Design of pavements has been based on experience and judgment in the past. But the design methodology has taken rapid strides towards rationality in the past few decades based on an understanding of the stresses and strains in the layers constituting the pavement and application of the concept of reliability in achieving the desired level of service during the pavement’s design life.
The methods of design are broadly classified as:
ii. Semi-empirical methods
iii. Analytical or mathematical methods.
iv. Mechanistic-empirical methods.
Some of the empirical and semi-empirical methods like the ‘group-index method’ (based on soil classification and traffic volume), California resistance value method (based on the results of Hveem’s stabilometer and cohesiometer tests), North Dakota core method (based on cone penetration test values), Mc Leod’s method (based on plate load test results), and even theoretical-empirical approaches like the triaxial test method and Burmister’s (two-layer) method have been practically superseded by more rational approaches based on long-term road tests on experimental stretches.
Hence, only the following methods will be presented here: 1. Group-Index Method 2. CBR Method (IRC: 37-1970 & 1984) 3. IRC Method [IRC: 37-2001 (Second Revision)] 4. IRC Guidelines for Low Traffic Volume Roads: (IRC: SP: 72-2007 and IRC: SP: 77-2008) 5. Boussinesq’s (Single Layer) Method 6. Burmister’s (Two-Layer) Method 7. Mechanistic-Empirical Method.
The group-index method was devised by the U.S. Highway Engineers (Highway Research Board) in 1945. The group-index of a soil, an arbitrary index assigned to different soil types based on the percent fines, liquid limit and plasticity index, is defined by the following equation –
GI = 0.2 a + 0.005 ac + 0.01 bd … (6.55)
The higher the group index, the weaker is the subgrade soil. The design curves are given in Fig. 7.5.
2. CBR Method (IRC: 37-1970 & 1984):
The California bearing ratio (CBR) value of the subgrade soil was the basis for the method of design of flexible pavements, developed originally by the California State Highway Department, and adopted by The Road Research Laboratory, London, for developing their own design procedure and design charts.
The advantage of the CBR method is that it can be used to find the total thickness of the pavement and that of the individual courses in addition to the thickness of the subgrade soil (provided the CBR-values of the materials of the courses are also known).
The thickness required could be got for known CBR-values from design curves for different wheel loads. The Indian Roads Congress has adopted this general procedure and developed design charts for the depth of construction versus the CBR-value for different traffic classifications based on the number of commercial vehicles per day.
The CBR-method is known to suffer from the following disadvantages:
(i) It is an arbitrary test which does not reflect the soil strength directly.
(ii) The moisture and soaking conditions are also arbitrary, and in arid zones, over-design.
(iii) The curves are applicable only under the conditions for which they were method is not reliable for high values of CBR.
The thickness of construction is given by a set of 7 curves, as shown in Fig. 7.6.
‘Traffic’ denotes the total number of vehicles in both directions (irrespective of whether the design is for a two-lane or a dual carriageway.) For estimation of future traffic, the growth rate is assumed as 7.5 per cent.
For single-lane roads, the traffic intensity is taken to be twice that for two-lane roads (because of the concentration of traffic on one lane only).
IRC Guidelines-revised in 1984 [IRC: 37-1970-Revised (1984)]:
The salient features of IRC guidelines revised in 1984 are:
1. New set of design curves relating the cumulative standard axles and the CBR value to the total pavement thickness.
2. Recommendations on the types of pavement materials suitable for various courses.
The method of computing the cumulative standard axles; the number of standard axles per commercial vehicle is designated as ‘vehicle damage factor (VDF)’.
The pavement design chart extrapolated up to 200 million standard axles is shown in Fig. 7.7.
The total thickness obtained from the chart is distributed into surface course of thickness x, base of thickness y, and sub-base of thickness z, as shown in Table 7.3.
DBM-Dense bituminous macadam
SDBC-Semi-dense Bituminous carpet
This is based on the revised guidelines incorporated in “IRC: 37-2001 – Guidelines for the Design of Flexible Pavements”. The design catalogue is based on the results of the Ministry of Surface Transport (MOST) study: “Analytical Design Approach for Flexible Pavements” implemented by IIT Kharagpur. The software PAVE was developed based on mechanistic -empirical (M-E) principles.
Salient features of the revised guidelines are:
(i) The flexible pavement is designed as a four- layer structure – bituminous surfacing consisting of two courses -wearing course and binder course, granular base and granular sub-base over the soil subgrade.
