Back to the Wooden Artifacts Page
Back to the Structural treatment Page
Wooden Artifacts Group Conservation Wiki
Contributors: Rian Deurenberg-Wilkinson
Your name could be here! Please contribute.
Copyright: 2012. The Wooden Artifacts Group Wiki pages are a publication of the Wooden Artifacts Specialty Group of the American Institute for Conservation of Historic and Artistic Works. The Wooden Artifacts Group Wiki pages are published for the members of the Wooden Artifacts Specialty Group. Publication does not endorse or recommend any treatments, methods, or techniques described herein.
Warped wood is a common problem in furniture conservation and one that is often hard to solve. The following article presents a basic outline of its causes and possible treatments.
- 1 Causes
- 2 Treatment
- 3 Direct Woodworking Techniques
- 4 Indirect Woodworking Techniques
- 5 Moisture Related Methods
- 6 Choosing a Method
- 7 Conclusions
- 8 Experiments
- 9 References
Warping is almost always related to the movement of wood in reaction to changes in moisture content and therefore, warping is often perpendicular to the grain. Due to the anisotropic character of wood, deformations can be hard to predict. The original orientation in the tree is the most important factor, but also species, width of the growth rings, defects, and drying techniques play a role. Kollman 1961 Of course, tensions can also present themselves in the formation of cracks or compression set instead of warping.
Sometimes a warp is temporary as in the case of the cupping of a tangential board during changes of relative humidity (RH). A warp can become permanent when a panel cannot move freely, for instance because it is contained in a tight frame or restrained by veneer on one side. Buck 1972
A less visible rigid construction is caused by an uneven distribution of moisture, during which the dryer side constricts movement. Veneer, a one-sided finish, or a one-sided wet treatment such as cleaning or dying, are well-known for causing an uneven moisture content in a panel.Buck 1972; Straub 1963; Straub 1968 A thinner panel will react faster and stronger than a thick panel.
This information is intended to be used by conservators, museum professionals, and members of the public for educational purposes only. It is not designed to substitute for the consultation of a trained conservator.
- To find a conservator, please visit AIC's Find a Conservator page.
- To learn how you can care for your personal heritage, please visit AIC's Resource Center.
Almost all treatments are based on applying forces of which the effects are hard to predict. Too much pressure can cause a reversed warp, while too little will not have enough effect. A panel can be forced flat in most cases, but over time, the tension left behind could cause ruptures in the panel, or craquelure and losses in rigid finishes like gilding or paint.
The following overview of techniques is not comprehensive, but will highlight some of the basic treatments. It should be kept in mind that treatments sometimes consist of a combination of multiple techniques, for instance using woodworking techniques like cleats after flattening a panel with heat and moisture.
The methods that make use of woodworking techniques are often very intrusive and are less fitting for modern day conservation practices. They will be mentioned briefly to create a background but will not be discussed in all its variations, and pros and cons.
Direct Woodworking Techniques
Direct woodworking techniques cover a wide range as far as removal of original material is concerned. The most direct method involves simply removing most of the thickness of the panel, making it flexible enough to flatten it and fixate it to a new substrate. In the slightly more sophisticated methods the panel is completely or partly spliced, then flattened, and finally the saw kerfs are filled with veneer or another material. von Reventlow 1978
A well-known variety is the method of grooves and wedges or ‘Sverzatura’. It is derived from the furniture making method of bending wood without steaming it. Emile-Male 1976 If the execution of the method is not done with precision and care, the wedges can project to the front side.
Indirect Woodworking Techniques
Indirect woodworking techniques are less intrusive because little or no original material is removed. A major disadvantage however is the presence of tensions left behind by forcing the panel in a flat position.
One of the methods most certain to succeed is the complete or partial addition of a substrate, for instance a sheet of plywood. The rigidity of this construction and sealing of the reverse side are important drawbacks. Counter-veneering is in fact a special type of the application of an additional substrate. In this case it’s not the new substrate but the strength of the drying adhesive that will straighten the panel. Hayward 1967
Cleats, Attached and sliding
Cleats or battens are a common furniture making technique to prevent or correct distortions. But even sliding battens can cause tears or distortions, as they often get jammed. In paintings conservation, the principle of moving battens has developed into a so-called ‘parquet’, a construction of vertical and horizontal battens, of which the vertical are attached and the horizontal battens slide. Since c. 1770, the parquet has evolved from wide wooden battens to aluminum blocks with synthetic ball-bearings. Nicolaus 1999
Moisture Related Methods
Moisture related methods make use of the properties of wood and often have a permanent effect without leaving tensions in the panel. They are based on two principles: one is the dimensional change of wood in reaction to moisture; the other is the fact that wood becomes more malleable at a higher moisture content, especially when heat is involved. Some treatments make use of both properties.
