given in bridge
documentation or in the standard valid at the time of bridge design may be taken as the characteristic
value of steel yield strength. When the year of the bridge fabrication is known and no doubts are about
the material classification (S 235, S 275 and S 355) the Table 1 could be used for estimation of steel
characteristics. The design cross-sectional and member resistances for ultimate limit state (without
fatigue) verification should be determined using values of design yield and ultimate strengths obtained
by dividing the characteristic values by partial factor γMi, which may be also taken from Table 1.
Unless bridge documentation is available or there are doubts about the quality of the material
used, results of diagnostic techniques and material tests should be applied. In this case, the properties
of steel material shall be determined according to standards [5], [6] or relevant provisions in Guideline.
The values of partial factors for resistances γMi should be calculated using the more precise approach
(see standards [5], [6]).
Unless more precise methods are considered for determining the values of partial factors, the
values from Table 1 may be used or the procedures given in Annex F of the Guideline may be applied
using statistical parameters of steel obtained by tests.
Unless the material for rivets and bolts used in steel bridge structure constructed before 1968 is
known, the characteristic values of their material properties may be considered according to Table 2 in
dependence on the quality of materials of connected elements. The value of the partial factor γM2 for
resistance of connections may be taken from Table 1. Characteristic values of material properties of
rivets and bolts and their partial factors for resistance in steel structures constructed after 1968 may
be considered in accordance with standard [12]. Rivets in steel structures of existing bridges
constructed before 1905 can be considered to be made of wrought iron. Characteristic values of the
material yield and ultimate strength and value of the partial factor γM2 for resistance of connection can
be taken from Table 1.

Unauthenticated
Download Date | 1/16/17 11:24 PM
Civil and Environmental Engineering Vol. 12, Issue 2/2016, 103-110

Table 1: Recommended material characteristics of steel and values of partial factors for member and
cross-sectional resistances.
Steel material/ grade Ultimate
Allowable stress Yield strength
strength
Year of bridge construction
Thickness
σadm fy
fu γM0 γM1 γM2
[MPa] [MPa]
[mm] [MPa]
before 1895 wrought iron 130 210 340 1.10 1.20 1.30
wrought iron 130 210 340 1.10 1.20 1.30
1895 - 1904
mild steel 140 230 360 1.10 1.20 1.30

1905 - 1937 mild steel 140 230 360 1.10 1.20 1.30
37 (S 235) 140 230 360 1.10 1.20 1.30
1938 - 1950
52 (S 355) 195 335 490 1.10 1.25 1.30

t ≤ 25 140 230 360 1.10 1.20 1.30

37 (S 235)
> 25 130 210 340 1.10 1.20 1.30
1951 - 1968
≤ 16 210 360 510 1.10 1.25 1.30
52 (S 355)
> 16 200 340 490 1.10 1.25 1.30

≤ 25 235 360
37 (S 235)
1969 - 1985 > 25 215 360 1.00 1.10 1.25

52 (S 355) ≤ 50 355 510

≤ 25 235 360
37 (S 235)
> 25 215 360
1986 - 1998 1.00 1.10 1.25
≤ 25 355 510
52 (S 355)
> 25 335 470

S 235 ≤ 40 235 360

S 235 40< t ≤ 80 215 360

S 275 ≤ 40 275 430

1998 - 2010 1.00 1.10 1.25
S 275 40< t ≤ 80 255 410

S 355 ≤ 40 355 510

S 355 40< t ≤ 80 335 470

Table 2: Characteristic values of materials for rivets and bolts.

Rivets Bolts
Material characteristics in structures made of material with yield strength
fy ≤ 300 MPa fy > 300 MPa fy ≤ 300 MPa fy > 300 MPa
fy [MPa] 200 245 300
fu [MPa] 310 440 500

