Mechanical response of wooden boards subjected to humidity step variations: climatic chamber measurements and fitted mathematical models

Abstract

This paper describes selected results from a research focused on the behaviour of wooden boards, simulating the supports of panel paintings, subjected to cyclic humidity variations. It describes the mechanical response of two boards made of Poplar ( Populus alba) wood, 400 × 400 × 40 mm (longitudinal × transversal × thickness), assumed to behave as “structural replicas” of true panel paintings, subjected to step variations of air humidity under controlled conditions. One of the two boards was free to deform, and its cupping was monitored; the other was prevented from deforming, and the forces it exerted against its constraints were also monitored by means of an expressly developed measuring apparatus, named monitoring cross beam (MCB). Free deformations of a third “dummy” board, smaller in size, were also monitored. Instantaneous values of forces and deformations, together with temperature and relative humidity of the controlled microenvironment, were monitored at 15 min intervals, and stored in a data logging system. Several adsorption–desorption cycles were carried out in a climatic chamber, keeping T = 30 °C and imposing RH step variations between approximately 35 and 50 and 65%. During some cycles both faces of the boards were free to exchange moisture with the environment, during other cycles one face was waterproofed, to simulate paint layers. Each cycle lasted approximately 3 months, in order to reach constant equilibrium moisture content throughout the whole thickness of the boards; in total, the tests lasted over 2 years. In response to each step variation of RH, both forces and deformations showed the same kind of response: 1) with both faces free, response was asymptotic, reaching final state after approximately 2–3 months; 2) with one face waterproofed, the asymmetrical moisture gradients produced—in addition to asymptotic ones—transient responses as well, culminating in about 15 days and fading out after about 6 months. Descriptive models of the board’s mechanical behaviour were developed by fitting experimental data to the following general exponential equation, made by the sum of a “short period” and a “long period” component: y = p 1 ( 1 - e ( - p 2 x ) ) p 3 + p 4 x p 5 e ( - p 6 x ) Parameters p 1 to p 6 were computed for all cycles. Although the time required to reach equilibrium to the new steady humidity conditions in the reported tests lasted approximately 3 months, very fast responses to hygrometric disturbances were also detected, lasting hours, or even minutes. Further analysis of the collected data needs to be performed, based on the constitutive equations of the involved phenomena, namely moisture diffusion and mechano-sorptive behaviour of wood. Research aim. – The main objective of our research is to develop an interpretative model, featuring both descriptive and quantitative characteristics, in order to predict the behaviour of panel paintings under environmental variations; this predictive ability should then open the way towards evaluating appropriate measures to improve their conservation.