Research ArticleAPPLIED SCIENCES AND ENGINEERING

Analysis of hygroscopic self-shaping wood at large scale for curved mass timber structures

See allHide authors and affiliations

Science Advances  13 Sep 2019:
Vol. 5, no. 9, eaax1311
DOI: 10.1126/sciadv.aax1311
  • Fig. 1 Self-shaping wood bilayer application at large scale.

    (A to D) Analysis and design process at laboratory scale. (E to G) Industrial scale, same thickness but increased length and width of bilayers toward plate geometries. (A) Parametric digital model and FE analysis for shape prediction of arbitrary configuration. (B) Example bilayer strip configuration (here, European beech wood) with passive and active layer components with wood anatomical directions radial (R), tangential (T), and longitudinal (L). (C) Laminated wooden bilayer strip in initial, wet, and flat shape. (D) Curved self-shaped bilayer after drying. (E) Industrial-scale manufacturing (here, Norway spruce wood) in plate and shell configurations. Plates can be air-dried or kiln-dried to achieve self-shaping. Multiple shaped bilayer plates can be stack-laminated and additionally laminated with a thin cold bent locking layer (with same thickness as a passive layer) to form shape-stable curved CLT. (F) Application example as barrel vaulted CLT roof structure with wood fiber direction of thick lamellas (active layers) in load-bearing direction. (G) Application example as curved CLT wall with wood fiber direction of thick lamellas (active layers) in vertical direction.

  • Fig. 2 Shape change after drying.

    Bilayer samples (configurations 1 to 3, depicted at the bottom) made out of beech (A) and spruce (B) wood during 900-hour drying time. Drying dynamics (WMC, denoted ω) as simulated by the FE models and measured on experimental samples (error bars denote ±SDs). Curvatures (denoted κ) versus square root of time, and curvatures versus moisture contents with comparison to model predictions.

  • Fig. 3 Shaping-induced stresses and strains.

    (A) Bilayer axial stress states (σxx, indices LL for passive and RR for active layer corresponding to local wood anatomical coordinate system) at four points of interest versus time (t) for configurations 1 to 3 and both wood species. Strain state (εxx) separated in individual contributors. (B) Elastic (εxxel). (C) Viscoelastic (εxxve). (D) Mechanosorption (εxxms). Irreversible plastic strains (εxxpl) equal zero for all configurations. Total strains (εtot) and hygro-expansion strains (εω) are not shown.

  • Fig. 4 Simulation model.

    (A) Representative elementary volume for wood material model (beech and spruce wood) with anatomical growth directions R, T, and L. (B) Schematic rheological model for bulk wood as in (35). (C) Setup and BC for the FE bilayer models (quarter model with XY and YZ symmetry). C, stiffness matrices of springs for respective deformation modes; H, hardening moduli; fc, yield functions; α, differential swelling coefficients; τi, characteristic retardation times; μj, moisture analogous to τi; σ and ε, stress and strain tensors.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/9/eaax1311/DC1

    Movie S1. Time-lapse video of large-scale wood bilayer actuation.

    Section S1. Sensitivity analysis

    Section S2. The sensitivity parameter Sitot

    Section S3. Statistical analysis of shrinkage coefficient in active layer

    Fig. S1. Results of sensitivity analyses.

    Fig. S2. Statistical analysis of shrinkage coefficient.

    Table S1. Input and output of uncertainty quantification.

    Table S2. Statistical test results on differential swelling coefficient measurements.

    References (4347)

  • Supplementary Materials

    The PDF file includes:

    • Legend for movie S1
    • Section S1. Sensitivity analysis
    • Section S2. The sensitivity parameter Sitot
    • Section S3. Statistical analysis of shrinkage coefficient in active layer
    • Fig. S1. Results of sensitivity analyses.
    • Fig. S2. Statistical analysis of shrinkage coefficient.
    • Table S1. Input and output of uncertainty quantification.
    • Table S2. Statistical test results on differential swelling coefficient measurements.
    • References (4347)

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Movie S1 (.mp4 format). Time-lapse video of large-scale wood bilayer actuation.

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article