2.1 Particle size analysis
After this lesson you should be able to:
- Describe the IS 2720 (Part 4):1985 methods for grain-size analysis: sieving above 75 µm (wet sieving for all soils; dry sieving only if there is no appreciable clay) and sedimentation below 75 µm (pipette standard, hydrometer subsidiary).
- State when the sedimentation analysis is not required (less than 10 percent passing the 75 µm IS sieve), and how the results are reported on a semi-logarithmic grain-size curve.
- Compute percent finer from a sieve record, interpolate D10, D30 and D60 on a logarithmic size scale, and compute the coefficients of uniformity and curvature used in IS 1498:1970 Table 3.
- Split a sample into the IS 1498 gravel, sand and fines fractions at 4.75 mm and 75 µm, and say why D10 cannot be read from sieves alone when more than 10 percent passes 75 µm.
Theory
A soil is a mixture of particles of many sizes. The grain-size distribution is the relationship between particle size and the percentage of the dry mass that is finer than that size. It is the first laboratory index test in this chapter, and it is the basis for the coarse side of soil classification (Lesson 2.3). Consistency limits, which describe the fines, are Lesson 2.2. The size boundaries used in Indian practice were introduced in Lesson 1.1 from IS 1498:1970: gravel from 80 mm down to 4.75 mm, sand from 4.75 mm down to 75 µm, silt-size from 75 µm down to 2 µm, and clay-size below 2 µm. This lesson is about how those percentages are measured.
IS 2720 (Part 4):1985 cl. 1.1 covers the quantitative determination of grain-size distribution in soils. Two methods are given for sizes larger than the 75-micron IS sieve (cl. 1.1.1). Wet sieving shall be applicable to all soils. Dry sieving shall be applicable only to soils which do not have an appreciable amount of clay: dry clay lumps would otherwise be counted as if they were coarse particles. For sizes smaller than 75 microns, cl. 1.1.2 gives the pipette method as the standard method and the hydrometer method as a subsidiary method. That sedimentation analysis shall not be applicable if less than 10 percent of the material passes the 75-micron IS sieve (the passing percentage being determined as in cl. 4). In outline, the test is done in three parts: dry sieve analysis of the fraction retained on the 4.75 mm IS sieve (cl. 3); sieving, including wet sieving, of the fraction that passes 4.75 mm and is retained on 75 µm (cl. 4); and analysis of the fraction that passes 75 µm (cl. 5), when that fraction is large enough to require it. A dispersing agent, sodium hexametaphosphate, is used so that the fine particles are separated rather than left as flocculated clusters (cl. 4.2).
The results are reported as in cl. 6.1: a grain-size distribution curve is drawn on a semi-logarithmic chart, with particle size on the log scale and the percentage finer than the corresponding size on the ordinary scale. A recommended record form is given in Appendix A of the standard. From that curve three particle sizes are read: D10, D30 and D60, the sizes at which 10, 30 and 60 percent of the dry mass is finer. D10 is also called the effective size. Because the size axis is logarithmic, interpolation between two adjacent sieves is done in log D, not in D itself.
IS 1498:1970 Table 3 (cl. 3.5.1) defines two coefficients from those sizes, in our own words: the uniformity coefficient Cu = D60 / D10, and the coefficient of curvature Cc = (D30)² / (D10 × D60). A well-graded gravel (GW) requires Cu greater than 4 and Cc between 1 and 3; a well-graded sand (SW) requires Cu greater than 6 and Cc between 1 and 3, together with the fines and plasticity rules of that table. Those classification limits are applied in Lesson 2.3. This lesson stops at computing Cu and Cc and at splitting the sample into gravel, sand and fines. ASTM D6913 / D7928 and EN ISO 17892-4 are named only as comparable test methods, not ranked.
Sedimentation of the fraction passing 75 µm rests on a textbook principle, Stokes’ law: a spherical particle of diameter D and specific gravity Gs, falling in a liquid of viscosity μ and unit weight γw, reaches a terminal velocity v = (Gs − 1) γw D² / (18 μ). The pipette and hydrometer methods time the concentration of solids at a known depth and convert that concentration, through a velocity and hence through D, into a percent finer. The law assumes a spherical particle, laminar settling, a known temperature (because μ depends on temperature), and a fully dispersed suspension. Real silt and clay particles are not spheres, which is one reason the result is a standard laboratory size, not a unique physical diameter. If less than 10 percent passes 75 µm, IS 2720 (Part 4) does not require this step, and D10 can often be read from the sieve curve alone. If more than 10 percent passes 75 µm, D10 (and sometimes D30) may lie below 75 µm and cannot be taken from the sieves.