Deformation of wood by compression set very effectively makes use of the dimensional change of wood in relation to moisture content. The underlying concept of the method is that the convex side is wider than the concave side and that the panel will be flat if both sides are equally wide. In order to make the convex side smaller, it is first moistened or steamed, causing it to swell. The panel tries to curve even more, but this is prevented by clamping it on a flat surface. The swelling cells are crushed in this process, making this side smaller and the panel straighter after drying. Crushing of the cells is an irreversible process. The technique makes use of the same mechanism that has been explained briefly in the causes of warping. The treatment can be executed in several stages for better control of the effect. Since only the convex side is moistened, tensions can arise in the concave side of the panel.
One Sided Impregnation
Another method based on the swelling of wood is impregnation of the concave side. The material most widely known for this method is Polyethylene Glycol (PEG). PEG causes more swelling of the wood than most other consolidants because it swells the cell cavities as well as the cell walls. PEG can be applied locally but elements that are not treated similarly will react differently to changes in RH. PEG diffuses via the free water in the wood, meaning that the moisture content needs to be fairly high: at least above the fiber saturation point of 30%. Therefore, the panel needs to be moistened first. Triboulot 1999; Boucher 1999 As always, the reversibility of a consolidant is debatable, if not simply impossible. It is also possible that because of the swelling, the panel will be damaged by compression set or other tensions, which would result in an inverted warp if the consolidant would be removed.
Plastic Deformation with Moisture Only
Tensions in a panel don’t always cause permanent deformations. Like many materials, wood deforms elastically until the proportional limit is reached, after which plastic or permanent deformation occurs. The proportional limit of wood is rather low, being under 0.6% strain for oak with the tension perpendicular to the grain. The relation between stress and strain is linear until the proportional limit and is described by the Modulus of Elasticity, E. Between the proportional limit and point of failure, wood is plastically deformed. Hoadley 1980 Intermolecular bonds are broken and are formed anew when the applied stress is removed. Buck 1970 The relation between stress and strain is not linear in this stage. In order to have complete plastic deformation it is important that the wood will react as uniformly as possible. The moisture content should be the same everywhere, as drier areas will react mostly elastically, which can result in retained tensions. Straub 1968
To make wood deform plastically, the moisture content needs to be raised. The swelling causes the cellulose chains to spread, lowering the binding forces. In fact, the density of the wood is lower, which means a decrease of strength. Panshin and de Zeeuw 1970 Around the fiber saturation point (FSP) of 30%, bound water starts to work as a lubricant. After the FSP, wood is not swelling any further, but it is not necessary to raise the MC this far for plastic deformation. In reality, the modulus of elasticity of wood is already much lower at an MC of 15-20%. Buck 1970 Normally dried wood has an MC of 10%. With this method, the concave side of the panel is moistened.
Plastic Deformation with Moisture and Heat
The plasticity of wood can be further enhanced by using heat in conjunction with moisture. The best known examples are perhaps the steam bent chairs of Thonet. The combination of heat and moisture lowers the modulus of elasticity Kollmann 1951 as well as the proportional limit Uterharck 1941 causing the wood to deform at lower stress values. Lignin and hemicellulose become softer and the cellulose chains can shift without damage during the deformation. Triboulot 1999, Hoadley 1980, Sandick 2000 However, steaming weakens the wood through hydrolysis of the wood contents. In an ideal situation, the panel is steamed through and through. When only one side is steamed (such as the concave side for this treatment), tensions can arise in the other side.
The differences between the methods can be explained by comparing wood to reinforced concrete, in which lignin is the concrete and cellulose the reinforcement. In this model, lignin (concrete) is the restricting factor under pressure, but not tension, because it is rigid even in moist conditions, only becoming soft at a higher temperature. Hoadley 1978 When pressure is applied nonetheless, the wood will be damaged rather than plastically deformed. This is what happens during compression set. Plastic deformation with only moisture expands the concave side while the convex side keeps the same dimension. Cellulose (the reinforcement) is the limiting factor under tension, but at a higher MC, the cellulose can be stretched without damage. During plastic deformation with both heat and moisture the wood can be stretched as well as compressed within certain limits, without damaging the structure because all components are malleable.
