### Filtern

#### Erscheinungsjahr

#### Dokumenttyp

- Zeitschriftenartikel (38)
- Beitrag zu einem Tagungsband (34)
- Vortrag (32)
- Beitrag zu einem Sammelband (17)
- Posterpräsentation (3)
- Buchkapitel (1)

#### Sprache

- Englisch (125) (entfernen)

#### Schlagworte

- Ground vibration (21)
- Railway track (9)
- Finite-element boundary-element method (8)
- Slab track (7)
- Track-soil interaction (7)
- Vehicle-track interaction (7)
- Layered soils (6)
- Mitigation (6)
- Field tests (5)
- Layered soil (5)

#### Organisationseinheit der BAM

- 7 Bauwerkssicherheit (54)
- 7.2 Ingenieurbau (54)

Abstract. Three measurement campaigns of train-induced ground vibrations are evaluated for the vehicle-track-soil interaction. Ground vibrations, track vibrations and vehicle vibrations have been measured for train passages and impulse excitation and compared with theoretical results.
The soil and the track-soil system are calculated by wavenumber integrals. The influence of the vehicle is introduced by a substructure method. By comparing theory and measurement the different components of excitation force and ground vibration can be analysed, the quasi-static excitation, track-alignment errors, the out-of-roundness of wheels, the wheel and rail roughness, and moreover, scattered axle impulses and ineffective high-frequency parts of the wheelset accelerations and forces.

Measured train passages and hammer impacts in combination with track-soil calculation have been successfully used for the detection of damaged slab tracks. This approach is now extended to intact slab and ballast tracks. The vibrations of many tracks have been measured at several levels from rail, sleeper, track plate, base plate, base layer to the subsoil by velocity or acceleration sensors. The time histories have to be integrated once or twice to get the displacements. The displacement signals include an arbitrary time-dependent shift which must be eliminated or respected in the interpretation. On the other hand, the calculation of slab and ballast tracks have been done in frequency-wavenumber domain. The displacements along the track and the frequency-dependent compliance transfer functions can be calculated. The latter can be compared with the results of the hammer impacts on the track. The deformation of the track can be transformed to time histories for a whole train and compared to the measured train passages. Many slab (and ballast) tracks have been measured at different sites. The displacements of the tracks are presented, and the following parameters have been analysed in the measurement results: slab track vs. ballast track, different types of slab tracks, damaged slab tracks, different trains, switches at different measuring points, an elastic layer, the mortar layer, different soils at different places. The soil should have the dominant influence on the track-plate displacements. Slab and ballast track yield also big differences in maximum displacement and width of deformation. Some of the preceding aspects will be analysed in comparison of measurement and theory.

Methods have been presented for detailed studies of railway vibration and for the fast prediction of train-induced ground vibration. The ground vibration is generated by static or dynamic loads. The main purpose of this contribution was to show the influence of inhomogeneous soils on the different vibration components.
Layered soils, namely a soft layer on a stiffer half-space, yield a quite specific transmission behavior. The low-frequency and sometimes also high-frequency cut-off of the transfer function of the soil is demonstrated in theory and by experiments at many sites of which the soil model is approximated from dispersion and transfer function measurements. The layer frequency divides the frequency range in a low-frequency range, where the stiff half-space rules the low amplitudes, and a high amplitude high-frequency range which is mainly determined by the softer top layer. A thick soft layer yields a very low layer frequency, so that the higher soft soil amplitudes have a wider range down to low frequencies. A thin layer yields a high layer frequency, so that the high frequencies above this layer frequency are dominant. The higher the contrast between the stiff half-space and the soft layer is, the stronger the increase between the half-space and layer amplitudes, the more characteristic are the spectra of the soil transfer function. The range of measured soils has been from vS1 down to 125 m/s, vS2 up to 1000 m/s and the layer frequencies are within 10 Hz < f0 < 75 Hz. Moreover, during this measuring campaign in Switzerland, all 11 sites showed clearly the layer-on-half-space behaviour. The transfer functions of inhomogeneous soils have been used to predict the ground vibration due to dynamic axle loads which is usually thought to be the most important component.
The passage of static loads, in the contrary, results in very small vibration amplitudes for low train speeds, which can only be found at near distances and at low frequencies. They attenuate very rapidly with distance and lose very rapidly the higher frequency content. The passage of static axle loads can be included in the prediction of railway vibration just for completeness.
Special attention should be given to the case if the train runs with the Rayleigh-wave speed of the soil (Rayleigh train). The Rayleigh-train effect is strongest for a homogeneous half-space: At the near-field of the track the amplitudes are raised strongly compared to normal trains, and in addition, little attenuation with distance is observed. In case of a layered soil, the low-frequency cut-off reduces the frequency range and the amplitudes of the homogeneous quasi-static ground vibrations. Therefore, the Rayleigh-train effects are clearly reduced by a layered soil and they disappear if the layer frequency (for example for a thin layer) is higher than the frequency band of the axle impulse. The Rayleigh-train effect could completely disappear in a randomly inhomogeneous soil, but this has not been analysed so far.
The axle impulses from static loads can have an additional, quite different effect. They can be scattered by a randomly inhomogeneous soil so that a part (the scattered part) of the axle impulse can reach further distances from the track. This can establish a certain mid-frequency component of the ground vibration which becomes dominant in the far-field, and this important component exists for all train speeds. Experimental results from BAM and international measurements show the importance of the corresponding frequency range.
The mitigation of train induced ground vibration by elastic and stiff track elements has been analysed threefold. The vehicle-track interaction yields the reduction at high frequencies above the vehicle-track resonance. This is the standard effect. The filtering of trackbed errors by the bending stiffness of the track yields a certain mid-frequency effect. An even stronger mid-frequency effect is predicted for the mitigation of the scattered axle impulses by the bending stiffness and elastic elements of the track.

The damage detection and repair control have become important tasks for ballast and slab tracks. Measurements which compare the damaged and the repaired status of the same track section at different times, or which compare a damaged and an intact track section at the same time, have been successfully performed at some sites in Germany. The loss of contact between the sleeper and the track plate, between the track plate and the base plate, and between the base plate and the base layer have been analysed. The soil properties of each site have been measured and have been used to establish realistic track-soil models. Theoretical results of the wavenumber domain and the finite-element boundary element method have been compared with the experimental results. The observed experimental and theoretical results, changes in the time histories of displacements and velocities due to train passages and in the transfer functions (receptances) due to hammer impacts, are encouraging that these measurements can be used to detect track damage.

