Flexible Pavement Design — CBR Method & IRC:37

1. Flexible Pavement — Structure & Layer Functions

A flexible pavement is a multi-layered structure in which each layer distributes traffic loads over a wider area as depth increases, eventually reaching the subgrade at stress levels the soil can safely bear. The term “flexible” refers to the pavement’s ability to deflect slightly under traffic load without fracturing — unlike rigid (concrete) pavements which resist load through beam action.

1.1 Layer Structure (top to bottom)

1.2 Load Transfer Mechanism

A standard axle load (80 kN) is applied by dual tyres on the road surface over a contact area of approximately 0.575 m diameter (for each dual tyre assembly). This localised load is distributed at an angle (approximately 1:1 horizontal to vertical) as it passes through each layer, so that at the subgrade level the stress is a fraction of the applied contact pressure. The design objective is to ensure this residual stress does not exceed the subgrade’s bearing capacity (related to its CBR).

2. California Bearing Ratio (CBR) Test

The CBR test (IS:2720 Part 16) measures the penetration resistance of a compacted soil specimen — it is the empirical index of subgrade strength used in IRC:37 pavement design.

2.1 Test Procedure

1. Prepare a soil specimen at 95% of Modified Proctor Maximum Dry Density (MDD) in a cylindrical mould (150 mm diameter, 175 mm height).
2. Soak the specimen in water for 96 hours (4 days) with surcharge loads on top (simulating overburden of pavement layers above subgrade).
3. After soaking, measure the swell and then apply a 50 mm diameter piston at a rate of 1.25 mm/min.
4. Record load at 2.5 mm and 5.0 mm penetrations.

2.2 CBR Calculation

2.3 Design CBR vs Field CBR

For pavement design, the soaked CBR at 95% MDD is used — this represents the worst-case condition (subgrade fully saturated). This is called the design CBR. It is typically lower than the in-situ (field) CBR of dry or partially saturated soil. Using soaked CBR ensures the design is conservative and the pavement performs adequately even during the monsoon season when subgrade is weakest.

3. Design Traffic — CVPD, VDF, and MSA

Pavement damage is caused almost entirely by heavy commercial vehicles — passenger cars cause negligible damage. The design traffic is expressed in terms of equivalent standard axle passes over the design life.

3.1 Standard Axle Load

3.2 Vehicle Damage Factor (VDF)

The VDF (also called LEF — Load Equivalency Factor) converts the actual mixed axle load spectrum of traffic to an equivalent number of standard axle passes. It is derived from the fourth-power law (empirical finding from AASHO road test):

3.3 Design Traffic (N) in Million Standard Axles (MSA)

3.4 Traffic Growth Rate

IRC:37 recommends using 7.5% annual growth rate for design when project-specific traffic studies are not available. For industrial and urban highways, higher rates (8–10%) may be used.

4. IRC:37 Design Method

4.1 Design Inputs

• Design subgrade CBR — from soaked CBR test at 95% MDD (design value = representative CBR for the project)
• Design traffic N — in MSA over the design life
• Layer material specifications — elastic modulus values from IRC:37 material catalogue

4.2 IRC:37 (2018) — Mechanistic-Empirical Approach

The 2018 revision of IRC:37 moved from the purely empirical CBR-chart method to a mechanistic-empirical (ME) approach. The design uses:

• Elastic layer theory (IITPAVE software) to compute critical strains: horizontal tensile strain at bottom of bituminous layer (ε_t) and vertical compressive strain at top of subgrade (ε_v).
• Empirical transfer functions to convert strains to allowable load repetitions before fatigue cracking and rutting.

4.3 Design Criteria (IRC:37:2018)

4.4 CBR-Based Pavement Thickness (Empirical Catalogue, for GATE CE)

For GATE CE purposes, the older IRC:37 (2001 / empirical) method is still widely used in questions. The total pavement thickness is read from design charts based on subgrade CBR and design traffic:

Values in mm. These are total pavement thicknesses (sub-base + base + bituminous layers) for the given CBR and design traffic.

4.5 IRC:37 (2001) — Empirical CBR Formula

5. Layer Materials & Thickness Ranges

5.1 Bitumen Grades Used in India (VG — Viscosity Graded)

Note: VG grades replaced the older penetration grades (80/100, 60/70) in India post-2006 (IS 73:2013).

6. Pavement Distress & Failure Modes

7. Overlay Design — Benkelman Beam Method

When an existing flexible pavement has deteriorated beyond acceptable levels but has not completely failed, a bituminous overlay can be applied to restore structural adequacy. The Benkelman Beam method (IRC:81) is the standard IRC method for overlay thickness design on existing bituminous roads.

7.1 Benkelman Beam Test

A standard Benkelman beam (3 m long, pivoting at 2:1 ratio) measures the rebound deflection of the pavement surface under a standard rear axle load of 80 kN (at each rear dual tyre position). The measurement gives the maximum surface deflection under load.