Derivation
- From a sieve record, the cumulative mass retained above each sieve is divided by the oven-dry mass of the sample (or the corrected mass used in the IS 2720 (Part 4) form) and multiplied by 100 to give the cumulative percent retained. The percent finer is 100 minus that cumulative percent retained.
- On the semi-log curve, D10, D30 and D60 are the particle sizes at 10, 30 and 60 percent finer. Between two adjacent sieves of sizes D_a (at P_a percent finer) and D_b (at P_b percent finer), a size D at percent finer P is interpolated by log10 D = log10 D_a + (log10 D_b − log10 D_a) (P − P_a) / (P_b − P_a).
- The IS 1498:1970 Table 3 definitions are then Cu = D60 / D10 and Cc = (D30)² / (D10 × D60). Both coefficients are dimensionless.
- The gravel fraction is the percent retained on the 4.75 mm IS sieve (of the material smaller than 80 mm, as in Lesson 1.1). The fines are the percent passing the 75 µm IS sieve. The sand fraction is the remainder. If the percent passing 75 µm is less than 10, IS 2720 (Part 4) cl. 1.1.2 does not require a sedimentation analysis.
- Stokes’ law, used only as the principle of the pipette and hydrometer methods, is v = (Gs − 1) γw D² / (18 μ), with D the equivalent spherical diameter. Rearrangement gives D from a measured settling velocity at a stated temperature.
Percent finer = 100 − cumulative percent retainedLog interpolation: log10 D = log10 D_a + (log10 D_b − log10 D_a) (P − P_a) / (P_b − P_a)Cu = D60 / D10; Cc = (D30)² / (D10 × D60) (IS 1498:1970 Table 3)GW (later, Lesson 2.3): Cu > 4 and 1 ≤ Cc ≤ 3; SW: Cu > 6 and 1 ≤ Cc ≤ 3 (same table, with the fines rules)Gravel / sand / fines split at 4.75 mm and 75 µm (IS 1498 ranges; sedimentation if 10 percent or more passes 75 µm)Stokes (textbook principle): v = (Gs − 1) γw D² / (18 μ)
- Percentages are on the oven-dry mass of the soil fraction being analysed, as in IS 2720 (Part 4). A water-content correction is needed if the test mass was not oven-dry.
- Logarithmic interpolation assumes a straight line between two adjacent sieves on the semi-log chart. It is not a physical claim about particles between those sieves.
- Cu and Cc are defined only when D10, D30 and D60 can actually be read. If more than 10 percent passes 75 µm and no sedimentation test has been done, D10 may be unknown.
- Stokes’ law assumes a spherical particle, laminar settling, a known temperature and viscosity, and a dispersed suspension. It is the textbook basis of the pipette and hydrometer methods, not a clause printed in IS 2720 (Part 4).
- Dry sieving of the fraction above 75 µm is applicable only when there is no appreciable clay (cl. 1.1.1). Wet sieving is the method that applies to all soils.
- The GW and SW limits are classification rules of IS 1498 Table 3. They are named here so that Cu and Cc have a use; a complete group symbol also needs the fines content and, where relevant, the plasticity tests of Lesson 2.2.
- ASTM and ISO methods are named for comparison only, not ranked, and are not used as a substitute for IS 2720 (Part 4) on an Indian specification.
Codes and standards
Indian Standards are cited by number and clause only; no clause text, tables or figures are reproduced. Other standards are listed for comparison, not ranked.
- IS 2720 (Part 4):1985 (reaffirmed 2006) — cl. 1.1, 1.1.1 and 1.1.2 (scope and the 10 percent rule); cl. 3 (dry sieving of the fraction retained on 4.75 mm); cl. 4 (sieving of the fraction passing 4.75 mm, including wet sieving); cl. 5 (fraction passing 75 µm); cl. 6.1 (semi-log grain-size curve and Appendix A proforma). Dispersant: sodium hexametaphosphate (cl. 4.2).