Offshore wind energy towers are dynamically loaded by waves and wind. Pile foundations provide stiffness and damping and should be properly calculated. A combined finite-element boundary-element method for the dynamic interaction of flexible structures and the soil has been developed. The flexible structures such as single piles or complete wind energy towers are modeled by the finite element method whereas the homogeneous or layered soil is modeled by the boundary element method which uses the Green’s functions for interior loads in the layered half-space to establish the dynamic stiffness matrix of the soil. Soils with a stiffness that is continuously increasing with depth can be modeled as multi-layer soils with step-wise increasing stiffness. The effects of different parameters such as the stiffness of the soil, the axial and bending stiffness of the pile, and the radius of the cylindrical contact area will be analysed for the different components of excitation (vertical, horizontal, rotation and coupling). The results can be determined as specific power laws which are different for the different load cases and for the different soil models (Winkler support, homogeneous continuum, continuum with increasing stiffness). The dynamic effect of radiation damping will be analysed by the frequency-dependent compliance functions. A clear layering of the soil can cause noticeable changes in the dynamic compliances as reductions of the stiffness and the damping in certain frequency ranges (below and around layer resonance frequencies). The distribution of the displacements along the pile help to explain the observed laws. An example of an offshore wind energy tower has been modeled and calculated for wind, wave and weight loads. The resonances of the tower are usually limited by the radiation damping which is strongest for a soft soil.

This contribution presents experimental methods to detect track damage. At BAM (Federal Institute of Material Research and Testing), a measuring car with a measuring system of 72 channels, geophones, mountings, cables, harmonic and impulsive exciters is used for dynamic measurements of the track, the soil and buildings. An instrumented hammer allows force measurements and to evaluate transfer functions of the track, and the soil. Wave measurements are used to identify the soil characteristics. Train passages are measured at the track and for the train induced ground vibrations. In addition to these in situ options, tests of tracks or track elements can be performed in a large laboratory.

Explosion-induced ground vibrations have been measured at several places. Results about the wave propagation are shown in this contribution. The particle velocities of the soil have been measured at up to 1000 m distance from the explosion and are presented as time records (seismograms) and one-third octave band spectra (transfer functions). The results are compared with the results of hammer impacts. The seismograms clearly show different wave types, compressional waves of the air, the water and the soil, and the Rayleigh wave. The hammer impacts yield good results up to 100 m and incorporate higher frequencies at about 50 Hz, whereas the explosion results in a ground vibration with frequencies around 10 Hz and a longer range of influence. Explosion and hammer excitations are evaluated for the wave velocities of the soil by using the wavenumber and the spatial auto-correlation method. The attenuation of the ground vibration amplitudes A with distance r can well be presented by a power law A ~ r -q. This type of amplitude-distance law and the corresponding power q > 1 are substantiated in the contribution. The influence of the charge weight W is evaluated as an additional power law A ~ W -p for each measuring site. The power is found quite similarly around q 0.6 as all sites have a medium soft soil such as sand and clay. The obtained amplitude-charge-distance law can be used to predict the explosion-induced ground and building vibrations at other sites.

Measurements of downburst wind loading acting on an overhead transmission line in northern Germany
(2017)

Along an overhead transmission line in Northern Germany, a unique instrumentation of anemometers and force measurements is installed. Details of this test line with wind measurements along a horizontal axis are given. A recent event of a presumable downburst wind event is analyzed by means of available data and precedent works on thunderstorm analysis. The measured response of the conductors at the suspension tower is investigated and compared with time domain simulation of a finite element model.

A combined finite-element boundary-element method for the dynamic interaction of the soil with flexible structures such as single piles or complete wind energy towers has been developed. Flexible piles in different soils are analysed in frequency domain. The different parameters such as the stiffness of the soil, the bending stiffness and the radius of the hollow pile are analysed for their influence on the complex compliances. The results have been determined as specific power laws which are different for the different load cases (horizontal, rocking, coupling) and for the different soil models (Winkler, continuum with constant, root-parabolic and proportional-linear stiffness variation). The strongest influence of the soil stiffness can be found for the homogeneous soil and the horizontal component. Winkler soils have a weaker influence than the corresponding continuous soils. An offshore wind energy tower has been modeled and calculated for wind and wave loads.

Three measurement campaigns of train-induced ground vibrations are evaluated for the vehicle-track-soil interaction. Ground vibrations, track vibrations and vehicle vibrations have been measured for train passages and impulse excitation and compared with theoretical results. The soil and the track-soil system are calculated by wavenumber integrals. The influence of the vehicle is introduced by a substructure method. By comparing theory and measurement the different components of excitation force and ground vibration can be analysed, the quasi-static excitation, track-alignment errors, the out-of-roundness of wheels, the wheel and rail roughness, and moreover, scattered axle impulses and ineffective high-frequency parts of the wheelset accelerations and forces.

Three measurement campaigns of train-induced ground vibrations are evaluated for the vehicle-track-soil interaction. Ground vibrations, track vibrations and vehicle vibrations have been measured for train passages and impulse excitation and compared with theoretical results. The soil and the track-soil system are calculated by wavenumber integrals. The influence of the vehicle is introduced by a substructure method. By comparing theory and measurement the different components of excitation force and ground vibration can be analysed, the quasi-static excitation, track-alignment errors, the out-of-roundness of wheels, the wheel and rail roughness, and moreover, scattered axle impulses and ineffective high-frequency parts of the wheelset accelerations and Forces.

Two measurement campaigns of train-induced ground vibrations are evaluated for the vehicle-track-soil interaction. Ground vibrations, track vibrations and vehicle vibrations have been measured for train passages and impulse excitation and compared with theoretical results. The soil and the track-soil system are calculated by wavenumber integrals. The influence of the vehicle is introduced by a substructure method. By comparing theory and measurement the different components of excitation force and ground vibration can be analysed, the quasi-static excitation, track-alignment errors, the out-of-roundness of wheels, the wheel and rail roughness, and moreover, scattered axle impulses and ineffective high-frequency parts of the wheelset accelerations and forces.

Measurements of ground and track vibrations have been performed at a high-speed line in northern Germany. Impacts on the track and the ground, and passages of different trains with different speeds on different tracks have been measured. Transfer functions of the soil are presented and approximated by theoretical soil models. By using these transfer functions, the measured ground vibration between 2 to 64 m distance from the track can be transformed into a load spectrum which can be used for predictions at other sites. The method is compared to the soil-dependent method of an emission spectrum at a certain distance (8 m for example). The influence of train type, speed and track type is discussed on the base of the different emission quantities and the original measurements. The strong influence of the track, ballast track and slab track, is analysed by a theoretical model in wavenumber domain. The response of the track to the passage of the static load is reduced by the stiffness of the slab, the deformation of the track as well as the impulse acting on the soil. Usually, the impulse on the soil should result in a slow quasi-static movement of the soil, slower at further distances. In a heterogeneous soil, however, the impulses from the static loads scatter and keep parts of the higher impulse frequency band. In this case the reduced impulse spectra of the slab track will yield reduced ground vibration in a certain frequency band. Additional (BAM and international) measurements will be used to discuss this and possible other explanations for the different ground vibration differences.