(ii) Soils having CBR-values in the range of 2 to 10% are covered.
(iii) Traffic volumes up to 150 million standard axles (msa) are considered in two ranges, 1 to 10 msa and 10 to 150 msa.
(iv) Pavement failure criteria or the critical strains considered are:
a. Vertical compressive strain on top of the subgrade under the centre of the dual wheels is correlated with ‘rutting’ or the vertical permanent strain of the pavement, the allowable rut depth being 20 mm.
b. Horizontal tensile strain at the bottom of the bituminous binder course under the centre of a wheel of the dual wheels is correlated with ‘fatigue failure’ or cracking, with the limit of fatigue being 20% cracked surface area.
c. In addition to these, vertical compressive stresses and strains within the bituminous course, granular base, and granular sub-base courses are also considered.
(v) The following empirical relationship are used to estimate the elastic modules of subgrade
Where, tg = thickness of granular layer in mm.
(vii) In case of arid regions where the annual rainfall is 500 mm or less, instead of a 4-day soaked CBR-value, the design subgrade CBR-value may be determined at the field moisture content prevailing immediately after the monsoon season.
(viii) Minimum subgrade CBR-value should be 2%; otherwise, the subgrade should be strengthened with a capping layer of at least 150 mm thickness having a CBR-value of 10%.
(ix) The granular sub-base layer should have a minimum CBR-value of 20% for traffic up to 2 msa and 30% for traffic more than 2 msa.
(x) Granular sub-base layer thickness should not be less than 150 mm for traffic up to 10 msa and 200 mm for traffic more than 10msa.
(xi) If the road is to be built by stage construction, the sub-base thickness required for all the stages should be provided in the first stage itself.
(xii) If a premix carpet of thickness up to 25 mm is used as wearing course, its thickness should not be counted as a structural layer.
Similarly, certain ‘drainage considerations’ are also specified to avoid adverse effects of ground water. Further details may be found in “IRC: 37-2001”. Design charts to determine the total thickness of pavement for subgrade CBR-values from 2% to 10% for the two traffic volume ranges 1 to 10 msa and 10 to 150 msa are shown in Figs. 7.8 and 7.9, respectively.
The pavement design tables consisting of recommended pavement composition and the total thickness for a traffic range of 1 to 10 msa for each CBR-value of subgrade from 2% to 10% are contained in IRC: 37-2001 (interpolation for intermediate traffic values may be made if needed). The recommended compositions for CBR-values 2%, 3%, 4%, 5%, and 6% are presented in Tables 7.4-7.8.
Similarly, pavement design tables for the traffic range 10 to 150 msa for each CBR-value of subgrade from 2% to 10% are also given in IRC: 37-2001.
The recommended compositions for 2%, 3%, 4%, 5%, and 6% CBR-values for this traffic range are presented in Tables 7.9 – 7.13. For all other details, IRC: 37-2001 has to be referred.
The design tables for the CBR-values 7%, 8%, 9%, and 10% are available in IRC: 37-2001, and may be used if necessary in a particular design problem.
IRC has published revised guidelines for the design of flexible pavements in its latest version – “IRC: 37-2012: Guidelines for the design of Flexible Pavements (Third Revision), IRC, 2012.” Although the principles and methodology of design set out in these guidelines are basically the same as those in the previous version – IRC: 37-2001 (Second Revision), certain improvements have been made in the light of changes in traffic pattern involving tandem, tridem and multi-axle vehicles and heavier axle loads, and the advent of new form of construction and materials.
Conventional construction materials such as aggregates are becoming progressively scarce on account of environmental concerns; this has shifted focus to the use of local, recycled and engineered marginal aggregates in construction – for example, stone matrix asphalt, modified and foamed bitumen, and cementitious bases and sub-bases. Attention is focused on fatigue resistant bituminous mixes with high-viscosity binders for heavy traffic with a view to constructing high-performance long-life bituminous pavements.
These have been included in the revised guidelines based on the experience abroad and the few successful trials in India.
The guidelines may require revision from time to time in the light of future field experience and feedback from the organisations using these guidelines to the Indian Roads Congress for further revision.
The salient provisions of these guidelines are given below, particularly where these differ from those of the previous version:
1. CBR-values of subgrade from 3% to 15% are covered.
2. Traffic volumes from 2 to 30 msa are dealt with exactly as per IRC: 37-2001; traffic volumes above 30 msa and upto 150 msa are to be dealt with using the guidelines of IRC: 37-2012.
