RAILROAD BALLAST SPECIFICATION AND EVALUATION
by Gerald P. Raymond
PREFACE Intrusive igneous rocks are formed when the magma cools within the
Ballast research financed by CN Rail, CP Rail and Transport Canada Earth. As such it cools slowly allowing coarser grained rocks to form. In
Research and Development Centre under Project 2.22 of the Canadian general, the closer the intrusion comes to the Earth's surface, or the smaller
Institute of Guided Ground Transport (CIGGT), and performed under the the size of the intrusion, the more rapidly it will cool, and the finer the
direction of the writer at Queen's University along with field related grain size of the minerals will be.
studies by CP Rail has led to the adoption of a new CP Rail ballast
specification. Some of the background information and reasoning are Sedimentary Rocks
presented herein. Under normally weathering processes, all rocks slowly disintegrate to
form clay, silt, sand and gravel, plus dissolved materials, which are
INTRODUCTION eventually deposited. One process of sedimentary activity is illustrated in
The selection of the top ballast (hereafter referred to as ballast) used for Figure 5. Over time, the unconsolidated deposits shown in Figure 5 may
railway track support is of major importance in establishing and become compacted and cemented together to form Clastic sedimentary
maintaining the characteristics of the track response and, consequently, the rocks, while the dissolved materials may precipitate to form Chemical
riding quality. For ballasted track, an elastic, non-cemented, stable and sedimentary rocks.
weather resistant ballast bed, well laid and compacted on a stable, compact
subballast and subgrade, is the first condition for low maintenance Metamorphic Rocks
expenditures. Rocks formed under one set of temperature, pressure and chemical
Ballast must be capable of withstanding many forces. Factors that conditions and then exposed to a different set of conditions, such as
cause deterioration of the ballast include extremely large cyclic loadings, illustrated in Figure 6, may undergo structural and chemical changes
vibrations of varying frequencies and intensities, repeated wetting and without melting that produce rocks with different textures and new
drying involving crystallization of rain dissolved soluble salts, and minerals. Typically, this process, which is known as metamorphism,
freezing and thawing in cold climates. Ballast must also be easy to handle results in the linear orientation of minerals along well-defined planes of
during maintenance. These requirements are invariably conflicting, weakness. These planes of weakness are detrimental to the choice of
requiring considerable judgement in aggregate selection for railway metamorphic rocks as ballast.
ballast. Herein are outlined some of the different requirements that should
be clearly understood in making proper economic selection from available Comments
aggregate sources of which the most common source is rock material. Although not an absolute guide, the rock formation processes outlined
above provide a useful rough-screening criteria for ballast applications.
ROCK MATERIAL In general, fine-grained igneous rocks are preferable to either sedimentary
Rock material consists of an intergrowth (or interbonding) of one or or metamorphic rocks. Medium to coarse grained igneous rocks and hard
more minerals. These minerals are chemical compounds and have both a mineral well-cemented sedimentary rocks are still preferable to most
specific crystal structure (or arrangement of atoms) and a specific metamorphic materials.
chemical composition. Note that two or more different minerals may have
the same chemical composition but will all have different crystal AGGREGATE SELECTION
structures, The way in which the minerals of the rock are intergrown is General
called the texture of the rock. The increasing cost of track and roadbed maintenance has made the
Rock names are based on two things. These are the minerals that selection of an appropriate aggregate for each ballast application a matter
constitute the rock and the texture of the rock. Thus two rocks of identical of considerable financial importance. It is clearly cost ineffective to haul
mineral composition having very different textures would have different a first class ballast long distances to surface a little used branch line, and
names. Mineral identification is generally based on simple tests which equally inappropriate to employ an inferior ballast material on a mainline
involve Hardness, Cleavage, Luster, Streak colour and Chemical track subject to a high density of heavy, fast traffic. The aggregate
composition. From a ballast performance viewpoint, mineral hardness, selection procedure must permit the decision maker to identify the
generally based on Mohs' hardness scale given in Figure 1, is of physical characteristics of a ballast so as to assess the differential aspects
considerable importance. Cleavage and fractures shown in Figure 2 may of the material with respect to other available materials with similar
also be important. properties and to evaluate, in financial terms, the expected costs and
Particular geological processes give rise to three rock types, Igneous, benefits arising from the use of each ballast.
