About Parma

Laboratory

Murals

Paintings

Contact

Home

Paint Layer

The paint layer in several of the murals was in weak condition. Tenting and flaking were evident in a few, very restricted areas. More extensive areas, however, presented a friable and weak paint layer. Merely touching these areas caused pinpoint losses of paint.

Because there was no specific record of what had been done to the murals in the past, we tried to make careful observations that would inform our process. Measurements were taken to test the pH exhibited on the surface, it was found it to be nearly neutral (pH 7.5). The canvas itself, a robustly woven linen, was found to be extremely brittle and weakly acidic (around pH 4). The ground layer, as well as the adhesive layer tested positive for lead.

Aesthetically the murals had a dull, flat appearance. To some extent this was due to imbedded dirt and grime. More evident in the dark colors, there appeared an obvious dullness. This was later identified to be the “starch layer” which was reportedly put on the the murals periodically, in order to enhance their appearance. In any event, this non-original material had discolored, obscuring the true colors and spatial relationships in the Knight murals.

Supports

The support panels for the murals were not perfect, but made with the best recommendation available to muralists in the 1920's. Pressboard was used as the substrate for all of the canvas murals. In the case of the smaller murals, the pressboard was in one piece. The large murals, however, were evidently made from three pieces of pressboard adhered to the retro canvas. These pieces were butted against one another, and a wooden cradle lay just beneath them, acting as reinforcement and providing a join for the butted ends of presswood. From the rear, the network of vertical and horizontal wooden braces hid all of the pressboard seams.

Wood expands and shrinks over time, depending on the weather. Because the canvas murals were glued directly onto the pressboard and cradle structure (with lead adhesive, which is notorious for its strength), the canvases simply followed the movements of the wood. While seemingly insignificant, this pattern was the source of serious consequences.

Since the large murals were adhered to panels that had lap-joined seams, much movement could be exerted at the seams. Eventually A microclimate between the seams of pressboard was created, further weakening the fabric at those sites. Laminar stress between the layers of brittle canvas, static lead adhesive, and the movement of the wooden support also accounted for the problem. Eventually, the canvas could no longer bear the tension, and it actually split at these sites.

Consolidation & Cleaning

Before it was even possible to clean the murals, or perform any structural work, they needed to be consolidated. The paint layer was simple too weak to work with in its current condition. A technique developed in Italy for secco murals was employed. The consolidant, which was delivered in low concentration via a carrier solvent, was administered by brush to penetrate the paint layer, and pressure was applied to the surface. Since some of the consolidant would be caught up in the starch layer, it was important to time the consolidation closely with cleaning. Waiting too long after consolidation would make removing the starch more difficult.

An EDTA gel was designed to remove the starch. Since it was a water-based gel, we could use time as a variable for weaker concentrations of the EDTA. The pH of the system was buffered to 8.25 with TEA and titrated with 1N HCl. A non-ionic surfactant was added to aid in soil removal.

The cleaning solution was brushed on and left on for a short period of time. (The use of a gel greatly reduced the penetration of the cleaning solution, and allowed time as a variable in cleaning). The gel and solubilized starch were then removed with dry cotton, and the area was rinsed with deionized water. After thoroughly dry, the paint layer received a final rinse of mineral spirits.

Structural Work

As mentioned previously, the structural supports are considered to have been skillfully crafted with the best recommendation in both technique and materials of the day. It is hard to fault the natural course of these materials, dealing with time, and the constant forces imparted upon them.

Ideally, a remedy would have been to remove the canvas murals from their existing wooden sub-structures, and to replace those structures with a modern solid supports, such as aluminum honeycomb panels. Because of time restraints and funding, it was decided to keep the wooden supports as they were. That meant that laminar stress would still occur at the join sites. Therefore, we needed to properly fill and close these sites with compatible compensating material which would react homogenously with the mural as a whole, while being strong and flexible enough to handle excessive force without cracking. Excellent wood epoxy systems were commercially available (they do not shrink and are made of 100% solids) and other lower viscosity systems were tested with the addition of inert materials like fiberglass and CaSO4.

A conservation product called woodepox was finally chosen to achieve this aim. The edges of the torn canvas were first protected with a layer of Beva 371 because epoxy will not stick to it. A low viscosity epoxy was first injected into the sites, to help consolidate any loose or oxidized wood particles in the gaps. Then the wood epoxy was plugged into to gaps with tiny spatulas. Care was taken to pack the epoxy in tightly, just to the level of the original presswood. A final compensation of gesso di Bologna (anhydrous gypsum) and rabbit skin glue was used to level the fill and bring it into plane with the painted surface.

Finishing

After the murals were consolidated, cleaned, and all structural compensation was achieved, the murals were varnished with a reversible conservation varnish that contains an ultraviolet light stabilizer. A non-glare finish was achieved through the blending matte and satin varieties of the varnish. Finally, any and all previous areas of loss, including the fill sites, were in-painted with Gamblin Conservation colors. Both the varnish and the pigments are compatible and completely reversible because of their low molecular weight.

Conclusion

This work would never have been accomplished without the care and support of the Field Museum's professional staff. Special thanks to the conservation department and exhibits department. Their passion and concern for the murals were guiding parameters at every stage of the project.

Pinpoint losses were evidenced in various murals.

Laminar stresses were released at the sites of the wooden support seams.

No longer able to bear the tension, the canvas had split at these sites.

View from the back of the mural, showing the original cradle support.

 

Parma Conservation, located in Chicago, is one of the country's leading resources for the conservation of paintings and murals.

about parma | laboratory | murals | paintings | contact | home

website copyright © 1998-2008, all rights reserved, Parma Conservation, Ltd.