figs.6a and b show tension change in the warp and weft directions during the two stages of the BEVA/sailcloth lining process. As the first layer of BEVA was applied to the sailcloth there was an immediate loss of 10N in both weave directions (see fig. 6a). The tension then stabilised at 37N in the warp and 45N in the weft as the solvent in the BEVA evaporated. Further applications of BEVA had a negligible effect on the tension in either weave direction. The prepared sailcloth was then re-tensioned and the painting positioned onto the lining fabric (fig. 6b). As the lining was heated an immediate reduction in tension from 38N and 35N to 15N and 7N occurred in the warp and weft respectively. This tension loss occurred as the BEVA softened at around 45–50° and continued until a temperature of 70°C was reached, at which point the lining was then cooled and the vacuum applied. The tension rose as the BEVA solidified and stable values of 25N in the warp and 43N in the weft were attained after 40 minutes.
BEVA and linen lining process
figs. 7a and b show tension changes in the warp and weft directions during the two stages of the BEVA/linen lining process. As the BEVA sizing layer was applied to the linen there was a loss in tension of 23N in the warp and 12N in the weft (see fig. 7a). It is difficult to distinguish between tension lost due to the application of the BEVA from that caused by creep, as both occurred simultaneously in the linen. Subsequent application of coats of BEVA produced smaller tension changes in the order of 10N in both warp and weft.
Once the lining canvas had been prepared it was re-tensioned and the painting positioned on top (fig.7b). As heat was applied, tension in the warp and weft reduced at a slower rate than that seen in the prepared lining canvas on its own because the vacuum slowed the rate of heating. After the picture surface had reached 70°C, the lining was cooled, producing a rise in tension, which stabilised after 40 minutes from 5N to 30N in the warp, and from 10N to 100N in the weft.
Interestingly, the BEVA/linen and wax-resin/linen lining processes result in a lining that has a large difference in the warp and weft tension when compared to the glue-paste/linen process. It is possible that this has occurred because these lining processes allowed a complete crimp redistribution, which was locked in when the adhesive solidified, rather than it being the sole action of the adhesive contracting. However, it is unclear why this should be the case.
Comparisons of stiffness before and after lining
Measurement of the thickness of the materials, using a micrometer, before and after lining show that the overall thickness after lining was less than the combined thickness of the lining and painting before lining. Hence, during the lining processes the materials had been compressed (see Appendix 3). This is because the pressure induced by the vacuum and the impregnation with adhesive altered the weave structure and packing of the yarns in the canvas. Calculations in terms of stress (force per unit area) are meaningless in this instance because the area of painting or lining canvas under tension cannot be accurately defined. Thus, comparisons have been made in terms of the tension experienced by each painting (all the same size) and the strain (extension compared to the original length). The biaxial tension versus strain, and the tension/RH versus time, have been compared for the unlined paintings, the lined paintings and the prepared linings.