JAIC 2003, Volume 42, Number 2, Article 9 (pp. 341 to 362)
JAIC online
Journal of the American Institute for Conservation
JAIC 2003, Volume 42, Number 2, Article 9 (pp. 341 to 362)

BEVA 371 AND ITS USE AS AN ADHESIVE FOR SKIN AND LEATHER REPAIRS: BACKGROUND AND A REVIEW OF TREATMENTS

LISA KRONTHAL, JUDITH LEVINSON, CAROLE DIGNARD, ESTHER CHAO, & JANE DOWN



4 ASSESSMENTS OF PAST AMNH TREATMENTS OF SKIN


4.1 ADHESIVES USED AND RATIONALE

BEVA 371 was introduced as an adhesive in the objects conservation laboratory at the American Museum of Natural History (AMNH) in 1985. Initially it was utilized in its solution-based form, either applied directly to the edge of the join if it was very small, such as in a butt join of a fringe element, or applied to a lining material and, upon drying, heat-set or solvent-reactivated in place. After BEVA film became available, it soon replaced the solution as the predominant BEVA 371 adhesive form used. A small selection of lining materials has been used with BEVA at the AMNH, primarily Reemay, Cerex, and goldbeater's skin.

Three other types of adhesives were used less frequently on skins during this period. These included polyvinyl acetate (PVAC) resins, Lascaux 498 HV acrylic dispersion, and Paraloid F-10 acrylic resin. Though the PVAC resins (grades AYAA and AYAC) were often favored among these because of their working properties, BEVA 371 heat-set continued to predominate over all choices. It was found that there was increased control over the degree of saturation of BEVA 371 into the substrate and therefore fewer tendencies for the adhesive to stain.

The focus of the evaluation to be described in the next section was to assess the long-term stability of repairs made using BEVA 371. These assessments were made by visual and physical examination of the repairs and the surrounding areas of skin. Some interesting trends were noted during the survey in relation to the other adhesives mentioned above, and these observations will be summarized as well.


4.2 ASSESSMENT OF BEVA-BASED TEAR REPAIRS

In 2002, 58 ethnographic skin objects treated during the years 1985 to 2000 using either BEVA 371 or other selected adhesives were reexamined in an attempt to determine the long-term stability of the repairs. The objects surveyed included skin clothing and apparel such as coats, leggings, aprons, bags, and shoes and musical instruments such as drums and harps. The types of skins examined included untanned (rawhide) and tanned (brain, oil and smoked) skins. A total of 49 of the 58 objects in the survey had tear repairs or fills made the BEVA 371 products.

The results of the survey demonstrated that an overwhelming majority of repairs and fills utilizing BEVA 371 were still intact and flexible from 2 to 17 years after treatment. A small number of these objects (5 out of a total of 49, or 10.2%) had repairs that had either failed or were beginning to exhibit minor changes, such as lifting or incipient lifting of the lining material from the substrate. It should be noted that the number of repair sites on an object varied from one to more than a dozen, and if just one backing out of several was beginning to separate, this treatment was counted as exhibiting changes. Of the five objects that exhibited changes indicating failure, three consisted of tear repairs made using BEVA film and two consisted of fills whose media was BEVA solution. No other types of deterioration, such as stiffening or related discoloration of the skin or adhesive, were observed. It is also important to note that all five of the objects exhibiting some degree of repair failure were included in traveling exhibits. As such, these objects were subjected to extensive handling engendered by repeated packing, unpacking, and placement of the objects on mounts.

Of the small number of objects with failed tear repairs (3 out of 40), the cause of the deterioration of each repair can be explained by choices made by the conservators, rather than by aging or deterioration of the adhesive. As previously outlined, many variables can affect bond strength, such as the choice of backing material, the quantity of adhesive applied, the temperature used, and the dwell time. A more precise cause of failure is sometimes difficult to pinpoint. Factors that appeared to promote weak or failed repairs in these cases included poor choice or preparation of the lining materials, inadequate size of lining material, and possibly inadequate preparation of skin surface prior to repair application.

Most commonly, the lining material may not have been properly matched with the weight and flexibility of the object, or the adhesive was insufficiently or inadequately applied to the backing. A thick-grade Reemay was often intuitively chosen as the lining material for large or heavy objects or for repairs on thick, relatively inflexible skins because of its heavier weight and strength. Such a choice was made, for instance, for the repair of a Mangbetu fur belt with numerous heavy dangling ornaments. The linings were applied with BEVA 371 film to stabilize the incipient tears on this object. Examination of the belt revealed that two of the five backings were in the process of completely separating from the surface, and another two linings exhibited lifting around their edges. These results may be partially due to incomplete application of the adhesive all the way to the edges of the backing as well as the use of cut rather than feathered edges. In retrospect, Reemay may sometimes be too thick or tough a lining material for particular objects. It is difficult to feather out and it loses adhesive during the feathering process, therefore providing less contour and grip at its edges. This loss becomes especially problematic for objects that are flexed during handling. Feathering the edges plus using an excess of adhesive at the edges would probably ameliorate much of the edge lifting seen within repairs using Reemay. Furthermore, observation of numerous repairs during the course of this survey indicated that for thick skin objects, such as the Mangbetu belt, application of two layers of Cerex, a thinner backing material, sandwiched with a second layer of BEVA film may provide better longevity.

