Multi-Stage Deformation and Sequential Construction Analysis

30 concepts22 papers1 proposal

Imagine building a massive brick tower. If you design it assuming the entire structure manifests instantly, your calculations will only account for the final, uniform weight on the foundation. In reality, you build it brick by brick. As the left side goes up, the foundation settles unevenly; when the right side catches up, it must adapt to a base that has already shifted and deformed. The final state of the tower carries the physical memory of every intermediate step.

This reality highlights the core limitation of traditional "state design" in structural and geotechnical engineering. Materials and structures do not experience forces in a vacuum; they evolve through a highly path-dependent sequence of states. To accurately predict how a tunnel will settle, a bridge will flex, or a material will fail, engineers use multi-stage deformation modeling. This paradigm breaks down a continuous physical process into discrete, chronologically ordered phases to track stress redistribution and cumulative strain.

The Fallacy of the Static State

In complex engineering projects, evaluating a system based solely on its final configuration can lead to catastrophic underestimations of stress. During excavation or loading, the geometry and boundary conditions of the system change constantly.

Hongjie Gao et al. address this by contrasting traditional static methods with a "process design load" calculation. By modeling tunnel excavation as a series of sequential steps, they demonstrate that each excavation stage alters the load profile of adjacent, unexcavated areas. The final load experienced by the structure is not a simple static snapshot, but the cumulative sum of these transient stress waves.

[Static State Design]   Initial State ──────────────────────────────► Final State (Misses transient stress spikes)

[Process Design Load]   Stage 1 (Excavate) ──► Stage 2 (Support) ──► Stage 3 (Next Bench) ──► Final Equilibrium

Simulating the Construction Sequence

To capture these path-dependent physics computationally, numerical simulations rely on the sequential activation and deactivation of elements.

Consider the Pile-Beam-Arch (PBA) method used in subway station construction. The ground is not hollowed out all at once; instead, pilot tunnels are dug, piles are poured, arches are cast, and then the interior soil is removed. Yao Bai et al. break down this intricate sequence in finite element models, demonstrating that ground deformation is highly non-uniform. Surface settlement spikes during pilot tunnel excavation, while horizontal displacement peaks during arch installation. A static model completely flattens these critical variations.

This time-stepping methodology is executed by systematically altering the model's mesh over time:

  1. Element Deactivation: To simulate excavation, soil elements are removed from the active matrix, causing the surrounding soil to instantly seek a new equilibrium.
  2. Stress Redistribution: The model calculates how the remaining soil mass absorbs the load previously carried by the removed elements.
  3. Element Activation: Structural elements (like concrete linings or anchor cables) are activated in the newly deformed space to provide support before the next excavation step.

This precise sequence is what allows researchers to simulate complex operations like benching techniques—where the tunnel face is excavated in staggered tiers—and immediate primary support installation, as modeled by Chun-lai Chen et al.. It also underpins simulations of shield tunneling, where the model must account for the temporary physical gap (the tail void) left behind a moving tunnel boring machine before the concrete segments are pressurized into place, a phenomenon explored in the authors' work on urban road structures.

The Phasic Nature of Material Response

It is not just human construction schedules that progress in stages; the physical response of the earth and engineering materials follows distinct, predictable phases. When monitored continuously, deformation curves reveal a "stepped" pattern that mirrors external triggers.

Using 3D laser scanning to capture high-resolution spatial data over time, Yimin Wu et al. and Hao Chen et al. independently identified that loess tunnel deformation resolves into a clear three-stage lifecycle:

  • Rapid Growth Phase: Immediate, high-velocity displacement triggered by the loss of confinement during excavation.
  • Continuous Growth Phase: A decelerating but steady creep as the surrounding ground transfers its load to the newly installed primary supports.
  • Slow Growth/Stabilization Phase: The asymptotic leveling off of displacement as the support system and the ground reach a cooperative equilibrium.