8. Worked Examples (GATE CE Level)

Example 1 — CBR Calculation from Test Data (GATE CE type)

Problem: A CBR test on a soaked subgrade sample gives the following loads: at 2.5 mm penetration, test load = 4.5 kN; at 5.0 mm penetration, test load = 6.2 kN. Compute the CBR value for design.

Answer: Design CBR = 33.5%

Example 2 — Design Traffic in MSA (GATE CE 2022 type)

Problem: An NH carries 3000 commercial vehicles per day (CVPD) in the year of opening. Annual traffic growth rate = 7.5%. Design life = 15 years. VDF = 4.5. The road is a two-lane undivided highway (LDF = 0.75). Find the design traffic in MSA.

Answer: Design traffic N = 97 MSA

Example 3 — Pavement Thickness from IRC:37 Chart (GATE CE type)

Problem: A new State Highway is to be designed for 20 MSA design traffic. Subgrade CBR = 5%. Determine the total pavement thickness and a suitable layer arrangement.

Answer: Total pavement thickness = 635 mm (WC: 40 + DBM: 60 + WMM: 250 + GSB: 285 mm)

Example 4 — Vehicle Damage Factor (LEF) Calculation (GATE CE type)

Problem: A traffic count shows 60% of commercial vehicles have an axle load of 100 kN and 40% have 60 kN. Compute the average VDF (use 4th power law; standard axle = 80 kN).

Answer: VDF = 1.59. The 100 kN axles dominate damage — despite being 60% of traffic, they contribute 92% of the damage.

Example 5 — Comparison of Pavement Thickness at Different CBRs (GATE MCQ type)

Problem: For a design traffic of 10 MSA, compare the total pavement thickness required for subgrade CBR values of 3%, 5%, and 10%. Comment on the sensitivity of pavement design to subgrade strength.

Answer: Thicknesses = 710 mm (CBR 3%), 565 mm (CBR 5%), 390 mm (CBR 10%) — strong sensitivity to subgrade CBR.

Example 6 — Traffic After n Years (GATE CE type)

Problem: A highway carries 1500 CVPD at opening. Growth rate = 8% per year. Find (a) CVPD at end of 10th year, (b) cumulative traffic over 10 years (no VDF/LDF adjustments — express in terms of vehicle passes).

Answer: CVPD at year 10 = 3238; Cumulative CVs = 7.93 million over 10 years

9. Common Mistakes

Mistake 1 — Using Field CBR Instead of Soaked CBR for Design

Error: Using the in-situ field CBR (tested without soaking) for pavement thickness design, giving a higher CBR and hence thinner (unconservative) pavement.
Root Cause: The field CBR at the time of testing may be under dry or partially saturated conditions — which will be the worst case only for a limited period. The soaked CBR represents the worst-case condition (fully saturated subgrade during monsoon), which IRC:37 mandates for design.
Fix: Always use soaked CBR at 95% Modified Proctor MDD for pavement design. If field CBR is given without soaking conditions being specified, assume it is the design CBR only if the problem explicitly states it is the soaked value.

Mistake 2 — Forgetting the Lane Distribution Factor (LDF)

Error: Using total CVPD across all lanes directly without applying LDF, overestimating traffic in the design lane.
Root Cause: Commercial vehicles distribute across lanes, but the design pavement only needs to carry the traffic in the most heavily loaded lane. On a two-lane undivided road, only 75% of CVs use any one lane on average.
Fix: Always apply LDF before computing MSA: single lane = 1.0; two-lane undivided = 0.75; four-lane divided = 0.40 (each direction carries 50%, and only 0.80 of that uses one lane). LDF is applied to the CVPD in one direction or across the section — check IRC:37 context carefully.

Mistake 3 — Applying the 4th Power Law with Load in Tonnes Instead of kN

Error: Computing LEF = (W tonnes / 8.16 tonnes)⁴ instead of (W kN / 80 kN)⁴, getting the wrong VDF.
Root Cause: The standard axle is defined as 80 kN = 8.16 tonnes. Both are equivalent but must be used consistently. If loads are given in tonnes, use 8.16 tonnes as reference; if in kN, use 80 kN. Mixing gives errors.
Fix: LEF = (W / W_std)⁴. Units of W and W_std must match. Standard: 80 kN or 8.16 t. Example: 10-tonne axle → LEF = (10/8.16)⁴ = (1.225)⁴ = 2.248.

Mistake 4 — Confusing Total Pavement Thickness with Bituminous Layer Thickness

Error: Treating the IRC:37 design chart value (e.g., 635 mm) as the thickness of bituminous layers only, ignoring the granular sub-base and base.
Root Cause: The IRC:37 chart gives the total pavement thickness including all layers above the subgrade — granular sub-base, granular base, and all bituminous layers combined. The bituminous layers alone might be only 100–150 mm.
Fix: Total pavement thickness (from chart) = GSB + WMM/Base + DBM/Binder + BC/Wearing. Always split the total into component layers per IRC:37 layer design guidelines. Bituminous layers (BC + DBM) typically account for 100–200 mm; the rest is granular.

