• No results found

Ballast Selection Criteria



2.2.3 Ballast Selection Criteria

Chapter 2: Critical review of granular media with special reference to railway ballast

with regard to different aspects of track condition.

Table 1. Selection criteria for railway ballast (after Selig and Waters, 1994) Selection

categories Selection criteria Test type Remarks/Purpose Ballast

source evaluation

1. Identification &


• Petrographic analysis

• Chemical analysis

• X-ray diffraction

Provide basic qualitative estimates of physical and chemical properties affecting performance.

2. Durability

• Los Angeles abrasion

• Mill abrasion

• Deval abrasion

• Crushing value

• Impact

Quantitative estimate of:



wearing by attrition;

resistance to crushing under static load;

resistance to sudden shock loading.

3. Weathering/En-vironmental resistance

• Cement value

• Permeability

• Freeze-thaw resistance

• Sulfate/Magnesium soundness

• Absorption

Quantitative measure of:

ability of material to remain free-draining and elastic;

drainage capacity;

disintegration in freezing conditions.

4. Stability

• Specific gravity

• Bulk unit weight

Quantitative estimate of ability to anchor ties against lateral and uplifting forces.

Ballast production

5. Shape and surface characteristics

• Flakiness

• Elongation

• Sphericity

• Angularity/roundness

• Fractured particles

• Surface texture

Quantitative estimate of resistance to instability under high and/or low frequency vibrations.

6. Gradation • Size

• Grain size distribution

• Fine particle content

Indication on the behaviour under high and/or low frequency vibrations.

7. Impurities • Clay lumps and friable particles

Estimate of the rate of degradation under loading.

Chapter 2: Critical review of granular media with special reference to railway ballast

Extensive research was carried out by Raymond et al. (1975-1979) in an attempt to predict field performance, as a means to measure and control material quality. The pertinent particle characteristics that affect the quality of ballast material were identified, and they are concisely reviewed in the following sections.

a. Identification and Composition

The engineering characteristics of a ballast type are, to a large degree, a direct function of the inherent physical properties of the parent rocks and/or minerals contained therein.

The purpose of the hand petrographic examination, and the subsequent chemical analysis and X-ray diffraction, is to determine the physical and chemical properties of the material, and to describe, classify and determine the relative amount of the constituents of the sample, as these have a bearing on the quality of the material.

Watters et al. (1987) presented examples of the relationships between petrographic results and rock particle properties. They also summarized characteristics identifiable by petrographic analysis, which, if present in abundance in a rock mass, may represent

‘fatal flaws’ and cause rejection of the rock for use as ballast (Table 2).

Raymond et al. (1976, 1979) and Greene (1990) examined ballast materials provided from various quarries in North America, and arrived at the conclusion that, in general, fine grained, hard-mineral and unweathered aggregates are best as ballast materials.

Therefore, the extrusive igneous rocks (rhyolite, andesite and basalt) form a primary source group, followed by the coarser-grained igneous rocks such as granite, diorite and gabbro, along with the hard-mineral grained and well-cemented sedimentary or metamorphic rocks such as quartzite (Raymond et al., 1979; Meeker, 1990).

Table 2. Petrographic properties that may represent ‘fatal flaws’

(Watters et al., 1987)

Properties of Ballast Source Rock Principal Deleterious Effect MINERALOGICAL

• General high content of very soft minerals (eg. clay, mica, chlorite)

• Argillaceous sedimentary rocks (eg. mudstone, shale)

• Mica-rich metamorphic rocks (eg.

slate, phyllite, schist)

• Igneous with deuterically altered feldspar

• Sulfide-rich (> 2 – 3%) (eg. pyrite, pyrhotite)

Rapid physical degradation, clay and fine, mica-rich fines;

Rapid physical degradation, clay-rich fines;

Rapid physical degradation, clay and fines, mica-rich fines;

Rapid physical degradation, clay-rich fines;

Oxidation of sulfide results in acidic conditions promoting chemical weathering of other mineral components.


• Poor consolidation (in sedimentary and volcaniclastic rocks)

• High porosity (> 5%) in sedimentary rocks

• Vesicularity (in volcanic rocks)

• Friable texture in crystalline rocks

Rapid physical degradation by abrasion;

susceptibility to freeze-thaw;

Degradation by freeze-thaw and abrasion if pores are large and abundant;

Degradation by freeze-thaw and abrasion;

Rapid physical degradation by abrasion;

susceptibility to freeze-thaw.


• Closely spaced joints, bedding partings, foliation

Rapid physical degradation by abrasion and to freeze-thaw; generation of unsuitable particle shapes.


• Smooth particle surfaces (often due to rock texture)

• Unsuitable particle shape

Poor mechanical stability;

Poor mechanical stability; load fracture of elongated or tubular particles.

Chapter 2: Critical review of granular media with special reference to railway ballast

b. Aggregates Durability and Strength Tests

Durability tests are designed to simulate load-induced degradation of ballast. The process of degradation due to wheel loading is known to occur in two modes: (a) particle fracture; (b) interparticle grinding or attrition. It was established that the dominant degradation mode existing at any particular point in time depends on the applied load, level of confinement, degree of compaction, particle size and particle shape (Schultze and Coesfeld, 1961; Raymond et al., 1985; Selig and Alva-Hurtado, 1981; Feda, 1982; Shenton, 1985). It is expected that granular materials having weak grains would exhibit higher degree of degradation, hence, larger compressibility and lower shear strength.

