Rigid Pavement Design — Westergaard’s Theory & IRC:58

1. Rigid vs Flexible Pavement — Comparison

2. Modulus of Subgrade Reaction (k)

The modulus of subgrade reaction k represents the reactive pressure per unit deflection of the subgrade/sub-base system — it is the spring constant of the foundation beneath the concrete slab (Winkler foundation model).

2.1 Plate Load Test (IS:1888)

The k value is determined from a plate load test using a 750 mm diameter rigid plate placed on the compacted subgrade/sub-base surface. Load is applied in increments and the settlement (deflection) is recorded. The k value is computed from the pressure at 1.25 mm deflection:

2.2 k Value on Prepared Sub-base

The k value used in slab design is determined at the top of the sub-base (not on the natural subgrade), because the sub-base provides additional support. IRC:58 provides correction factors or equivalent k values based on the subgrade CBR and sub-base type/thickness.

DLC = Dry Lean Concrete (M10 grade used as sub-base below rigid pavement)

3. Radius of Relative Stiffness (l)

The radius of relative stiffness l characterises how the concrete slab and the subgrade interact — it is the length over which the slab transfers load to the subgrade. A large l means the slab is stiff relative to the subgrade (load spreads widely); a small l means the foundation is stiff relative to the slab.

3.1 Physical Significance of l

• l represents the effective slab length over which the load influence diminishes to near zero.
• For slab dimensions (length L and width B) much larger than l: slab behaves as an infinite plate (Westergaard’s formulas apply directly).
• For slab dimensions comparable to l: finite slab effects become significant (corner stresses governed by slab size).
• Slab dimension: typical IRC:58 slab is 4.5 m × 3.5 m; l ≈ 0.6–0.9 m → L/l ≈ 5–7 (semi-infinite behaviour).

4. Westergaard’s Stress Equations

H.M. Westergaard (1926) derived closed-form solutions for the maximum tensile stress in a concrete slab on a Winkler (spring) foundation for three load positions: interior, edge, and corner. These remain the foundation of rigid pavement design worldwide.

4.1 Load Assumption

A single tyre contact area is approximated as a circular load of radius a:

4.2 Westergaard’s Interior Loading Stress (σ_i)

Load applied far from any edge — fully surrounded by slab on all sides:

4.3 Westergaard’s Edge Loading Stress (σ_e)

Load applied at the slab edge — one side of the load is at the free edge:

4.4 Westergaard’s Corner Loading Stress (σ_c)

Load applied at or near the slab corner — two free edges are nearby:

4.5 Comparison of Westergaard’s Stresses

For a typical highway slab (a = 0.15 m, l = 0.75 m, h = 0.25 m), approximate relative stresses: σ_e ≈ 1.5–2.0 MPa; σ_i ≈ 0.8–1.2 MPa; σ_c ≈ 1.5–2.5 MPa.

5. Critical Load Position — Edge Loading

IRC:58 adopts edge loading as the critical (design) condition for the following reasons:

• Edge loading produces a higher stress than interior loading for the same slab thickness and subgrade conditions (σ_e ≈ 1.8 × σ_i).
• Vehicles frequently travel close to the pavement edge — especially trucks which tend to track near the lane edge for operational reasons.
• The concrete slab edge has no lateral support (no adjacent slab or shoulder), whereas interior loading benefits from slab continuity on all sides.
• Tied concrete shoulders (TCS) or widened outer lanes can convert edge loading to interior loading — IRC:58 allows a reduction in design thickness when TCS is provided.

Corner loading is separately addressed through dowel bar design and corner reinforcement — the Westergaard corner stress formula is used to check whether dowels are required and their spacing.

6. Temperature Warping Stress — Bradbury’s Analysis

Concrete slabs warp due to temperature differentials between the top and bottom surfaces. During the day, the top surface is hotter than the bottom (curling upward — ends lift, centre depresses). At night, the top is cooler (curling downward — ends press down, centre lifts). These warpings are restrained by slab self-weight and subgrade friction, inducing additional stresses.