Sedimentary and Metamorphic. Rock specimens may be used to classify To perform well in track, the aggregate for ballast must be tough
the rock type and also give information about the geological history of the enough to resist breakdown through fracturing under impact, and must be
area it is obtained from. This information is valuable to the ballast hard enough to resist attrition through wear at the ballast particle contacts.
selection process. It must be dense enough so that it will have sufficient mass to resist lateral
forces and anchor the ties in place. The aggregate must be resistant to
Igneous Rocks weathering action so that weakening of the ballast doesn't occur from
Igneous rocks are formed from a cooling magma (a very hot, molten crystallization or acidity of impurities dissolved in rainwater or from daily
liquid of silicates and other compounds) from the type of activity or seasonal fluctuations in temperature or other weathering processes. The
illustrated in Figure 3. The rate at which the magma cools determines the aggregate must be free of secondary minerals which may weather
texture of the igneous rock formed. The composition of the magma detrimentally such as pyrites that oxidize to give ferric compounds and
determines mineralogical constituents of the rocks. These two properties, thence acids which are highly corrosive to the metallic parts of the track
texture and composition, are the basis for the classification of igneous structure.
rocks of which Figure 4 is a simplified illustration. They provide a basis It must also be resistant to the chemical degradation resulting from the
for their identification. Note that on Figure 4 two rocks having the same action of rainwater on foreign source fines. For example trace elements
composition may have different names, depending on their grain size such as sulphur in coal are highly likely to increase the acidity of any
(texture). moisture trapped within the ballast. This acidity will cause solution
Extrusive igneous rocks are formed when the magma is poured out weathering of the aggregate particularly limestones.
onto the Earth's surface. Extrusive magma solidifies rapidly to form a All aggregate material may be expected to degrade to some extent with
glassy rock or an extremely fine-grained rock. time. Because of this fact it is strongly recommended that aggregates be
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 1
subjected to some form of petrographic examination in order to assess the to the environmental pollution of the locale where the aggregate is
long term effect of the degradation and production of ballast origin fines to be used.
on the ability of the ballast to remain free draining and elastic. Where (g) Properties of degraded fines; including shape, probable effect on
aggregate is composed of fine grained minerals whose identification is permeability and their susceptibility to solution and precipitation
difficult to identify in a hand held examination then petrographic thin weathering including cementability.
section analysis is strongly recommended. Thin section examination as (h) Additional engineering tests not in specification; suggestions with
well as giving a conclusive determination of rock type, mineralogy and explanations if made.
structure, also establishes whether microfractures exist within the (i) Estimation of engineering test results; explanation in terms of
aggregate source and whether former microfractures have been weakly petrological features.
cemented with secondary minerals that might weather and soften quickly. (j) Recommendations and summary; report important petrological
Origin of Aggregate
Quarried stone ballast should be obtained from competent strata of Engineering Assessment of Weathering
reasonable thickness. The extent of the rock deposit should be sufficient A number of tests are available to assess the potential of aggregates to
for economic ballast production. A large variety of rock types are used as degradation or weakening due to weathering. The two tests commonly
ballast. In general the fine hard mineral grained unweathered aggregates specified are;
make the best ballast. These include igneous rock types such as rhyolite, (a) Soundness Testing
andesite and basalt. Second best are the coarser grained igneous rocks Soundness testing where rock particles are alternatively immersed and
such as granite, diorite and gabbro, along with the hard mineral grained then dried using a salt solution. The AREA specification calls for a
well-cemented sedimentary rock and hard mineral grained metamorphic solution of Sodium Sulphate although the writer has found Magnesium
(or transformed) rock such as quartzite. Less satisfactory but often more Sulphate more appropriate and this has been adopted in the CP Rail
commonly used because of their cheaper production cost and wider Specification. The test is considered by many to be only applicable to
availability are the sedimentary rock types such as limestone, dolomite, assessing the expansive pressures generated from freezing water, however
sandstone and siltstone. it also gives an indirect assessment of the resistance of the aggregate to
Rock types such as shale and slate which result in flaky or elongated crystal growth from dissolved pollutants in rainwater through the use of
particles, should not be permitted since these shaped particles do not result a standard salt solution (e.g. Magnesium Sulphate). For example the
in good interlocking particularly when subjected to vibrations. Similarly soundness test is the main test used in the United Kingdom to assess the
sedimentary and metamorphic rock types that contain visible quantities of resistance of building stone to the crystallisation of soluble salts contained
secondary minerals which weather quickly should be rejected. in rainwater that penetrates within the pores of the stone (Leary 1983). In
Where cobble and pebble size gravel is to be used to produce broken this regard building stones are observed to degrade on their surfaces
stone it is recommended that cobble size (plus 75 mm) particles should depending on their degree of exposure to rainwater, the extent of
first be sorted by the use of a coarse bar mesh. The material retained is atmospheric pollution and their climatic exposure (inland or coastal and
then used as the source for crushing. frost or no frost locations). Note the test has been found applicable in
Where slag material is permitted, air cooling is generally required. No non-frost zones of the United Kingdom subject to different degrees of
molten or meltable material should be present and slags from hematite atmospheric pollution. Clearly the importance of the soundness test will
castings are normally prohibited. Slag should also be free from splintered be more significant in regions where both atmospheric pollution and
or glassy components. Prior to the use of slags it is generally worth freeze-thaw are greatest.
