Go back to article: Doping at the Science Museum: the conservation challenge of doped fabric aircraft in the Flight gallery
Doped fabric materials
Understanding how and why particular materials and techniques were chosen for making doped fabric is very important from a conservation perspective for two reasons. The first is due to material identification. In order to manage an object effectively it is crucial to know, or at least be able to make a well-informed assumption, about the types of materials present. This is because identifying the material type may determine conservation treatment decisions, such as the selection of adhesives and cleaning agents, as well as influencing long-term decisions about care and storage conditions.
Knowing how far generalisations can be made about the materials used in constructing doped fabric aircraft will, in addition, have major implications for determining how far conservation treatment and collections care decisions can be applied across an entire collection as opposed to on a case-by-case basis.
Beyond the material type, moreover, understanding why aircraft designers and manufacturers selected certain materials provides conservators with a much better insight into the outcomes and results they were hoping to achieve. As noted above, shrinkage of the fabric to make it taut was the primary purpose of applying dope to the fabric, but it was by no means the only consideration. Other requirements of doped fabric are summarised in Table 1, but this list is not exhaustive, and other considerations could be added, such as economy and ease of use.
Table 1 Summary of the requirements of doped fabrics and examples of materials used in their construction
These requirements are important to consider from a conservation perspective since they inform our understanding of how and why an object is significant, both in terms of its tangible and intangible values. They are the intentions and values of the original makers as represented through the materiality of the object itself, and for conservators it is important that there is an awareness of how interventions might affect the way these intentions and values are manifested in the object. A conservation treatment, for example, that sought to stabilise a tear by removing the tautness throughout the doped fabric might be very effective at preventing further tearing, but would have serious ethical implications because of its impact on one of the most fundamental thought processes underpinning the creation of doped fabric aircraft, namely the desire of the original makers to create as taut a material as possible.
The section below discusses in more detail some of the specific materials identified from the historic literature as components of doped fabric, and considers why these were selected by aircraft designers and manufacturers. Furthermore, it considers how and why the presence of these materials may have implications from a conservation perspective.
The selection of fabric was an important consideration for doped-aircraft manufacturers as historic studies found that the properties of the doped fabric were strongly related to the type of fabric yarn selected and its processing to create a finished textile. It seems that there was initially a general preference for Irish linen (Linum usitatissimum), since this was found in experiments to have the greatest tensile strength (Walen, 1918; Esselen, 1918; Turner, 1920). This tensile strength, however, came at the cost of increased stiffness, and some authors argued that cotton (Gossypium hirsutum) would provide a better alternative since, although not as strong as Irish linen under tensile loading, it did have greater elasticity, enabling it to resist tearing under the types of load actually experienced during flight (Walen, 1920; Ramsbottom, 1924).
A third alternative was mercerised cotton. Mercerisation is a process in which the yarn fibres are soaked in strongly alkaline conditions, and recent studies have shown that this causes a reorganisation of the crystalline structure of the cellulose chains that make up the fibre yarns, increasing the amount of hydrogen bonding between polymers (Wertz et al, 2010). Mercerised cotton was, therefore, found to have a greater tensile strength than non-mercerised cotton while still retaining much of its elasticity.
The debate as to which fabric was better appears to have never been truly settled and remained an open question. The USA, in particular, seems to have preferred mercerised cotton over Irish linen, though this may in large part have been down to problems of supply and the greater costs involved in the import of Irish linen to the USA (Catoe, 1962).
The only significant alternative to cotton or linen was silk, although this seems to have been dropped relatively quickly as a material in most countries before the First World War. There may have been a number of factors behind this, such as its high cost and problems of its rapid deterioration in exterior conditions. Another material investigated was ramie (Boehmeria nivea), although this seems to have never received any significant attention or use beyond experimental research (Turner, 1920).
The fabrics discussed above, barring silk, are all derived from naturally occurring plant materials, whose principal component is cellulose. Cellulose is a long-chain polymer molecule in which the separate chains are bonded together by hydrogen bonding (see Figure 4). The properties of cellulose have a major impact on the final characteristics of the different plants that it constitutes. Variations, such as in the length of the cellulose chains, how well ordered and closely packed together these chains are, and whether the cellulose chains are present in a matrix with other materials, mean that different plant materials can have very different material properties (Hon, 1994).
© Ben Regel
Structure of two cellulose polymers. The dotted red lines show the location of hydrogen bonds joining the two molecules.
This variation in fabric selection may be of significance to conservation because of the differences in the behaviour and deterioration one might expect from these materials. Cotton, for example, is primarily made of cellulose with relatively few impurities compared with linens, which often contain a much higher proportion of other compounds, such as lignin, hemicelluloses, waxes and oils, depending on how they are prepared and processed (McDougall, 1993).