Mistake 5 — Using Penetration Grade Bitumen Values Instead of Viscosity Grade (VG)

Error: Specifying 60/70 or 80/100 penetration grade bitumen in a design problem for Indian highways after 2013.
Root Cause: India switched from penetration grading to viscosity grading (IS 73:2013), replacing 60/70 with VG-30 and 80/100 with VG-10/VG-20. Older textbooks and question papers still use penetration grades.
Fix: For GATE CE: use VG grades (VG-10, VG-20, VG-30, VG-40). VG-30 is the most common for Indian conditions; VG-40 for heavy traffic/high temperature zones. VG-30 ≈ old 60/70 pen; VG-20 ≈ old 80/100 pen (approximate equivalence).

10. Frequently Asked Questions

Q1. Why does doubling the axle load cause 16 times more pavement damage instead of 2 times?

The fourth-power law (LEF = (W/W_std)⁴) was derived empirically from the AASHO Road Test (1958–1960, USA), where thousands of axle load applications were applied to test sections and the number of passes to failure (rutting, cracking) was counted for each axle load level. The test found that pavement damage increases with approximately the fourth power of the axle load ratio. The physical reason involves the non-linear nature of both fatigue damage in bituminous materials (which follows a power law with tensile strain) and rutting accumulation in subgrade (power law with compressive strain) — and since strain is roughly proportional to load, the compounding of two power-law relationships gives an overall fourth-power dependence. The practical implication is profound: an axle carrying 160 kN (twice the standard 80 kN) causes 2⁴ = 16 times more damage. A single heavily overloaded truck can cause as much damage as dozens of legally loaded trucks. This is why India’s Motor Vehicles Act specifies strict axle load limits and why overloading is such a severe problem for road infrastructure — it destroys pavements designed for legal loads in a fraction of the intended design life.

Q2. What is the difference between the IRC:37 (2001) empirical method and the IRC:37 (2018) mechanistic-empirical method?

The 2001 version of IRC:37 was purely empirical — it provided design charts (total pavement thickness vs CBR and MSA) developed from observed performance of Indian roads and calibrated against AASHO data. The designer simply read off the total thickness and split it into layers using standard proportions. The advantage was simplicity; the disadvantage was that it could not account for different material properties, temperature zones, or new materials like modified bitumen and recycled aggregates. The 2018 revision introduced the mechanistic-empirical approach: the pavement structure is modelled as an elastic multi-layer system (using IITPAVE software), critical strains are computed at the bottom of the bituminous layer (fatigue criterion) and top of the subgrade (rutting criterion), and empirically calibrated transfer functions convert these strains to allowable traffic repetitions. This method can handle any combination of materials, layer thicknesses, and climatic conditions. For GATE CE, the older empirical chart approach (reading thickness from CBR vs MSA charts) is still the primary tool for numerical problems, but questions about the mechanistic criteria (ε_t, ε_v, and the fatigue/rutting equations) are increasingly appearing.

Q3. How is the design CBR determined when the subgrade soil is non-uniform along the project?

When the subgrade varies significantly along a highway project (e.g., alternating clay and sand sections), a single design CBR cannot represent the entire road. IRC:37 recommends a statistical approach: collect CBR tests at regular intervals (typically every 500 m or more frequently in variable terrain), compute the mean and standard deviation, and use the characteristic CBR = mean – 1 standard deviation for design (equivalent to the 16th percentile — i.e., 84% of the subgrade is at or above this CBR). This is more conservative than using the mean and avoids designing for the absolute worst-case point, which would result in an over-designed pavement for most of the project. In practice, road sections with very different CBRs (e.g., CBR 2% sections through swampy ground vs CBR 10% sections through rocky areas) are designed separately with different pavement thicknesses — a technique called “pavement zoning” or “variable cross-section design”.

Q4. What is the role of the granular sub-base (GSB) in flexible pavement design, and why is its drainage function important in India?

The GSB serves three structural functions in flexible pavement: (1) it distributes the traffic load from the granular base course over a wider area of subgrade, reducing the stress on the subgrade; (2) it acts as a separation layer preventing the migration of fine subgrade particles upward into the coarser base course (which would clog the granular base and reduce its drainage and structural performance); (3) critically in India, it provides a drainage layer that rapidly conducts water away from within the pavement structure. This drainage function is particularly important during the Indian monsoon, when subgrade soils can become fully saturated and their CBR drops to the soaked design value. A GSB with adequate permeability (typically ≥ 30 m/day for Type I GSB) and connected to edge drainage allows water entering through cracks in the surface to escape laterally rather than ponding at the base of the bituminous layers or on the subgrade surface. Inadequate drainage is one of the most common causes of premature pavement failure in India — roads that perform well in dry conditions deteriorate rapidly in the first monsoon after construction if the drainage system is poorly designed or blocked.