3. The cracking and rutting of the pavement have been restricted to 10% of the area for design traffic exceeding 30 million standard axles, the models being the same as those in IRC: 37-2001. (For traffic up to 30 msa, 80% reliability equations are used; 90% reliability equations are used thereafter.)
4. These revised guidelines include alternative materials such as cementitious and reclaimed asphalt materials for analysis using the software IITPAVE, a modified version of FPAVE for layered system analysis, developed by IIT Kharagpur.
5. A flexible pavement covered according to these guidelines consists of different layers from bottom to top as given below:
(i) Subgrade/Stabilised subgrade
(ii) Sub-base (cemented/unbound)
(iii) Base (cemented/unbound)
(iv) Aggregate interlayer for cemented base/stress-absorbent membrane inter layer (SAMI)
(v) Bituminous layer
6. Perpetual pavement is one which has a life of 50 years or more. If the tensile strain caused by the traffic in the bituminous layer is less than 70 micro strains, the endurance limit of the material, the bituminous layer never cracks (Asphalt Institute, MS-4, 7th edition, 2007). Similarly, if vertical subgrade strain is less than 200 micro strains, there will be a little rutting in the subgrade.
Different layers are so designed and constructed that only the surface layer is the sacrificial layer which is to be scrapped and replaced with a new layer from time to time. There is now enough evidence worldwide to show that deep strength bituminous pavements suffer damage only at the top and nowhere else.
7. Pavement Design Procedure:
(a) Using IITPAVE:
Any combination of traffic and pavement layer composition can be tried using IITPAVE. The designer will have full freedom in the choice of pavement materials and layer thickness. The traffic volume, number of layers, the thickness of individual layers and the layer properties are user specified inputs in the program, which gives strains at critical locations as outputs. The adequacy of design is checked by the program by comparing these strains with the allowable strains as predicted by the fatigue and rutting models built into the program. A satisfactory design is achieved through iterative process by varying layer thicknesses or, if necessary, by changing the layer materials.
(b) Using Pavement Design Catalogues:
Design catalogues are provided giving pavement compositions for various combinations of traffic, layer configuration and assumed material properties; such material properties must be validated by means of simple laboratory tests and used.
Five different combinations of traffic and material properties have been considered and pavement composition has been suggested in the form of design charts:
(i) Granular base and granular sub-base, (for CBR-values of 3, 4, 5, 6, 7, 8, 9, 10, and 15 percent)
(ii) Cementitious base and cementitious sub-base of aggregate interlayer for crack relief. Upper 100 mm of the cementitious sub-base is the drainage layer (for CBR-values – 3, 5, 10, and 15 percent)
(iii) Cementitious base and sub-base with stress absorbing membrane interlayers (SAMI) at the interface of base and the bituminous layer, (for CBR-values – 3, 5, 10, and 15 percent)
(iv) Foamed bitumen/bitumen emulsion-treated reclaimed asphalt pavement or fresh aggregates of over 250 mm cementitious sub-base (for CBR-values – 3, 5, 10, and 15 percent).
(v) Cementitious base and granular sub-base with crack-relief layer of aggregate overlay above the cementitious base (for CBR-values – 3, 5, 10, and 15 percent)
The appropriate design chart is used for the CBR of the subgrade, from which the traffic in msa, the pavement thickness, the layer thicknesses and composition are read off.
The allowable strains can be computed using the modulus values and the composition of the layers, and the predicted strains obtained from the software IITPAVE. These two values are compared with the allowable values for ensuring safety of the adopted section.
A typical design chart is shown in Fig. 7.10 for the first case of granular sub-base and granular base over subgrade (with bituminous surface course) for CBR 9 and 10% and traffic volumes of 2, 5, 10, 20, 30, 50, 100 and 150 msa-
I. Granular sub-base (GSB)
II. Granular base (GBASE)
III. Dense bituminous macadam (DBM)
IV. Bituminous concrete (BC)
(Semi-dense bituminous concrete (SDBC) up to 5 msa)
The IRC guidelines contained in “IRC: SP: 72-2007 – Guidelines for the design of flexible pavements for low volume rural roads, India Roads Congress, New Delhi, 2007” and “IRC: SP: 77-2008 – Manual for the design, construction and maintenance of gravel roads, Indian Roads Congress, New Delhi, 2008” are specifically directed to the design of low traffic volume rural roads with flexible pavements. These methods of design are based on the ‘serviceability concept’ adopted in the “AASHTO: Guide for Design of Pavement Structures (1993)”.
