JAIC , Volume 39, Number 1, Article 9 (pp. to )
JAIC online
Journal of the American Institute for Conservation
JAIC , Volume 39, Number 1, Article 9 (pp. to )

DISASTER RECOVERY AT THE UNIVERSITY OF ALBERTA, OR, EVERY FLOOD HAS A SILVER LINING

SHIRLEY ELLIS



3 3. RECOVERY

The recovery operation for the flood-damaged clothing and textiles included treatments such as wet-cleaning; spot removal using suction, absorption, and blotting techniques; and altering the pH of the solution in order to attempt the reversal of dye transfer. To illustrate some of the techniques employed to treat these water-damaged textiles, they will be presented as case studies.

Certain procedures were followed when treating the frozen artifacts. Any existing records were retrieved prior to removing them from the freezer, since in most cases it was close to impossible to view the artifact properly. Time was of the essence. These records included student documentations or earlier conservation reports. They provided information such as descriptions, fiber and material identification, and diagrams or photographs to reduce the potential for surprises. If the artifact was dry, low-suction vacuuming generally preceded treatment.

A number of factors determined how and whether a treatment strategy would proceed—for example, the conservator's level of experience, the degree of difficulty of the treatment, and the ramifications of the treatment on the object. Consultation with the head conservator was usual. Those cases that posed some uncertainty in the outcome or a permanent alteration to the artifact required that the curator be consulted to make the final decision.


3.1 3.1 CASE STUDY 1 (ACC. NO. 79.8.8)

An off-white low-twist crepe silk 1920s dress with black silk embroidery floss suffered from dye transfer. The dress had been frozen but was only damp when retrieved from the freezer. The black embroidery floss was not fast in water and subsequently transferred a purple-colored dye to the off-white ground fabric. The transfer was evident in many areas throughout and near the embroidered sections. For this reason, spot-cleaning did not appear to be an option. Full immersion also did not appear to be an option because the black dye was so obviously not fast. With testing, it was found that the dye was easily removed with an anionic detergent solution of 0.2% Orvus WA Paste in distilled water. The curator was called for a consultation. Given the positive results of the spot test, it was decided to proceed with a full-immersion wet-clean.

In anticipation of unfast dyes, the drying area was prepared with fans and hand-held blow-dryers set out so that the dress could be dried quickly. The dress was immersed in the bath using the same solutions as the test. Dye loss was evident only during the final rinsing with distilled water. In an attempt to encourage the dye away from the dress, the rinse water was swirled by hand, creating a slight current. The dress was removed from the bath and blotted with white cotton towels, where dye transfer was visible. Once blotted, the dress was immediately hung on a polyethylene-covered padded hanger, and drying was accelerated by the fans and blow-dryers set on a cool setting.

Results of the wet-cleaning were very positive. The dye transfer that had occurred during the flood, in most areas, was removed from the white silk with only two lightened stains remaining. There was no additional transfer of dye to the white silk. Despite gentle stretching during drying, there was some shrinkage of 4% in the lengthwise direction due to the tightening of the crepe yarn.

The reason the dye was easily removed with wet-cleaning might be that the conditions for dye uptake were less than ideal. And the reason the newly bled dye did not penetrate or diffuse into the white silk fibers upon drying might be that the embroidery was dried quickly. Given the sample size, simple testing methods noted in E. R. Trotman (1970, 603–4) were inconclusive except to say that the black dye is not of a basic class but is possibly an acid, premetallized, or mordanted dye. Further investigation is warranted.


3.2 3.2 CASE STUDY 2 (ACC. NO. 77.5.101)

A 1930s beige crepe silk dress with a multicolored floral print was frozen due to water damage along the back hem. Fugitive dyes on this dress included the navy blue, green, yellow, and magenta. The dye transferred into both the beige areas and the adjacent dyed areas. When this dress was retrieved from the freezer, it was still wet. In order to prevent drying of those areas that were not immediately treated, they were covered with polyethylene sheeting.

There were two objectives for the treatment of this dress. The first was to remove or reduce the dye transfer that had occurred, and the second was to prevent the formation of a tide line along the wet edge. In attempting the removal of the dye transfer, the most innocuous solution of water was tried first. Water was judiciously sponged only onto the areas of dye transfer, with disposable diapers underneath to absorb the water. Finding no success, the conservator sponged on a 0.2% anionic detergent solution of Orvus in the same manner. As there was still no success, a more aggressive approach was taken.

