EU PROJECTS‎ > ‎Craft West‎ > ‎ARKÈ PROCESS‎ > ‎

Arkè process - part 2

3.3 Impregnating agents selection

3.3.1 Introduction

In the present chapter, we study the starch as consolidating material for the structure of the waterlogged wood. For its different uses its form vary and also some important characteristics that made of it an important material for the consolidation and conservation. For this reason we have prepared a detailed study of it. Its possible sources, its constitution, its properties and the methods of its characterisation will be here described.

3.3.2 Choice of the starch for the consolidation of the waterlogged wood

The different treatments methods for the conservation of the object of waterlogged wood have essentially the same aim. It consists in the extraction of the water from the structure without causing a notable deformation of the objects. Compared to the initial state of the fresh wood, after being waterlogged it lost a relatively important amount of the initial matter. The causes of this loss have been already described. Many possible solutions to recovery the waterlogged wood avoiding its deformation have been already mentioned . The common thought is that the best solution is to reinforce the wood structure, so the deformation in the structure will be decreased. Stamm (1942) demonstrated that the wood species that contain many water extractable matter should have less strong shrinkage. The author justify this with the fact that the water extractable matter, localized in the fibrilles of the wall cell, avoid the normal retraction of the wood.

The aim of the impregnation in the waterlogged wood conservation is to assure a dimensional stabilization. To do this it’s important to let the consolidation material go inside the interstices of the wall cell. The rigidity of the wood structure comes from the same sites. But we know that only water and the substances which molecules have a size similar water can reach these interstices.

It’s normal that the wood undergoes some deformation - even if very small- when the water is taken out from its structure. This is what happen in the green wood when it is dried. More, the shrinkage level is variable in function of the axial, radial and tangential dimensions. It’s possible to assure a certain reduction of the shrinkage if we bring some substances inside the cavities of the waterlogged wood. We will reintroduce the substances that already existed in those cavities of the wall cell of the wood, for example the water extractable as the starch and sugars; doing this we could bring the waterlogged wood shrinkage similar or equal to the fresh wood one.

The different experiences showed that to reinforce the waterlogged wood it’s convenient to transfer the matter through a liquid support. It’s important that the consolidation substance is soluble in the water and that it doesn’t cause any danger to the remaining wood structure. A priori, the cellulose and lignin should be the two constituents ideally indicated to reach that aim. But they are soluble only in strong solvents that usually provoke some damns to the constituents of the wood structure, particularly of the cellulose and hemicelluloses. On the contrary, the celluloses and lignin are not soluble in water, only a neutral fluid could play a fundamental role at this level.

So it’s necessary to demonstrate that only a few substances can satisfy the ensemble of the ideal conditions. The selection of the substances of consolidation must be done to satisfy a certain number of important conditions for the validation of the procedure. So the main criteria to chose the impregnation material are the following:

✔  the availability of the consolidation material; the starch answer perfectly to this request

✔  the low cost of the chosen material: the starch is cheaper than the PEG

✔  it should allow to stabilize as much as possible the dimension of the wood objects

✔  it should allow to conserve the state of the surface, included the colour and the absence of crackings , the starch doesn’t give any particular coloration to species of wood containing a notable quantity. It’s coloration, generally white, it’s visible only in the fruits, grains and tubercles where it represent the main component. As it constitute the reserve of the wood parenchyma, it can play the role against the tensions that cause the cracking during the drying phase.

✔  it should be able to produce chemical links with the wood constituents; the starch, in its macro molecular structure, is very similar to the cellulose structure, the main constituent of the wood. From this point of view some chemical links can be created between wood constituents and added starch.

✔  It doesn’t have to present any manipulation risk for men, neither bring to any degradation of the wood structure: elaborated for current uses, as the alimentation of humans and animals or for the production of industrial paper, textiles and glues, all the commercialised starches used in the present study don’t contain any toxic product and don’t degrade the cellulose.

✔  It will not have to need a prolonged heating during all the impregnation phase: the starch in a well defined concentration in water, after being gelatinised and solubilized, don’t need any heating if the suspension is continuously stirred. A certain mechanical stirring it’s sufficient to avoid the sedimentation of the starch.

✔  It mustn’t be a volatile substance because it must stay inside the wood structure during the rest of the treatment: the starch, because of its macro molecules, is not volatile. It must be completely carbonised to be combined with oxidising agents and to produce volatile products.