According to these guidelines, the present pavement performance and its condition is rated by a panel of experienced drivers, rather subjectively, on a scale of 0 to 5, 0 and 5 representing the extreme conditions ‘very poor’ and ‘very good’, respectively. This number is called the “present serviceability rating (PSR)”.
Pavement performance measured in terms of objective indicators – slope variance, cracking, patching and rutting. This idea of quantifying the present condition of pavement serviceability is known as the ‘serviceability concept’.
A regression equation was developed by AASHTO using AASHTO road test data for the ‘present serviceability index (PSI)’ with the following ratings-
PSI rating- 4-5 very good, 3-4 good, 2-3 fair, 1-2 poor, 0-1 very poor.
The numerical difference between the initial PSI for a newly constructed road and the final or terminal PSI (at failure) is termed ‘Loss of PSI (APSI)’. This is specified as 1.7 and 2.2 for high-volume and low-volume roads, respectively.
Salient features of ‘IRC: SP: 72-2007’ and ‘IRC: SP: 77-2008’ are given here briefly.
(i) Low-traffic rural roads with flexible pavements are designed as unpaved roads (gravel/aggregate surface roads) according to IRC: SP: 77-2008, or as paved roads (flexible pavements) designed according to IRC: SP: 72-2007.
(ii) For the design of very-low-traffic roads, the initial and terminal serviceability index values are considered to be 4.0 and 2.0 respectively. (The loss of PSI, APSI, is taken to be 2.0.) The rut depth should not exceed 50 mm under a 3 m-straight edge.
(iii) For the design of low -traffic flexible pavements (paved roads) (IRC: SP: 72-2007), the ‘structural number (SN) design concept of AASHTO (1993) at 50% reliability level pertaining to Indian climatic zones, is used.
The structural number, which ranges from 0 to 6, is an index of the overall strength of the pavement, derived from the stiffness of individual layers/courses.
(iv) For both these types of roads, the subgrade strength is classified into 5 groups based on the CBR-value:
S1: Very poor (CBR = 2%)
S2: Poor (CBR = 3% to 4%)
S3: Fair (CBR = 5% to 6%)
S4: Good (CBR = 7% to 9%)
S5: Very good (CBR = 10% to 15%)
(v) Traffic volume is categorised into 7 groups based on ‘equivalent single axial load (ESAL)’ applications:
Tl: 10, 000 to 30, 000
T2: 30,000 to 60, 000
T3: 60,000 to 1,00, 000
T4:1, 00, 000 to 2,00,000
T5: 2, 00, 000 to 3,00,000
T6: 3, 00, 000 to 6,00,000
T7: 6, 00, 000 to 10, 00, 000 (maximum up to 1 msa)
The traffic groups T1 to T3 are considered to be for very low traffic volume roads (IRC: SP: 77-2008), and the traffic groups T4 to T7 are considered for design of flexible pavements. (IRC: SP: 72-2007).
(vi) Seasonal variations in traffic volume on rural roads are considered to be in accordance with peak harvesting and lean seasons. Single phase harvesting seasons may also be considered for estimating traffic in arid zones in India.
(vii) For traffic volume estimation, the count survey must include details of unladen, laden and overloaded commercial vehicles (with laden weight greater than 3 tonnes), along with their axle load configuration.
Commercial vehicles are classified as ‘Heavy (HCV)’and ‘medium (MCV)’; the former class comprises heavy trucks, lorries and buses, while the latter includes medium heavy trucks, mini-buses, and agricultural tractor-trailers.
For design of new roads, traffic estimation may be based on traffic on a similar road in the vicinity and must include generated and diverted traffic.
Design traffic is expressed in terms of cumulative ESAL of 80 KN repetitions carried during the design period of 10 years with an annual traffic growth rate of 6%. In the absence of complete field data, the average annual daily traffic (AADT) after considering seasonal variations, can be determined from the formula (IRC: SP: 72-2007)-
T = Average number of all types of vehicles per day, during lean season
n = Number of times the traffic (no. of. vehicles/day) at a peak harvesting season increases more than the traffic estimated in a lean season,
t = Duration of a harvesting season in days.
Design traffic volume, Nf, (msa) can be obtained using Equations (7.1), (7.2), (7.4), and (7.6), AADT being taken as P in Eq. (7.1).