Because the fiber content of this dress was silk, it was speculated that the dye might be of an acid class. An alkaline solution therefore might be successful in stripping the unwanted dye. An alkaline solution was prepared using a 0.1% solution of sodium carbonate in distilled water. The pH was lowered to 8.9 by adding 4:1 glacial acetic acid in water dropwise. This alkaline solution was applied onto the areas of dye transfer using a dropper and absorbed in a disposable diaper underneath. While the diapers worked reasonably well in pulling the liquid directly downward so as not to spread to adjacent areas, the suction table was also tried but was found to have an inadequate draw for this purpose. Removal of the navy blue and magenta dyes was moderately successful, while the yellow and green dyes tended to leave a yellow halo around the motif that only lightened with repeated applications of all three solutions but did not disappear.

Areas that were wet but had no dye transfer were sponged with the distilled water and detergent solutions as above, then rinsed. They were dried quickly using hand-held blow-dryers set on a cool setting. A slight tide line formed upon drying, so an attempt to remove this by sponging with only a slightly dampened sponge while feathering the edge was tried. There was some success with this technique, although applying the same solutions while using the suction table resulted in greater success. The tide line could be moved to the seam line or hidden by a motif. Treatment on the suction table also prevented the crepe yarn from further twisting and promoted a flattened texture.


3.3 3.3 CASE STUDY 3 (ACC. NO. 84.40.1B)

A bustle skirt and drape, dated from the mid-1880s, was damaged during the flood. It is constructed from unweighted shot-silk satin, which incorporates two different yarns in its structure. The brick-red warp yarn is visible on the right side, while the teal weft yarn is only visible on the wrong side when the light catches it.

The skirt and drape had been frozen because they were wet when discovered during the salvage operation. The front area of the skirt hem was wet, and dark tide lines were evident on the right side of the skirt where it had been stained with the floodwater but had already dried. The right corner of the skirt drape was also wet. The drape was easily wet-cleaned in a bath of distilled water and a 0.2% anionic detergent solution using Orvus. The skirt, however, was not so easily cleaned because of its multiple layers.

Two techniques were used to flush the wet area and reduce the formation of tide lines. With the first technique, the wet areas were sponged with distilled water through to a disposable diaper that was placed beneath each section of pleats. With the second, the suction table was used. Where there were layers of fabric as with the pleats, the diaper technique was more successful. However, with only single layers of fabric, the suction table worked well.

Initially, a saturated wet sponge was used to supply a greater quantity of distilled water to flush the area. This area was blotted with white cotton towels, then encouraged to dry quickly using a hand-held blow-dryer set on a cool setting. Meanwhile, the perimeter of the stain was feathered out by using a slightly damp sponge. The black tide line softened but continued to move. It appeared as if more than a watermark were staining the fabric. The interfacing beneath the pleats was composed of a black cotton fabric, the dye of which exhibited very poor washfastness. It was thought that this is likely the source of the black tide line. This theory was supported by the following case study.


3.4 3.4 CASE STUDY 4 (ACC. NO. 73.15.16 A-D)

A second silk twill-weave bodice and skirt with a brown cotton interfacing, also believed to date from the mid-1880s, was frozen in a solid block following the flood. The salvage comments read, “Bodice soaked entirely, upper half and all of skirt back soaked, two of the three small fragments are wet, one fragment dry, brown dye fugitive, boning in bodice.” These comments, as well as the previous treatment, were found to be useful for proposing a treatment strategy.

The degree of wetness of all the garment components was extensive, so it was decided that an immersion might be the best treatment to remove the floodwater so as to eliminate the formation of tide lines. A number of the pieces had cream-colored silk insets decorated with metallic threads, so it was decided to test only one fragment first with the full immersion. It was known from the salvage comments that the brown dye was fugitive. The treatment of the fragment in the test immersion was successful, and although the golden brown dye did discharge as expected, it did not transfer.

The bodice and skirt were immersed in a bath of distilled water at 33C. Initially, there was the golden-brown dye discharge as had been seen with the fragment. However, once the lining on the skirt thawed and became wetted, there was a severe dark brown dye discharge. It seemed that “clouds” of dye were moving into the bath. The bodice was quickly removed and blotted with white towels. In an attempt to minimize the aftereffects of the dark brown dye from the skirt, distilled water was allowed to run into the bath while it was draining. The skirt was soon removed and blotted with white cotton towels.

Both the bodice and the skirt were set up to dry quickly with the aid of numerous fans. Because the bodice had iron-based metal stays, which had been identified by the previous rust stains on the inner bodice and by a magnetic pull, rapid cool air from blow-dryers assisted the fans with drying. There was no additional iron corrosion evident and no dye transfer to the cream-colored silk. The positive results of being able to hasten the drying process of a wet treatment on an object with iron-based findings has led to subsequent treatments of such objects, also with positive results.

