Technical Note

Static and Dynamic Testing of High-Speed Rail Bridges in Spain

Miguel A. Vicente1; Dorys C. González2; and Gongkang Fu, M.ASCE3

Abstract: There has been a rapid development of high-speed rail in Spain. Many bridges have been constructed to carry these lines. The rel-
evant Spanish design codes require static and dynamic testing of these structures for acceptance of the design and construction before opening
Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 01/16/18. Copyright ASCE. For personal use only; all rights reserved.

for service; such testing is considered too costly in some other countries. This paper presents the concept, practice, and results of such testing for
119 precast and posttensioned concrete girder bridges. Their static and dynamic properties and responses to loads have implications for their
structural safety, serviceability, and possibly durability. Inaccurate models used in design for static and dynamic parameters (such as stiffness,
first natural frequency, and damping ratio) may result in incorrect responses and possibly unsafe structures; therefore, in situ testing is required.
The test results of these structures also lead to an empirical relation between the first natural frequency and span length and a range of damp-
ing ratios for these types of structures, which can be useful for bridge engineers, particularly at a preliminary design stage of similar structures.
DOI: 10.1061/(ASCE)BE.1943-5592.0000654. © 2014 American Society of Civil Engineers.
Author keywords: High-speed rails; Bridges; Load testing; Natural frequency; Damping ratio.

Introduction According to the Spanish specifications (Spanish Ministry of

Public Works and Transport 2007), any structure with a span length
During the last 20 years, there has been a significant development longer than 10 m (32.8 ft) is subject to static and dynamic load
of high-speed rail in Spain. As a result, Spain currently has about testing. Testing load is designed to verify that the calculation models
3,100 km (1,926 mi) of high-speed rail in service, making it the first in for design are appropriate, with respect to load-response relations
Europe and the second in the world, after China, in terms of total length and dynamic characteristics. Such testing also conceptually addresses
in service. Given the topography and design requirements in Spain, it construction quality up to the point when testing is performed.
has been necessary to construct approximately 1,200 bridges for these The ITPF-05 (Spanish Ministry of Public Works and Transport
lines. According to the governing specifications (Spanish Ministry of 2005) specifically requires that the measured responses be for
Public Works and Transport 2005, 2007), these structures are required deflection and tilt for static load testing and for the first natural
to be load tested, statically and dynamically, before they are accepted for frequency, damping ratio, dynamic deflection, and acceleration
rail traffic. This paper presents the concept, practice, and results of such for dynamic load testing. This paper addresses the practice in
testing for 119 precast and posttensioned concrete girder bridges. Spain for bridges in high-speed rail according to the relevant
specifications.
Specification Requirements for Testing
Dynamic Effects Considered in Design
The Spanish high-speed rail is subject to the European design speci-
fications (European Commission 1996, 2001, 2007) and, more spe- For the design of high-speed rail bridges in Spain, the IAPF-07
cifically for interoperability of the trans-European high-speed rail (Spanish Ministry of Public Works and Transport 2007) repre-
system, adaptation of the European directives into Spanish instruc- sents the governing code for loading. It is a transposition of Part 2 of
tions (Spanish Ministry of Public Works and Transport 2005). Eurocode 1 [European Committee for Standardization (CEN) 2002]
In particular, the bridges and viaducts are subject to static and for practice in Spain. This code explains how the dynamic behavior
dynamic load testing before opening for service. Although such of railway loads must be taken into account in bridge design, similar
testing is required at both stages of the supporting structure only and to the UIC Leaflet 776-1 (International Union of Railways 2006) and
the entire completed system, this study focused on the former, the China high-speed railway design code (People’s Republic of
namely, the testing of the structure without the track system of China Ministry of Railway 2009).
ballast, sleepers, and rails, as shown in Fig. 1. Both the natural frequency and damping ratio have a significant
influence on the dynamic response of the bridge, as well on the
1 critical speed of the bridge. During the early stages of high-speed rail
Professor, School of Civil Engineering, Univ. of Burgos, 09001 Burgos,
Spain (corresponding author). E-mail: mvicente@ubu.es bridge design, it is very useful to have an approximate relationship
2
Professor, School of Civil Engineering, Univ. of Burgos, 09001 between the first natural frequency and the span length. This in-
Burgos, Spain. E-mail: dgonzalez@ubu.es formation can help estimate the critical speed of the train. The de-
3
Professor and Chair, Dept. of Civil, Architectural, and Environmental signer can then check whether the critical speed is close to the
Engineering, Illinois Institute of Technology, Chicago, IL 60616; formerly, operating and limit speeds of the rail service. If that is true, separation
Changjiang Scholar Chair Professor, Chinese Education Ministry, Tongji
of the two is advised strongly by redesigning the span arrangement.
Univ., Shanghai 200092, China. E-mail: gfu2@iit.edu
Note. This manuscript was submitted on July 8, 2013; approved on The damping ratio also directly influences the magnitude of dynamic
May 19, 2014; published online on June 13, 2014. Discussion period response of the structure, such as deflections and internal forces. In
open until November 13, 2014; separate discussions must be submitted for turn, it influences the impact factor to be used in design. Therefore,
individual papers. This technical note is part of the Journal of Bridge dynamic load testing of these bridges focuses on these two important
Engineering, © ASCE, ISSN 1084-0702/06014006(3)/$25.00. parameters.