The dynamic compliance of different railway tracks on different layered soils have been calculated using a combined finite-element and boundary-element method:
conventional track on homogeneous soil with different stiffness,
extreme normal and inverse layering of the soil,
plates in and on top of the soil (slab track on various soils),
conventional track with elastic rail pads.

This contribution presents some principles and some examples of the mitigation of railway-induced ground vibrations. The principles are different for the mitigation measures at the track, in the soil or at the building. Force transfer functions of isolated and un-isolated track-soil systems, reflected and transmitted wave amplitudes at walls and trenches in the soil, and the transfer of the (free-field) vibration amplitudes to the foundation amplitudes of the building are analysed. The mitigation effect can be calculated by exact or simplified formulas. Some examples with 3D (finite-element boundary-element), 2D (beam-on-support), and 1D track models, 2D and 1D soil models, detailed 3D building models and finite or infinite 1D wall-floor models are investigated to find out if simple models can be used for a satisfactory prediction of the mitigation effect. The 1D track examples show that the force transfer of the track without vehicle can be exactly calculated, whereas the total force transfer can be calculated approximately if appropriate wheelset masses per track length are used for the isolated and the un-isolated track. The mitigation effect of a filled trench is calculated by a 2D finite element model and the results compare with simple transmission formula if the stiffness per area rather than the wave impedances are used for the infill material. The base isolation of a building is analysed by a detailed 3D model and the results are similar to the analytic results of a single wall with floors on the soil. Other reduction measures as different floor and column dimensions are usually less effective so that the clearly best mitigation solution at a building is a partly or a complete base isolation.

Measurements on the vehicle-track interaction and the excitation of railway-induced ground vibration
(2017)

Two railway measurement campaigns have been performed in Germany and Switzerland which yield insight in the vehicle-track-soil interaction. The campaign in Germany has included simultaneous measurement of vehicle, track, and soil vibrations during train runs with 16, 25, 40, 63, 80, 100, 125, 140, 160 km/h, and impulse measurements of the passenger car, three track sections and the soil. Two ballast tracks, one on the soil surface and one on a concrete bridge, have been investigated as well as a slab track in a tunnel. Ten different sites in Switzerland have been measured for soil properties and train-induced ground vibrations, which allow to determine the excitation forces of the railway traffic. New axle-box measurements at some of the Swiss sites have been analysed to get further experimental evidence. All these measurements have been evaluated to characterize the excitation processes. Relations between vehicle vibration and ground vibration can be observed. The vehicle vibrations, namely the accelerations of the wheelsets, yield the dynamic forces due to the passage over the irregularities of the vehicle and the track. The ground vibrations are correlated to these dynamic forces to a certain extent. Some mid-frequency ground vibration amplitudes, however, are higher than expected from the dynamic excitation forces. The experimental observations can be explained by an irregular response to the passage of the static loads, that means the passage of the static loads over an irregular ballast or soil. This correct understanding of the excitation processes is important for the prediction as well as for the mitigation of railway induced ground vibrations.

The damage detection and repair control have become important tasks for ballast and slab tracks. Measurements which compare the damaged and the repaired status of the same track section at different times, or which compare a damaged and an intact track section at the same time, have been successfully performed at some sites in Germany with slab tracks and ballast tracks and compared with the theoretical behaviour of intact and damaged tracks. The loss of contact between the sleeper and the plate, between the plate and the base layer, and some problems with soft or weakened soil have been analysed. The observed results, changes in the time histories of displacements and velocities due to train passages and in the transfer functions (compliances) due to hammer impacts, are encouraging that these measurements can be used to detect track damage. In addition, calculations with the combined finite-element boundary-element method have been used to confirm the conclusions about intact or damaged railway tracks.

Transfer admittance spectra of technically induced surface wave fields are analyzed in theory and experiments. Theoretical admittance spectra of layered soils are obtained by integration in wavenumber domain and compared with experimental admittances due to hammer or vibrator excitation. The admittance spectra are strongly influenced by the layering and damping of the soil. Deep stiff-soil layers yield a low-frequency cutoff, whereas a strong damping yields a high-frequency cutoff. A sharp cutoff in a narrow frequency band, which is measured at some sites, can be explained by a damping that increases with frequency, such as viscous material or scattering damping.

Measurements on the vehicle-track interaction and the excitation of railway-induced ground vibration
(2017)

Experiments have been performed at a test site with six different tracks with under-ballast plates. Hammer excitations of the soil and the tracks as well as train passages have been measured. The experimental observations are as follows. 1. The natural soil is stiff gravel whereas the railway dam consists of softer material. 2. The track compliance indicates a soft ballast if no train is present to provide a confining pressure. 3. The track response to the train passages can be split into a low-frequency region which is ruled by the static loads and a high-frequency region which is ruled by dynamic loads. 4. The track responses to hammer and track excitation indicate the presence of many voids between the sleepers and the ballast. 5. The ground vibrations are highly influenced by the soil. Due to the stiff soil at the site, the hammer and train induced spectra have a considerable high-frequency content. 6. A reduction of the ground vibration has been observed in a low-frequency range. The mitigation effects of an under-ballast plate are also investigated by calculations of a wavenumber domain model. The under-ballast plate has an effect at low frequencies where it distributes the static load over a longer track section. The impulse of the axle passage is longer and the frequencies are lower due to the plate stiffness. The axle impulses could yield a low-frequency ground vibration in an irregular soil with a randomly varying stiffness. This low-frequency part of the ground vibration (the scattered axle impulses) seem to be reduced by the under-ballast plate.