✔  It will not have to generate tensions that can cause shrinkage or other structural modifications in opposition with the aim of reaching a dimensional stabilization.

✔  The consolidating agent will favour the elimination of the water contained in the wood without completely substituting it. So the water could be used as support for the transfer of the consolidating agent that should be solubilized; the solubility increases the consolidation agent diffusion in the deep structure of the wood. We underline that the starch, that is insoluble in the native nature, once modified, can acquire the chemical-physic characteristics that give him solubility

So it’s clear that the starch satisfy the major part of these criteria, and can show a higher performances than the different resins and substances traditionally used in the waterlogged wood sector.

The introduction of a foreign material as substitute of the cellulose (better, hemicellulose plus cellulose) and of the natural lignin is an approach method that will modify the natural structure of the wood. So it’ll be necessary to choose some substances that will not be too different from the wood original constituents.

It’s probably for this reason that the most part of the researchers are oriented towards the general use of molecules done with C, H and O.

In the present restoration-conservation methodology, there is a strong reason to choose the starch: its structure is very similar to the cellulose in comparison with other synthetic resins. Moreover the it’s a natural substance coming from a vegetal matter. The starch is a biopolymer very abundant in nature, so ecologic, renewable and very cheap (between 0.3 and 0.6 euro/kg – industrial production). The PEG of high molecular mass are insoluble as the starch in cold water. These solid resins need some heating during the impregnation phase. Some precautions must be observed during the starch impregnation. It’s important to not introduce, together with the starch, other substances that could provoke some dangerous reaction with the substances existing in the waterlogged wood.


Figure 1 : Schematic representation of the molecular conformation of the amylosium (A), amylopectin (B) and cellulose via cellobiose (C).

The most part of the solvents already used in the multi-stages methods of substitution as the acetone (acetone - PEG or acetone - colophon) are dangerous for the wood constituents.

The acetone causes a lignin dissolution, that are known as origin of the wood rigidity. The alcohols as the methanol and ethanol have been used to extract the peptic substances that protect the wood. Other solvents (benzylic alcohol, dioxane, cyclo hexanone, diethylic ether, isopropylamine, etc) dissolve the cellulose and its derivatives. It’s clear that the best solvent, compatible with the water that is inside the wood, is the water itself.

3.3.3 Physical properties of the starch

The raw starches, also named native starches, have many different chemical and physical behaviours. In fact the grains of raw starch seems to vary in function of the sources. For each kind of plant the size and the shape are distinct. The SEM observation showed that is possible to characterize and to recognize the source of the starch simply by its grains. In the following table some dimension and characteristic shapes of some utilized raw starches have been reported. Some of the dimensions have been found in literature. A complete list of the starch sources has been written by Moe (1979). Other varieties of starch, usually the granular starches used in this study, have been directly visualized and measured starting from the SEM analysis, confirming the values of the literature.

Sources (%)

Mini (μm)

Maxi (μm)

Average (μm)


Maize (71-74)




Granules nearly uniform and angular

Potatoes (65-85)




Small circular grains, big oval and irregular

Wheat (66-69)

From 1 to 5

From 35 to 50


Circular grains sensibly spherical but without intermediate shapes (Brautlech 1953)





Composed grains of different shapes (Moe 1979, Brautlech 1953)


From 10 to 20

From 50 to 70


Grains of oval and ellipsoid shape (Moe 1979, Brautlech 1953)


From 1 to 3

From 5 to 10

From 3 to 5

Grains more angular similar to the maize ones (Lelievre 1974, Brautlecht 1953)

Sweet potatoes




Grains similar to the maize ones and of different sizes (Moe 1979)

Sugar cane




Simples grains but bigger  (Moe 1979)





Simple grains, irregular an  oval shape (Moe 1979)


From 3 to 5

From 10 to 12


Assembled grains separated in function of the treatment




From 25 to 30

More big than small grains in %, Oval shape and spherical (Moe 1979)





Small and big grains (Moe 1979)

Table 1: Size and shape of starch grains from different sources (dimension in μm). 
In some cases there are the % of starch near the source of it.

3.3.4 Hydro thermal properties of the starch

When suspended in the water, the starch raw grains are insoluble at room temperature. If the starch suspension is heated, the grains start to gelatinise at a temperature superior or equal to 60°C, absorbing the water, swelling, and only partially solubilizing [French 1971, Lelievre 1973, Donovan 1979, Biliaderis 1980].