examining their chemistry since their properties as ballast aggregate may In a study of the soundness tests based on 5 cycles of immersion and
often be cheaply improved by the addition of silica sand prior to cooling. drying the Magnesium Sulphate test correlated best with 60 cycles of
freeze thaw using a 24 hour half cycle. This correlation is shown in
Petrological and Geological Requirements Figure 7. The significance of 60 cycles may be seen in Figure 8 and
A visual petrological analysis using a hand lens or low powered represents less than one year's exposure for much of North America.
(stereo) microscope of freshly broken rock samples is of major value in If weathering occurred independently of other factors, a higher
selection of a suitable quarry. In addition, examination of the sand and standard for aggregates would need to be specified for branch lines than
smaller sized particles produced during physical testing, such as the Los for main lines where ballast breakdown from loading on a branchline is
Angeles Abrasion Test, gives an indication of the angularity and clearly less. Unfortunately these processes are not independent and any
permeability of future ballast breakdown. Where the minerals are fine weakening from weathering allows accelerated breakdown from the
grained and hard, supplemental examination by means of thin section loading environment. Thus higher standards are generally specified for
analysis and possible chemical analysis may be required. In general the ballast used in mainline tracks than those used in branchlines. This is
petrographer, if experienced, should be given freedom to decide the extent considerably more demanding then presently (1985) required by the
of the testing required. AREA Manual specification (i.e. 7% for use with both wood and concrete
The information obtained from the petrological analysis should be ties).
documented under those of the following headings that are appropriate to (b) Absorption Testing
the aggregate under examination - Absorption testing where oven dry ballast particles are immersed in
(a) Rock type; with percentages where more than one rock type is water to measure their surface absorption. The test indicates the ability of
present. the particles to retain water which would freeze and cause degradation
(b) Mineralogy; including proportions present in each rock type. during daily freeze/thaw cycles or could react chemically with minerals.
(c) Texture; include comments on grain size, shape orientation plus Its importance is related to (a) the rainfall during that time of the year that
mutual relationships and matrix material between minerals. freeze/thaw cycles occur plus the extent of such weather, and (b) the
(d) Structure; identifying bedding, fracture, cleavage, and foliation impurities of the local rainwater where the material is to be used. For
planes. example the reaction of low pH (acid) rainwater with calcite the main
(e) Mechanical properties from a geological viewpoint; including mineral in limestone. As seen in Figure 9, the acidity of rainwater falling
Mohs' hardness of the minerals, induration or compactness in 1974-75 on Eastern North America was quite high (pH less than 7 is
including porosity of the rock, possible strength and brittleness acidic and each one unit drop represents a tenfold increase in acidity) and,
including comments on possible or existing types of fracture as seen in Figure 10, it is changing with time. Greater acidities exist for
weaknesses, and shape and roundness of particles. Estimate of 1984 as evidenced by the death of lakes and forests in areas like the
specific gravity of rock. Adirondack Mountains. Their exact values should be checked at the time
(f) Chemical properties; defining existing chemical weathering and of aggregate selection. The 1984 CP Rail Specification permits a
potential chemical weathering. Where known, this may be related maximum increase in weight of dry aggregate particles after submersion
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 2
in water of 1/2% for primary mainline track rising to 1% for minor performance, where better performance of ballasts with equal LAA values
branchlines. Again considerably more demanding than the present (1985) is noted from the ballasts composed of harder mineral grains. This is
AREA Manual specification (unlimited for use with wood ties and 1.5% particularly evident with the use of concrete ties, where the concrete is
for use with concrete ties). made from silica sands having minerals of Mohs' hardness of six or more
and are thus more abrasive than wood, and hard mineral ballasts have now
Stability been generally adopted (Weber, 1975). Indeed, the present (1985) AREA
The properties of an aggregate that are most commonly associated with Manual concrete tie specification specifically excludes carbonate ballasts
the holding power of a ballast are measured by - (limestones composed of calcite, Mohs' hardness = 3; and dolomites,
(a) The Bulk Specific Gravity Mohs' hardness = 3.5-4) because of their recorded poor performance with
The bulk specific gravity of the rock which is related to the unit weight concrete ties. Because rocks are composed of minerals each having
of the processed ballast. The unit weight is a major factor in determining different hardness values it was necessary to develop a method of
both the vertical and lateral holding capacity of the track; the holding assessing the rock's overall hardness. This was achieved by the adaption
capacity increasing as the track mass increases. The 1984 CP Rail of an autogenous grinding test known as the Mill Abrasion (MA) that is
Specification requires a minimum of 2.65 for primary mainlines and 2.60 commonly used in the mining industry to access the grindability of ores
for other lines. (e.g. McIntyre and Plitt, 1980). Rocks having a predominance of hard
(b) Particle Shape and Texture minerals were noted to have low MA values. Rock having similar
The shape and texture of the particles is theoretically a production minerals were also noted to give a variation in values based on their
factor. Clearly if the aggregate source is pebbles and cobbles the source degree of induration or compactness which added to the significance of
particles must be large enough to result in sufficient fractured faces after the test in terms of assessing rock hardness.
