ALONG with absorbing residual vibrations and dynamic impact from running trains, track subgrade is exposed to seasonal soil variations from frost/heave and thawing/settlement cycles. It is therefore the most vulnerable part of the track which must be protected and strengthened wherever possible.


Russian Railways (RZD) currently uses a number of railway subgrade bearing capacity improvement methods. These include installing a cover layer over the sub-ballast zone using organic binders, and installing sub-ballast mats and expanded polystyrene plates under the main platform which also serve as insulators.

RROne of the most effective methods is the installation of a protective sub-ballast layer (PSL), which offers improved durability and prevents subgrade deformation. However, the main disadvantage of PSLs, particularly for live railway lines, is the need to excavate the existing railway embankment to a depth of up to 1m, which is a disruptive, expensive and lengthy process as it requires a long-term suspension of traffic.

One way to reduce the required excavation depth is to combine an PSL with one or several geosynthetic interlayers (geogrids) depending on the strength parameters of the underlying soils. Using geogrids helps to reduce the overall roadbed and track maintenance costs by increasing the line capacity and axleloads.

From 2007-2010 a comparative research study was undertaken for RZD of track sections using PSL and geosynthetic reinforced/stabilised sub-ballast (with no PSL) on the St Petersburg - Moscow October Railway and at the Shcherbinka test centre in Moscow.

The study commenced by obtaining data on the current condition of the ballast layer and the subgrade. The compaction level was measured with a "water balloon" densitometer, static plate loading test and dynamic plate loading test, with the plate diameter set at 300mm and for ballast testing, the thickness of stone below the plate was set at 200mm. No PSLs were installed under the ballast but instead, single layers of various geosynthetic materials, including geogrid, geotextile and expanded polystyrene, were inserted below the ballast at various locations.

Russia fig 1The deformation properties of the ballast layer and the subgrade were measured using a plate loading test in accordance with a method described in the Russian National Standard GOST 20276-99 and Norms ICS 46-83, which determines the deformation modulus E0 and elasticity modulus Eelas under applied pressure values of up to 0.2MPa (2kgf/cm2) below the plate. Figure 1 shows a comparative diagram of the resulting deformation modulus E0 for the tested structure.

In 2008 a study was carried out at a site on the second track of the St Petersburg - Moscow line before and after reconstruction. The upgraded track included a protective sub-ballast layer installation, stabilised with a bi-axial composite geogrid SS30G interlayer supplied by Tensar. The protective layer comprised a crushed-stone-sand-gravel aggregate mix with a grain size of up to 40mm with its composition meeting local requirements for:

  • compaction
  • vibration stability
  • frost heave resistance, and
  • no suffusion of the protective layer aggregate with the crushed-stone grains above and the soil particles beneath.

Russia fig 2As was the case in 2007, the measurements of sub-ballast and subgrade deformation properties were carried-out using a plate loading test in accordance with the procedure described by GOST 20276-99 and IBS 46-83. The results of these tests are shown in Figure 2.

In 2009 the test method for subgrade and protective sub-ballast layers on the Moscow - St Petersburg line was updated and endorsed as an official application used by RZD engineers. The updated method is based on two loading stages and is in accordance with the German standard DIN 18134. The German Standard is very close to its Russian counterpart as it uses Ev1 modulus ("modulus of primary load"), which is in fact similar to E0, and Ev2 modulus ("modulus of secondary load"), which is similar to Eelas modulus. In accordance with the new method results were processed according to DIN 18134 as well as per the GOST standard.

In 2009, a year after installation, new studies in accordance with the updated method commenced at the reconstructed section of the St Petersburg - Moscow line. The results of these studies revealed:

  • the value of the measured primary load deformation modulus E0 was in the range of 48-103MPa
  • value of secondary load deformation modulus Ev2 was in the range of 113-211MN/m2
  • the ratio of Ev2/Ev1 on average was 1.9-2.3, which means good compaction quality of the SPL layer, and
  • the deformation properties of the sub-ballast layer had not deteriorated after one year of service life.

In 2010 the study was conducted on two further experimental track sites: on the Shcherbinka test loop and on the 2km Sablino - Tosno section of the October Railway's second running track. The tests were conducted in two stages: before and after reconstruction of the line, with the main purpose of the test to check and compare the effectiveness of the installation of a Tensar TriAx 170G hexagonal geocomposite geogrid, with a Neoweb geocell, within the sub-ballast protective layer.

Russia fig 3During the reconstruction of the test loop at Shcherbinka station, the stone ballast was removed and the subgrade was tested with a first plate loading test. A geocomposite Tensar TriAx 170G geogrid was subsequently installed within a sand and gravel sub-ballast protective layer, with a second plate loading test taking place on the surface of this layer. Figure 3 shows data from these tests, with an improvement in deformation properties below the ballast after reconstruction of the test loop.

The tests were conducted in two stages during the reconstruction of the test site on the Sablino - Tosno section. During the first stage, loading was applied at the bottom of exploratory holes on the subgrade surface and the second stage comprised testing the finished surface of the stabilised PSL. Testing was carried out in both locations, with one stabilised using a geocomposite geogrid Tensar TriAx 170G and the other with Neoweb geocells from PRS. The comparative test results for the section with the hexagonal Tensar TriAx 170G geogrid and the section with a geocell at the base of PSL are shown in Figure 4.

Figure 4 shows that the tensile stress-strain characteristics improved by installing both Tensar TriAx 170G and geocell Neoweb. The results are based on plate loading tests and show that Tensar TriAx 170G outperforms Neoweb geocell on each of the three sections.

From the results of the studies investigating the effectiveness of different geosynthetic materials installed on the subgrade surface, we can derive a nominal improvement factor for each method compared with the control section which does not use geomaterial and is equal to 1.

Russia fig 4Let us assign to the unstabilised section a factor of 1. From the figures in the table we can conclude that with a performance improvement factor lower than 1, the polystyrene addition has a negative effect. Indeed the table shows that the preferred solutions are the bi-axial, with an improvement factor
2.8-2.9 and three axial, with an improvement of 3.3-3.6, composites. However, the three axial clearly shows the highest efficiency.

The "improvement factor" values indicated herein are elementary and are based on a number of factual experiments. They are a good indication of the improvement that can be achieved but also indicate the wide variation in improvement depending on the method used. This table does not depict final or ultimate results because tests on are still in progress. Further work is needed for different types of subgrade improvement using various methods and materials and tested repeatedly under different climatic conditions, different loads and for a wide range of soil conditions.

Subsequent work includes research conducted in 2013-2014 by the Moscow State University of Railway Engineering, the St Petersburg State Transport University and the research and production company, Stroy-Dynamics. Their experiment included a full-scale test of the Tensar TriAx geogrid application to measure the efficiency of the railway subgrade's performance by conducting multicycle loading to simulate rolling stock vibration.

Based on the results of their tests, it was concluded that a significant performance improvement is possible by applying Tensar TriAx geogrids to stabilise and reinforce the track subgrade. The tensile stress-strain characteristics of the stabilised track sections are 2-2.5 times higher, than the non-stabilised sections and are essentially isotropic. The settlement of the track panel at the impact forces is also 2.5-3 times lower in the sections with Tensar TriAx application.