ABSTRACT

Large-span roof systems (space frames, long-span trusses, cable-net and membrane roofs, and hybrid steel–concrete solutions) are increasingly used for stadiums, terminals, exhibition halls, and logistics centers. However, their construction is frequently constrained by limited crane capacity, tight site logistics, high sensitivity to geometric tolerances, time-dependent structural behavior during erection, and elevated safety risks during temporary stability phases. This article synthesizes evidence from published case studies and technical literature on large-span roof erection (including strand-jack lifting, block assembly, sliding, and in-situ element erection), and proposes an optimization framework for step-by-step assembly.

АННОТАЦИЯ

Для стадионов, терминалов, выставочных залов и логистических центров все чаще используются крупнопролетные системы кровли (пространственные каркасы, длиннопролетные фермы, кабельно-сетевые и мембранные крыши, гибридные сталебетонные решения). Однако их строительство часто ограничено ограниченной пропускной способностью крана, жесткой логистикой участка, высокой чувствительностью к геометрическим допущениям, временным поведением конструкций при возведении и повышенными рисками безопасности при фазах временной устойчивости. В данной статье обобщены данные из опубликованных кейсов и технической литературы по возведению крупнопролетных крыш (включая подъем струнных джокеров, сборку блоков, скольжение и возведение элементов в грунте), а также предложена схема оптимизации поэтапной сборки.

Keywords: large-span roofs; erection engineering; step-by-step assembly; strand jacking; space frames; temporary works; tolerances; safety; productivity.

Ключевые слова: кровли больших пролетов; монтажное проектирование; поэтапная сборка; продольный монтаж; пространственные каркасы; временные сооружения; допуски; безопасность; производительность.

1. Introduction

Large-span roofs are defined here as roof structures whose primary load-carrying span is large enough that construction staging becomes a governing design and delivery constraint. Typical triggers include: (a) spans exceeding the practical reach or capacity of available cranes; (b) structural systems (cable nets, tension rings, space frames) whose stiffness and stability depend strongly on final geometry; and (c) projects where schedule pressure demands prefabrication, modularization, or heavy-lift operations. In many such projects, the “design” cannot be separated from the “build sequence”: structural members and connections that are safe and efficient in the final state can be highly stressed, unstable, or geometrically incompatible during intermediate erection stages.

In practice, step-by-step assembly (sometimes called staged erection, incremental erection, or multi-stage lifting/positioning) is used to manage these challenges. The roof is subdivided into modules, blocks, or subassemblies; these are assembled at ground level or on temporary supports; and then incrementally lifted, slid, or connected into the final configuration. Heavy-lift techniques such as strand jacking are widely used to raise large roof rings or trusses, reducing the extent of high-level work and enabling better control of tolerances and safety. Case literature describes strand-jack lifting for large roof components and the importance of analyzing erection stages, tolerances, and temporary works (e.g., on Heathrow Terminal 5) and for stadium roofs where tension rings are lifted and then cable systems are installed (e.g., London Stadium).

Despite decades of large-span construction experience, projects still face recurring problems: (1) rework due to fit-up and tolerance mismatch; (2) safety incidents during temporary instability phases; (3) schedule disruption due to equipment constraints, weather, and logistics; (4) cost growth from temporary works overdesign or redesign; and (5) coordination failures between fabrication, transport, and installation. These issues are amplified in developing markets and in regions with high seismic demand, strong winds, and extreme seasonal temperature swings—conditions that can be relevant in Central Asia and many continental climates.

1.1. Contributions and novelty

• A unified optimization framework for step-by-step assembly that links method selection, staged structural analysis, tolerance management, digital planning (4D BIM), and risk-based control.
• A process map and checklists for industrial implementation, emphasizing measurable deliverables (staged analysis outputs, tolerance budgets, lifting plans, inspection points).
• Comparative evaluation of common erection strategies (in-situ element erection, block erection, lift-up, sliding) for space frames and truss/cable systems.
• A critical synthesis of evidence from published case studies and research on BIM-enabled optimization and virtual pre-assembly for large steel spatial structures.

2. Methods

From these sources, we derive a set of “governing constraints” (equipment, geometry, temporary stability, tolerances, logistics, and safety). We then formulate an optimization framework that treats step-by-step assembly as a sequence planning problem with constraints and objective functions. The framework is presented at two levels:

Level A (strategic): selection among erection method families and module partitioning, based on boundary conditions and risk appetite.

Level B (operational): optimization of the sequence, resources, and control points using measurable metrics: duration, cost, risk exposure, and quality probability.