These impurities have been found to be far less chemically stable than the cellulose content, and increase the likelihood of reactions causing undesired changes in fabrics, particularly under exposure to light radiation and/or oxygen. The main observable alterations might include colour changes, such as yellowing, and a decrease in the tensile strength and flexibility of the material (Hallett and Bradley, 2003). Of particular concern is the release of acidic compounds by the breakdown of the impurities under exposure to light, since this can lead to acid hydrolysis of the cellulose, which shortens the cellulose chains, leading to a marked reduction in strength (Seery, 2013).
Exposure to light was observed to be a major cause of deterioration in doped fabrics during the early years of flight, when it was noted that aircraft fabrics deteriorated much faster during summer months compared with winter ones. Further testing using more controlled light sources as well as sending test pieces out to places with high light levels, such as Iraq and Egypt, gradually established that it was the more energetic UV rays of sunlight that were causing most damage (Ramsbottom, 1924; Atkins and Woodcock, 1917; Wendt, 1922; Padfield, 1969).
Besides light, another problem encountered with aircraft fabrics was moisture exposure. During periods of high humidity, moisture uptake would cause fabrics to ‘slacken off’, the tautness being restored once the relative humidity (RH) dropped again and the fabric was able to dry out. This indicates that expansion and contraction cycles of the material occur, which are dependent on cycling of the RH conditions. This type of cycling has been established to be a cause of damage and deterioration in various types of museum materials that conservators deal with, such as ivory, wood and those in paintings, and it seems probable that it would play a part in the deterioration of doped fabric (Michalski, 2011).
Exposure to high RH levels, in addition, risks organic growth developing, such as mould. This is a health and safety issue for those who might have to work with the objects owing to the release of spores into the atmosphere. It can also lead to a weakening of the material, as studies have shown that organic growth on objects can release enzymes that break down the cellulose (Edwards and Falk, 1997). Such a process causes the long-chain molecules of the cellulose to break down, reducing the structural integrity of the material.
The selection of the polymer to be used in dope formulations was very important as it was this polymer that would form a film within and around the fabric fibres, causing it to shrink and grow taut. Early, pre-doped fabric aviators experimented with a number of readily available options, such as casein, glues and resins, but none were found to be particularly effective (Drinker, 1921; Smith, 1919).
The main group of materials discovered to be effective tauteners was cellulose esters. These are semi-synthetic plastics, in which the cellulose chain is modified by replacing a proportion of the hydroxyl groups attached to the main chain with a new chemical group (see Figure 5). Depending on the type of group used to replace the hydroxyl group, a very large number of products is theoretically possible, whose properties may further vary depending on factors such as how many hydroxyl groups are replaced, the distribution of the substituent groups along the chain, the length of the polymer molecules and the presence of impurities from the production mechanism (Zugenmaier, 2004).
© Ben Regel
Structure of cellulose derivative products used in dopes. R represents the location of hydroxyl groups that may be substituted.
Cellulose nitrate was the first cellulose ester to be used in aircraft dopes in the years before the First World War. It was a major step forward, having very good tautening powers, and set the standard against which alternative polymers would be judged for years to come (Kline and Malmberg, 1938; Reinhart and Kline, 1939; Inoue, 1941). It was, however, also highly flammable, and so more stable alternatives were sought, which led to the development of dopes based on cellulose acetate, cellulose acetate butyrate and cellulose stearate.
Each of these cellulose derivatives has different properties and characteristics that provided them with various advantages and disadvantages. No definitive decision was ever made as to which polymer was best for tautening and all – except cellulose stearate, which appeared too late in the history of doped fabric to be widely investigated or adopted – were used in many different contexts and parts of the world (Weissberg and Kline, 1949; Meyer and Gearhart, 1944; Dreyfus, 1949; Jones and Hockney, 1944).
The presence of these cellulose-derived materials is a potential challenge from a conservation perspective. All of these polymers are potentially unstable, undergoing chemical degradation on exposure to various conditions, such as UV light, heat and moisture. The result of such chemical changes might be the loss of substituent side groups from the polymer chain or even scission of the main polymer chain itself, leading to a drop in molecular weight (Berthumeyrie et al, 2014; Quye et al, 2011; Selwitz, 1988). These reactions can have a significant effect on various properties of the material, such as its colour, tensile strength, flexibility and acidity.
As with cellulosic fabrics, light is particularly damaging to objects containing cellulose derivatives. This is because the UV energy is sufficient to break the bonds holding the side groups – the acetate, nitrate or butyrate groups – to the cellulose polymer backbone. Once released, these chemical groups are then able to react with moisture to form acidic compounds. These acidic compounds then react with the main polymer cellulose chain, causing it to break down and shorten.
This has serious consequences for the properties of the material, since it affects the length of the polymer chain molecules. As these shorten, the tensile strength of the material is decreased as there is a greater number of weak points in the material where failure might occur. It can also result in a less extensible and more brittle material, further increasing the risk of failure. This is because the areas most vulnerable to degradation reactions tend to be amorphous portions of the material from where much of the material’s flexibility is derived, and the reactions may also lead to cross-linking of the polymer chains, which again limits their ability to move past and around each other (Hon and Gui, 1986).