3.2 Global analysis

To enable the more precise approximation of the actual bridge behaviour and allowing for the
effects of the possible imperfections or damages of the structural members, the spatial computational
models should be used for global analyses of steel railway bridges. Elastic methods shall be used to
determine internal forces and moments of the bridge structural elements. Usually, the elastic first order
method is applied using the initial structural shape with respect of the relevant criterion for application
of the first order theory (see standard [10]). In the global analysis of existing bridge, its actual condition
shall be respected. The failures and deteriorations including relevant imperfections of structural
members or bridge parts shall be implemented into the bridge computational model by the appropriate
way to allow for effects of them on the final bridge response to all loads. In addition, the effects of the
failures and deteriorations shall be considered in calculations of the bridge structural member and
cross-sectional resistances. The shape and size of imperfection significantly affecting the member
Unauthenticated
Download Date | 1/16/17 11:24 PM
Stavebné a Environmentálne Inžinierstvo Vol. 12, Issue 2/2016, 103-110

load-carrying capacity is recommended to be obtained by means of measurements on the real bridge

structure, especially in the case of riveted bridge structural elements where the effects of residual
stresses is negligible. The following damages and imperfections should be considered in the global
analysis:
- noticeable global buckling of the member systems creating one unit (arch bridges,
compression chord of truss bridges with open arrangement),
- significant deformations of the structural members and parts of bridge steel structures due to
vehicle impacts,
- absenting structural members or bridge parts,
- significantly corroded cross-sections of structural members.
In the case of bridges with welded cross-sections, the equivalent geometrical imperfections
according to standards [10, 11] may be used. The assumed imperfection shape of the steel bridge
structural member or part is recommended to be derived from the shape of elastic critical buckling
mode of a structure as a unique global and local imperfection. The amplitude of this imperfection
should be determined by means of 5.3.2 in the standard [10]. In the case of plated structures, the
shapes and amplitudes of equivalent geometrical imperfections may be determined in accordance with
Annex C in standard [11]. When method of equivalent member for resistance verification according to
standard [10] is used, local imperfections of cross-sections and members should be allowed for into
their resistances using relevant reduction factors for flexural, lateral-torsional or plate buckling.
In the global analysis of bridge plated structures, the effect of shear lag shall be taken into
account. Unless more precise approach is used, the effect of this phenomenon may be considered by
means of an effective width. The effect of plate buckling in the elastic global analysis may be allowed
for by effective cross-sectional areas of the elements in compression.
When the load-carrying capacity of the bridge structure should be significantly limited by
resistance of very slender compression member, there is a possibility to omit this member from global
analysis upon reaching their compression resistance, provided that the elastic redistribution of internal
forces and moments is admissible and the remaining part of the structure is allowable to carry the
acting loads.

3.3 Determination of load-carrying capacity for ultimate limit states

The reliability of existing bridge structural elements shall be verified and their load-carrying
capacities shall be determined using partial safety factors method, by which the fulfilment of conditions
of relevant ultimate limit state regarding the appropriate design situation shall be assessed.
Welded cross-sections should be classified with respects to slenderness of their partial parts in
accordance with Table 2 in the standard [10]. Longitudinally stiffened compression webs and flanges
should be classified respecting the rules defined in the standard [11]. For classification of the riveted
cross-sections, the widths of relevant cross-sectional parts defined in Fig. 1 should be taken into
account. Except for transversal direction, the classification should also respects the rivet distances
parallel to the acting compression stress. Due to unknown behaviour of riveted cross-sections in
plastic area of stressing, the elastic resistance should only be considered for assessment of the cross-
sectional resistance.

Fig. 1: Definition of widths for classification of riveted cross-sections.

Unauthenticated
Download Date | 1/16/17 11:24 PM
Civil and Environmental Engineering Vol. 12, Issue 2/2016, 103-110

From the ultimate limit states viewpoint, the load-carrying capacity of the bridge structural
member should be expressed by the general equation defining the Rating Factor (RFLM71) of variable
traffic load represented by the LM71 in the form as follows:

n −1
 
RFLM 71 =  Rd − ∑ Ers , Ed ,i  / ELM 71, Ed , (1)
 i =1 
where:
Rd is the cross-sectional resistance of the bridge structural member,
ELM71,Ed represents design value of vertical variable rail traffic load effects represented by the LM71
including dynamic factors,
ΣErs,Ed,i is design, combination or group values of the other load effects acting concurrently with the
vertical rail traffic load.