crushing. Because of the instability and increased maintenance noted by After considerable analysis of laboratory and some field data an
CP Rail in their field studies of poorly crushed ballast a stringent Aggregate Index was proposed equal to the Los Angeles Abrasion (C-535
definition of a crushed particle was developed. The 1984 CP Rail grading 3) plus five times the Mill Abrasion (both tests on freshly crushed
specification defines a fractured face as a freshly exposed surface whose aggregate). The 1984 CP Rail specification renames this index an
maximum length is at least one-third the length of the maximum particle Abrasion Number (Na) so that
dimension and whose maximum width is at least one-quarter of the
maximum particle dimension. A definition of the crushed surface's
maximum width is not given. A simple definition would be the maximum Ia ' N a ' LAA % 5 MA (1)
distance between two parallel lines on either side of and in the same plane
as the crushed face. A fractured particle is defined as having not less than
three fractured faces whose planes, if intersecting, subtend an angle of less Correlation of this index with plastic axial strains from two stage triaxial
than 135 degrees. These requirements are considerably more stringent tests using a confining pressure of 5 psi and deviator stresses of 20 psi and
than the present (1985) AREA Manual specification. CP Rail's minimum 30 psi each repeatedly applied for 500,000 cycles are shown in Figure 11.
requirement for percentage of crushed particles are; 60% for branchlines The plots for stage 1 only, stage 1 and 2 together, and stage only all
and 90% for mainlines while their largest particle main grading established significant relationships. Figure 12 shows the particle
(equivalent to AREA no. 3) requires 100% crushed particles. breakdown measured at the end of testing and also from one dimensional
In addition good ballast particle interlock requires a minimum of (consolidometer) testing using 80 psi repeatedly loaded for 500,000
elongated or flaky particles and thus rejection of rock sources exhibiting cycles. Breakdown has been measured as a percentage change in the
schistosity or slate like structure. ¯
grading fineness modulus A and as the percentage of particles passing the
No. 8 sieve (the initial value being zero). The correlations are significant
Load Resistant Characterization for the triaxial data and mildly so for the particles generated passing the
Without exception every ballast specification attempts to assess the No. 8 sieve in the one dimensional test. The one-dimensional test data
quality of the ballast particles under loading. Ideally this quality measure was probably less significant because of the use of such a high repeated
should reflect both the hardness and toughness of mineral/bonding matrix loading. Finally examined and plotted in Figure 13 are results obtained
making up the ballast particles or parent rock. The typical tests that are from ten ballasts placed in adjacent quarter mile sections of CN Rail track.
performed on a mix of the particles by railroads world-wide include A subjective field breakdown rating established by Gaskin and Raymond
Impact testing, Crushing Value testing, Los Angeles Abrasion testing, and (1976) and based on maintenance requirements was noted to be mildly
the like. These tests unfortunately measure mainly the toughness of an significant while breakdown as reported by Dalton (1973) of particle
aggregate and are only slightly affected by the hardness of its minerals. grading curves of extracted samples resulted in disappointing results. The
Good correlation has been obtained from a comparison of the results of latter however is not surprising as initially gradings in any mass produced
any two of the tests (Gaskin and Raymond, 1976). To make comparative stockpile are highly variable. In addition breakdown was more than load
measures of the toughness of different rocks it is only necessary to use one environment produced with several of the aggregates showing
of these tests. The Los Angeles Abrasion (LAA) test is almost universally considerable weathering.
accepted in North America. Even so, as seen from the results in Table 1, Research by CP Rail has related the aggregate index or abrasion
this test performed to specification using different size particles results in number to the observed life of ballast from the loading environment alone.
different LAA values as does the use of track used particles rather than Ballasts which weathered or were badly fouled from foreign sources were
freshly crushed aggregate. It is therefore important to compare ballasts on eliminated from the CP Rail study. In addition the subgrade support was
a standard size of freshly crushed particles, say 19 mm - 38 mm (0.75 in - relatively good. For AREA No. 4 graded ballast the cumulated short tons
1.5 in), irrespective of the maximum particle size of the ballast being used. of 33 tons axles to result in breakdown to the point where the ballast
For this reason the new CP Rail specification requires the exclusive use needed renewal was found to be given by
of ASTM C535, Grading 3 for the LAA evaluation.