The gelatinisation could be defined as a endothermic reaction during which a phase transition, from crystalline to amorphous phase, resulting from the fusion of the crystalline structures of the starch in presence of the water. The starch forms a colloidal suspension with the hot water. Its solubility is function of the temperature [Biliaderis & al 1980, Bowler & al 1980] and of its concentration in the water [Eliasson 1986, 1985, 1983, 1980; Colonna & Mercier, 1985]. The three-dimensional structure of the amylosium allow to retain, in the lattice, the water molecules to form a gel: it’s the jellification that is the characteristic to form punctual junction zones. A starch paste is thus obtained and, after a prolonged cooling, tends to a partial re - crystallization: this phenomenon its called retrogradation [Keetels 1995, Moe 1979]. It’s recognized that the retrogradation of the starch paste during the cooling is provoked by the precipitation of the amylosium inside the starch solution and the realignment of the free molecules trough the hydrogen bonds. The retrogradation is emphasized by a consequent realignment of the amylosium molecules with the participation of the long chains of amylopectin. It’s one of the reasons why some authors said that the waxy starch doesn’t contain nearly at all the amylosium (amount lower to 2%) and don’t produce gel (Biliaderis et al. 1986). This probably explains why the starch paste rich in amylopectin are surely more stable, and why some starch varieties are more required by the paper, textile and glue industry. To stabilize the starch paste towards the retrogradation phenomenon, some additives as some derivatives of the fatty alcohol called “ethoxyles” are used. These substances are obtained by reaction of the ethylene oxide (9 moles) with the alcohols in C12-14 and C12-15 and are generally utilized in concentration ranging between 0,1 and 1,0% for starch suspensions in water with a concentration ranging from 5 to 40%.

3.4 The impregnation process with starch and derivatives

3.4.1 The impregnation kinetic

The modelization of the impregnation kinetic, mainly based on the diffusion process of Fick type, is based on three principal parameters: the concentration of the «solution» (expressed, as g/l), the diffusivity D of the «solution» considered within the material and the sample shape.

In these analysis we shall dedicate a different approach if the impregnation is made with a real solution (soluble compounds, mixtures of different soluble compounds, dextrin, sugars, etc. used before their solubilization in water) or a suspension (suspended starch grains in a impregnation environment constituted by water only, or in presence of additional sugars, dextrin and different starch soluble compounds, raw or degraded).

In this second case, at different stages and speed, the sedimentation phenomena could occur; for this reason the stirring will be necessary to avoid that phenomena and to maintain an homogeneous «suspension» at it nominative concentration. The Ddiffusivity depend on the nature (physical-geometric) of the sample and on the nature of the «solution» and mainly on the viscosity. Impact of the starch concentration

From the first characterisation we tried to find the impact of the different concentration of different starch varieties and of the combination of starch and derivatives on the rate of impregnation of the different samples of wood within a fixed duration. The three series of test impregnation have been developed, each on a period of 10 days. The Figure 1 represents an example of the obtained results.

Figure 1: Impact of the starch concentration on the rate of impregnation at a room temperature with a duration of 10 days. 
The impregnation rate is determined starting from the mass and is confirmed by the measures of the dry matter content of the wood.

These measures realized at room temperature show that the evolution in the dry matter content depends firstly from the concentration of the starch in the solution independently by the treated wood samples. The impregnation rate at a fixed concentration will vary depending on the shape, porosity, dimensions and also from the essence of the samples. This results, more than others, allow to confirm the existence of a starch concentration for which the impregnation is at its maximum for any kind of waterlogged wood considered. In fact, the concentration must be accompanied by a great viscosity of the mixture, such reducing the value of the diffusivity D.

The new obtained dry matter content (near and sometimes higher than 10 g/100 g of waterlogged wood) means a good efficacy of the impregnation.

In the case of all starch solution, the optimal concentration seems to be from 65 to 75 g of dry starch for litre of water. The value for the experimented treatment is 70 g of dry starch / litre of water. In the case of mixtures of starch, the optimal concentration seems to decrease from 60 to 70 g/l probably because of the important parameter of the viscosity. The results giving the better impregnation rate for the majority of the wood samples are due to a medium concentration of 66 g of dry starch per litre of water; this is the chosen value for the following tests. Impact of the degradation level of the samples

These tests show a difference in the behaviour regarding the absorption of the starch depending on the nature of the samples of the waterlogged wood considered.