Vehicle damage factors for laden HCV, unladen HCV and overloaded HCV and similar categories of MCV can be obtained by applying the fourth-power damage rule knowing the complete details of the axles and axle loads of these vehicles, in relation to the standard axle load of 80 KN.
(viii) Subgrade strength in terms of CBR-value is the next crucial design parameter after design traffic volume. At least three samples per kilometre length must be tested. The effect of seasonal variations of the ground water table (GWT) must govern the soaking criteria for obtaining the CBR-value of the subgrade soil. CBR may be obtained at the ‘equilibrium moisture content’ value, the balance value between dry and wet seasons, rather than under soaking condition.
(ix) Thickness Design of Pavements:
For very low volume unpaved roads (IRC: SP: 77-2008), the first three categories of traffic – T1 to T3 are considered along with the subgrade strength categories – S1 to S5, and the corresponding design charts are used to determine the thickness of the wearing course, base course and sub-base courses.
For flexible pavements or paved roads (IRC: SP: 72-2007): The remaining categories of traffic (T4 to T7) are used along with all the five classes – SI to S5 – of subgrade strength. Black-topped roads should be designed for a minimum of 1 00 000 ASL applications, along with a base course of 150 mm minimum thickness for a design period of 10 years.
Suitable drainage arrangements should be made for all type of pavements. Selection of materials: Maximum utilisation of locally available materials should be aimed at for rural roads. Water-bound macadam (WBM) and wet-mix macadam (WMM) may be used depending on the needs of the traffic and availability of materials. The usage of secondary materials like industrial waste, steel slag, lime and fly-ash stabilised fine-grained soils may also be considered for economy.
This is a theoretical approach based on Boussinesq’s theory for calculating the stresses and strains in a homogenous, isotropic, semi-infinite, and elastic soil medium bounded by a horizontal surface, under the influence of a point load applied at the surface.
The vertical stress, σz, at a point at a radial distance, r, and a depth z below the surface, has been derived as
Where, E is the elastic modulus of soil.
The stress distribution under a flexible plate is uniform since the plate bends uniformly under loading. Similar to a rubber sheet, therefore, the deflection under the plate is non-uniform.
If v is taken as equal to 0.5,
Eq. (7.12) can be used for the design of a flexible pavement by limiting the value of A, the deformation, to a desired value.
(q can be evaluated and its distributed value at the bottom of the course can be limited to the bearing power of the subgrade.)
Donald Martin Burmister (1943) of Columbia University, New York (USA) developed a method of analysis for a two-layered soil system, similar to a flexible pavement having its top layer stiffer than the bottom layer. Later, extended the theory to a three-layered system.
Many researchers have developed different analysis for multi-layered systems based on elastic, non-linear, visco-elastic and finite element methods. These solutions involve complex mathematical operations, which could be dealt with by the use of computers. Burmister’s two- layered system.
The assumptions are:
1. Each layer is elastic, homogenous and isotropic, with unique values of modulus of elasticity and Poisson’s ratio.
2. The elastic modulus of the top layer, E1 is greater than that of the bottom one, E2
3. The layers extend infinitely in the horizontal direction.
4. The interface is rough.
5. The applied stress is uniform over a circular area.
6. The top layer is of finite thickness, h, while the bottom layer is infinitely thick.
The system is shown in Fig. 7.11.
By introducing additional pavement layers over the subgrade, the vertical stress induced on it can be significantly reduced (to the extent of 30% to 68%). For this to happen, the thickness of the additional layer has to be equal to the radius of the loaded area and its elastic modulus should be approximately ten times that of the subgrade; also, the Poisson’s ratio has to be nearly 0.5.
For this reason, multiple layers are preferred to a single thick layer in the construction of flexible pavements. For multi-layered systems, several software programs have been developed for linear elastic, non-linear, and visco-elastic analysis using finite element method, which give accurate results. However, the computational effort required is relatively very high.
Theoretically calculated pavement response is compared to the allowable maximum values of stresses, strains, or deformations. The theoretical calculations require the principles of mechanics to be applied for given loadings, material characteristics and environmental conditions. The common deformation criteria used are rutting, cracking, patching and fatigue for the comparison of the theoretical and experimentally determined values from long-term road Tests on test stretches.
These results are incorporated in the design tables and charts.
This method, devised by the AASHTO, has been suitably modified in the IRC Guidelines for the Design of Flexible Pavements, already given (IS: 37-2001).