A complicating factor when analyzing the cause of repair failure as seen on the Mangbetu belt is the fact that the belt was oil-tanned during manufacture and may have retained oily residues on its surface. While BEVA 371 has been recommended as an adhesive for oily or waxy surfaces, it is possible that the residues left on the skin surface of the belt may have compromised the integrity of the mends. Repairs of this type would probably be more successful if the skin were superficially degreased by swabbing with an appropriate solvent for oils prior to application of the BEVA and backing. Additionally, past treatments on objects at AMNH also exhibiting oily surfaces, such as Chinese shadow puppets, indicate that stronger bonds may be obtained through direct application of BEVA 371 solution to the surface of the repair area and, upon drying, heat-setting BEVA film with the appropriate lining.

Examination of numerous repairs and fills applied to a highly deteriorated shaman's apron from the Luba culture of Africa revealed a Cerex backing lifting from the skin surface at one repair site, with a strip of inadequately heated BEVA film remaining as a separate intact layer. Treatment procedures outlined previously such as temperature and exposure time to heat are important factors in these kinds of repairs. As with the choice and preparation of lining materials, conservators should be meticulous about ensuring that a complete bond is achieved during execution of the treatment. Additionally, this object had a number of cut rather than feathered backings that were lifting at their edges. Upon examination of the reverse side of the backings, it was evident that the BEVA film had not been applied all the way to their edges. Even if the edges of a backing are a significant distance from the area requiring support, an inadequately adhered lining edge could lead to further lifting of the lining if the object were to undergo excessive handling.

Another example of repair failure occurred as a result of handling during the course of this survey. A Blackfoot shirt was fit quite snugly on a mannequin. Manipulation to examine a Reemay–BEVA film repair caused the backing to completely detach from one side of the join. It was not clear what the source of this join failure was. While the surface area of the join was adequate, the appearance of the adhesive on the detached backing seemed to indicate that the repair may not have been carried out to a sufficient degree of saturation (as controlled by the dwell time) or with enough heat. Both of these treatment elements would be necessary to ensure a sufficiently strong bond to withstand the handling and continual stress of years on a closely fitted mannequin.

A final source of join failure was observed on another of the artifacts surveyed. In this case a detached tassel of a Blackfoot bag that was repaired prior to travel in a multistop exhibit clearly failed, as neither the tassel nor the lining was extant at the time of the survey. For this repair, the lining material used for the join was probably too small for the requirements or the use of the object. Other remaining tear repairs at less vulnerable locations on the object were extant but were very tiny in size. It should be noted that the lining material must span a sufficiently large area beyond the tear to provide ample surface for the adhesive mend, thus allowing for handling and manipulation of the object.

Most of the failed repairs described above were executed 12 to 17 years ago. It is clear that a history of the use of BEVA over this period of time engendered experimentation with techniques that led to greater finesse in skill. As mentioned, none of the BEVA 371 tear treatments appeared to fail as a result of chemical problems within the adhesive, but instead from specifics concerning inappropriate treatment decisions and procedures chosen for the particular object.


4.3 ASSESSMENT OF BEVA 371–BASED FILLS

Nine of the 58 objects in the survey had applied to them fills that used BEVA 371 solution as the adhesive element. The composition of the fills made to skin or leather objects utilizing BEVA 371 was of two types: those that included glass microballoons as the inert filler and those that included cellulose powder or pulp as the inert filler. Both types of fills were internally colored by the inclusion of dry pigments in their mixtures. These fill mixtures were empirical. They were not formulated according to weighed-out amounts; hence, their properties could vary widely. Too much inclusion of the inert fillers could result in less flexible fills or fills that could separate from the edges of the skin due to inadequate tack or adhesion.

Earlier fills at AMNH consisted of mixtures of BEVA 371 solution mixed with glass microballoons and dry pigments applied over Reemay backings (Kaminitz and Levinson 1988). These fills were often used to replace losses or cracks in rawhide coverings that were stretched over the wooden bodies of musical instruments. Where possible, Reemay or another type of backing was slipped under the skin and held in place using PVAC resin (chosen for its high degree of tack). At times narrow surface cracks in the skin were filled with the mixture in order to achieve visual integration; this was necessarily accomplished without the backing. More often, the BEVA mixture, of thick putty consistency, was applied over the Reemay with a spatula, and upon setting was shaped and textured with a heated spatula. Because there is a high degree of tension in the stretched leather or skin, these instruments are particularly reactive to fluctuations in temperature and relative humidity. Therefore, the fills required sufficient flexibility to allow for movement of the skin over the rigid core and adequate cohesive strength and tack to adhere to the thin edges of the remaining skin.