Train passages induce forces on the track, train-induced vibrations propagate through the soil and excite neighbouring buildings. The emission, which is the first part of the prediction of vibrations near railway lines, is presented by focusing on the dynamic axle loads. The calculation of the axle loads is based on the vehicle-track-soil interaction. This interaction calculus utilises the dynamic stiffness of the vehicle (the inertia of the wheelset) and the dynamic stiffness of the track-soil system. Based on various time consuming finite-element boundary-element calculations, an approximate track-soil model has been established. The vehicle-track-soil analysis yields several transfer functions between the various geometric or stiffness irregularities and the axle loads of the train. Geometric irregularities of the vehicle (the wheels) and the track (rail surface and track alignment) are the simplest components. Geometric irregularities of the subsoil (trackbed irregularities) have to be transferred to effective irregularities at rail level. The bending stiffness of the track is filtering out the short-wavelength contribution. Stiffness irregularities occur due to random variations in the ballast or the subsoil, which must also be transferred to effective track irregularities, and due to the discrete rail support on sleepers. All necessary transfer functions for the prediction of axle-load spectra are presented as general formula and as specific graphs for differing vehicle and track parameters. The prediction method is applied to a ballast track and a slab track and compared with corresponding axle-box measurements. Moreover, ground vibration measurements at numerous sites are exploited for the axle-load spectra and the validation of the prediction method. All theoretical and experimental results confirm that the dynamic axle-load spectra have an approximate value of 1 kN per third of octave and increase with train speed, track stiffness and around the vehicle-track resonance.

The vibrations of soil and foundations are demonstrated for different types of loading. Train-induced ground vibrations are studied in a measurement campaign where a test train has run with regularly varied speeds. The measured train-induced soil vibration at 2 to 100 m distance from the track is compared with the wave propagation due to hammer excitation and with the theoretical wave field. The strong influence of the soil and the train speed on the amplitudes and frequencies of the vibration has been analysed for passages of the locomotive and the carriages. - The generation of ground vibration by strong explosions has been studied on a large testing area with sandy soil. The propagating waves were measured in a regular grid of measuring points in 10 to 1000 m. Therefore, the dominance of certain waves at certain distances and the changes of compressional waves and Rayleigh waves could clearly be observed. The results are compared with impulse hammer measurements in the range of 5 to 50 m. - A drop test facility has been built on the testing area of the Federal Institute of Materials Research and Testing (BAM). Heavy masses (containers) of up to 200 t can be dropped from 10 m height on a big reinforced concrete foundation. The foundation was instrumented by accelerometers, strain gauges and pressure cells to give information about the loading condition and by geophones to measure the vibration of the surrounding soil and building. Both excitation processes, the release of the mass and the impact, produce high vibration amplitudes. On a smaller drop foundation, the influence of the drop height and the target stiffness has been studied more systematically.

A complex measuring campaign has been performed including the simultaneous measurement of vehicle, track, and soil vibrations during train runs at 16, 25, 40, 63, 80, 100, 125, 140, 160 km/h, and impulse measurements of the passenger car, three track sections and the soil. A ballast track on the soil surface and on a concrete bridge have been investigated as well as a slab track in a tunnel. The evaluation and comparison of all these data shows a generally good agreement for all components if the strong low- and high-frequency cut-off characteristics of the layered and damped soil are incorporated. There is a strong causal correlation between the vehicle and the soil by the dynamic excitation forces and a weak relation between the track and the soil by the axle-sequence spectrum of the train. However, the similarity between the axle-impulse spectrum observed at the track and the spectra of the ground vibration lead to the special excitation component of “scattered axle impulses” which is pre-dominant at the far-field points of the soil.

Some results of our measurements of the Intercity experimental are presented together with simulated results and their interpretation. Three frequeny ranges could be measured and calculated: low frequencies which decrease very rapidly with distance (the quasi-static part), high frequencies (mainly due to the sleeper passage), and a mid-frequency which has the weakest attenuation with distance and is therefore dominant at the far field. Mono-frequent excitations get a wide frequency band due to the Doppler effect of the moving high-speed train.

Ground vibrations created by running high-speed trains at speeds between 100 and 320 km/h are calculated in detail using transfer functions to model the effects of the moving loads. These transfer functions for layered soils are obtained by integration in the wavenumber domain. The train-induced vibrations in a soil that is considered to consist of single layers of two slightly different soils are analysed for different excitations: for their spectra, attenuation laws and amplitude-speed relations. An important mid-frequency component is shifted through the cut-on region of the layered soil with an increase in the train speed. The cut-on frequency divides the response of the layered soil into a low-frequency low-amplitude range and a high-frequency high-amplitude range. This leads to completely different train speed dependencies for the two soil layers with strongly increasing amplitudes around the cut-on frequency and almost constant amplitudes beyond this frequency. All calculated results closely agree with ground vibration measurements at two corresponding sites, especially if the mid-frequency component is calculated by axle impulses.

A variety of isolation measures exists to reduce the vibration in the neighbourhood of railway lines. They can be roughly classified as elastic or stiffening systems. There are the following elastic elements, rail pads or resilient fixation systems between rail and sleeper, under sleeper pads or sleeper shoes under the sleepers, and ballast mats under the ballast. Stiffening systems (plates) are used as slab tracks, floating slab tracks, or mass-spring systems. In the EU project “Railway induced vibration abatement solutions (RIVAS)”, elastic under sleeper pads have been investigated. The dynamic behaviour of the track and the surrounding soil has been calculated by the combined finite-element boundary-element method in a systematic parameter study. It has been shown that the mitigation effect can be improved by soft under sleeper pads or by heavy sleepers. Consequently, such track elements (soft under sleeper pads and heavy sleepers) have been thoroughly investigated in laboratory tests to establish the static and dynamic parameters as well as their serviceability. Finally, field tests at and near railway tracks with and without under sleeper pads have been performed. To determine the reduction effect of the isolated track, the ground vibrations excited by trains or artificial sources have been measured. The soil properties at the different sites have also been measured so that the comparison of the isolated and un-isolated track can take into account possible differences of the soil parameters. The contribution shows how the different (numerical, laboratory and field) methods and results can be combined to achieve an improved mitigation solution with soft under sleeper pads and heavy sleepers for ballasted and slab tracks.

The Federal Institute of Material Research and Testing (BAM) has collected some experience with the testing of damaged, repaired and newly constructed railway tracks. The experimental methods are hammer testing of the track at different positions, hammer testing of the soil, measurement of train passages, and in all cases, measurements with geophones at different positions. The measured signals are evaluated for wave velocities (dispersion of the soil or the track-soil system), for transfer functions (transfer admittances of the soil, compliances of the track in amplitude and phase), and one-third octave band spectra of the track response to hammer and train excitation. These methods are applied at different stages of the track construction. Before track construction, wave velocities and transfer functions of the sub-soil can indicate problems with soft soils. After track construction, a check of the acceptable state of the track can be done by comparison of many excitation positions and track sites. After a track damage (a lose sleeper or a lose plate of a slab track) and after its repair, the sensitivity of the different measurement quantities to different track errors and the achieved improvement of the repair can be determined. The contribution shows examples of all these track situations.

