Track stiffness assessment Alain Robinet, Engineering Department, SNCF, France Email: alain.robinet@sncf.fr Mohsen Hosseingholian, Civil Engineering Department, University of Caen, France Email: mhosseingholian@yahoo.com Alain Quibel, Centre d Experimentation Routière, CETE NC, France Email: alain.quibel@equipement.gouv.fr Michel Froumentin, Centre d Experimentation Routière, CETE NC, France Email: michel.froumentin@equipement.gouv.fr Abstract A good track support and its homogeneity are very important parameters for long-term stability of track geometry and a good life cycle cost of the track. The overall vertical support is provided by various substructure layers including the ballast, subballast, and subgrade layers. Weak modulus and non-uniform support from these layers can cause poor track geometry under traffic loading, which in turn will accelerate track component degradation and lead to the degradation of vehicle performance. Track reliability has nowadays become particularly important due to increased traffic densities, heavier axle loads, and shorter maintenance periods. To improve track maintenance efficiency as well as railroad operational performance, a continuous vertical track support measurement method was to be tested. We will present a feasibility research dealing with the suitable parameters to develop a new, non destructive, continuous method to measure track stiffness have been tested. The aim of the research was to adapt the Portancemeter to the track support measurement. This technique is implemented to measure the platform modulus currently in France. The improvements would be achieved through identification of high-priority track repair locations in order to aid in optimizing maintenance, and to observe track strength degradation. These improvements facilitate maintenance planning. This technique will be added to others tools used by SNCF for the assessment of the track support as the ground penetrating radar and platform drillings. Introduction For a long time, railroad engineers have wanted to develop a reliable method for measuring the response of the track structure under the traffic loading. One of the most important methods is based on measuring the vertical track stiffness or track modulus, to define the mechanical condition of the structure, in order to localize the track defaults and to optimize the maintenance periods. There have been many attempts to determine track stiffness or modulus from experimental data such as the ASCE-AREA Special Committee on Stresses in Railroad Track in 1918, the Swedish test performed at Säbylundmossen during the 1920s, and the efforts that have been done by track loading vehicles (TLV) to measure the standstill track modulus in the 1970s and 1980s. The ability to specify the bearing capacity of a track (or track stiffness), in order to determine the resulting rail stresses and accompanying track deformation, is essential proper track design and maintenance. Tthe railway structure can de divided into the superstructure, which consists of rails, fastening system, sleepers and ballast and the substructure consisting of the subballast layers, embankment fill (if needed) and subsoil. Since many substructure problems are related to water, the drainage system may be considered as part of the substructure. Historically, most attention has been paid to superstructure inspection techniques, which has resulted in a few standard measurements techniques that are used worldwide. The substructure has been given much
less consideration, especially the subballast and subsoil components, even though it has a major influence on the cost of track maintenance. Most of the substructure investigation techniques are not standard measurements and are not performed as regular measurements [1]. Vertical track stiffness can be defined as the ratio between the track load and the track deflection as a function of time. The stiffness varies with frequency, dynamic amplitude and applied load. It is an important parameter of wheel-rail interaction, which will monitor the mechanical behaviour of the track as well as track degradation. Track stiffness can be measured both at standstill and while rolling along the track. The method proposed in this paper for rolling measurements uses dynamic excitation of the track through a vibrating rolling wheel, and can thereby investigate the dynamic properties of a railway track without any severe restrictions. Test Method Presentation of implemented structure The objective was to study the response of a railway structure element, which has been reconstructed in a test site of the CER (Centre d Expérimentations Routière de Rouen) under specific dynamic stresses carried out mainly by means of a gantry of loading to variable parameters (forces and frequencies). The prepared structure was of the same type as classical railway structures. An overview of the structure and gantry of loading is given in Figures 1a and 1b, and sections in Figure 2. The dynamic loads were in the range of 10 to 75 kn and in a frequency band of 5 to 35 Hz. The aim was the test of various combinations of force-frequency to evaluate the dynamic stiffness of the structure for three different points of applied loading: right over a sleeper (A), in halfway between two sleepers with good conditions of coupling ballast-sleeper (B), and right over a sleeper in a case of non-contact between sleeper and ballast (C).
Figure 2: Schematic sections of the structure Based on the experimental design, the installation of sensors (stress gages, accelerometers, and probes of water content measurement) at various levels of the structure was provided. The measurement of displacement during the loading was carried out by a contact-free laser type sensor. In addition, the rail was equipped with seven axial accelerometers in different places on the bottom of the rail. In order to control moisture conditions the structure as whole was wrapped by a film of polyane. This article relates the results obtained from the first series of the tests, which were carried out under the normal moisture content conditions of the structure. The second phase of measurements will be accomplished after decreasing the modulus by watering the silty sand under controlled conditions. Material characteristics The standard proctor dry density of the silty sand (10.8 % < 80 µm) of the subgrade was, ρ d OPN = 1.86 kg/m3 at W OMC = 9.3 %. This material was classified as B2 type according to the French standard [2]. The average water content during the compaction was about 6.4%, which corresponded to a dry state. The UGM formation layer was a 0/31.5 entirely crushed quartzite (LA= 20, MDE= 3.4) with 5.6% of elements < 80 µm, whose modified proctor dry density was ρ d OPM = 2.23 kg/m3 at WOMC = 6.7%. The ballast was of the class 31.5/50 mm and conforms to SNCF specifications for the LGV (High SpeedTrack) according to XP P 15/545 article 12. The geotechnical and mechanical properties of the material are resumed in Tables 1 and 2.