Of the nine objects with fills that were examined, six were musical instruments treated with BEVA solution–glass microballoon fills. The instruments were on exhibit for 2 years, traveled to several venues, and then spent 10 years in a stable storage environment. During travel, one of the objects with narrow, unbacked fills exhibited separation of the fill from the surrounding skin. It was easily remodeled with a heated spatula to the edges of the skin. During the recent survey, this harp again exhibited slight separation of the same fill from the edges of the skin. Another harp that initially showed severe distortion of the skin and numerous repairs was stable during the length of the exhibit. However, with time the skin has contracted, causing the fill to separate to an unacceptable degree. Thus, the survey revealed that two out of six repairs to objects of this type exhibited some degree of instability. The release of the edges of the fill from the skin appears to be a failure of a physical nature and is not attributed to the chemical stability of the resin. Given that there is so little of the actual adhesive present along the edges of these fills, and given the inordinate amount of strain and flexing that such a repair can undergo, this degree of failure is not surprising. This is especially the case with the fluctuations in temperature and relative humidity posed by travel. Perhaps pretreating the edges of the loss with BEVA 371 solution and then applying the fill material would prevent such failure.

In another early repair, the BEVA 371–glass microballoon fill material was applied directly to the surface of a couple of tassels on the vegetable-tanned skin of an African club, thinly filling surface losses that had resulted from insect grazing. The tassels, though fragile because of such extreme loss, were not backed to lend more strength. Currently, they are intact, though the fills are separating around their edges from the skin. Again, the instability of these fills appears to have nothing to do with a chemical interaction of the adhesive and skin. Instead, it appears to be an example of inadequate choice of repair technique by the conservator in which a simple backing would have better addressed the issue of stability. No matter how much flexibility or tack the adhesive may have, application of a small fill to a very mobile, flexible element will most likely result in separation of fill edges from the substrate.

The second type of fill mixture, consisting of BEVA 371 in solution with cellulose powder or pulp and dry pigments, was developed to be a more flexible fill material. It was used in the laboratory from 1997 to the present, specifically to fill losses in two objects composed of deteriorated skins with sueded surfaces—a Luba shaman's apron and a Yakut shaman's coat. The mixture was used over Reemay or Cerex backings and, in these cases, was applied to objects that required a considerable amount of handling prior to exhibition. The two objects thus treated have remained remarkably stable. The fills currently retain their flexibility, and the adhesion of the fills to the edges of the skin has been sufficient throughout the handling imposed during mount making, packing, and exhibit. It should be noted that the flexibility of this fill material can vary from fill to fill because of the empirical nature of the mixtures. This inconsistency could be easily addressed by using weighed-out formulations to create facsimiles, obtaining more reproducible properties.


4.4 ASSESSMENT OF AYAA AND AYAC (PVAC) RESINS AND LASCAUX 498 HV REPAIRS

Of the 58 objects reexamined in the survey, 10 included repairs made using PVAC resins (AYAA or AYAC) or the acrylic dispersion Lascaux 498 HV consisting of 56% butylacrylate-methylmethacrylate copolymer (Down et al. 1996). Though considerable treatment failure was noted in this group, no conclusions have been made about the reasons for the failure since the sample size is too small and it is outside the scope of this study. However, interesting trends were noticed. Of the seven objects that were repaired using one of the PVAC resins, two had mends that had failed. Visual inspection indicated that the reasons for these failures could be similar to some of the failures seen with BEVA 371 repairs. In one case, the backing size was possibly inadequate for the repair demands of a highly responsive drum skin; in the other, the repair was made using discrete dots rather than a continuous bead, which has been shown to be a weaker repair method. None of the adhesives at these failed joins appeared particularly degraded or yellowed.

Two objects repaired using Lascaux 498 HV were also reexamined, and both exhibited some degree of failure. In one case, the adhesive was chosen because the conservator wanted a weak bond that would release if the area of the mend was under too much pressure. In that sense the treatment could be considered successful, as the repair backing had totally released from the drum skin. In another situation, the synthetic web backing material employed appeared too thin for the weight and flexibility of the object— a large, extremely heavy tipi. Additionally, the adhesive may not have had enough strength for the necessary repairs. The high percentage of failure in this small sample size suggests that further examination and study may be needed.


4.5 SUMMARY OF ASSESSMENT RESULTS

A survey was conducted of nearly 60 skin and leather artifacts at AMNH treated between 1985 and 2000, where the majority of repairs used BEVA 371 film or solution as the adhesive. Results indicated that in most cases the BEVA 371 tear repairs and fills have remained intact: they are visually unchanged, and the mends continue to provide the stability required. Of the 10% that showed any degree of failure, the changes could usually be explained by unsuitable treatment decisions often closely related to how the object would be used in the future. Several factors that could influence the long-term success of these repairs include the nature, size, and thickness of the backing material, the degree of heat used and the time it was applied, and the degree of saturation of the adhesive into the skin and the backing. As expected from its known stability, chemical deterioration of the adhesive does not appear to be at cause. In the few instances where the adhesive failed, or was intentionally reversed to provide a view of the adhesive under the backing material, no yellowing or other discoloration of the adhesive was observed. It is clear that individual treatment procedures and choices are key factors in the longevity of repairs made with this substance.


Copyright 2003 American Institution for Conservation of Historic & Artistic Works