Analysis of time-dependent deformation of CFRP considering the anisotropy of moisture diffusion

Yoshihiko Arao*, Jun Koyanagi, Hiroshi Hatta, Hiroyuki Kawada

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

17 Citations (Scopus)


The moisture absorption behavior of carbon fiber-reinforced plastic (CFRP) and its effect on dimensional stability were examined. Moisture diffusivity in CFRP was determined by measuring a specimen's weight during the moisture absorption test. Three types of CFRP specimens were prepared: a unidirectionally reinforced laminate, a quasi-isotropic laminate and woven fabric. Each CFRP was processed into two geometries - a thin plate for determination of diffusivity and a rod with a square cross-section for the discussion of two-dimensional diffusion behavior. By solving Fick's law expanded to 3 dimensions, the diffusivities in the three orthogonal directions were obtained and analyzed in terms of the anisotropy of CFRP moisture diffusion. Coefficients of moisture expansion (CMEs) were also obtained from specimen deformation caused by moisture absorption. During moisture absorption, the specimen surfaces showed larger deformation near the edges due to the distribution of moisture contents. This deformation was reasonably predicted by the finite element analysis using experimentally determined diffusivities and CMEs. For unidirectional CFRP, the effect of the fiber alignment on CME was analyzed by micromechanical finite element analysis (FEA) and discussed.

Original languageEnglish
Pages (from-to)359-372
Number of pages14
JournalAdvanced Composite Materials
Issue number4
Publication statusPublished - 2008 Nov 1


  • CFRP
  • Dimensional stability
  • Fiber alignment
  • Water diffusion

ASJC Scopus subject areas

  • Ceramics and Composites
  • Mechanics of Materials
  • Mechanical Engineering


Dive into the research topics of 'Analysis of time-dependent deformation of CFRP considering the anisotropy of moisture diffusion'. Together they form a unique fingerprint.

Cite this