A second factor that needs understanding in relation to toughness
evaluation is that for rocks of similar field rating, the impact from the Life ' 106 exp (8.08 & 0.0382 N a) short (2)
LAA steel ball charge will increase slightly as the hardness of the mineral
grains increases, resulting in a higher LAA breakdown. However, the
field breakdown from the harder mineral rock is often slower because less CP Rail also noted that AREA No. 3 graded ballasts had about a 20%
powdering occurs at the points of contact between particles and the broken increase in life while AREA No. 5 graded ballast only had 50% the life of
particles are more angular and coarser resulting in a slower rate of track the No. 4 graded ballast.
fouling. Comparisons based on LAA alone that measures primarily rock Figure 14 shows some numerical values of the above equation in
toughness would therefore result in decisions contrary to field relationship to laboratory test data from the writer's work on different
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 3
aggregates showing an interesting relationship for carbonate materials. In the case of quarried material it is always possible for clay seams or
Also identified on Figure 14 are ballast types whose petrological seams of soft rock to be present within the deposit. Similarly, clay or soft
description is given in Table 2 allowing an estimate of aggregate life for particles may be present in gravel sources. During production of ballast
those organizations not having a Mill Abrasion apparatus. It must be it is necessary to periodically check for purity. Three tests are normally
clearly understood that weathering or foreign source fouling of ballast required although they do not form part of the CP Rail Specification since
dramatically reduces the life times given in Figure 14. For example they would be noted in the petrographic examination. These tests are used
excessive fouling from highly plastic clay fines may result in as much as in the 1985 AREA Manual specifications whose limits are: (1) Soft and
an order of magnitude drop in the observed life. friable pieces < 5% (< 3% for concrete ties) (2) Material finer than No.
The correlation of aggregate index noted in Figure 11 when load 200 sieve 1% (0.5% for concrete ties) (3) Clay lumps 0.5% (0.5% for
ranking ballast for the permanent settlement observed from triaxial tests concrete ties).
suggests that the same aggregate index may be used to estimate the
settlement performance between tamping cycles provided there is no Gradation
fouling from foreign source fines. Confirmation of such a relationship The particle gradation of a ballast selected for track use is clearly
may be inferred from the data obtained by Hay et al (1977). This data has independent of the aggregate source. Ballast gradings are usually close
been plotted in Figure 15. It may be seen that the fine grained hard to uniformly graded with field productions based on the use of two or
mineral ballasts (basalts and sulphide slags) clearly outperform the soft three sieves. For example the AREA No. 4 grading which is one of the
mineral grained and large harder grained aggregates (limestones and most often used gradings, permits as an extreme 100% passing the 38 mm
granites). Additional confirmation has been presented by Bing and Gross (1-1/2") sieve with 80% retained on the 25 mm (1") sieve. The 1985
(1983) where the aggregate index formed part of a ballast index that was AREA Manual and new CP gradings are shown in Table 3.
found to be the most significant parameter in predicting track degradation Single sized ballasts have larger void volumes than broadly graded
for maintenance planning. ballasts and thus where ballast fouling from aggregate breakdown is the
Apart from having limits of the LAA, MA and Na the CP Rail major source of contamination they are generally to be preferred. Broader
Specification suggests a minimum life for ballast of 30 years and requires graded ballasts are generally stronger and where track stability is a major
a life of more than 20 years based on expected line tonnage. concern such as on high curvature track or track with high grades, the
broader graded AREA 24 ballast or an even more broadly graded ballast
PRODUCTION TESTING may be beneficial provided the aggregate quality is high and aggregate
General degradation estimated to be minimum. Such a grading has been used by
Once the aggregate source for the ballast has been selected and a B.C. Rail on their mountain territory that has up to 12 degree curves
contract has been signed for ballast production it is necessary to monitor combined with 2.2 percent grades. The aggregate source was a basalt.
its production. While it is advisable to check the quality of the aggregate Prior to the adoption of a broad grading an AREA No. 4 grading was in
from time to time by performing the same tests as used for the aggregate use and the maintenance cycle in this territory was as low as 3 months on
selection more important is the performance of tests to monitor the ballast the worst curves. After adoption of the grading shown in Figure 16 the
properties that are production variable. lowest maintenance cycle rose to 2 years.