The Federal Institute of Material Research and Testing (BAM) has collected some experience with the testing of damaged, repaired and newly constructed railway tracks. The experimental methods are hammer testing of the track at different positions, hammer testing of the soil, measurement of train passages, and in all cases, measurements with geophones at different positions. The measured signals are evaluated for wave velocities (dispersion of the soil or the track-soil system), for transfer functions (transfer admittances of the soil, compliances of the track in amplitude and phase), and one-third octave band spectra of the track response to hammer and train excitation. These methods are applied at different stages of the track construction. Before track construction, wave velocities and transfer functions of the sub-soil can indicate problems with soft soils. After track construction, a check of the acceptable state of the track can be done by comparison of many excitation positions and track sites. After a track damage (a lose sleeper or a lose plate of a slab track) and after its repair, the sensitivity of the different measurement quantities to different track errors and the achieved improvement of the repair can be determined. The contribution shows examples of all these track situations

This article presents an integrated model for the computation of vehicletrack interaction and the ground vibrations of passing trains. A combined finite element and boundary element method is used to calculate the dynamic compliance of the track on realistic soil whereas multi-body models are used for the vehicle. The dynamic stiffness of the vehicle and that of the track are combined to calculate the dynamic axle loads due to the irregularities of the vehicle and the track as well as those due to sleeper passing excitation. These loads serve as input for the calculation of ground vibration near railway lines in the time and frequency domains.
The theoretical methods and results have been proven by experiments in several respects and at several instances. First, on the occasion of the test and record runs of the Intercity Experimental, there was a very good quality of the vehicle and of the newly built track so that the deterministic parts of the excitationthe static load and the sleeper-passing componentcould clearly be identified, the first being of minor importance apart from the track. Second, simultaneous measurements of the vehicle, the track and the soil at three different track situations were performed where we could verify the different parts of the stochastic excitation and their importance for the ground vibrations. The irregularities of the vehicle are dominant at high frequencies whereas the irregularities of the track are more important at lower frequencies. The comparison of the theory and the measurements also points to the phenomena of the vehicletrack resonance and the scattering of the quasi-static axle impulses by randomly varying soil.

The maintenance of the transport infrastructures and their further development are going to remain focal points for investment and research in Germany in future. According to the latest development forecasts made by both the federal government and Deutsche Bahn, even if rail´s percentage share of the market were to remain unchanged, growth of around 50% would be expected in the next ten years, especially in freight traffic. This growth is necessitating considerable development both in the technical design of the tracks and in the abatement of the noise and vibration caused by railway traffic.

The dynamics of un-isolated and isolated ballast tracks have been analysed by multi-beam models for the track and by a layered half-space model for the soil. The solution is calculated in frequency-wavenumber domain and transformed back to space domain by a wavenumber integral. This is a faster method compared to other detailed track-soil interaction methods and almost as fast as the widely used Winkler-soil method, especially if the compliances of the soil have been stored for repeated use. Frequency-dependent compliances and force transfer functions have been calculated for a variety of track and soil parameters. The ballast has a clear influence on the high-frequency behaviour whereas the soil is dominating the low-frequency behaviour of the track. A layering of the soil may cause a moderate track-soil resonance whereas more pronounced vehicle-track resonances occur with elastic track elements like rail pads, sleeper pads and ballast mats. Above these resonant frequencies, a reduction of the excitation forces follows as a consequence. The track deformation along the track has been analysed for the most interesting track systems. The track deformation is strongly influenced by the resonances due to layering or elastic elements. The attenuation of amplitudes and the velocity of the track-soil waves change considerably around the resonant frequencies. The track deformation due to complete trains have been calculated for different continuous and Winkler soils and compared with the measurement of a train passage showing a good agreement for the continuous soil and clear deviations for the Winkler soil model.

Ballast mats are an efficient measure to reduce the vibrations near railway lines. The vehicle-track system gets a low eigenfrequency due to the insertion of an elastic ballast mat under the ballast. For frequencies higher than this low vehicle-track eigenfrequency, the forces, which are generating the vibration of the soil, are considerably reduced. In this contribution, a combined finite-element boundary-matrix method is used to calculate a number of completely three-dimensional track models with and without ballast mats. The influence of the important parameters such as the stiffness of the ballast mat, the unsprung vehicle mass, the mass of the track, and the stiffness of the subsoil is investigated. The numerical results are presented as the transfer functions of the total force that is acting on the soil and generating the vibration of the environment. The effectiveness of ballast mats is achieved by division of two of these force functions. The general tendencies for this insertion loss are discussed and a comparison with measurements is given. To come to an improved practical tool for the design of ballast-mat tracks, the finite-element method results are approximated by a simple two-dimensional model of which the solution is given explicitly.

The reduction of train-induced ground vibration by elastic elements such as rail pads and sleeper pads has been analyzed by a combined finite-element boundary-element method. The dynamic compliance of the track, the transfer function of the total force on the ground and the ground vibration ratios have been calculated for a variety of isolated and un-isolated track systems. It has been found that the soil force transfer, which describes the excitation force of the soil, is an appropriate quantity to predict the reduction of the ground vibration and the effectiveness of isolated tracks. All force transfer functions of isolated tracks display a vehicletrack resonance where the wheelset on the compliant track is excited by wheel and track irregularities. At higher frequencies, considerable reductions of the amplitudes are observed as the benefit of the resilient element. The influence of the stiffness of the rail or sleeper pads, the ballast and the soil, and the mass of the sleeper and the wheelset on the resonance frequency and the reduction has been investigated. Sleeper pads are advantageous due to the higher mass that is elastically supported compared to the rail-pad track system. The combination of elastic rail and sleeper pads has been found to be disadvantageous, as the second resonance occurs in the frequency range of intended reduction.

The computation of the wave propagation in homogeneous and layered soils can be performed by a numerical integration in wavenumber domain. The numerical difficulties of an infinite integral and an integrand with poles can be solved. But if this computation must be repeated for many distances, many frequencies, many loads, or many soil models, it becomes a time consuming task which is not acceptable for a user-friendly prediction tool for railway induced ground vibration. Therefore, an approximate method for the computation of the wave field has been developed. The computation consists of several steps. At first, an approximate dispersion profile is calculated according to rules which have been derived from exact solutions. Secondly, the dispersion is used to achieve the amplitude for a certain frequency and a certain distance by calculating the approximate solution of a corresponding homogeneous half-space. Thirdly, three layer corrections are added which include lowfrequency near-field effects, high-frequency far-field effects, and a resonance amplification around the layer frequency. This procedure yields the wave field due to a point load. For a train load, many of these point-load responses have to be summed up, and a frequencydependant reduction factor has to be multiplied to incorporate the effect of the load distribution along and across the track. - The prediction method is applied to real sites, and the appropriate soil models are identified by approximating the measured transfer functions (frequency-dependant amplitudes) which is presented as an alternative to the approximation of the dispersion (frequency-dependant wave velocities). These examples demonstrate the general behavior of layered soils: the low amplitudes of the stiff half-space at low frequencies, the high amplitudes of the softer layer at high frequencies, the strong increase of amplitudes and a possible resonance amplification at mid frequencies. The material damping of the layer yields a strong attenuation of the amplitudes with the distance for high frequencies. The response depends strongly on the resonance or layer frequency which is shown for different layer depths and velocities always in good agreement with measurements. The layer frequency can be of immense influence if train-speed effects are analysed in a layered soil. The good agreement with many measurements in this contribution as well as in the references validates the prediction of ground vibration based on the theory of a layered half-space.