This history of the material itself, in terms of how it was produced and its quality, is also highly relevant to the issue of its conservation. When making cellulose derivative products, sulphuric acid is often used as a catalyst in order to substitute the hydroxyl groups of the cellulose chain with the required chemical group. If the final product is not then properly purified, trace sulphuric acid can remain in the material, again leading to acid hydrolysis and chain scission of the cellulose backbone.
Besides the polymer of the dope responsible for forming a film, numerous other compounds might be added to manipulate the film in various ways. These could include plasticisers, fire retardants, pigments, UV reflectors and absorbers, and fungicides (see Table 1). The list of materials that might be added therefore includes, but is not limited to, items such as triphenyl phosphate, castor oil, aluminium powder, iron oxide powder, a boric acid and borax mixture, and inorganic pigments such as carbon black and yellow ochre. The proportions of the mixtures would also vary, some having higher quantities of the different components, depending on the particular manufacturer and dope purpose.
Use of these additives significantly complicates the potential long-term behaviour and predictability of the material, as it introduces further potential reactions and processes. Metallic compounds, for example, have been found to act as catalysts in the degradation of other cellulosic materials, and triphenyl phosphate degradation is believed to potentially release phosphoric compounds which may form acids, again causing and increasing the rate of degradation reactions (Daniels, 1999; Tsang et al, 2009; Groom, 1999).
As well as the dope ingredients, the solvent mixture in which it was all dissolved in order to be brushed or sprayed onto the fabric could also vary significantly. Drinker (1921) reports that over fifty solvent formulations were in use at various times by the Allies during the First World War alone, which could result in unintended and unanticipated consequences. Tetrachloroethane, for example, was a common solvent in use in most countries prior to 1916, but research at the Royal Aircraft Factory indicated that when trapped within the material it could react in sunlight, releasing acidic volatile compounds which then reacted with the fabric substrate, increasing its rate of deterioration (Robertson, 1916; RAF, 1915). This, coupled with the fact it was linked with several fatalities and numerous health issues, meant it was phased out of British dopes in 1917 (Hamilton, 1918).
Other products, such as varnishes and paints, might also be combined with the doped fabric to act as a protective coating against environmental conditions. These varnishes created a waterproof finish, and could be pigmented with metal powders or other compounds to absorb and reflect sunlight before it could reach the doped fabric. There were numerous varnishes developed during the First World War, such as Protective Coating 10 (PC10), which was a nitrocellulose-based varnish containing carbon black and yellow ochre pigments, and V84, which was a nitrocellulose varnish containing aluminium powder to absorb and reflect the energy from the Sun’s rays (British Engineering Standards Association, 1918).
The documents stored at the NA make clear, however, that this was an area of constant change and development as new ideas and supply issues meant that new materials had to be developed. In 1918, for example, two new products named pigmented oil varnish (POV) and clear oil varnish (COV) were introduced which were made with tung oil (Vernicia fordii) and intended to replace varnishes such as PC10 and V84 because of a shortage of nitrocellulose.
Towards the end of the First World War, the British also began experimenting with pigmenting dopes directly by adding aluminium powder straight into the dope formulation. This obviated the need to add any varnish or protective coat over the doped fabric, and in time appears to have become the most widely accepted method for providing environmental protection. Such pigmented dopes, however, would only usually be applied in the very top two coats to save on weight, with the first few coats below remaining unpigmented.
The dope, therefore, not only made the fabric taut, but protected it from environmental factors that might cause it to weaken and fail. In this role it acted as a sacrificial material designed to protect the more expensive fabric below, and was expected to deteriorate in the environment. A time frame of 60–90 days was suggested before the dope should begin to show any negative signs of ageing, such as peeling away from the fabric and discolouration, when it would require replacing (Esselen, 1918). This raises an interesting conservation question as to whether conservators should necessarily attempt to preserve dope for as long as possible, or should be willing to renew it periodically given the transient nature of the material.
Doped fabric summary
From the above discussion it appears that some general assumptions can be made about doped fabrics, primarily that they will consist of some type of cellulosic fabric impregnated with a cellulose ester-based compound. It is also important to be aware of some recurring causes of deterioration. Exposure to light in particular can be highly damaging to doped fabric materials, as can exposure to fluctuating RH. More research, however, is required to establish the deterioration mechanisms taking place and how these alter the characteristics of doped fabrics.
On the other hand, these generalities still mask a very high degree of variability, with different materials and processes used in different times and locations. A few examples of dope recipes proposed and used in various times and parts of the world are provided in Table 2. These examples help to indicate the range of materials used just within the dope, but it is also worth remembering that this variability extends to other decisions such as the choice of fabric and the application of protective varnishes.
From a conservation perspective, this means that caution is required when devising and applying treatment methodologies. Differences between doped fabrics may mean that a treatment that is effective in one context will not work as well in another. It may also mean that types of deterioration that occur for one object will not necessarily be common or expected across all types of doped fabric aircraft.
Table 2 Examples of different dope formulations*
Component DOI: http://dx.doi.org/10.15180/160605/011