This multi-stage signature is a universal characteristic of materials under stress. Under cyclic traffic loading, granular subgrade fillers exhibit an identical three-stage permanent deformation curve (initial rapid growth, deceleration, and stabilization), which Zhang Zong-tang et al. unified into a single predictive mathematical model. Similarly, Dechao Chi et al. map the full displacement of mudstone highway tunnels to a predictable four-stage curve, providing a generalized framework for identifying whether a tunnel is safely stabilizing or approaching catastrophic collapse.

Why Phasic Modeling Works: Isolating Nonlinear Regimes

The primary reason multi-stage modeling is so effective is that it allows engineers to apply different mathematical rules (constitutive equations) to different physical regimes. A single linear equation cannot accurately describe a material that transitions from a solid state to a fractured, flowing state.

Zhongkai Tong et al. leverage this via a tri-phasic modeling framework for stabilized soils. Instead of treating the soil's resistance as a constant, they segment its lifecycle into three distinct mathematical regimes:

$$\text{Total Lifecycle} = \underbrace{\text{Linear Elastic Phase}}_{\text{Hooke's Law}} \longrightarrow \underbrace{\text{Nonlinear Hardening Phase}}_{\text{Micro-crack development}} \longrightarrow \underbrace{\text{Post-Peak Softening Phase}}_{\text{Residual strength retention}}$$

By isolating the post-peak softening phase, their model can calculate the residual strength of the soil after it has technically failed. This prevents engineers from over-designing structures, as it accounts for the material's remaining capacity to bear loads even when cracked.

Isolating specific phases also allows researchers to uncover the hidden mechanics of composite materials. In asphalt modification, Rongxin Guo et al. utilize a three-phase deformation analysis (elastic, yielding, and creep) during force ductility testing. By mathematically isolating the final creep phase, they can strip away the behavior of the primary asphalt matrix and directly quantify how chemical additives improve the material's long-term flexibility and crack resistance.

Dynamic Feedback and Back-Analysis

The ultimate utility of breaking deformation down into chronological stages is the ability to perform back-analysis—using real-world data captured during early stages to rewrite the design assumptions for future stages.

┌────────────────────────────────────────────────────────┐
│ Stage 1: Measure Real-World Deformation via Sensors    │
└───────────────────────────┬────────────────────────────┘
                            ▼
┌────────────────────────────────────────────────────────┐
│ Back-Analysis: Back-calculate actual soil/rock stiffness│
└───────────────────────────┬────────────────────────────┘
                            ▼
┌────────────────────────────────────────────────────────┐
│ Stage 2: Update Model & Optimize Next Support Phase    │
└────────────────────────────────────────────────────────┘

Because a multi-stage model tracks the evolving geometry and constraints of a project step-by-step, it can be updated dynamically. Lixin Zhang et al. demonstrate a load calculation method for soft rock tunnels that uses measured structural deformations from the initial excavation stages to back-calculate the actual, highly variable rock pressure. This updated load profile is then used to optimize the support parameters for the upcoming excavation stages.

This dynamic feedback loop is equally vital in structural engineering. In the construction of long-span cable-stayed bridges, Bo Li et al. use stage-by-stage modeling to track how the physical towers and beams deflect as each new concrete segment is cast. By calculating the precise shift in anchorage points at each step, they can continuously adjust the installation angles of the protective cable ducts, ensuring perfect alignment before the bridge reaches its final, locked-in state.

Key Foundational Work

Finite element simulation of shallow - buried and mining tunnel ling in adjacent frame structures (2014) — contributed 2 concept elements to this direction.
Evaluation of low temperature performances of SBS+PPA compound modifying asphalt by force ductility test (2015) — contributed 1 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_stepped_deformation_law_3465ad6b,
  title = {stepped deformation law},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/stepped-deformation-law/rdr_3465ad6baecc46ce97ed4fc37a6181c9},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_stepped_deformation_law_3465ad6b_v1,
  title = {stepped deformation law},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/stepped-deformation-law/rdr_3465ad6baecc46ce97ed4fc37a6181c9/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake