Creep and Shrinkage in Concrete: Long-Term Deformation in Structural Design

Elastic analysis predicts how a concrete structure deflects the instant a load is applied. It cannot predict what happens over the following months and years. Two time-dependent deformation mechanisms — creep and shrinkage — continue to shorten, sag, and redistribute forces in concrete structures long after construction ends. For most reinforced concrete buildings, long-term deflections exceed the instantaneous elastic deflections by a factor of two to three. For prestressed concrete, creep and shrinkage prestress losses can consume 15 to 25 percent of the initial jacking force. Ignoring these effects leads to cracked finishes, ponded roofs, and misaligned facades.

Creep: Mechanism and Magnitude

Creep is the increase in strain under sustained stress at constant temperature and moisture. It originates in the cement paste, where water molecules in the gel pores migrate and redistribute under sustained compressive load. The aggregate skeleton does not creep; it restrains paste deformation, which is why higher paste-to-aggregate ratios produce more creep.

The creep strain ε_cr is expressed as a multiple of the initial elastic strain ε_i through the creep coefficient φ:

ε_cr(t, t₀) = φ(t, t₀) × ε_i

where t is the current age and t₀ is the age at loading. The ultimate creep coefficient φ_u ranges from about 1.2 to 3.0 for typical concrete, depending on the factors below. ACI 209R and the Eurocode 2 B3 model both express φ as the product of a time-development function and a set of correction factors.

Factor	Effect on creep	Typical range of influence
Age at loading t0	Earlier loading increases creep	Loading at 3 days vs. 28 days can double φu
Relative humidity	Drier environment increases creep	RH 40% vs. 80%: roughly 50% more creep
Volume-to-surface ratio	Thicker members creep less (slower drying)	V/S ratio doubles: 15–25% creep reduction
Concrete strength	Higher f′c reduces creep	f′c 3000 vs. 6000 psi: 30–40% reduction
Stress-to-strength ratio	Creep is linear below ~0.5f′c	Above 0.6f′c, nonlinear creep accelerates

Shrinkage: Mechanisms and Types

Shrinkage is volume reduction unrelated to applied load. Three types matter for structural design:

• Drying shrinkage: Moisture loss from exposed surfaces causes the paste to contract. This is the dominant type in most structures, developing over months to years. The ultimate drying shrinkage strain ε_sh,u typically falls between 400 and 800 microstrain for normal-weight concrete.
• Autogenous shrinkage: Chemical hydration consumes water, reducing the overall volume of cement paste without any moisture loss to the environment. Autogenous shrinkage is negligible for water-cement ratios above 0.42 but becomes significant for low-w/c high-strength mixes (< 0.35) where it can reach 100 to 200 microstrain.
• Plastic shrinkage: Rapid surface drying before the concrete sets produces surface cracking within the first few hours. This is a construction-phase issue managed by curing and windbreaks rather than a structural analysis problem.

Differential shrinkage between a new concrete pour and an older adjacent section — such as in composite deck slabs cast on older concrete beams — induces parasitic stresses that must be checked against the tensile capacity of the interface.

Structural Consequences

The structural implications of creep and shrinkage extend beyond simple additional deflection:

Deflection amplification: ACI 318 Section 24.2.4 uses a multiplier λ_Δ = ξ/(1 + 50ρ′) applied to the immediate live-load deflection to estimate additional long-term deflections, where ξ is a time-dependent factor (peaking at 2.0 after five years or more) and ρ′ is the compression reinforcement ratio. Compression steel resists creep shortening, explaining why doubly-reinforced beams deflect significantly less over time than singly-reinforced sections.

Prestress loss: In prestressed concrete, creep shortening of the member directly shortens the prestressing tendon, reducing the prestress force. Combined creep and shrinkage losses typically total 30 to 60 ksi for low-relaxation strand, requiring that initial prestress be elevated to compensate. Designers use the AASHTO LRFD refined method or PCI Design Handbook estimates to quantify these losses at transfer, at service, and at 50-year service life.

Redistribution in indeterminate structures: In continuous beams and frames, creep redistributes bending moments from regions of high initial stress toward regions of lower stress — a phenomenon called creep redistribution. This can be beneficial (reducing peak moments) or detrimental (increasing moments at already-cracked sections). Long-term analysis with age-adjusted effective modulus methods captures this redistribution.

Mitigating Long-Term Deformation

Practical measures to limit creep and shrinkage effects include: using higher-strength concrete (which has higher modulus and lower creep coefficient), delaying loading until concrete has gained substantial strength, maximizing compression reinforcement in zones of high sustained moment, providing adequate control joints to allow free shrinkage movement without restraint cracking, and using shrinkage-compensating cements or fibers for slabs-on-grade where uncontrolled cracking is unacceptable. Long-term monitoring of deflection-sensitive structures using embedded gauges or periodic survey allows early detection of creep rates that exceed design assumptions before serviceability is compromised.