At present, there are three attrition tests commonly used by the railways authorities worldwide. Furthermore, in order to quantify the toughness of ballast particles, their resistance to crushing is measured by tests such as Aggregates Crushing Value and Wet/Dry Strength Variation. It was suggested that a better estimate of in-track performance would be obtained from impact tests used by European railway authorities and somehow simulated by the Los Angeles attrition test. Nevertheless, higher values of attrition would indicate poor in track performance (eg rapid degradation).

c. Stability

The specific gravity, together with the shape and texture of the grains, are the major characteristics of an aggregate commonly associated with its ability to control track performance. The shape and texture are features that could be, to some extent, controlled by the production process although they represent intrinsic characteristics of the source rock, therefore, they are discussed in the next section.

Raymond et al. (1979) showed that the holding capacity of ballast, which relies on both its shear strength and bulk unit weight, followed an increasing linear function of density for a given ballast. The specific gravity (SG) of an aggregate provides a reasonable measure of the relative ability of the material to provide a high ballast density (for a given grading and level of compaction), hence, the higher the SG value, the greater the holding capacity of the ballast and the lower the degradation (Raymond et al., 1978).

The ability to provide lateral stability is a governing requirement for curved sections of tracks, especially when they are expected to carry rolling stock with heavy axle loads (Raymond, 1985). Furthermore, Raymond et al. (1983) reported that ballast having a higher SG effectively had a higher capacity to damp out low-frequency vibrations (believed to be the cause of occurrence of the differential settlement). Therefore, he recommended the use of ballast having high SG for tracks carrying high-speed passenger traffic. Ballast Production

Raymond (1985) stressed that there is a definite distinction between ballast production and source evaluation due to the ability to control certain parameters during ballast production. The particle shape, grains surface/texture, gradation and purity of the aggregates are the feature varied during production of ballast.

a. Particle Shape and Surface Characteristics

The shape and the texture of the grains are mainly dictated by the nature of the deposit itself and during the production process only a limited degree of control is achievable.

Field observations have showed that materials that fracture into relatively equi-dimensional, angular fragments provide high stability to the ballast layer, i.e. higher

Chapter 2: Critical review of granular media with special reference to railway ballast

lateral resistance to the lateral ballast flow, hence, lower settlement. The highest attention during the ballast production should be given to the content of flaky particles as research reported by ORE (1970) found that lower elastic moduli were associated with the ballast specimens containing flat particles. However, a limited quantity of flaky particle added to the ballast specimens resulted in increased shear strength during triaxial tests (Eerola, 1970; Dunn and Bora, 1972; Gur et al., 1978; Siller, 1980) and a lower rate of settlement accumulation and a reduced overall settlement during semi-confined compression tests (Jeffs and Marich, 1987; Jeffs and Martin, 1994).

Nevertheless, Gur et al. (1978) reported that addition of flaky grains caused increased degradation during compaction and higher level of deformation during tests.

It has been established that independent of grain size, materials having angular grains were associated with higher shear strength viz. materials having subrounded to rounded particles, (Holtz and Gibbs, 1956; Holubec and D’Appolonia, 1973; Norris, 1977; Thom and Brown, 1989). However, higher grains angularity was associated with an increase of the failure strain and a decrease in the ballast stiffness (Holubec and D’Appolonia, 1973). The effect of grains roughness on railway ballast behaviour is further discussed in a later section.

b. Gradation and Grains Size

The selection of the particle size distribution of ballast layer has a marked effect on its in-situ performance, and also on the economic evaluation of track design. Particle size distribution is commonly represented by a cumulative frequency distribution plot as percent of material passing a given set of sieves versus logarithm of the particle diameter.

Current thinking is that a narrow gradation would best meet the requirements for railway ballast. However, there is a disagreement within the railway system with regard to the maximum and minimum particle size that would offer the best performance for a ballast material, i.e. lower deformation and degradation of ballast layer. For example, the European and Australian railway systems recommend larger size ballast with a maximum particle size of 63 mm and a minimum particle size of 13.2 mm (Fig. 9.a).

On the other hand, most North American railway systems prefer a smaller size ballast with a particle size distribution (PSD) varying from 4.76 mm up to a maximum of 51 mm (Fig. 9.b). The lower boundary of gradation is intended to control the permeability (drainage capacity) of a given material (Lambe and Whitman, 1969).

In contrast to the uniform gradation specified for railway ballast worldwide, several researchers showed that within the provided PSD boundaries, independent of type of ballast, a broader gradation is associated with reduced deformation and degradation (Klugar, 1978; Roenfeld, 1980). Further discussion on the effect of PSD on the deformation and degradation of granular media viz. railway ballast is to be presented in a later section.

The effect of particle size, however, appears to be inconclusive. Holtz and Gibbs (1956) and Vallerga et al. (1957) found that particle size had little effect on shear strength. Marachi et al. (1972) showed that the strength increases as the particle size decreases, whereas Dunn and Bora (1972) reported that the shear strength increases with increasing grain size.

Chapter 2: Critical review of granular media with special reference to railway ballast

10 100

Particle size, (mm)

0 20 40 60 80 100

% P ass ing

RIC Specifications SNCF Specifications BR Specifications CFR Specifications


1 10 100

Particle size, (mm)

0 20 40 60 80 100

% Pass ing

AREA (coarser) Specifications AREA (finer) Specifications CNR Specifications

CP Specifications


Figure 9. Main line particle size distribution boundaries adopted by some railway systems: (a) Europe and Australia; (b) North America (after SNCF, 1979; Gray,

1983; TS 3402-83; STAS 3197/1-84; Chrismer, 1985; Raymond et al., 1987)