6.1 Temperature Stress Formula (Bradbury)

6.2 Bradbury’s Warping Stress Coefficient C

C depends on the ratio L/l (slab length to radius of relative stiffness) and must be read from Bradbury’s chart. Key values from the chart:

For large slabs (L/l ≥ 10), C approaches 1.0 and σ_t = EαΔT/[2(1–μ)] — the maximum possible warping stress.

6.3 Design Temperature Differential (IRC:58)

7. Combined Stress and Design Criteria

7.1 Critical Stress Combination

The total design stress at the slab edge is the sum of the load-induced edge stress and the temperature warping stress. The most critical condition is daytime summer — edge loading from a heavy vehicle coincides with maximum positive temperature differential (top hotter than bottom, inducing tensile stress at the slab bottom under the wheel load):

7.2 Flexural Strength of Concrete (Modulus of Rupture)

8. IRC:58 Design Procedure

8.1 Design Inputs (IRC:58:2015)

• Design traffic: cumulative number of standard axles (CSA) over 30-year design life
• Subgrade CBR → modulus of subgrade reaction k (from IRC:58 table or plate load test)
• Sub-base type (granular or DLC) → effective k
• Concrete grade: M40 minimum for NH/SH; M35 for lower-category roads
• Climate zone: temperature differential ΔT from IRC:58 table
• Axle load spectrum (from weigh-in-motion data or IRC:58 default)

8.2 Design Steps

1. Determine effective k from subgrade CBR and sub-base details.
2. Assume a trial slab thickness h.
3. Compute radius of relative stiffness l = [Eh³/(12(1–μ²)k)]^1/4.
4. Compute load-induced edge stress σ_e for each axle load category using Westergaard’s edge formula.
5. Compute temperature warping stress σ_t using Bradbury’s formula with ΔT for the zone.
6. For each load group, compute stress ratio SR = (σ_e + σ_t)/f_r.
7. Determine allowable fatigue repetitions from Teller-Sutherland equation.
8. Compute cumulative fatigue damage CFD = Σ(n_i/N_i).
9. If CFD ≤ 1.0: the trial thickness h is adequate. If CFD > 1.0: increase h and repeat.

8.3 IRC:58 Design Catalogue (Indicative Slab Thicknesses)

Values for M40 concrete with tied concrete shoulder (edge loading reduced). Without tied shoulder, add 10–20 mm.

9. Joint Design

Concrete slabs must be divided into panels by joints to control cracking due to temperature and moisture volume changes, shrinkage, and differential settlement.

9.1 Types of Joints in Rigid Pavement

9.2 Dowel Bars vs Tie Bars

10. Worked Examples (GATE CE Level)

Example 1 — Radius of Relative Stiffness (GATE CE 2022 type)

Problem: A concrete slab has the following properties: E = 30,000 MPa, h = 0.25 m, μ = 0.15. The modulus of subgrade reaction k = 50 MN/m³. Find the radius of relative stiffness l.

Answer: l = 0.945 m

Example 2 — Westergaard’s Edge Stress (GATE CE 2021 type)

Problem: Compute Westergaard’s edge loading stress for a concrete pavement slab with: wheel load P = 50 kN, slab thickness h = 0.25 m, radius of relative stiffness l = 0.80 m, and radius of contact a = 0.15 m. (Use h > 1.724a check; assume b = a for simplicity)

Answer: σ_e = 1.52 MPa

Example 3 — Temperature Warping Stress (GATE CE type)

Problem: A concrete slab (L = 4.5 m, B = 3.5 m) has l = 0.85 m. E = 30,000 MPa, α = 10 × 10⁻⁶/°C, μ = 0.15, ΔT = 16°C. Find Bradbury’s C for L/l and B/l, and compute the temperature warping stress.

Answer: C = 0.78; σ_t = 2.20 MPa

Example 4 — Combined Stress and Flexural Strength Check (GATE CE type)

Problem: For the slab in Example 2 and 3, check whether the design is adequate if concrete is M40 (f_r = 4.43 MPa). Total load stress σ_e = 1.52 MPa; temperature warping stress σ_t = 2.20 MPa.

Answer: SR = 0.84; only 37 repetitions allowed — design is NOT adequate. Increase h or concrete grade.