© ASCE 06014006-1 J. Bridge Eng.

J. Bridge Eng., 2015, 20(2): 06014006

Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 01/16/18. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Test loading for a multispan bridge (image by Miguel A. Vicente)

Experimental Program

The experimental campaign reported here consisted of static and

dynamic load testing of 119 bridges over waterways, roads, and
land, on several high-speed rails in Spain. These tests were con-
ducted over the last 10 years or so. Static load–induced deflection
and dynamic load–demonstrated natural frequency and damping
ratio were acquired during testing. A wide range of span lengths from
20 to 90 m (65.6 to 295.2 ft) was included, for various superstructure
systems of precast prestressed I-beams and U-beams, cast in situ
posttensioned concrete slabs, and concrete box girders. The main
objectives of testing are summarized as follows:
1. To obtain correlation between the theoretical and measured
deflections and between the theoretical and measured first
natural frequencies, to establish an understanding regarding
the quality of the design calculation models.
Fig. 2. Span length to measured deflection versus span length multi-
2. To obtain an empirical relation between the measured first
plied by number of truck lanes for 119 bridges (1 m 5 3:3 ft)
natural frequency and also the damping ratio to the bridge span
length, depending on the structure type. Such relationships are
useful at an early stage of structural design to avoid resonance.
Four superstructure types were included in the testing program:
(1) isostatic spans with precast, prestressed concrete superstructure
(IPP); (2) hyperstatic spans with precast, prestressed concrete su-
perstructure (HPP); (3) isostatic spans with posttensioned cast in situ
concrete superstructure (IP); and (4) hyperstatic spans with precast,
posttensioned cast in situ concrete superstructure (HP).

Experimental Program Results

Measured Vertical Deflection

Fig. 2 shows the ratio of span length multiplied by the number of
truck lanes to measured maximum deflection versus span length for
119 bridges tested in this program. Each dot in the figure represents
one bridge. The deflection was measured at the midspan section. It
appears that this ratio decreased with an increase in span length. The Fig. 3. Measured first natural frequency versus span length for 119
same also was evident when the dots were plotted separately for each bridges (1 m 5 3:3 ft)
of the four superstructure types.

Measured First Natural Frequency and Damping Ratio

An effort was made to extract this relation, which can be useful
Fig. 3 displays the measured first natural frequency versus span for bridge designers. The following is given to describe this relation:
length for all 119 bridges tested. Each dot represents one bridge. In
the case of multispan bridges, the natural frequency was for the f ¼ a1 × L2a2 (1)
longest span, because it was expected to be the minimum natural
frequency or very first natural frequency of all spans. One sees in where f 5 first natural frequency (Hz) as defined previously; L
Fig. 3 that there was a trend of decreasing natural frequency with 5 span length (m); and a1 and a2 5 model coefficients. For higher
span length, which does not appear to significantly depend on the fidelity, the model coefficients were identified, respectively, for each
structure type. of the four superstructure types included in the test program, using

© ASCE 06014006-2 J. Bridge Eng.

J. Bridge Eng., 2015, 20(2): 06014006

 Table 1. Coefficients a1 and a2 Found for Eq. (1) Table 2. Statistical Parameters of Damping Ratio for 119 Tested Bridges
Typology a1 a2 R 2
Typology Number of specimens Mean (%) SD (%)
All 37.79 0.54 0.337 IPP 27 2.41 0.64
IPP 266.97 1.09 0.732 HPP 28 2.04 0.55
HPP 65.77 0.70 0.688 IP 7 2.24 1.19
IP 25.12 0.48 0.290 HP 57 1.76 0.58
HP 22.00 0.39 0.142

type, with respective statistical significance as shown in

Table 1. The correlation may be used preliminarily to esti-
mate the bridge’s behavior with regard to resonance, impact
Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 01/16/18. Copyright ASCE. For personal use only; all rights reserved.

factor, and performance.