No proprietary project data are used. Numerical values are based on typical ranges reported in literature and on illustrative planning parameters; they are intended for methodology demonstration rather than universal prescriptions.

2.1. Key definitions

Table 1.

Values

3. Results

3.1. Taxonomy of erection/assembly methods for large-span roofs

Evidence from space-structure references and practice indicates that most large-span roof erection strategies can be grouped into a limited number of method families. For space frames and space grids, method families are commonly described as: (1) element erection (high-level assembly of individual members and nodes), (2) block erection (ground-assembled blocks lifted into place), (3) lift-up (assembling large portions at a low level then lifting as a whole), and (4) sliding or incremental launching (assembling on one side and sliding into position). These families appear across truss roofs and cable-supported systems, with adaptations depending on whether the roof relies on tension rings, prestress, or membrane behavior.

Table 1 summarizes typical applicability, advantages, and constraints. The step-by-step philosophy can be implemented within any family, but block erection and lift-up methods generally reduce work-at-height and enable better quality control by shifting assembly to the ground. In contrast, element erection can be appropriate where modules are small, access is good, and temporary works can be standardized; however, it tends to raise safety exposure and sensitivity to weather and fit-up issues. Sliding methods can reduce crane requirements but introduce friction control, path stability, and temporary support design challenges.

Table 2.

Comparative summary of erection method families for large-span roofs

3.2. Step-by-step assembly optimization framework

The core result is an optimization framework that treats step-by-step assembly as a system rather than a sequence of isolated tasks. The framework has five interacting pillars:

P1. Structural staging and load-path integrity: every assembly step must have a verified load path and stability margin. This requires erection engineering analysis and explicit modeling of temporary works. Case-oriented literature emphasizes that staged behavior (e.g., during lifting, prestressing, and load transfer) can govern the design and the execution plan.

P2. Modularization and interface design: partition the roof into modules whose size is compatible with lifting/transport constraints while minimizing critical interfaces. The aim is to maximize work done under controlled conditions (ground assembly, workshop fit-up) and minimize on-site welding/bolting in difficult access.

P3. Tolerance management: define a tolerance budget that links fabrication tolerances, assembly tolerances, support deflections, temperature effects, and measurement uncertainty. Tolerance budgets must be “closed” numerically so that the final geometry is achievable without excessive forcing.

3.3. Quantitative metrics and planning calculations

To make optimization actionable, planners need measurable metrics. We propose a balanced scorecard that includes: (I) duration and critical path sensitivity, (II) direct cost (labor, equipment, temporary works) and indirect cost (delay exposure), (III) quality probability (probability that fit-up and geometry meet tolerances without rework), and (IV) risk exposure (probability × consequence for major hazards).

3.3.1 Duration (T)

T = Σ (tI / rI) + Δweather + Δinterfaces,

where tI Is base work content of activity I, rI Is productivity rate, Δweather accounts for weather disruption sensitivity, and Δinterfaces accounts for additional time for interface closure and survey-based corrections. Block and lift-up methods generally reduce Δweather for assembly because more work is done at ground level.

3.3.2 Tolerance closure probability (PQ) Let the final closure gap g be a random variable driven by fabrication error ef, assembly error ea, support movement es, and temperature-induced movement et. For a simplified one-dimensional closure:

g = ef + ea + es + et.

If g must satisfy |g| ≤ g_allow for fit-up, and the errors are modeled as independent normal variables with standard deviations σf, σa, σs, σt, then g has σg = sqrt (σf^2 + σa^2 + σs^2 + σt^2). The probability of successful closure without corrective work is:

PQ = P(|g| ≤ g_allow) = 2Φ(g_allow/σg) − 1,

where Φ is the standard normal CDF.

This simple model supports trade-off decisions: increasing module size may reduce interfaces (fewer closures) but can increase σs due to higher deflection during lifting; improving survey control reduces σa.

3.3.3 Heavy-lift synchronization and control (strand jacks)

For lifting a large roof ring or truss with n lifting points, each jack i records load Fi and stroke si. The objective is to keep differential displacement within a limit:

|max(si − sj)| ≤ δ_allow for all i, j,

to avoid inducing unintended torsion or secondary stresses. Computer-controlled strand jack systems are commonly used to control loads and strokes in heavy lift operations; published project descriptions highlight the importance of monitoring and staged analysis during lifting.

3.3.4 Risk exposure (R)

For each major hazard k (e.g., loss of temporary stability, dropped load, wind gust during lift), define:

Rk = Pk × Ck,

where Pk is probability and Ck is consequence (scaled cost, injury potential, or schedule impact). The assembly plan should minimize Σwk Rk subject to time/cost constraints (wk are stakeholder weights).