Portancemètre MLPC This research aimed to test the feasibility of an apparatus that continuously measures the stiffness of the structure by rolling along the rail. Portancemètre MLPC [3], (operational version for measuring the modulus of platforms ) was used especially within the test framework in order to study the response of the structure (Figure 3). The characteristics at the wheel contact are, 10 kn of static weight, 0.5 mm of theoretical amplitude at 35Hz (Figure 4). During the tests, the vibration frequency of Portancemètre has been manually regulated between 15 and 35 Hz (by 5 Hz intervals). ] The expression of FTA is (equation 1):. FTA = M1.g + M 0.Γ b + (M 1 -M 0 ).Γ c + m e.ω² cos (ϕ) (1)
In which M1 is the total mass of the unit (trailer), M0 is the mass of vibrating wheel,.γ b is the vertical acceleration of the vibrating wheel, Γ c is the vertical acceleration of the suspended mass, ω = 2n.frequency and ϕ is the angle of rotation. The double integration of the wheel vertical acceleration,. vibrating wheel, d, (equation 2). b determines the vertical displacement of the Γ. b (2) d(t) = () t. dt dt The calculated total force, FTA, and the corresponding deflection, d, allow one to determine the stiffness, k, over an average period on platforms. The apparatus establishes a continuous profile of stiffness versus the measured distance along the swept trace. Definition and calculation of track stiffness under hydrodynamic excitation jack In the time domain, track stiffness (k) is generally defined as the ratio of track load F(t) and track deflection d(t) (Equation 3) in the linear part of the rising phase of the load. It is a function of frequency as well as of load. If track stiffness is studied in the frequency domain, the track receptance, á(f) or dynamic flexibility is preferable, which is the inverse of the dynamic track stiffness, and is often displayed with magnitude and phase. F() t k ( t) = (3) d() t Figure 5 displays a force-deflection diagram where the rail is loaded to 50 kn while the corresponding deflection is measured. Nearly every time, the curve is non-linear with a damping factor (hysteresis). In simplifying, the stiffness is often considered as a constant, which may be not valid in the certain cases of force-frequency. However, in order to use the concept of transfer functions, we assume that the system is linear in a limited part of the rising phase of the force-deflection diagram between 0.3 Fmax to 0.9 Fmax.
Experimental Results For evaluating the total dynamic stiffness of the structure, the various types of force-frequency have been selected with the hydrodynamic jack. The maximal forces were 10, 30, 50 and 75kN and the frequencies varied between 5 and 35Hz by intervals of 5. Before executing each experimental mode, the structure undergone a stabilization phase by generating a 500 000 loading cycles, with a force of between 20 and 50 kn at a frequency of 35 Hz. Behaviour of the structure under the tests The attenuation in acceleration from rail to subgrade was very fast (Figure 6). The ratio of the measured accelerations between the surface of the rail and the subgrade was about 10, depending on the applied force and frequency. According to the measured parameters in the structure (induced accelerations and pressures), the distributions of these parameters show the damping factor of the ballast layer. Under the force of 75 kn the measured stresses in the formation layer and subgrade were respectively about 30 and 20 kpa. At this mode, the recorded values of acceleration were respectively about 13 and 9 m/s². Definition of an operational zone for a specific vibrating wheel on tracks Figure 7 represents the results of track stiffness (the values in blue) and amplitudes (the values in red) in the various modes of force-frequency combination obtained by the hydrodynamic jack of loading. These results show the average values of four series of tests for each mode. The results of Portancemètre on the rail (the values on green background) and the result of RSMV-type (the value on pink background) are also added in the figure. Based on the force deformation curves like Figure 5, and the diagram Figure 7, the operational zone of suitable solicitations related to the characteristics of the track can be defined. This zone was obtained according to the reconstructed structure in CER.
By this figure, it is clear that the points of actual Portancemètre are not well adjusted in frequency for this purpose. On the track, the maximum level of dynamic force remained approximately low (about 18 kn) compared to the one on platforms (twice on a 150 MPa). A significant part of the energy also was lost at these low forces and high frequencies. The second phase of research on the different type of structures which is going to start, permits us to develop the new version of Portancemètre (operational version for railway structure) and to determine its operational mode. Conclusion The research of Portancemètre on the rail was very different from that known on a classical soil. Notably, we can add here the operation mode of this apparatus should be stay in the low frequency and high dynamic load ranges. These results will be completed by the second series of tests carried out under hydrous condition of saturated support. Then the calculated stiffness by different excitation modes will lead us to determine the sensibility of this parameter under degradation of the modulus of support. After these feasibility tests, a new design of the Portancemètre within the operational zone will be achieved in order to setup a specific vibrating wheel, which be able to check the characteristics of the substructure with the suitable calibration for maintenance applications.
References [1] Dynamic Track Stiffness Measurement, A new tool for condition monitoring of track substructure, Thesis of Eric Berggren 2005 [2] Norme NF P 11 300, Classification des matériaux utilisables dans la construction des remblais et des couches de formes d infrastructures routières [3] Le Portancemètre Mesure continue de la portance des plates formes, Descriptif technique, CETENormandie-Centre - Centre d Etudes et de Construction de Prototypes