The load-carrying capacity of cross-sections classified into the classes 1, 2 and 3 subjected to
combination of bending, shear and normal force, provided that shear force VEd fulfils the following
condition

VEd V
η3 = ≤ 0,5 resp. η3 = Ed ≤ 0,5 , (2)
Vpl ,Rd Vbw, Rd

should be determined using equation as follows:

1 − η1,rs
RFLM 71 = , and (3)
η1, LM 71

N rs , Ed M y ,rs , Ed M z ,rs , Ed
η1,rs = + + , (4)
A· f y γ M 0 Wel , y ⋅ f y γ M 0 Wel , z ⋅ f y γ M 0

N LM 71, Ed M y , LM 71, Ed M z , LM 71, Ed

η1, LM 71 = + + , (5)
A⋅ fy γ M 0 Wel , y ⋅ f y γ M 0 Wel , z ⋅ f y γ M 0

where:
Vpl,Rd is the design plastic shear resistance,
Vbw,Rd is the design shear resistance respecting the web shear buckling,
NLM71,Ed, My,LM71,Ed, Mz,LM71,Ed are the design values of internal forces and moments caused by the
vertical variable rail traffic load represented by LM71 including dynamic
factors,
Nrs,Ed, My,rs,Ed, Mz,rs,Ed are the design, combination or group values of the internal forces and
moments due to other load effects acting concurrently with the vertical
rail traffic load,
A, Wel,y, Wel,z are the cross-sectional characteristics,
γM0 is the partial safety factor for the cross-sectional resistance.

Because of dependence of the shear force VEd on the load-carrying capacity in equation (2), the
calculation of RFLM71 should run using an iterative approach. When for shear force VEd > Vpl,Rd (Vbw,Rd),
i.e. η3 ≥ 0.5 or η3 ≥ ≥ 0.5 respectively is valid, then verifying the elastic cross-sectional
resistance of classes 1, 2 and 3 and determination of their load-carrying capacities RFLM71 should be
performed by means of the following quadratic equation

4 ⋅ k ⋅ Z LM 71 ⋅η 3, LM 71 + Z LM 71 ⋅ (η1, LM 71 + 8 ⋅ k ⋅η 3, LM 71 ⋅η 3, rs − 4 ⋅ k ⋅η 3, LM 71 ) +
2

Unauthenticated
Download Date | 1/16/17 11:24 PM
Stavebné a Environmentálne Inžinierstvo Vol. 12, Issue 2/2016, 103-110

+η1,rs + 4 ⋅ k ⋅η3,2 rs − 4 ⋅ k ⋅η3,rs + k − 1 = 0 or

or (6)
4 ⋅ k ⋅ Z LM 71 ⋅η3, LM 71 + Z LM 71 ⋅ (η1, LM 71 + 8 ⋅ k ⋅η 3, LM 71 ⋅η3,rs − 4 ⋅ k ⋅η 3, LM 71 ) +
2

2
+η1,rs + 4 ⋅ k ⋅η3,2rs − 4 ⋅ k ⋅η3,rs + k − 1 = 0 respectively,
respectively (7)

where η1,rs and η1,LM71 are defined by the equations (4) and (5) and

Vrs,Ed Vrs,Ed VLM 71,Ed

η3,rs = , η3,rs = , η3,LM 71 = ,
Vpl ,Rd Vbw,Rd Vpl , Rd
(8)
VLM 71, Ed  M f , N , Rd 
η3,LM 71 = , k = 1−  
Vbw, Rd  M pl , N ,Rd 
 

VLM71,Ed is the design shear force caused by the vertical variable rail traffic load represented by LM71
including dynamic factors,
Vrs,Ed is the design, combination or group values of shear force due to other load effects acting
concurrently with the LM71,
Mf,N,Rd is the design plastic moment resistance of the cross-section consisting of effective areas of
flanges (with respect of shear lag) reduced due to effect of normal force NEd according to
6.2.9 in the standard [10],
Mpl,N,Rd is the design plastic moment resistance of the cross-section composed of effective areas of
flanges and fully effective web of cross-section without respect to cross-section classification
reduced due to normal force NEd according to 6.2.9 in the standard [10].