Shape and Surface of Particles The major problems associated with broadly graded ballasts is
As already commented on in relation to (track) stability in the section segregation of the particles which commonly occurs during handling and
on Aggregate Selection the particle shape and its surface is of utmost transportation. Special precautions need to be enacted to minimize such
importance and has long been recognized as having a major effect on track segregation if broadly graded ballasts are adopted.
stability. High quality ballasts are normally required to have a high
percentage of fractured faces and be cubical. European practice, as given TESTING
by the International Union of Railways (1967), is primarily based on A distinction has been made between aggregate selection and its
limiting the percentage of particles whose ratio of longest dimension to production into ballast. This distinction is recognized in the CP Rail
least dimension (the least dimension if measured by passage through an Specification although no distinction has been made in relation to the
infinite slot) exceeds three, with no single particle having a ratio greater frequency of testing during production. The CP Rail specification
than 10. The percentage limit varies from railroad to railroad ranging requires tests to be performed every 1000 tones compared with the AREA
from 5% - 20% for tolerance 'A' ballast and 20% - 33% for tolerance 'B' recommendation of every 200 tons. If the 200 ton requirement is followed
ballast. The desire for cuboid shape particles is unquestionable. CP Rail, then the writer would suggest that only production related testing as
based on their experience with their ballast, have chosen in their outlined herein be done every 200 tons while aggregate selection tests be
specification to have no direct restrictions on particle shape but do have performed every 1000 tons. Based on the writer's knowledge of railroad
very stringent controls on ballast particle surface. practice many would consider a 1000 ton test requirement excessive.
The Queen's University study pointed out the importance of clarity
regarding the definition of a crushed face. This led CP Rail to do an CONCLUDING COMMENTS
extensive study of the maintenance costs associated with poorly crushed The evaluation of ballast requires a two stage process involving first an
gravel ballast compared with well crushed gravel ballast. The results of aggregate selection and then a monitoring of the processed material. The
these studies are reflected in their specification where they require not less reasoning for the specification of each test and the explanation of what it
than 60 percent crushed particles for branch lines rising to not less than 90 evaluates is given. The data presented forms part of the background
percent for mainline track built with continuous welded rail. In addition, material relating to the 1984 CP Rail specification whose main
their largest particle grading (No. 5) is required to have 100 percent requirements are given in Figure 17. Generally not understood is the
crushed particles. Their definition of a crush particle is also very stringent value that can be obtained from perological analysis since no mention is
requiring three crushed faces whose planes, if intersecting, subtend an made of its use in the ballast specification of the present (1984) AREA
angle less than 135 degrees for a particle to be termed a crushed particle. Manual for Railway Engineering. This was a major recommendation of
A crushed face is defined as being a fresh crushed surface having a length the Queen's research and has been made a major requirement of the new
not less than one-third the maximum particle dimension and a maximum CP Rail specification which states (Item 4c of their specification)
width not less than one-quarter the maximum particle dimension. A "Where a discrepancy arises between the estimated results from the
definition of the crushed surface's maximum width is not given. A simple petrographic analysis shall have precedence; provided the petrologist
definition would be the minimum distance between two parallel lines on reviews all test results and identifies the reasons for the discrepancy".
either side of and in the same place as the crushed face. Even on quarried Since ballast is generally made from rock and Petrology is the study of
rock these requirements will limit elongated and flaky particles. rocks such a requirement is nothing more than "common sense".
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 4
The original work formed part of a general study on the geotechnical
problems of railroad support being financed by CN Rail, CP Rail, and
Transport Canada Research and Development Centre under Project 2.22
of the Canadian Institute of Guided Ground Transport (CIGGT), at
Queen's University under the direction of the writer. Additional financial
support was obtained from a National Research Council of Canada Grant.
Bing, A.J., and Gross, A., "Development of Railroad Track Degradation
Models", Transportation Research Record No. 929, Transportation
Research Board, Washington 1983, pp. 27-31.
"Classification of Ballast Prescriptions", Report No. 7, International
Union of Railways, Office for Research adn Experimentation, Question
D71, Stresses in the Rails, the Ballast and in the Formation resulting from
Traffic Loads, Utrecht, the Netherlands, June, 1967, p. 62.
"C.P. Rail Specification for Evaluating Processed Rock, Slag and Gravel
Ballasts", Canadian Pacific Ltd., Montreal, 1981.
"C.P. Rail Specification for Processing Rock, Slag and Gravel Ballasts",
Canadian Pacific Ltd., Montreal, 1981.
Dalton, C.J., "Field Durability Tests on Ballast Samples as a Guide to the
Significance of teh Specification Requirements", Canadian National
Railways Technical Research Centre, St. Laurent, Quebec, Canada, 1973,
Gaskin, P.N. and Raymond, G.P., "Contribution to Selection of Railroad
Ballast", Journal of teh Transportation Engineering Division, ASCE, Vol.
102, No. TE2, Proc. Paper 12134, May, 1976, pp. 377-394.