Despite of the degradation heterogeneity of the different wood samples, we can state that the harder is the wood (i.e. in the case of the lots of woods 6 E and 15 E), the worse it’s impregnated. We can expect an effect of re-uniformation of the dry matter content of the different type of samples. 

It’s possible to notice that some wood samples, characterized by an elevated water content and thus supposed more degradable, can present a low impregnation rate (samples ECBL of the Figure 1 ). Despite of their advanced degradation level, these samples seem to be different because of a particular structure (cell walls forms a barrier to the impregnation) that don’t help absolutely a impregnation process. The analysis of the internal structure of these varieties of wood allow to say that the modalities of link of the cells, particularly well conserved, constitute a barrier hardly over passed by the starch particles. The impregnation, so, happens when the starch fractions are completely soluble in the water. These fractions, because of their low molecular mass and of their perfect solubility are able to permeate inside the wood to remain there after desiccation and after the water elimination. The analysis of the impact of the nature and of the varieties of the «solutions» of starch is the objective of the following paragraph.

3.4.2 Selection of the starch types

The selection among the various kind of starch and of starch mixtures studied has been carried out on the basis of the level of absorption obtainable and first of all on the basis of the stabilization of the dimensions and of the colour of the sample. Three kind of granular starch and four kind of starch derivatives has been selected. Among selected granular starch, all with a low level of solubility, are characterized by an higher content of amylopectin (between 94 and 98%). These are starch types commonly used by the industries of paper, textiles and glues production.

The selection of the four kind of dextrin has been carried out on the basis of their wood impregnation capacities and of their synergic potential towards the granular starch selected.

3.4.3 Starch distribution in the thickness of the treated samples

In order to study the distribution of the impregnation level in the sample thickness, we have used the analytical procedure for the determination of the total dry matter content (TMS) on adjacent specimen obtained from a single sample. The results obtained with the different wood categories have been collected in Table 1.

Table 1

An example of the values determined is reported in Figure 1 and is related to sample ECBL.

Figure 1: Starch distribution on the sample thickness (sample ECBL). [The wood sample is divided in two adjacent specimens: the first is analysed before and the second after the impregnation. For the analysis of the dry matter content each specimen is cut in parallel sub samples Tx (x = 1,.., 5) with a thickness of 8-10 mm, successively dived each one in three following the radial direction R1, R2 et R3 for the non impregnated sample and R1 imp., R2 imp. and R3 imp. For the impregnated samples]. 
The impregnation level: T imp % = [TMS beforet – TMS afters] * 100 / TMS befotre.

We can observe that for very spongy wood (ultra degraded wood) it’s possible to increase the dry matter of 200% with reference to the initial mass: as concerns the hard and semi-hard wood (mean and minimum degree of ammaloration) we can observe values ranging between 90 to 194% as function of the degradation level on a similar section of wood sample.

As partial conclusion of these tests we can observe that the results obtained with premixed mixtures of granular starch and dextrin give the values of highest impregnation level. This fact seems to confirm our previous observations.

These results are perfectly coherent if we refer them to the presence of micro fissures that can be easily reached by starch granules. The higher density of the granule toward the soluble part of starch suspension drives these granules inside the fissures and allows the complete filling of the volume. The gelatinisation of the granules creates new bridges and bindings to consolidates the layers these unstable structures and minimizes the risks of cracks formation. The complete filling of these fissures is not possible for small molecules like the dextrin that, differently from the starch granules, are able to enter the layers structure reinforcing it from the internal part: this fact explains the better behaviour of starch/dextrin mixtures.

3.4.4 Selected impregnation procedure

The results of the tests carried out with several mixtures of starches and derivatives allowed us to identify this method as the preferable one to obtain a good impregnation level in order to consolidate its structure and to fill all the open fissures, cracks, pores, insect holes, etc.

The preferable procedure is based on the use of a mixture of granular starch and dextrin prepared and homogenised before to start the impregnation process (premixed). A mechanical agitation is necessary in order to avoid the sedimentation of the insoluble part of the starch. The process improvement is clearly visible with the increasing of the agitation speed till to a limit in which the level of impregnation is maintained or slightly reduced; for the purposes of our study this limit on speed is about 200 rotations/min.