Plastic Deformation with Microwaves
Recently an alternative for steaming has been found in the use of a microwave, which is based on the same principle of heat and moisture. A microwave induces heat by changing the polarity of the electromagnetic field of the microwaves. Polar molecules like water turn with the changing direction of the electric field. This gives friction in the material and thus a rise in temperature. Materials with a lower density, simple form and homogeneous composition heat up fast and evenly. Wood may have a relatively high density and heterogeneous composition, but water is omnipresent in the cell walls, making it possible to be heated in the microwave. The power setting dictates the strength of the field and can increase the intensity of the molecules’ reaction. At high power the material is heated quickly; at lower power it is heated more slowly, but more evenly. In a microwave the object is heated through and through and not just from the outside like in an oven. Because of this, less time is required, which is a clear advantage for bad heat conductors such as wood which dehydrate during prolonged exposure to heat. Another advantage of the microwave is that the panel does not need to be moistened, because the microwaves heat up water that is present in the cells. The most important disadvantage to microwave heating is that it cannot be restricted to the wood of the panel alone: finishes and adhesives in joints will also soften. Lifting of veneer can be avoided by using tape, appropriate clamping materials and careful clamping, but finishes cannot be protected. Another obvious drawback is that the size of the microwave oven is a limiting factor.
The treatment is fast and no tension need remain. By wrapping the wood in moisture resistant microwave-proof foil and leaving it in the foil while cooling and drying, the moisture content can be kept stable, making the result of the treatment easier to control.
Choosing a Method
The following table is based on three aspects: In almost all cases there will be at least one side of the warped panel that cannot be treated because of the presence of veneer, finish or patina. This is called the viewing side. Treatment options are obviously less limited when the panel is not directly visible.
- The table is based on panels with a warp in just one direction. A panel with twists in multiple directions can sometimes be treated locally.
- Only those methods that do not remove original material are included.
This overview indicates that there are possibilities to treat warped wood but is also clear that it is a complex problem that does not fit a standard approach. As is always the case in conservation, it is important to be prudent and try to reduce a warp in several steps rather than creating an inverted warp by a treatment that is too strong.
The following five methods were tested in 2001 by the author: Deurenberg 2005
- Deformations by compression set
- Impregnation of the concave side with PEG
- Plastic deformation with moisture
- Plastic deformation with moisture and heat
- Plastic deformation with a microwave
The test boards were cut from an old oak panel with an orientation of the grain of roughly 45 degrees (neither radial, nor tangential). They were bent by compression set with the widest dimension across the grain. Each technique had four variations considering clamping time, amount of PEG solution or the power setting of the microwave. Each variation was tested on three panels. The tests took between one day and nine weeks in time, plus two weeks of observation after treatment. During and after treatment, the curvature was measured with a dial depth gage. At these moments in time, the panels were also weighed to monitor the changes in moisture content and the possible dimensional changes relating to it.
Results of experiments
It could be concluded that the treatment is very fast: clamping panels on a moist cloth for one day and then letting them dry freely was already sufficient to reduce the warp. Through measuring the weight and the curvature, it could be determined that the panel had absorbed moisture during the first day. It subsequently showed a strong reversed warp, but was much flatter after drying. Compression set through steaming of the convex side and subsequent drying and clamping for 6-9 weeks, resulted in a strong reversed warp. The measurements indicated that weight and curvature became fairly stable during the 2 week observation period, indicating that the result of the treatment would not change much anymore.
Impregnation of concave side with PEG
Because it was unclear from published methods how much PEG solution was needed, this was the focus of the impregnation tests. It was assumed that the amount of moisture that is needed to straighten the panel is the same amount of solid material that is left behind after drying. A PEG solution of 25% PEG 1500 in water was used. Because the solution had the tendency to run off the surface of the panels, mylar ridges were adhered to the sides with hotmelt adhesive. Only the calculated amount of PEG caused a reduction of the warp, but only with about 34% decrease of the curvature. The panels did not seem to be able to absorb more PEG solution. Less solution had hardly any effect. After drying, the surface appeared bleached and felt wax-like. Also the panels gained 1-2 mm in width.
Plastic Deformation with Moisture Only
During the tests, the concave side was moistened until the panels were flat, after which they were clamped down and dried slowly. This resulted in a reduction of the warp of 16-70%, but no consistent differences between the clamping periods of 6-9 weeks could be discerned.
Plastic Deformation with Moisture and Heat
Tensions can arise in the side that is left untreated, which was probably the case in my experiments, in which I only steamed the concave side. The curve partly returned after treatment with a final reduction of 30-40%. The clamping periods of 6-9 weeks did not seem to be of influence.
Plastic Deformation with Microwaves
The panels that were heated at high power appeared to react more uniformly and were more stable according to their weight and warp. The veneered panels reacted well to the inverted warp that was deliberately established. The veneer remained well attached, but the finish had clearly suffered. When the panels are heated for a prolonged period, air bubbles underneath the veneer will expand and can eventually erupt. All finishes became soft, but the appearance of thinner layers did not change much. The method was also applied to a panel from the architectural model with veneer on both sides. The warp was reduced from 3.0mm to 0.5mm.
Conclusions of Experiments
All treatments were effective, but compression set and plastic deformation in the microwave were clearly the most successful. The results of plastic deformation with moisture and heat were not as good, presumably because steaming the panels on one side did not heat the panels through and through as in the microwave. Impregnation with PEG had very little effect and was also far from ideal in its execution. Plastic deformation with only moisture was effective, but it did not become sufficiently clear which factors influenced the results.