Hay, W.W., Baugher, R., Reinschmidt, A.J., "A Study of Railroad Ballast
Economics", Federal Railroad Administration, U.S. Department of
Transport, Report FRA/OROD-77/64, Sept. 1977, p. 91.
Leary, E., "The Building Limestones of the British Isles", Building
Research Establishment Report, Her Majesty's Stationery Office, London,
1983, p. 91.
Manual for Railway Engineering, Vols. 1 and 2, American Railway
Engineering Association, Washington, D.C., 1984-1985.
McIntyre, A., and Plitt, L.R., "The Inter-relationship Between Bond and
Hardgrove Grindabilities", Canadian Institute of Mining and Metallurgy
Bulleting 63, June 1980, pp. 149-180.
Raymond, G.P. and Diyaljee, V.A., "Railroad Ballast Load Ranking
Classification", Journal of the Geotechnical Engineering Division, ASCE,
Vol. 105, No. GT10, Oct. 1979, pp. 1133-1153.
Raymond, G.P., Boon, C.J. and Lake, R.W., "Ballast Selection and
Grading - A Summary Report", Canadian Institute of Guided Ground
Transport, Queen's University, Kingston, Ontario, Report No. 79-4, April
1979, p. 50.
Weber, J.W., "Development of the Prestressed Concrete Tie in the USA",
Symposium on Railroad Track Mechanics and Technology, A.D. Kerr,
ed., Pergamon Press, Elmsford, N.Y., 1975, pp. 265-282.
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 5
TABLE 1 LOS ANGELES ABRASION RESULT FINDING
TEST METHOD RESULT AGGREGATE
C535 GRADING 3 13.9 Andersite freshly crushed
C131 GRADING A 17.2 Andersite freshly crushed
C131 GRADING B 16.6 Andersite freshly crushed
C131 GRADING C 19.6 Andersite freshly crushed
C131 GRADING D 24.6 Andersite freshly crushed
C535 GRADING 3 34-39* Granite freshly crushed
C535 GRADING 3 25 Granite particle already used in LAA test to
obtain above result
* Rnage of three tests on same source.
TABLE 2. Part 1 off 2. HAND SPECIMEN-PETROLOGICAL EXAMINATION
Kenora Granite Granodioritic gneiss -- mostly plagioclase feldspar, quartz horneblende. Hard but weak to medium toughness. Prone to
fracture on foliation.
Sudbury Slag Two-phase material from smelting of nickel ore -- mostly silicates; very hard and relatively tough. Fine material is about
90 percent angular fragments; should remain highly permeable.
Noranda Slag Very hard, tough material -- a by-product from the smelting of copper-zinc sulphide ores. Fine material is predominantly
angular fragments -- should remain highly permeable.
Medicine Hat Gravel containing quartzite, diabase, and granite predominant; tough, medium hard, isotropic material; should retain
Suicide Creek Creek gravel, mostly granodirite with some gabbro, granite; of medium hardness and toughness; fractures on grain
boundaries; material should be permeable, but could produce clays from weathering of feldspars.
CEM-1 Initially classified as fine to medium grained igneous rock then identified by thin section as metamorphosed Quartz
Diorite Porphyry. Two sets of joints present with substantial weathering and secondary mineral formation on the joint
planes. Fractures along joints would produce 25 to 75 mm (1" to 3") blocky fragments. Mineral matrix appears sound.
If main body of rock is sound and unweathered aggregate should make excellent ballast.
Marmora Trap Actually epidote skarn; generally hard, but composition varies considerably. Otherwise quite tough. Mostly calcium-
Rock magnesium-iron silicate. Chemical weathering of fines could produce clay minerals. Otherwise should remain
CP-2 An isotropic, equi-granular grey granite, composed of plagioclase feldspar, potassium feldspar, biotite, horneblende and
quartz. Material is tough and hard and should remain highly permeable.
Steel (OH) Slag About 20% of the slag is a massive grey stone material which, although the surface is sometimes coated with a relatively
soft scale is relatively hard, tough, of high density, angular shape and exhibits no obvious planes of weakness. The other
approximately 80% consists of pieces exhibiting either one or more of the following features: (a) widespread vesicles
(b) substantial fine grained crystal growth (c) inclusions of several materias including a soft carbonacious substance of
about the consistency of coke (d) spherical particles about 1/2 mm in diameter resting within the larger vesicles, and (e)
a substantial degree of rust. Although much of the material grains are hard, the aggregate is weak, crumbling on impact
and unsuitable on its own for ballast. About 5% of the O.H. material is either brick or other nondescript material
presumably some flux derivative. These materials, will not make good ballast.
