Thread-by-thread tear mendings in conservation of canvas paintings: a problem of reproducibility in bonding qualities

ABSTRACT

The paper introduces the technique of thread-by-thread tear mending in the conservation of canvas paintings and is dedicated to the investigation of difficulties in the reproducibility of bonding qualities. An insight is given into the basics of the bonding technique as well as the requirements on the adhesives. Based on initial test series, the cause of scattering is investigated and examined on the basis of scale effects, thread geometry and general bond qualities. Selected results from uniaxial tensile tests on single thread and canvas strip specimens are presented and discussed. As a result, it can be stated that the scattering is inherent in the material and technology. Therefore, the aim must be to show the range of possible results within further tests to provide reliable data for practical conservation and restoration treatments.

KEYWORDS:

• Thread-by-thread tear mending
• canvas painting
• conservation
• restoration
• adhesive selection
• bonding
• reproducibility
• scattering
• uniaxial and biaxial testing

Acknowledgements

This complex subject is topic of a cooperative PhD research project, executed by the author Hannah Flock M.A. The project is located at Cologne Institute of Conservation Sciences (CICS) at TH Köln and the Chair of Applied Mechanics (LTM) at Saarland University (UdS), supervised by Prof. Dr. Elisabeth Jägers (CICS) and Prof. Dr.-Ing. Stefan Diebels (LTM) as well as Prof. Dr. Wulff Possart, specialist for adhesion and interphases in polymers at UdS.

The co-authors and supervisors of the doctoral project as well as Dipl.-Rest. Petra Demuth and Prof. Hans Portsteffen from CICS should be expressly thanked at this point.

Notes

¹ Type of Fabric: Pattina L 13 (puro lino), Co. Tessitura Enrico Sironi.^[15]

² Own pretests show: With 0.6 µl adhesive the fibres are wetted by all chosen adhesives. Lower amounts (<0.6 µl) may form weaker bond qualities (as proven for 0.3 µl), while higher amounts (>0.9 µl) show significant loss in strength due to over-dosage of adhesive.

³ Main components of BEVA 371: 45% EVA (with 32% VA), 27% cyclohexanone-resin, 15% EVA (with 15% VA), 9% oil-free paraffin wax, 4% phthalic acid esters.^[1]

⁴ Uniaxial testing of canvas strips according to DIN EN ISO 13934-1,^[19] biaxial testings related to uniaxial setup and in consideration of standard (JIS) MSAJ/M-02 1995.

⁵ Testing according to DIN EN ISO 527-3, ASTM D 823 and ASTM D 2370^[14,15].

⁶ Testing of single threads according to DIN EN ISO 2062, testing of canvas strips according to DIN 13934-1^[15].

⁷ Unwoven linen yarn: Type Nm12/1, Co. TextilTrade^[15].

⁸ Linen ply yarn: Col. 4011, 100% Linen, LZ 50 m, CA 02776, Co. Gütermann.

⁹ Polyester ply yarn: Col. 111, 100% Polyester, M782 100 m, CA 02776, Co. Gütermann.

¹⁰ Arithmetic mean of 50 single measurements per 1 m thread length^[15].

¹¹ The preliminary investigations worked only with engineering stress, relating the force to the initial cross-sectional area before testing. More precise is true stress, relating the force to the reduced cross-sectional area due to elongation of sample during testing. So true stresses may become higher than engineering stresses when samples elongate. Consequently, the observed tendencies could differ when using true stresses and maybe present larger decreases of variations. Unfortunately, trials of tracking the measures of thread sample during testing were not successful. However, it must be pointed out once again that here we can only speak of effective tensile stress, as there is no uniform continuum in the continuum mechanical sense.

¹² Two thread materials (M1 and M2), consisting of twisted flax fibres, have probably not undergone much post-processing, only a minimal amount of water-soluble size was applied. Whereas the other two thread materials (M3 and M4) are ply yarns, consisting of two twisted threads made from twisted fibres. The linen ply yarn (M3) has been bleached and shows a water-insoluble, distinct size. Whereas the polyester ply yarn is made from only artificially produced synthetic fibres, probably even filament fibres (>1000 m length of single fibres).

¹³ Regarding the linen ply yarn (M3), the nonpolar size layer probably leads to declined wetting and adhesion of adhesive on substrate. The same is valid for the nonpolar synthetic polyester thread (M4).