The above presented calculation of RFLM71 should run using an iterative approach. Verifying the
plastic resistance of welded cross-sections of classes 1 or 2 subjected to combination of bending,
shear and normal force and their load-carrying capacities determination can be conservatively
performed using equation (3), but the plastic resistances instead elastic ones should be used in
equations (4) and (5). Verifying resistances of slender cross-sections of class 4 shall respect effects of
the shear lag and plate buckling, which may be taken into account by means of effective cross-
sectional characteristics. Provisions given by 6.2.9.3 in the standard [10] and chapter 3 and 4 in the
standard [11] should be respected. Load-carrying capacity of the slender cross-section subjected to
bending, shear and normal force can be determined using equations (3), but the effective cross-
sectional characteristics should be substituted into equations (4) and (5) and possible shift of the
centroid of the effective cross-sectional area Aeff relative to the center of gravity of the gross cross-
section according to 6.2.2.5 in the standard [10] shall be allowed for.
Buckling resistance of the compression member and resistance against lateral-torsional
buckling of members subjected to major axis bending should be estimated in accordance with 6.3.1
and 6.3.2 in the standard [10]. The load-carrying capacities of those members may be determined by
means of the following equations:

RFLM 71 = ( N b , Rd − N rs , Ed ) / N LM 71, Ed , RFLM 71 = ( M b , Rd − M rs , Ed ) / M LM 71, Ed , where (9)

Nb, Rd = χ A ⋅ f y / γ M 1 , M b, Rd = χ LT W ⋅ f y / γ M 1 , (10)

where χ is the reduction factor for the relevant buckling mode of the flexural buckling and χLT is the
reduction factor for lateral-torsional buckling. To determine the load-carrying capacity of the member
subjected to compression and biaxial bending, the condition of reliability verification in accordance
with 6.3.3 and Annex B valid for method 2 in the standard [10] shall be applied. An iterative approach
should be applied to determine the load-carrying capacity of this member.

Unauthenticated
Download Date | 1/16/17 11:24 PM
Civil and Environmental Engineering Vol. 12, Issue 2/2016, 103-110

4. Conclusions
The new Guideline for determination of railway bridge load-carrying capacity is introduced in
this paper. Therefore, the paper presents brief description of general parts of the new Guideline and it
also pays attention to the methodology of estimation of load-carrying capacities of steel bridge
structural members and their cross-sections.

Acknowledgement
The paper presents results of the research project APVV-14-0772 supported by the Slovak
Research and Development Agency and the project SK-PL-2015-0004 in frame of bilateral
cooperation.

References
[1] VIČAN, J. et al.: Determination of load carrying capacity of railway bridges, Guideline of Railways
of the Slovak Republic, 2015 (in Slovak).
[2] STN EN 1991-2, Eurocode 1: Actions on structures, Part 2: Traffic loads on bridges, Slovak
Office of Standards, Metrology and Testing, 2006.
[3] STN EN 15528: Railway applications – Line categories for managing the interface between load
limits of vehicles and infrastructure, Slovak Office of Standards, Metrology and Testing, 2013.
[4] STN EN 1991, Eurocode 1: Actions on structures, Part 1-1: General actions. Densities, self-
weight, imposed loads for buildings, Slovak Office of Standards, Metrology and Testing, 2007.
[5] STN EN 1990, Eurocode: Basis of design, Slovak Office of Standards, Metrology and Testing,
2009.
[6] STN ISO 13822 Bases for design of structures, Assessment of existing structures. Slovak Office
of Standards, Metrology and Testing, 2012.
[7] KOTEŠ, P. - VIČAN, J.: Recommended reliability levels for evaluation of existing bridges
according to Eurocodes. In: Structural Engineering International, Vol. 23 (2013), p. 411 - 417.
[8] KOTEŠ, P. - VIČAN, J.: Reliability Levels for Existing Bridges Evaluation According to Eurocodes.
In: Procedia Engineering, Vol. 40, 2012, p. 211 - 216.
[9] CREMONA, C. : Assessment of existing structures in France: Standard and advanced practices,
Bridge Maintenance, Safety, Management and Life Extension – Chen, Frangopol & Ruan (Eds),
2014 Taylor & Francis Group, London, 978-1-138-00103-9.
[10] STN EN 1993-1-1 Eurocode 3: Design of steel structures – Part 1-1: Design of steel structures,
General rules and rules for buildings, Slovak Office of Standards, Metrology and Testing, 2006.
[11] STN EN 1993-1-5 Eurocode 3: Design of steel structures – Part 1-5: Plated structures, Slovak
Office of Standards, Metrology and Testing, 2006.
[12] STN EN 1993-1-8 Eurocode 3: Design of steel structures – Part 1-8: Design of joints, Slovak
Office of Standards, Metrology and Testing, 2007.

Unauthenticated
Download Date | 1/16/17 11:24 PM