The increasing of the temperature of the impregnating solution does not give any improvement to the process: the increased viscosity of the starch granules under gelatinisation has a negative effect from this point of view. On a long term basis, an high temperature will damage the wood quality and is thus absolutely not envisaged to use heating in the impregnation phase.

Also the application of a preliminary drying process, before to start the impregnation has to be avoided because of the high risk of closing the wood punctuations.

The duration of the impregnation process will be selected on the basis of the dimensions of the samples to be treated and especially of their thickness.

The results of the study on the starch impregnation process showed us the existing variability among the different types of waterlogged wood sampled. Normally we can observe that the most degraded wood samples seems to adsorb better starch than the less degraded. But some kind of wood has not showed this behaviour.

An ideal impregnation treatment will allow a complete saturation of the porous wood structure in order to assure the minimization or he elimination of the shrinkage during drying. Unfortunately there is not a technique able to solve the problem posed by the elimination of the water contained in the microfibrilles of wood by substitution with a resin able to solidify without any shrinkage.

This is the main cause of the presence of shrinkages in all the treated waterlogged wood samples, that we are trying to minimize. The effects of the intermediate thermal treatment coupled with the effects of a good drying process will probably allow us to reach this very important objective.

3.4.5 Conclusion on the results of the impregnation tests

In this part of the research we have tried to demonstrate the feasibility of the waterlogged wood impregnation with different kind of starch (granular starch and dextrin, in solution and/or in suspension in pure water or in a solution of ordinary sugar). The optimisation of the operative conditions has been carried out in order to obtain the highest level of dry matter content in the impregnated sample.

If compared with the PEG impregnation process, largely used in the archaeological wood sector, with the starch impregnation process is possible to obtain samples with lower dry matter content; in fact, the starch impregnation process aims to the impregnation of the deteriorated waterlogged wood structure with binding and consolidating agents while the PEG based process aims to the saturation pf the structure with a filling agent. Differences are really sensitive as concerns final density of the impregnated sample, that is closer to fresh wood for starch impregnated wood and strongly higher for PEG impregnated wood.

The large amount of starch based products present on the market will allow to meet very well the requirements of wood samples, both as concerns the geometrical properties and the degradation level, both as concerns the density of the impregnation agent to be used. Thus will be possible to select the starch based product better useful in each specific consolidation process.

The visual aspect (colour) and the sensorial impression (weight, texture, etc.) is largely better when the starch impregnation process is used; the mechanical properties and the geometrical properties (deformation, shrinkage, cracking, etc.) are not comparable at level of simple impregnation: it’s necessary to carry out all the steps of the consolidation process in order to obtain a measurable final product.

3.5 Intermediate thermal treatment

The intermediate thermal treatment of the waterlogged wood has an impact at three different complementary levels: structure of the wood, starch structuration and eventual debacterization of the material.

In this chapter we present the results of different thermal treatments tested. It consists on a fundamental point in the proposed process, we tried to study its influence on the structure but also, more in general, on the quality of the wood samples; the evaluation of the obtained quality can be determined only after the drying treatment. The results of the different treatment variation allowed us to determine the impact of the treatment at microscopy (SEM et NMR), at macroscopic level (resistance of the materials, aspect, deformation, etc), on the final structure of the wood and of the starch and at microbiological level. It’s on the basis of the “quality” parameter, that we went on optimizing the treatment from the temperature, heating conditions and duration point of view; the speed increment and specially the decreasing temperature (decompression, etc.) has been as well determined.

3.5.1 Variants of the thermal treatment Variant 1 of thermal treatment 1

The first variant of thermal treatment is characterized by the use of water steam at high pressure (<5bar) and high temperature (<165°C) as continue heat source for all the duration of the operation. It’s possible to note that, within this variant, two ways of decompression have been studied: one instantaneous and the other slow. The phenomena already cited has been confirmed in the treatments according to the variant 1. So the duration of the real treatment for samples which dimensions vary from 20 mm to 90 mm ranges from 40 to 60 minutes.

Finally, all the experiments performed for the thermal treatment according to variant 1 confirmed that the equilibrium time for the temperature between the heart and the walls doesn’t exceed 20 minutes for the thickest waterlogged wood samples (80/2=40mm). So, the use of 30 min as heating time during the thermal treatment has been confirmed.