The skirt treatment was less successful, however. The cream-colored silk had grayed somewhat and there were numerous tide lines throughout. It appeared that the dye from the dark brown cotton lining had transferred to the silk. Further treatment strategies were reviewed at this point. Removal of the graying seen on the cream-colored inset showed some success with a controlled sponging with the anionic detergent solution (0.2% Orvus in distilled water), followed by a rinse with distilled water. Detaching this inset and wet-cleaning with detergent was contemplated, but it was decided that subjecting the metallic threads to a second wet treatment was not justified. The improvement to the cream-colored silk would only be moderate and damage to the metallic threads a higher risk. There was a possibility that the detergent solution and agitation of the sponging action would cause a loss of adhesion of the metallic foil to the paper substrate.

The skirt required further treatment to reduce the tide lines. A systematic approach was taken by isolating the skirt pleats with disposable diapers underneath for absorption of the wet solutions. Removal of the black tide lines was successful only by flushing the line with the 0.2% anionic detergent solution of Orvus in distilled water, followed by a distilled-water rinse over the entire area, and only when the entire pleat was rewet. Basically, the tide line was moved into the seam or buried beneath the next pleat.

Treatments of the above two artifacts exhibited similar results with regard to the black tide lines. The dye from the dark brown or black cotton interfacing had extremely poor washfastness, and, when in direct contact or close to the outer fabric, the dye migrated to the silk. Because further treatment to the tide line did exhibit removal or movement, it is likely that it could be removed if the two layers could be separated. Identification of this dye is not confirmed, but it is suspected from its poor washfastness to be a direct dye or possibly a mordanted basic dye.


3.5 3.5 CASE STUDY 5 (ACC. NO. 89.26.1)

A green-and-beige plaid, silk-taffeta dress with a dark beige cotton lining dated from the 1860s suffered water damage along the back sleeve edge and proper left back hem. This dress was missed during the initial salvage operation and had dried, leaving crisp tide lines in these areas and a softened hand below them.

In testing for colorfastness, the green dye from the dress's outer fabric and dark beige dye from the skirt lining did transfer in both distilled water and a 0.2% anionic Orvus detergent solution. The skirt lining had no tide lines visible, while that of the sleeve did. The treatment objective for this artifact was to remove the tide lines in a controlled manner, while not causing the green dye to transfer or excessively wetting the lining. Therefore, flushing was not an alternative.

Conservator Yolanda Olivotto perfected a tamping technique using blotting paper to remove tide lines. Where the tide line was faint, only distilled water was applied by using a slightly damp corner of a small square (2 cm sq.) of blotting paper. The moist blotting paper was pressed against the silk, which was then immediately blotted with dry blotting paper to absorb all the water from the silk. Where the tide line was darker, the technique incorporated a 0.2% anionic Orvus solution followed by a distilled water rinse. This technique successfully removed the tide lines while not wetting out the lining. The crispness of the line was also removed.


3.6 3.6 CASE STUDY 6 (ACC. NO. 87.13.7)

A late-19th-century silk, diamond-patterned quilt was water-damaged along one side. It had been frozen in the walk-in freezer at the Provincial Museum of Alberta. A conservation team treated some of these rolled textiles at the museum within weeks of the flood. However, when the team examined this quilt, they decided that it required a more time-consuming, specialized treatment, and they returned it to the freezer. Once it was transported to the university, another team of conservators examined the quilt and proceeded to treat it with success.

This quilt was fragile, with many silk patches in a degraded state. There was dye transfer from the flood, and a student documentation report had noted pre-existing dye transfer as well. Full-immersion wet-cleaning was not an option.

The tide line and slightly beyond in the patterned portion of the quilt and to the edge of the border were flushed with the anionic detergent solution followed with distilled water. The perimeter of the flushed area was feathered out using just slightly dampened blotting paper as with the previous treatment. As an alternative, clean white cotton gloves were used instead of blotting paper. Fans were positioned nearby to hasten drying. Once the obverse was treated, tide lines on the reverse were reduced by using dampened blotting paper sparingly, as with the previous case study.


3.7 3.7 CASE STUDY 7 (ACC. NO. 85.15.1)

A turn-of-the-century wool-weft and cotton-warp cream-colored dress with an embroidered net trim over the yoke and cuffs had water damage along the proper left back hem. The water damage had left a brown tide line with a crisp edge.

Treatment of this dress proceeded as a standard wet-cleaning in a 0.2% Orvus anionic detergent solution in distilled water at 27C. The tide line was easily removed, resulting in a successful treatment.