Brandon Gravel River gravel; mostly gneiss of varying types; weak, medium hard, will fracture along gneissosity; permeability will be
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 6
TABLE 2. Part 2 off 2. HAND SPECIMEN-PETROLOGICAL EXAMINATION
Steel (BOF) About 60% of the slab is a massive grey stone material which although the surface is sometimes coated with a relatively
Slag soft scale is relatiely hard, tough, of high density, angular shape and exhibits no obvious planes of weainess. The other
approximately 40% consists of pieces exhibiting either one or more of the following features; (a) widespread vesicles
(b) substantial fine grained crystal growth (c) inclusions of several materials including a soft carbonacious substance of
about the consistency of coke (d) spherical particles about 1/2 mm in diameter resting within teh larger vesicles, and (e)
a substantial degree of rust. Although for much of this 40% portion the material grains are hard, the aggregate is weak,
crumbling on impact and unsuitable on its own for quality ballast. About 2% of the B.O.F. material is either brick or
other nondescript material presumably some flux derivative. These materials will not make good ballast, but because of
their relatively minor occurrence should have little impact on ballast performance.
Kinmberly Float Metasediments with highly variable mineral assemblages; or medium hardness and toughness, but showing pronouned
planar fabrics due to metamorphism. Should be permeable and relatively resistant to chemical weathering.
CP-1 Creek gravel; 60 per cent horneblende-biotite skarn, 30 per cent feldspar horneblende intrusive, and 10 per cent limestone;
materials generally tough but soft; long-term permeability could be a problem.
N. Walachin Pit Gravel; about 65 per cent skarn, 25 per cent marble, 10 per cent hornfels. Materials are soft, weak-marble especially prone
to fracturing on cleavage; shoudl remain permeable.
Alberta North Limestone, abundant crinoidal fossils; weak and soft; should remain permeable but will be prone to solutional weathering
by weakly acidic waters.
PAR-1 Fine to medium-grained rather massive limestone or dolomite. Rather 'dirty' with substantial portion of clay-sized
particles among fines. May be expected to abrade easily although aggregate appears tough.
Coteau Tough, very fine-grained dolomite; relatively soft, but no preferred directions of fracture. Fines 95 per cent powder,
Dolomite should remain relatively permeable.
Saint Isidore Very dirty, fossiliferous limestone; soft, weak, will evolve considerable clay and sand on breakdown, which will foul
S. Joliette Limestone, fossiliferous; soft, weak; fractures on bedding planes and cleavage of CaCO3 crystals. Fine material about
Limestone 95 percent powder, 5 per cent angular cleavage rhombs. Should remain permeable if track environment is wekaly acidic.
Montreal Coarse-grained, clayey lime-stone; soft, weak; cleavage in coarse crystals forms planes of weakness encouraging
Limestone fracturing. Fines about 90 percent powder, initially permeable, but could foul with clays as weathering proceeds.
Megantic Compact, isotropoic limestone, soft and weak; shoudl remain permeable, but will be prone to solution weathering by
Limestone weakly acidic water.
Grenville Non-ballast used for laboratory test comparisons. Coarsely crystalline calcite. Soft mineralled and weak. Not suitable
Marble as a ballast material.
Saint Marc Coarsely-crystalline, fossiliferous limestone; soft, weak; fractures on cleavage planes. Clayey material, some quartz
Limestone present in fines. Permeability likely to decrease over time in track.
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 7
TABLE 3 AREA AND CANADIAN RAILWAY'S RECOMMENDED BALLAST GRADATIONS
Limits of Percent Passing Each Sieve (Square Openings) Percen by Weight
Grading Norminal 3" 2 ½" 2" 1 ½" 1" ¾" ½" d" #4
Area 24 2½" - ¾" 100 90-100 25 - 60 0 - 10
AREA 3 2" - 1" 100 95 - 100 30 - 70 0 -15
CP 5 2" - 1" 100 90 - 100 35 -70 0-5 0-3
AREA 4 1½ - ¾ 100 90 - 100 20 - 25 0 15
CP 4 1½ - ¾ 100 90 - 100 20 - 55 0-5 0-3
CP 3 1½ - ½ 100 90 - 100 70 - 90 30 50 0 - 20 0-3
CP 2 1½ -d" 100 90 - 100 70 - 90 50 - 70 25 - 45 10 - 25 0-3
CN 1 -3/16" 100 75 - 90 50 -75 20 - 55 0-5
AREA 5 1 -d 100 90 - 100 40 - 75 15 - 35 0 - 15
AREA 57 1 -3/16" 100 95 - 100 25 - 60 0 - 10
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 8
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 9
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 10
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 11
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 12
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 13
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 14
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 15
RAILROAD BALLAST SPECIFICATION AND EVALUATION by Gerald P. Raymond - Lecture notes 16