Concrete Creep & Shrinkage Simulator (Eurocode 2 / fib MC2010)

Prestressed concrete (PSC) bridges and buildings: In post-tensioned and pre-tensioned members the prestress in the strands keeps relaxing because of concrete creep and shrinkage. This is the famous "prestress loss". fib MC2010 gives the long-term φ and ε_sh used to estimate the remaining prestress as initial-prestress × (1 − loss-fraction). Total losses are typically 15-25%, and can exceed 30% on very long span bridges. Eurocode 2 explicitly requires this evaluation.

Long-span and balanced cantilever bridges: For long span bridges and balanced-cantilever construction with precast segments, creep and shrinkage cause mid-span settlement that must be predicted stage by stage, otherwise the deck alignment will not match. Construction monitoring pre-cambers the precast segments by the expected long-term deformation so the final geometry converges to the design line. A tool like this one, where the time axis can be moved freely, gives quick intuition for the stage-by-stage camber design.

High-rise column shortening: In RC, SRC and composite high-rises above 200 m, lower-floor columns shorten by tens of centimetres to a metre due to gravity load combined with creep and shrinkage. The Burj Khalifa (828 m) shortened by about 1 m after completion, and the Shanghai Tower (632 m) by about 400 mm. Designers must "build oversize" by compensation to allow for future shortening, and use differential-shrinkage measures so interior and perimeter columns do not move apart.

Mass concrete thermal and shrinkage cracking: Dams, bridge piers and underground walls with thick sections (h₀ > 500 mm) suffer thermal stress from hydration heat, followed by drying-shrinkage cracking at the surface. Designers pick low-heat (S type) cement to set α = −1 and rely on stress relaxation from creep. With the cement set to S and h₀ near 1000 mm in the tool, the slow time progression is clearly visible and a large early φ is what relieves the thermal stress.

The first and biggest misconception is to confuse creep and shrinkage as the same phenomenon. They are physically distinct. Creep is mainly the rearrangement of water inside the calcium-silicate-hydrate (C-S-H) gel under sustained stress, and is partially recovered when the load is removed. Shrinkage is the bulk volume change of the C-S-H gel itself due to hydration and drying, and is independent of stress. This is exactly why Eurocode 2 / fib MC2010 formulates the two separately. Mixing them up in a PSC loss calculation easily produces 50% errors. This tool deliberately exposes six stat cards — elastic, creep and shrinkage — to keep the two ideas distinct.

The second pitfall is to assume "the higher the applied stress, the larger the creep coefficient φ". In the linear creep theory used by Eurocode 2, as long as the stress level stays below σ / f_cm = 0.4, φ is independent of the stress and is given by the formula above. Above 0.4·f_cm you enter the non-linear creep regime and need an additional k_σ factor. This tool only implements the linear theory, so it is accurate at the default 8 MPa < 14 MPa (0.4 × 35) but understates creep when σ is pushed to 20 MPa. That is why the verdict warns when σ exceeds 0.4·f_cm.

The third trap is the idea that "only the t = ∞ creep coefficient matters". In practice PSC designers use different φ values for immediate losses (right after stressing), short-term losses (a few months) and long-term losses (decades). A typical PCa girder may be jacked up at t = 30 days when φ ≈ 0.5-1.0, then keep losing prestress as φ grows to about 2.0 at t = 10 yr and 2.3 at t = 100 yr. Each stage must be evaluated with its own φ, or construction tolerances and remaining prestress will be wrong. The evaluation age t is a free slider here precisely so multiple snapshots can be inspected.