Train-induced ground vibration can be excited by wheel and track irregularities and by two kinds of irregularities of the soil, by geometric irregularities or by the spatially varying soil stiffness. For both types of irregularities, the effective track irregularity on top of the track is calculated in wavenumber domain and with wavenumber integrals. For a general multi-beam track model, the wavenumber integrals are solved numerically. The irregularities of the soil are filtered by the track when transferred from the bottom to the top of the track. The high-wavenumber irregularities are strongly reduced due to the bending stiffness of the track and the compliance of the support. In addition, soft track elements reduce directly the stiffness variation of the support. Therefore, the mitigation effect of elastic track elements for these excitation components seems to be important. For under-sleeper pads and slab tracks, calculation and measurements are presented including additional excitation components and the dynamic vehicle–track interaction, and the relevance of the excitation mechanisms is discussed based on the dynamic forces which are acting on the ground. Due to the restricted amplitudes, the parametric excitation by the stiffness variation seems to be less important than the geometric irregularities. The calculations yield the correct trends of the measurements and many details of the measured ballast, slab, and under-sleeper-pad tracks.

A survey of the phenomena and methods for floor vibrations is presented. Experimental results of floor vibrations are shown for many floors in six different buildings. The signals have been evaluated for waves and modes by simple procedures. General rules have been established between the material and the area of a specific floor, and its local eigenfrequency. The damping values of the floor vibrations have been found between D = 1 and 10 % where somewhat higher values have been measured for wooden floors, and a weak correlation with the eigenfrequency has been established. The velocities of bending waves propagating in a storey and the attenuation with distance in the building have been analysed. A considerable transfer of vibration from one room to far away parts of the building has been found in the studied buildings with concrete and wooden floors. An example building has been analysed for modes of coupled floor bays. The strong coupling of similar neighbouring floor bays would yield a wide band of global resonance frequencies. The measured wooden floor exhibits a weak coupling of the neighbouring floor bays and a narrower band of eigenfrequencies. A special method has been tested with the impulse measurements to estimate the coupled eigenmodes in presence of the high damping. From the ambient measurement, a low-frequency vibration mode has been detected which includes the vibration of the whole building and the soil. The coupling of floors to other floors and the whole building is an important phenomenon of structural dynamics which should be observed for the prediction of vibration due to internal and external sources.

Train-induced ground vibrations are generated by static and dynamic axle loads which can be calculated by vehicle-track-soil models and the vehicle and track irregularities. A fast prediction method has been developed which uses approximate transfer functions of layered soils. In the present contribution, this prediction method is used for the inverse calculation of the axle-load spectra from the measured ground vibration. The layered soils of some measuring sites show very differing ground vibration spectra in the amplitude range of 0.0001–1.0 mm/s as a consequence of the soft layer and stiff half-space, differing layer frequencies, as well as the far- and near-field measuring points. The back-calculation, however, yields axle-load spectra within a single order of magnitude around 1 kN. Axle-box measurements confirm the amplitude level of the axle loads. This standard axle-load spectrum can be used for a basic prediction at a new site. The separation of train and site-specific components allows a better evaluation of railway vibrations, for example, of different trains and different tracks. By eliminating the effects of differing soil characteristics, an important mid-frequency component has been found which lies between 8 and 32 Hz depending on the train speed. The origin of this dominant mid-frequency component is discussed using advanced prediction methods like moving constant loads, scattered axle impulses and axle-sequence spectra.

Ground vibrations due to different technical sources are analysed in theory and experiment for the dispersion of Rayleigh waves and the admittance spectra. Both tasks are theoretically based on the same concept: The admittance function in frequencywavenumber domain yields the dispersion as its maxima, and the admittance function in space domain is obtained by integrating it over the wavenumbers. On the experimental side, many signal processing methods have been applied to many sites and have been developed by the authors in the last 35 years, i.e., time-domain methods, including the cross-correlation method, and frequency-domain methods such as the spectral analysis of surface waves with two or multiple sensors, the wavenumber-transform method, and the spatial autocorrelation method. All methods are presented by their basic formula and by at least one example site. Different sensor arrays and deterministic and stochastic sources have been tested for the spatial autocorrelation method and the wavenumber-transform method at several sites. In addition, all frequency-domain methods are presented for a specific layered site comparing their quality. The evaluated dispersion curves are very similar, but a somewhat higher frequency range has been found for the fastest method, i.e., the multi-sensor spectral-analysis-of-surface-waves method. The theoretical solutions have been used for the inversion of the measured dispersion to the soil profile of the specific layered soil. The theoretical soil model has subsequently been used to predict the ground vibration spectra of hammer and railway excitation that exhibit a good agreement with the corresponding measurements. Thus, the contribution shows the benefit of active and passive seismic methods for the prediction of railway vibration, including a new version of the spatial autocorrelation method for technical vibrations. On the other hand, technical and namely railway vibrations are considered a seismic source for the exploration of near surface soils.

The soilstructure interaction of elastic plates on homogeneous or layered soils excited by horizontally propagating waves is analysed. Large plates are modelled by a combined finite-element boundary-element method (FEBEM), whereas the response of infinitely long plates is calculated by a numerical integration in the frequencywavenumber domain. The finite-element boundary-element method yields the complete soilplate transfer function of frequency and distance whereas the frequencywavenumber solution of the infinitely long plate can serve as an approximation for long distances on a finitely long plate. The soilplate transfer function starts to decrease strongly at the coincidence frequency, where the bending stiffness equals the plate inertia. A strong decrease follows at mid frequencies and a strong reduction of less than 0.1 of the ground vibration is reached at high frequencies. Rules for the characteristic frequencies are derived from the numerical results clearly indicating the strongest influence of the soil stiffness and the weaker influence of the bending stiffness of the plate. The influence of the mass, length and width of the plate are shown to be limited in case of realistic parameters, but it should be noted that the reduction effects are less effective for layered soils and for nearer observation points.