- IS 1498:1970 Table 3 (cl. 3.5.1) — definitions of Cu and Cc, and the well-graded limits Cu > 4 and 1 ≤ Cc ≤ 3 for GW, Cu > 6 and 1 ≤ Cc ≤ 3 for SW. Size ranges of Lesson 1.1 (gravel, sand, silt-size, clay-size).
- IS 460 (test sieves) is the sieve specification named from IS 2720 (Part 4) cl. 3.1.2. Preparation of dry samples: IS 2720 (Part 1), named in cl. 3.2.
- ASTM D6913 / D7928 and EN ISO 17892-4 are named for comparison only, not ranked.
Worked examples
Problem. An oven-dry sample of 500 g, all passing the 4.75 mm IS sieve, is wet-sieved and then dry-sieved on the 2.36, 1.18, 0.600, 0.300, 0.150 and 0.075 mm IS sieves. Masses retained are 40, 80, 120, 130, 80 and 35 g respectively; the pan (passing 0.075 mm) holds 15 g. (a) Tabulate the percent finer. (b) Find D10, and interpolate D30 and D60 on the log-size scale. (c) Compute Cu and Cc. (d) Split the sample into sand and fines, and say whether IS 2720 (Part 4) requires a sedimentation analysis.
- The retained masses sum to 40+80+120+130+80+35+15 = 500 g, so the percentages are on a 500 g base. Cumulative masses retained: 0 g on 4.75 mm; then 40, 120, 240, 370, 450 and 485 g; pan 500 g.
- Cumulative percents retained: 0, 8.0, 24.0, 48.0, 74.0, 90.0 and 97.0 percent. Percents finer: 100, 92.0, 76.0, 52.0, 26.0, 10.0 and 3.0 percent at 4.75, 2.36, 1.18, 0.600, 0.300, 0.150 and 0.075 mm.
- D10 is the size at 10 percent finer. That point falls on the 0.150 mm sieve, so D10 = 0.150 mm.
- D30 lies between 0.300 mm (26.0 percent finer) and 0.600 mm (52.0 percent finer). log10 D30 = log10(0.300) + (log10(0.600) − log10(0.300)) × (30 − 26) / (52 − 26) = −0.522879 + 0.301030 × 4/26 = −0.476566, so D30 = 10^(−0.476566) = 0.3338 mm.
- D60 lies between 0.600 mm (52.0 percent finer) and 1.18 mm (76.0 percent finer). log10 D60 = log10(0.600) + (log10(1.18) − log10(0.600)) × (60 − 52) / (76 − 52) = −0.221849 + 0.293731 × 8/24 = −0.123938, so D60 = 10^(−0.123938) = 0.7517 mm.
- Cu = D60 / D10 = 0.7517 / 0.150 = 5.01. Cc = (0.3338)² / (0.150 × 0.7517) = 0.1114 / 0.1128 = 0.988.
- The sample has 0 percent gravel (nothing retained on 4.75 mm), 3.0 percent fines (passing 75 µm), and 97.0 percent sand. The fines are less than 10 percent, so cl. 1.1.2 does not require a sedimentation analysis. D10 was readable from the sieves, which is consistent with that rule.
- Classification is not completed here. For a later Lesson 2.3 comparison only: a clean sand would need Cu greater than 6 and Cc between 1 and 3 to be well graded (SW) under IS 1498 Table 3; these figures would not meet those two limits.
Answer. Percent finer: 100, 92.0, 76.0, 52.0, 26.0, 10.0, 3.0 at 4.75 to 0.075 mm. D10 = 0.150 mm, D30 = 0.3338 mm, D60 = 0.7517 mm, Cu = 5.01, Cc = 0.988. Sand 97 percent, fines 3 percent; no sedimentation analysis required under cl. 1.1.2.