Example 5 — Corner Stress (GATE CE type)

Problem: Find Westergaard’s corner loading stress for: P = 40 kN, a = 0.15 m, l = 0.85 m, h = 0.25 m.

Answer: σ_c = 1.085 MPa (less than edge stress of 1.52 MPa — edge governs)

Example 6 — Flexural Strength from Concrete Grade (GATE MCQ type)

Problem: A rigid pavement is constructed with M45 concrete. Find the modulus of rupture (flexural strength) f_r and the maximum permissible stress ratio for unlimited fatigue life.

Answer: f_r = 4.70 MPa; unlimited fatigue life when design stress ≤ 2.12 MPa (SR ≤ 0.45)

11. Common Mistakes

Mistake 1 — Using the Interior Stress Formula for Design Instead of Edge Stress

Error: Computing Westergaard’s interior stress σ_i and using it as the design stress, giving a thickness smaller than required.
Root Cause: Interior loading is intuitively the standard position for a wheel, but IRC:58 mandates edge loading as the design condition because it produces higher stress and wheels frequently operate at lane edges.
Fix: Always use σ_e (edge stress formula with 0.572 P/h²) for design. σ_i (interior, 0.316 P/h²) and σ_c (corner, 3P/h²) are used for checking specific conditions — corner stress is checked for dowel bar design.

Mistake 2 — Confusing k (Modulus of Subgrade Reaction) with CBR

Error: Using CBR value directly in the radius of relative stiffness formula l = [Eh³/(12(1–μ²)k)]^1/4 instead of converting CBR to k first.
Root Cause: Both CBR and k measure subgrade strength but in completely different ways — CBR is a penetration resistance ratio (dimensionless %), k is a pressure-per-unit-deflection (MN/m³). They are not interchangeable.
Fix: Convert CBR to k using IRC:58 Table 2 (or the table in this page) before using k in any formula. CBR is used for flexible pavement (IRC:37); k is used for rigid pavement (IRC:58).

Mistake 3 — Using the Wrong Formula for b (Equivalent Contact Radius)

Error: Always using b = a (without checking the 1.724h condition) or always using the full b = √(1.6a² + h²) – 0.675h formula regardless of the condition.
Root Cause: Westergaard derived two expressions for b depending on whether the contact radius a is less than or greater than 1.724h:
— If a < 1.724h: b = √(1.6a² + h²) – 0.675h
— If a ≥ 1.724h: b = a
The condition arises because for large-radius contact areas (a ≥ 1.724h), the Boussinesq correction for slab stiffness becomes negligible.
Fix: Always check: compute 1.724 × h; if a < this value, use the full formula for b. In most GATE CE problems with h = 0.25 m and a = 0.15 m: 1.724 × 0.25 = 0.431 > 0.15, so the full formula applies.

Mistake 4 — Not Accounting for Temperature Differential When Checking Design Adequacy

Error: Checking whether σ_e alone is within permissible limits without adding the temperature warping stress σ_t.
Root Cause: Temperature stresses in rigid pavements can be as large as or larger than traffic load stresses — in some cases σ_t > σ_e. Ignoring temperature stress gives a significantly unconservative design.
Fix: Total design stress = σ_e + σ_t (in summer daytime — the worst combination). IRC:58:2015 explicitly requires both stresses to be included in the fatigue damage analysis for each load group.

Mistake 5 — Confusing Dowel Bars with Tie Bars

Error: Specifying deformed (ribbed) bars at transverse contraction joints or smooth bars at longitudinal joints.
Root Cause: Both are called “bars” in pavement joints, and both are steel reinforcement — but their functions are opposite. Dowel bars at transverse joints must allow horizontal movement (expansion/contraction) while transferring vertical shear — so they must be smooth to allow slip. Tie bars at longitudinal joints must prevent lateral separation between lanes — so they must be deformed to bond to both slabs.
Fix: Transverse joints (expansion, contraction, construction) → smooth dowel bars (plain round bars). Longitudinal joints → deformed tie bars (ribbed bars, bonded to both slabs).

12. Frequently Asked Questions

Q1. Why is edge loading critical for rigid pavement design while interior loading governs flexible pavement?