These calculations are not replacements for detailed engineering; they are decision-support tools that help compare alternative step-by-step sequences and select where to invest in controls (extra survey, improved temporary works, enhanced monitoring).

3.4. Illustrative example: comparing two step-by-step sequences

Consider a hypothetical 120 m span steel roof over a sports hall using a spatial truss/space frame hybrid. The roof can be divided into eight ground-assembled blocks, each weighing 30–45 t, or into four larger blocks weighing 70–85 t. Two erection strategies are compared:

Strategy A (eight blocks): assemble blocks on the ground, lift with a 250–300 t crane, place on temporary shoring towers, connect blocks at height, then remove shoring after closure and final bolting/welding.

Strategy B (four blocks + lift-up): assemble four larger blocks on the ground with greater pre-assembly, connect them at low level into two half-roofs, then lift each half using strand jacks supported by temporary lifting frames, perform final closure at ridge, and transfer load to permanent supports.

4. Discussion

4.1 What the evidence implies about “optimal” step-by-step assembly.

The reviewed evidence supports several robust patterns. First, large-span roofs are rarely erected safely and efficiently without explicit erection engineering analysis and carefully designed temporary works. Case-oriented publications on major roofs emphasize build sequence, tolerance management, temporary works, and controlled lifting (including strand jacking) as central elements rather than secondary details. Second, method families (element, block, lift-up, sliding) are not competing “brands” but rather interchangeable tools whose suitability depends on constraints: site access, crane capacity, allowable ground assembly footprint, sensitivity of the structural system to intermediate states, and the project’s risk management maturity.

A common failure mode is to treat the erection plan as a linear schedule exercise rather than a coupled structural-logistical system. When this happens, the project is forced into reactive redesign of temporary works, unplanned fit-up corrections, and schedule recovery measures that increase risk. The optimization framework proposed in this article pushes decision-making earlier: choose module boundaries, connection details, and monitoring points while there is still design freedom.

4.2 Critical and logical viewpoint: why “step-by-step” sometimes fails

Step-by-step assembly is sometimes presented as a universally safer and faster approach, but it can fail for logical reasons:

- Too many stages: if the roof is divided into excessive small modules, interfaces multiply, and cumulative tolerance/rework risk increases.

- Overreliance on temporary works: if shoring dominates the system, schedule becomes hostage to temporary tower erection and removal, and the structure may behave differently than expected.

- Competence mismatch: heavy-lift and synchronized lifting can reduce work-at-height, but only if the team has proven capability and monitoring. Without competence, the risk shifts rather than decreases.

- Data gaps: staged analysis needs realistic stiffness assumptions for temporary works, connection behavior, and boundary conditions. If these are not validated, optimization outputs can be misleading.

The practical implication is that optimization must be “context-aware.” A method that is optimal in a high-capability environment may be suboptimal where equipment or monitoring systems are unavailable. Thus, the optimization framework includes a competence and resource assessment as an explicit constraint.

5. Conclusion

This article presented an evidence-based optimization framework for the technology of step-by-step assembly of large-span roof structures.

Across erection method families, block erection and lift-up/heavy-lift strategies commonly reduce work-at-height and weather sensitivity, improving quality control—provided that lifting competence and monitoring are in place. Element erection and sliding methods remain valuable in constrained contexts, but they demand heightened attention to temporary stability and tolerance accumulation. The paper’s illustrative calculations show why reducing the number of closure interfaces can significantly reduce cumulative rework probability and schedule drift.

References:

1. McKechnie, S. et al. (2004/2021 upload). Design and construction of Terminal 5 roof. The Structural Engineer. (ResearchGate full-text upload).
2. LUSAS. Erection engineering analysis for the London Stadium roof (case study).
3. Lan, T.T. (Handbook chapter). Space Frame Structures – Methods of erection (element, block, lift-up, etc.).
4. Sejati, P.A., Isvara, W. (2023). Identification of roof truss work activities for the Jakarta International Stadium project with heavy lifting using strand jack system. AIP Conference Proceedings.
5. Han, D. et al. (2021). Research on BIM and virtual pre-assembly technology in construction management (PDF).
6. Zhang, W. et al. (2023). Application Research of BIM Technology in Airport Terminal Construction. Buildings (MDPI).
7. Ruby + Associates (Constructability Part IV). Strand Jack heavy lift technology and temporary works considerations (PDF).
8. DLT Engineering. Brasilia Stadium cable net roof erection using computer controlled strand jack system (project note).