Problem. A second oven-dry sample of 500 g has 80 g passing the 75 µm IS sieve (16.0 percent fines) after wet sieving. (a) Does IS 2720 (Part 4) cl. 1.1.2 require a sedimentation analysis? Can D10 be read from the sieves alone? (b) As a textbook illustration of Stokes’ law, how long does an equivalent spherical particle of diameter 0.075 mm and Gs = 2.70 take to settle 100 mm in water at 20 °C? Take γw = 9.81 kN/m³ and μ = 1.0016 × 10⁻³ Pa·s (the ISO/IAPWS reference viscosity of Lesson 1.1). State the limits of this picture.
- Yes. 16.0 percent passes the 75 µm IS sieve, which is not less than 10 percent, so cl. 1.1.2 requires a pipette (standard) or hydrometer (subsidiary) analysis of that fraction. D10 lies somewhere below 75 µm and cannot be read from the sieve curve alone.
- Diameter D = 0.075 mm = 7.5 × 10⁻⁵ m. In SI units, γw = 9 810 N/m³, so v = (2.70 − 1) × 9 810 × (7.5 × 10⁻⁵)² / (18 × 1.0016 × 10⁻³) = 16 677 × 5.625 × 10⁻⁹ / 0.0180288 = 9.3808 × 10⁻⁵ / 0.0180288 = 5.203 × 10⁻³ m/s.
- Time to settle 100 mm = 0.100 / 5.203 × 10⁻³ = 19.22 s.
- This is a textbook Stokes calculation for a single equivalent sphere in still water. It is not the IS 2720 (Part 4) pipette or hydrometer procedure (which uses a dispersed suspension, a temperature record, and a standard computation form), and real silt particles are not spheres. It is used only to show why a 75 µm particle settles in seconds, while a 2 µm clay-size particle, whose D² is about 1 400 times smaller, takes many hours.
Answer. Yes, sedimentation is required and D10 cannot be taken from the sieves. A 0.075 mm equivalent sphere at the stated values settles 100 mm in 19.22 s on this Stokes picture, which is a teaching estimate only.
Step-by-step procedure
- Prepare the sample as in IS 2720 (Part 1), record the oven-dry mass, and split the material on the 4.75 mm IS sieve (cl. 3 and cl. 4).
- Sieve the fraction retained on 4.75 mm by the dry method of cl. 3, on the coarse IS sieves required by the standard.
- For the fraction passing 4.75 mm, use wet sieving unless the soil has no appreciable clay and dry sieving is therefore permitted (cl. 1.1.1). Wash, dry and sieve down to 75 µm as in cl. 4.
- If 10 percent or more passes the 75 µm IS sieve, disperse the fines (sodium hexametaphosphate, cl. 4.2) and run the pipette method (standard) or the hydrometer method (subsidiary) as in cl. 5. If less than 10 percent passes, stop at the sieves.
- Compute percent finer, plot the semi-log grain-size curve (cl. 6.1), and read or interpolate D10, D30 and D60 in log size.
- Compute Cu = D60/D10 and Cc = (D30)²/(D10 D60), and the gravel, sand and fines percentages. Leave the IS 1498 group symbol to Lesson 2.3, after the consistency limits are known if they are needed.
- Report the results on a form of the kind given in Appendix A of IS 2720 (Part 4), including the method (wet or dry; pipette or hydrometer) and the temperature of any sedimentation test.
Site notes
- A sample that is air-dried into hard lumps, and then dry-sieved, will look coarser than it is. Soils with any clay should be wet-sieved, as cl. 1.1.1 requires.
- The 75 µm wash water must not be thrown away if a sedimentation test is required: that water carries the fines that the pipette or hydrometer has to measure.
- Record the water temperature during a hydrometer or pipette test. Viscosity, and therefore the inferred diameter, changes with temperature.
- A grain-size curve is not a classification by itself. Two soils can have the same Cu and still behave very differently if one has plastic fines and the other does not. Do the consistency limits (Lesson 2.2) before issuing a group symbol.
- On site, a jar test or a feel-and-wash estimate is a field clue only (IS 1498 field identification, Lesson 1.1). It does not replace the IS 2720 (Part 4) test when a specification asks for a grading.
- Dry-sieving a clayey soil and treating the lumps as gravel or coarse sand.
- Interpolating D30 or D60 linearly in millimetres instead of in log D on a semi-log chart.
- Computing Cu or Cc when D10 has not been measured, because the fines were not analysed and more than 10 percent passed 75 µm.
- Using Cu > 4 as the well-graded limit for a sand, or Cu > 6 as the limit for a gravel (IS 1498 Table 3 uses 4 for GW and 6 for SW).
- Reporting “percent passing 75 µm” as the clay fraction. That figure is the fines (silt-size plus clay-size). Clay-size is the portion finer than 2 µm, from a sedimentation test.
- Forgetting that the hydrometer method is the subsidiary method in IS 2720 (Part 4); the standard method for the fines is the pipette method.
- Treating a Stokes teaching calculation as the official IS 2720 (Part 4) result, or ranking ASTM / ISO test methods against Part 4.
Sieves above 75 µm (wet sieving for all soils; dry only if no appreciable clay). Pipette (standard) or hydrometer (subsidiary) below 75 µm, if 10 percent or more passes 75 µm (IS 2720 (Part 4)). Plot percent finer against log size. Cu = D60/D10, Cc = (D30)²/(D10 D60). SW: Cu > 6 and 1 ≤ Cc ≤ 3; GW: Cu > 4 and 1 ≤ Cc ≤ 3 (IS 1498 Table 3, used in Lesson 2.3).
Summary
- IS 2720 (Part 4) measures grain size by sieving above 75 µm and by pipette (standard) or hydrometer (subsidiary) sedimentation below 75 µm. Wet sieving applies to all soils; dry sieving of the fraction above 75 µm is only for soils without appreciable clay.
- Sedimentation is not required if less than 10 percent passes the 75 µm IS sieve. Results are plotted as percent finer against particle size on a semi-log chart.
- D10, D30 and D60 are read from that curve, interpolating in log size. IS 1498 Table 3 defines Cu = D60/D10 and Cc = (D30)²/(D10 D60), and uses them in the well-graded rules for GW and SW (Lesson 2.3).
- Gravel, sand and fines are split at 4.75 mm and 75 µm. Percent passing 75 µm is the fines, not the clay-size fraction.
- Stokes’ law is the textbook principle behind sedimentation; the official computation is the one in IS 2720 (Part 4), at the recorded temperature, on a dispersed sample.
Practice questions
Tap a question to see the answer.
1Which method does IS 2720 (Part 4) cl. 1.1.2 name as the standard method for sizes smaller than 75 µm?
The pipette method. The hydrometer method is subsidiary.
2A sample has 7 percent passing the 75 µm IS sieve. Is a sedimentation analysis required under cl. 1.1.2?
No. The method is not applicable if less than 10 percent passes the 75 µm IS sieve.
3Percent finer is 18 at 0.150 mm and 42 at 0.300 mm. Interpolate D30.
log10 D30 = log10(0.150) + (log10(0.300)−log10(0.150))×(30−18)/(42−18) = −0.823909 + 0.301030×12/24 = −0.673394, so D30 = 0.2121 mm.
4D10 = 0.12 mm, D30 = 0.35 mm, D60 = 0.90 mm. What are Cu and Cc?
Cu = 0.90/0.12 = 7.50. Cc = (0.35)²/(0.12×0.90) = 0.1225/0.108 = 1.13.
5Would those coefficients meet the IS 1498 Table 3 well-graded limits for a clean sand (SW), leaving fines and plasticity aside?
Yes. Cu = 7.50 is greater than 6, and Cc = 1.13 is between 1 and 3. The full group symbol still belongs in Lesson 2.3.
6Percent retained on 4.75 mm is 22, and percent passing 75 µm is 11. What are the gravel, sand and fines percentages?
Gravel 22 percent, fines 11 percent, sand 100 − 22 − 11 = 67 percent.
References
Standard texts, listed in no order of merit. Our explanations, examples and diagrams are original; nothing is copied from these books.
- K. Terzaghi, R. B. Peck, G. Mesri, Soil Mechanics in Engineering Practice, John Wiley and Sons, 3rd edition, 1996.
- R. F. Craig, Craig's Soil Mechanics, Spon Press, 8th edition, 2012.
- B. M. Das, Principles of Geotechnical Engineering, Cengage Learning, 9th edition, 2017.
- IS 2720 (Part 4):1985 (reaffirmed 2006), Methods of test for soils — Grain size analysis, cited by clause only. IS 1498:1970 Table 3, cited by clause and table number only.