The attenuation of technically induced surface waves is studied theoretically and experimentally. In this paper, nineteen measurements of ground vibrations induced by eight different technical sources including road and rail traffic, vibratory and impulsive construction work or pile driving, explosions, hammer impulses and mass drops are described, and it is shown that the technically induced ground vibrations exhibit a power-law attenuation ν ~ r -q where the exponents q are in the range of 0.5 to 2.0 and depend on the source types. Comparisons performed demonstrate that the measured exponents are considerably higher than theoretically expected. Some potential effects on ground vibration attenuation are theoretically analyzed. The most important effect is due to the material or scattering damping. Each frequency component is attenuated exponentially as exp(-kr), but for a broad-band excitation, the sum of the exponential laws also yields a power law but with a high exponent. Additional effects are discussed, for example the dispersion of the Rayleigh wave due to soil layering, which yields an additional exponent of 0.5 in cases of impulsive loading.

Construction work, such as pile driving and soil compaction, or road and railway traffic excite nearby buildings, and the perceptible or audible vibration can be a nuisance for nearby inhabitants. A simplified building model has been created for these situations, which includes the effects of soil-structure interaction, the low-frequency amplification along the height of the building as well as the high-frequency reduction and the floor resonances. The model consists of one wall for all supporting structures (walls and columns) and one floor for each storey. The effect of different floor resonance frequencies is included in a stochastic procedure. The soil is modelled by a spring and a viscous damper, and the free-field amplitudes of the soil are applied under this soil element.
The model can be calculated by transfer matrices or in a continuous wave-type version where an analytical solution can be evaluated numerically. The building response in the high-frequency (acoustic) region is calculated as mean values over wider frequency bands. The approach to an infinite building model can be found for these high frequencies and the corresponding soil-structure transfer can be described by the ratio of impedances at foundation level.
The rules for choosing the parameters to obtain realistic results are derived from complex calculations for example, for the stiffness and damping of building foundations and many measurements as for the damping of floor resonances. The influences on the floor resonance from the soil (damping) and the supporting structure (detuning) are important. Some more effects will be discussed by the simplified and detailed models and by measurements to establish a good understanding of ground-induced building vibrations.

Excitation force spectra are necessary for a realistic prediction of railway-induced ground vibration. The excitation forces cause the ground vibration and they are themselves a result of irregularities passed by the train. The methods of the related analyses - the wavenumber integration for the wave propagation in homogeneous or layered soils, the combined finite-element boundary-element method for the vehicle-track-soil interaction - have already been presented and are the base for the advanced topic of this contribution. This contribution determines excitation force spectra of railway traffic by two completely different methods. The forward analysis starts with vehicle, track and soil irregularities, which are taken from literature and axle-box measurements, calculates the vehicle-track interaction and gets theoretical force spectra as the result. The second method is a backward analysis from the measured ground vibration of railway traffic. A calculated or measured transfer function of the soil is used to determine the excitation force spectrum of the train. A number of measurements of different soils and different trains with different speeds are analysed in that way. Forward and backward analysis yield the same approximate force spectra with values around 1 kN for each axle and third of octave.

Long wooden floor beams above a ball room in an old historical palace have been analysed experimentally. The eleven beams are weakly coupled by three layers of floor boards. It has been investigated if the state (the stiffness) of the wooden beams can be determined by vibration measurements of global or preferably local modes. Hammer, heel-drop and ambient excitations have been used. The vibration modes of the structure show dominating local deformations if an impact excitation is applied. This is understood as the positive superposition of several modes which yield the maximum at the excitation point but a cancellation at more distant points. Natural modes have been estimated from these vibration modes by standard and special methods which were necessary for the high damping of the wooden floor. It has been found that all floor beams contribute to each natural mode even for a weak coupling of the beams. In addition to the modal discussion, the impact tests have also been analysed for the wave propagation and amplitude attenuation with distance. The coupling of floor beams has been studied theoretically by an analytic multiple-beam model where the coupling by translational or rotational springs and by a common support motion has been assumed.

Train passages induce static and dynamic forces on the track, the train-induced vibrations propagate through the soil and excite neighbouring buildings. The problem of train vibrations is divided into the parts emission, which is the excitation by railway traffic (the present contribution), transmission, which is the wave propagation through the soil, and immission, which is the transfer into a building, - The calculation of the axle loads are based on the vehicle-track-soil interaction. This interaction uses the dynamic stiffness of the vehicle (the inertia of the wheelset) and the dynamic stiffness of the track-soil System. Based on various time consuming finite-element boundary-element calculations, an approximate track-soil model has been established. The vehicle-track-soil analysis yields several transfer functions between the various geometric or stiffness irregularities and the axle loads of the train. Geometric irregularities of the vehicle (the wheels) and the track (rail surface and track alignment) are the simplest components. Geometric irregularities of the subsoil (trackbed irregularities) have to be transferred to effective irregularities at rail level. The bending stiffness of the track is filtering out the short-wavelength contribution. Stiffness irregularities occur due to random variations in the bailast or the subsoil, which must also be transferred to effective track irregularities, and due to the discrete rail support on sleepers. The axle loads due to the effective track errors from stiffness variations have their specific vehicle-track transfer function. - All necessary formula for the prediction of axle-load spectra will be presented. The prediction method is compared with axle-box measurements at a Standard ballasted track. Moreover, ground Vibration measurements at numerous sites are exploited for the axle-load spectra and the Validation of the prediction method.

Vibration of normal apartment, office and production buildings, which are excited by technically induced ground vibrations are considered. Many wavelengths of the Rayleigh waves of the soil fit into the foundation dimensions. The related high discretization effort can nowadays be realized with detailed soil-structure interaction method. The combined finite-element boundary-element method is used here as a detatiled method. Simplified method can be used with less computation time, but these methods must be calibrated by exact results. One simplification is to extent the structure to infinity and to solve the problem by wavenumber domain methods. Another simplification is the use of a Winkler soil instead of the continuous soil. Usually, the Winkler parameters are not only soil parameters but depend also on the rigid or flexible foundation structure. Substructure methods use commercial FEM software for the building part. The contribution will show some detailed and some simplified results on large structural elements such as foundation plates, walls, storey plates on columns as well as results on complete buildings. The reduction of the ground vibration by stiff elements and the amplification due to floor or building resonances are discussed which are the most important phenomena of the soil-building interaction.

A combined finite element boundary element method has been developed to calculate the dynamic interaction of the railway track and the underlying soil. The track-soil results are coupled with a simple vehicle model to establish the force transfer function of the vehicle-track-soil system. Mitigation measures at the track, namely three different tracks with under-sleeper pads, are analysed. The un-sprung vehicle and heavy track masses on soft track elements yield a lower vehicle-tracksoil resonant frequency and a better reduction of the excitation forces at higher frequencies. If the effectiveness of the mitigation is measured as a vibration ratio between the isolated and an un-isolated reference track, the vehicle-track-soil resonance of the reference track has an important influence on the mitigation effectiveness. Therefore, different un-isolated reference tracks are analysed. It is shown how the frequency and amplitude of the vehicle-track resonance are influenced by the stiffness of the rail pads (strongest), the ballast (medium) and the soil (weakest). The reference track models have been compared with shaker tests on two railway tracks with strong resonances. The damage detection and repair control have become important tasks for ballast and slab tracks. Measurements which compare the damaged and the repaired status of the same track section at different times, or which compare a damaged and an intact track section at the same time, have been successfully performed at some sites in Germany and compared with the theoretical behaviour of intact and damaged tracks.

The ground vibrations, which are generated by trains on different tracks, have been calculated by finite-element boundary-element models. The ballasted track is modelled in detail by the finite element method. The infinite soil is modelled by the boundary element method as a homogeneous or layered half-space. The track-soil system is coupled to a simple rigid mass model of the vehicle so that the vehicle-track interaction is completely included. Transfer functions are calculated in frequency domain without and with vehicle-track interaction, the compliance of the track and the mobilities of the soil at different distances from the track. Finally, the ratios between the ground vibration amplitudes with and without mitigation measures are calculated to quantify the effectiveness of the mitigation measures.
Tracks with under-sleeper pads have been investigated in a wide parameter study for the RIVAS project. The main parameters that influence the reduction of ground vibration are the stiffness of the under-sleeper pad, the mass and the width of the sleeper. The softest sleeper pad yields the best reduction of the ground vibration. The influence of the sleeper mass is not so strong, as the characteristic frequency is ruled by the mass of the sleeper and the mass of the wheelset as well.

The layered soil is calculated in the frequency wavenumber domain and the solutions for fixed or moving point or track loads follow as wavenumber integrals. The resulting point load solutions can be approximated by simple formula. Measurements yield the specific soil parameters for the theoretical or approximate solutions, but they can also directly provide the point-load solution (the transfer function of that site). A prediction method for the train-induced ground vibration has been developed, based on one of these site-specific transfer functions. The ground vibrations strongly depend on the regular and irregular inhomogeneity of the soil. The regular layering of the soil yields a cut-on and a resonance phenomenon, while the irregular inhomogeneity seems to be important for high-speed trains. The attenuations with the distance of the ground vibration, due to point-like excitations such as vibrator, hammer, or train-track excitations, were investigated and compared. All theoretical results were compared with measurements at conventional and high-speed railway lines, validating the approximate prediction method.

Vehicle, track and ground vibration as well as their interaction are considered in a combined finite-element boundary-element (FEBEM) approach. The layered soil is calculated in frequency wavenumber domain and the solution for fixed or moving point or track loads follow as wavenumber integrals. The soil results from the measurements and the detailed models are approximated by simple formula which are used for the prediction of train-induced ground vibration. The influence of the track and the soil on the train induced ground vibration is analysed by the detailed models. The ground vibrations strongly depend on the regular and random inhomogeneity of the soil. The regular layering of the soil yields a cut-on and resonance phenomenon while the random inhomogeneity yields a scattering of the axle impulses which proved to be important for high-speed trains. The attenuation with distance of the ground vibration due to the point-like excitations such as vibrator or hammer excitations and the train-track excitation are investigated and compared. All theoretical results are compared with measurements at conventional and high-speed railway lines.

The dynamics of slab tracks and floating slab tracks are analyzed by multibeam models for the track and by integration in the wave-number domain for the soil, which is modeled as a layered half-space. Frequency-dependent compliances and force transfers are calculated for a great variety of track and soil parameters. The distribution of the load and the displacements along the track is investigated as well as the wave propagation perpendicular to the track and the ground vibration amplitudes. The floating slab track has a dominating plate-mat resonance and a strong high-frequency reduction. A track-soil resonance can also be recognized for an unisolated slab track in the case of layered soils. Generally, there is a strong damping of the track by the soil. The reduction effect of the slab mat is mainly owing to the elimination of this strong damping. The continuous soil yields slightly different rules for the displacements and force densities than those of a Winkler support. The total force transfer from the rail to the soil is the best criterion to judge the effectiveness of a floating slab track in reducing the ground vibration at some distance from the railway line. The total force transfer is easier to calculate than the double Fourier integrals of the ground vibration amplitudes, namely in the far field, and it has the best correlation with the reduction of the ground vibration.

The propagation of waves through homogeneous or layered soil is calculated based on half-space theory. The moving dynamic loads of a train are approximated by fixed dynamic loads and the wave field can be calculated if the spectrum of the dynamic train loads is known. In addition to this dynamic wave field, there are three different components at three different frequency ranges which are caused by the passage of the static loads:
the regular static component at low frequencies,
the irregular static component at medium frequencies,
the sleeper-passing component at high frequencies.
For each of these components, an approximate solution is presented. The calculated wave field is compared with measurements of different trains at different sites. The measurement of impulse and harmonic point load excitation verifies the soil dynamic base of the method.

Seismic response of three-dimensional structures using a green's function approach to the soil
(1986)

An integral method to calculate the solution of a homogeneous or layered soil due to a harmonic point load is described. An infinite plate at the surface of the soil can be introduced in this integration in wavenumber domain, too. Finite structures on the soil are calculated by a combined finite element and boundary element method, which makes use of the point load solution of the soil. The compliance functions for a vertical point load and some vibration modes are calculated for realistic parameters of the plate and the soil and for a wide range of frequencies. The influence of the stiffness of the soil and the foundation is investigated, showing that the soil mainly affects the low-frequent response whereas the structural properties are more important at higher frequencies. A rigid approximation of flexible plates is only found at low frequencies, if the elastic length is used as the radius of a rigid disk. At higher frequencies, a characteristic behaviour of the flexible plate of approximately is observed, what is in clear contrast to the compliance of rigid foundations. A plate on a visco-elastic support (Winkler soil) shows similar displacements as a plate on a homogeneous half-space, but the maximal stresses between the plate and the soil are considerably smaller which is found to be more realistic for a plate on a layered soil. For practical applications, a normalized diagram and some explicit formulas of the exact and the approximate solutions of an infinite plate on a homogeneous half-space are given, which is a useful model to represent the soil-structure interaction of flexible foundations.