3. Three of the four superstructure types tested were found to
have an average damping ratio above 2%, with the fourth at
1.8%. It appears that a 2% average value can be used as a
reasonable estimate for design purposes.
4. This work also offers the Spanish experience for comparison
with practices in other jurisdictions, which is needed by bridge
test engineers.

Acknowledgments

The authors express their deepest appreciation to the Spanish Min-

istry of Public Works for allowing publication of the data presented
Fig. 4. Measured damping ratio versus span length for 119 bridges herein. They also extend their gratitude to INDAICO INGENIEROS
(1 m 5 3:3 ft) S.L. for assisting in these tests and providing relevant information.

statistical regression. The results are shown in Table 1, with re- References
spective R2 values. Note that this statistical relation is similar to that
proposed by Paultre et al. (1992) for highway bridges. Some low R2 European Commission. (1996). Council directive 96/48/EC of July 23, 1996
values in Table 1 are attributable to these variables: (1) construction on the interoperability of the trans-European high speed rail system,
method (precast or cast in place), (2) material properties, (3) to- Brussels, Belgium.
pology other than span length, (4) span type (simple versus con- European Commission. (2001). Directive 2001/16/EC of the European
tinuous), and (5) bearing supports. Parliament and of the council of March 19, 2001 on the interoperability
For the damping ratio associated with the first natural frequency, of the trans-European conventional rail system, Brussels, Belgium.
Fig. 4 shows the result plotted against span length as well. Nev- European Commission. (2007). Commission directive 2007/32/EC of June
1, 2007 amending annex VI to council directive 96/48/EC on the in-
ertheless, no correlation between them was observed, whereas the
teroperability of the trans-European high speed rail system and
damping ratio varied between 1 and 4%. The same also was observed annex VI to directive 2001/16/EC of the European Parliament and of
for the damping ratio grouped by superstructure type. For the 119 the council on the interoperability of the trans-European conven-
bridges tested, Table 2 displays the average and SD for each su- tional rail system, Brussels, Belgium.
perstructure type. Except for the hyperstatic spans with precast, European Committee for Standardization (CEN). (2002). “Actions on
posttensioned cast in situ concrete superstructure (HP), all structure structures—Part 2: Traffic loads on bridges.” Eurocode 1, Brussels,
types had an average damping ratio a little more than 2%. Belgium.
International Union of Railways. (2006). “Loads to be considered in railway
bridge design.” UIC Leaflet 776-1, Paris.
Conclusions Paultre, P., Chaallal, O., and Proulx, J. (1992). “Bridge dynamics and
dynamic amplification factors. A review of analytical and experimental
The main conclusions of this work may be summarized as follows: findings.” Can. J. Civ. Eng., 19(2), 260–278.
1. Results of static load testing showed that the ratio of span People’s Republic of China Ministry of Railway. (2009). “Code for design of
length multiplied by the number of truck lanes to the measured high speed railway.” TB 10621-2009/J 971-2009, China Railway Press,
deflection appeared to decrease with span length. The isostatic Beijing (in Chinese).
spans with posttensioned cast in situ concrete (IP) superstruc- Spanish Ministry of Public Works and Transport. (2005). “ORDEN FOM/
1951/2005, de 10 de junio, por la que se aprueba la instrucción sobre las
ture showed smaller values among the four types tested. The
inspecciones técnicas en los puentes de ferrocarril.” ITPF-05, Madrid,
span type contributed to this behavior. Spain (in Spanish).
2. The presented experience identified a regression relation be- Spanish Ministry of Public Works and Transport. (2007). “ORDEN FOM/
tween span length and first natural frequency for high-speed 3671/2007, de 24 de septiembre, por la que se aprueba la instrucción
railway bridges in Spain. Options were offered for such corre- sobre las acciones a considerar en el proyecto de puentes de ferrocarril.”
lation to be used or not, depending on bridge superstructure IAPF-07, Madrid, Spain (in Spanish).

© ASCE 06014006-3 J. Bridge Eng.

J. Bridge Eng., 2015, 20(2): 06014006