TDIL

Timber drying, shrinkage and joints

Posted Sep 4, 2026 ・ Checked Oct 5, 2026

The dimensions of timber change with its moisture content. A tsugite or shiguchi is formed by fitting members together, so a change in the dimensions of the members affects the joint[1]. This article covers how shrinkage on drying occurs, how it relates to loosening and cracking of tsugite and shiguchi, and what the references say about limiting loosening. How joints fail is covered in a separate article.

Moisture content and the start of shrinkage

Moisture content is expressed as the mass of water as a percentage of the oven-dry mass of the timber[1][2]. Freshly felled green timber holds a large amount of water, and for sugi (Japanese cedar) many pieces exceed 100%[2]. On drying, the free water in the cell cavities is lost first. The state in which the free water has gone is called the fiber saturation point, at a moisture content of about 30%[2][3]. Sato states that for many species it lies in the range of 25 to 30%[1].

Shrinkage begins only after the moisture content falls below the fiber saturation point and the bound water in the cell walls starts to decrease[1][3]. The timber then reaches the air-dry state, in which it is in balance with the humidity of the air. This equilibrium moisture content is given as roughly 14 to 16% in Japan[2], and one source explains that it stabilizes at about 15%[3].

Shrinkage differs by direction

The amount of shrinkage differs by direction. It is greatest in the tangential direction, along the growth rings; it is smaller in the radial direction, across the growth rings; and there is almost no shrinkage along the grain[1][2][4]. According to material from the Ishikawa Prefectural Forestry Experiment Station, the approximate shrinkage of sugi down to the air-dry state is 0.03% in the axial direction, 1.1% in the radial direction and 3.5% in the tangential direction, and a flat-sawn board 10 cm wide shrinks by 3 to 4 mm in the tangential direction[2]. Sato gives the tangential shrinkage from green to air-dry as 3 to 8%, depending on the species[1]. The values vary considerably with species and density, so the figures are only a guide[1].

This difference between the tangential and radial directions is the cause of cracking and warping during drying[2].

Cracking and sewari

A boxed-heart timber, which includes the pith, shrinks more in the tangential direction than in the radial direction, so cracks open radially from the outer surface toward the pith[1][5]. In a thick member the surface dries first. The surface tries to shrink while the interior stays at a high moisture content and does not shrink, so tensile force acts on the surface and cracks form[1][2]. In Sato's explanation, in large-section members such as beams and girders these cracks can advance into the interior over several months to more than a year[1].

Sewari (a relief kerf) is a treatment that fixes in advance where the timber will crack, so as to limit cracking on the surface[1][5][6][17]. After sawing, a saw cut reaching the pith is made from the top end (suekuchi, 末口) to the butt end (motokuchi, 元口)[1][6]. It is normally made on a face that is not easily seen[5]. One source says it is not made in heart-free timber, which excludes the pith[6], and another says it may be omitted in some cases[5].

However, even a boxed-heart timber with a sewari will continue to shrink after construction if it was used without being dried sufficiently. Sato gives an example of a 12 cm square post in which the face without the sewari narrows by about 2 mm while the sewari opens by about 6 mm, so the section changes from a square to a trapezoid[1].

Effect on tsugite and shiguchi

According to Sato, if members are precut with tsugite and shiguchi while undried and are assembled after an interval, they shrink during that interval and no longer fit together. If undried members are cut and assembled immediately, shrinkage after construction can cause loosening or cracking at the joint[1].

There is also an explanation, concerning a connection made with metal fittings, that when the timber becomes thinner and a gap opens at the joint, its ability to transmit load may fall[7]. For the nuki (a penetrating tie beam)[18] joints of traditional construction as well, Kitamori states that if gaps arise from drying shrinkage or construction tolerances, the structure cannot be made rigid[8].

What the references confirm goes no further than the fact that shrinkage opens gaps in joints. No reference describing how the loosening differs by type of tsugite was found.

Ways of limiting loosening

The measures confirmed in the references are as follows.

On pegs (sen) and shachi, the explanation of the basic forms set out by Kanasaki says that they are made so that the tsugite or shiguchi tightens as they are driven in[9]. No statement of them as a measure against drying shrinkage was found.