Boucher, N., Polyethylene glycol (PEG) in furniture conservation. In: Van Duin, P., van Loosdrecht D. and D. Wheeler (eds.), Proceedings of the Fourth International Symposium on Wood and Furniture Conservation, Amsterdam, Rijksmuseum/VeRes, 1999, pp. 13-22
Buck, R.D., Some Applications of Rheology to the Treatment of Panel Paintings. In: R.D.Buck, The Behavior of Wood and the Treatment of Panel Paintings, The Upper Midwest Conservation Association, Minneapolis Institute of Arts, Minneapolis, 1978, first published 1972, pp. 52-62 (received 12/1/71)
Buck, 1972; Straub, R.E., Über die Erhaltung von Gemälden und Skulpturen, Fretz & Weismuth Verlag, Zürich/Stuttgart, 1963, p.p. 152-164; Straub, R.E., Das Problem des verwölbten Holztafelbildes. In: Nachrichtenblatt der Denkmalpflege in Baden-Württemberg, 11, Heft 3, 1968, pp. 71-77
Buck R.D., The Dimensional Stabilization of the Wood Supports of Panel Paintings. In: R.D. Buck, D. Kolch en J.S. Horns, The Behavior of Wood and the Treatment of Panel Paintings, The Upper Midwest Conservation Association, Minneapolis Institute of Arts, Minneapolis, 1978, first published 1970, pp. 48-51
Buck R.D., The Dimensional Stabilization of the Wood Supports of Panel Paintings. In: R.D.Buck, D. Kolch en J.S. Horns, The Behavior of Wood and the Treatment of Panel Paintings, The Upper Midwest Conservation Association, Minneapolis Institute of Arts, Minneapolis, 1978, first published 1970, p.p. 48-51
These experiments were part of a final thesis and were presented and published as a paper in 2005: Deurenberg, R.M.H., Buigen of Barsten [Straightening of Warped Wood], in: Handelingen, Zevende Nederlandse Symposium Hout- en Meubelrestauratie [Postprints of the Seventh National Symposium on Wood and Furniture Conservation] , Stichting Ebenist, Amsterdam, 2005
Emile-Mâle, G., The Restorer’s Handbook of Easel painting, Van Nostrand Reinhold Company, New York, 1976, pp. 24-30 (Translation of La restauration des peintures de chevalet)
Hayward, C.H., Furniture Repairs, Evens Brothers Limited, London, 1967
Hoadley, R.B., Understanding Wood, A craftman’s guide to wood technology, The Taunton Press, Newton, Connecticut, 1980
Hoadley, R.B., The dimensional response of wood to variation in relative humidity. In: Conservation of Wood in Painting and the Decorative Arts, Preprints of the Contributions to the Oxford Congress 17-23 September 1978, International Institute for Conservation of Historic and Artistic Works (IIC), London, 1978, p.p. 1-6
Kollman, F, Rheologie und Strukturfestigkeit von Holz. In: HOLZ als Roh- un Werkstoff, 19, 1961, Heft 3, 73-80 (In verkürzter englischer Fassung auf dem V. Welt-Forst-Kongreβ in Seattle, Wash., USA, am 2.9.1960 vorgetragen)
Kollmann, F., Technologie des Holzes und derHolzwerkstoffe, Erster Band Anatomie und Pathologie, Chemie, Physik, Elastizität und Festigkeit, Springer Verlag, Berlin-Göttingen-Heidelberg, J.F. Bergmann, München, 1951, p. 606
Nicolaus, K., Handboek voor het restaureren van schilderijen, Könemann Verlagsgesellschaft mbH, Köln, 1999
Panshin A.J. & C de Zeeuw, Textbook of Wood Technology, Volume 1, Structure, Identification, Properties, and Uses of the Commercial Woods of the United States and Canada, New York: McGraw-Hill Book Co., 1970, p. 223
Reventlow, V. von, Use of B72 in the restoration of a marquetry surface – case history. In: Brommelle, N.S., Moncrieff A. And P. Smith (eds.), Conservation of Wood in Painting and the Decorative Arts, Preprints of the Contributions to the Oxford Congress, 17-23 September 1978, International Institute for Conservation of Historic and Artistic Works (IIC), London, 1978, pp. 37-39
Triboulot, 1999; Hoadley, 1980; Sandick, Ending the bending – on the straightening of warped wood: a general overview. In: Conservation News, nr. 73, 2000, pp. 31-34
Uterharck, F., Handbuch der holzverarbeitende Industrie, Dr. Sändig Verlagsgesellschaft, Leipzig, 1941, pp. 110-112