In flexible pavement, load is transferred through granular layers at all positions — the layer geometry is the same whether a wheel is at the edge or in the centre of the lane. The critical design location is the bottom of the bituminous layer at the maximum tensile strain, which is approximately beneath the wheel regardless of lateral position. In rigid pavement, the concrete slab acts as a plate on a spring foundation. The moment (bending stress) induced in the slab depends on edge and boundary conditions — at the slab edge, the lack of lateral support from adjacent slab material means the bending moment is not shared with the surrounding plate area. The edge loading produces a cantilever-like stress concentration. The ratio σ_e/σ_i ≈ 0.572/0.316 ≈ 1.81 — edge loading produces 81% more bending stress than interior loading. Additionally, tyres often travel close to the edge of the outer lane (within 0.5 m of the edge), making edge loading a realistic and frequent load case. When a tied concrete shoulder is constructed adjacent to the traffic lane, it provides lateral support and effectively converts edge loading to interior loading — this is why IRC:58 allows a reduction in slab thickness (about 10–15%) when TCS is provided.

Q2. What is the physical interpretation of the radius of relative stiffness l?

The radius of relative stiffness l = [Eh³/(12(1–μ²)k)]^1/4 represents the characteristic length over which the concrete slab deflects under a concentrated load before the deflection becomes negligible. Physically, it is the distance from the load point to where the slab deflection reduces to approximately 1/e (37%) of the maximum deflection — analogous to a characteristic length in beam-on-elastic-foundation theory. A large l means the slab is stiff relative to the foundation (either very thick slab, high E, or low k) — the load influence spreads over a wide area, stresses are lower, but deflections are higher. A small l means the foundation is stiff relative to the slab — load influence is concentrated near the load point, stresses are higher. The practical significance for design is that all Westergaard’s stress formulas are expressed in terms of l/a or l/b ratios — the slab’s stress response depends on how the load contact area (a or b) compares to the slab’s natural stiffness scale l.

Q3. Why do contraction joints form at specific locations while expansion joints are placed only at structures?

When concrete cures and subsequently cools during cold nights, it contracts and tends to crack. If the crack location is not controlled, cracks form randomly — which looks unsightly and is harder to seal and maintain than neat joints. Contraction joints are saw-cut to a depth of 1/3 to 1/4 of the slab thickness within 12–24 hours of concrete placement (before full strength develops). These weakened planes are where cracks preferentially form, controlling the crack location to a clean, straight line that can be sealed with bitumen sealant. Contraction joints are spaced at 4.5–6.0 m because this is approximately the spacing at which uncontrolled cracking would naturally occur for typical slab dimensions and thermal conditions. Expansion joints are needed only at structural discontinuities (bridges, culverts, underpasses) where differential movement between the rigid slab and the structure requires accommodation of large expansion — the gap allows the slab to expand without developing buckling forces. On long continuous concrete pavement sections without structures, thermal expansion is accommodated by the small gaps at multiple closely-spaced contraction joints rather than requiring separate full expansion joints.

Q4. How does IRC:58 (2015) differ from the older Westergaard method in practice?

The classical Westergaard approach (used in older IRC:58 versions) compared the maximum combined stress (σ_e + σ_t) against the permissible stress (f_r/safety factor) for a single design axle load — essentially an elastic design approach. The 2015 revision adopts a fatigue damage approach: instead of checking one critical load, the designer analyses the entire axle load spectrum (from weigh-in-motion surveys or default IRC:58 data), computes the stress and stress ratio for each axle load group, determines allowable fatigue repetitions from the Teller-Sutherland equation, and sums the fatigue damage fractions across all axle groups. The design is adequate when cumulative fatigue damage CFD = Σ(n_i/N_i) ≤ 1.0 (Miner’s hypothesis). This approach explicitly accounts for the fact that not all passes are at the maximum design axle load — lighter loads cause less fatigue damage than heavy loads, and this credit should be taken. In practice, the 2015 IRC:58 method typically gives slab thicknesses 20–30 mm thinner than the older prescriptive method for the same design conditions, while achieving a more reliable design life.