No reference describing a procedure or frequency for retightening an assembled joint with pegs or wedges was found.

See them in 3D

Sources

Sources not cited by number in the text were consulted during fact-checking.

  1. Sato, "Changes in the dimensions and shape of wood caused by changes in moisture", Mokuzai Hozon Vol.22-2 (1996), Japan Wood Protection Association (Japanese)
    https://www.jstage.jst.go.jp/article/jwpa1975/22/2/22_2_74/_pdf/-char/ja
  2. "Understanding Ishikawa's forestry and forest technology No.6: drying sugi timber", Ishikawa Prefectural Forestry Experiment Station (2004) (Japanese)
    https://www.pref.ishikawa.lg.jp/ringyo/publish/documents/sugi_kansou_1.pdf
  3. "Forest Research Institute newsletter: timber and moisture", Niigata Prefecture (Japanese)
    https://www.pref.niigata.lg.jp/site/shinrin/rin-nii-201308.html
  4. "The great property of wood: shrinkage", Maruhon (Japanese)
    https://www.mokuzai.com/LearningWood/in_di-63
  5. "What is sewari in timber? Differences from boxed-heart and heart-free timber, and its effect", Kozo Sekkei Navi (Japanese)
    https://www.kouzou-navi.com/articles/sewari.html
  6. "On Yoshino posts (special grade, with sewari)", Mokuzai Tonya (Japanese)
    https://www.mokuzai-tonya.jp/blog/processing/2131.html
  7. "Loosening of a hane-ita bolt caused by drying shrinkage of timber", arch-assist (Japanese)
    https://arch-assist.net/2026/05/06/wood-shrinkage-bolt-loosening/
  8. Kitamori, "A thousand years of timber construction technology", Research Institute for Sustainable Humanosphere, Kyoto University (Japanese)
    https://www.rish.kyoto-u.ac.jp/logos/wp-content/uploads/2015/11/Kitamori_Kokai_2015.pdf
  9. Kanasaki, "WOOD JOINERY SYSTEM", Keio University SFC (2011) (Japanese)
    https://fab.sfc.keio.ac.jp/paper/files/2011_kanasaki.pdf
  10. "Key points of timber drying", Mie Prefecture Forestry Research Institute (Japanese)
    https://www.pref.mie.lg.jp/ringi/hp/84196046149.htm
  11. Sugita and Ono, "How can the traditional tsugite and shiguchi be passed on?", Nihon University College of Engineering lecture No.40 (Japanese)
    https://www.cit.nihon-u.ac.jp/kouendata/No.40/4_kenchiku/4-018.pdf
  12. "Kanawa-tsugi", Namba Kenchiku (Japanese)
    https://www.nanbakenchiku.co.jp/kiji/blog/9726/
  13. "For those aiming to be architects! Explaining the kusabi (wedge)", Miyashita (Japanese)
    https://miyashita-lww.jp/2026/02/21/%E5%BB%BA%E7%AF%89%E5%A3%AB%E3%82%92%E7%9B%AE%E6%8C%87%E3%81%99%E4%BA%BA%E5%90%91%E3%81%91%EF%BC%81%E3%80%8C%E6%A5%94%E3%80%8D%E3%82%92%E8%A7%A3%E8%AA%AC%EF%BC%81/
  14. "Traditional shiguchi, kumite and tsugite", Shimizu Mokkobo (Japanese)
    https://www.kiri-shimizu.com/fitting/
  15. "Japanese Agricultural Standard for Sawn Timber, JAS 1083", Ministry of Agriculture, Forestry and Fisheries (Japanese)
    https://www.maff.go.jp/j/jas/jas_standard/attach/pdf/index-481.pdf
  16. "Chapter 2: shrinkage and cracking", OI Innovation (Japanese)
    https://www.oin-s.co.jp/dry/basic/02.html
  17. "The Sewari Kerf", Covington & Sons
    https://covingtonandsons.com/2020/01/11/the-sewari-kerf/
  18. "nuki", JAANUS
    https://www.aisf.or.jp/~jaanus/deta/n/nuki.htm
  19. "Hanasen" (鼻栓), architectural terms, Takumi Sogo Jimusho (Japanese)
    https://kenchikuyogo.com/?page_id=8981