We note that, within the procedure of this variant 1 of thermal treatment, the duration of the considered operation has been measured from the instant of beginning of the steam introduction inside the treatment chamber till to the discharging of the steam outside from the chamber. The declared duration exceed very much the such determined limit of the real treatment of the wood. The total duration of the treatment in this variant is inferior to one (01) hour for the waterlogged wood samples here cited. Variant 2 of thermal treatment 

We remind that, in this variant, the treatment is performed with relatively dry air: at atmospheric pressure or under pressure near 1,5 bar. The heating was realized by irradiation from the walls at fixed temperature, but a convection phenomena from the air introduced will help very much the heating of the sample. We note that the temperature level of the wood remains clearly under the one needed, and the speed to reach the temperature is largely lower to the one found during the first variant. This is caused by the partial dehydration of the waterlogged wood, due to the presence of a steam pressure lower than the equilibrium pressure at considered temperature; this is in fact the measurement of the «humid temperature» of the sample that we measured within this variant.

Because of the slow speed in reaching the right temperature, the level of the effective duration of the operation is clearly superior to the one found in variant 1. For a duration of the real treatment of 30 minutes, it will be necessary to foresee a global time (in addition to the instauration of the needed temperature of the wall) de plus de 90 minutes. A pre-heating of the treatment chamber with steam before the introduction of the sample and the beginning of the treatment will make easier the set up of the installation and reduce the total duration of a cycle of thermal treatment. Variant 3 of thermal treatment 1

The variant 3 is characterized by the use of a humidified atmospheric air under pressure systematically inferior to 1,5 bar. The humidity of the environment is obtained both by the intermediation of a steam injection at the beginning of the operation, and by the presence and maintenance of a certain amount of liquid water in the treatment chamber.

3.5.2 Results and quality of the samples Shape preservation

The analysis of the results of the different thermal treatments under the three exposed variants has been realised principally under the aspect of the preservation of the initial shape (ASE). 

Figure 1 : Results of the thermal treatment on wood sample6E (a series of samples impregnated with granular starch and two with not impregnated ). The Conditions Operatives COP are reassumed in Table 1 

The obtained results show the enormous impact of the thermal treatment on the global quality of the final wood. The samples «well» treated have a highly improved quality with, for example, the ASE at about 80-90%. The effect of the thermal treatment is visible also in the case of the not impregnated samples. Considering the three variants, V2 and V3 seem the most interesting ones with a distinction, at this level, for the variant V2. The thermal treatment under humid air pressure give systematically the best results.

Figure 2: Results of thermal treatment on wood n. 15 E (a series of samples impregnated with starch mixed with 
two series of not impregnated wood). The Operative Conditions COP are reassumed in Table 1.

The results of comparative treatments with different operative conditions are presented for three varieties of samples of wood (Figure 1, Figure 2, Figure 3). The operative conditions for each series of samples are reported in Table 1. 

From these figures results very clear that the impact of each of the different treatments is similar, with no reference to the kind of sample treated. Because the insufficient amount of specimen obtained from samples coded 6 E and ECBL, these samples have a number of COP of 3-4 instead of the 4-5 of the other samples. The case of sample 15 E of the lot, impregnated with starch and treated with humidified air with steam show values of A. S. E particularly interesting also for the future next chapters. The relative important number of treated samples (six to eight per COP) confirmed the reproducibility of this operation. The best results have been obtained with treatment temperatures of about 140°C. On the contrary, the waterlogged wood samples treated with water steam pressure between 3,5 and 4,5 bar at temperature between 155 and 160°C gave not good results regarding quality. The more the temperature is elevated, the more the samples coming out from the reactor tend to flakes. It’s not possible to perform the analysis calculating the ASE for those samples of wood. And the obtained results, in those conditions, seem to denote the presence in the liquid leached after the treatment both the absorbed starch, and the substances that maintain bound the structure of the wood in its waterlogged state. More, the great degradation of the global quality of the wood treated with variant 1 is linked to the continuous steam injection that cause the leaching of the starch contained in the impregnated wood samples. The test has been performed starting from the coloration test with iodine on the condensates obtained from the steam generated during the treatment. 

So we can understand that the different tests realized in variant 1 of thermal treatment gave results absolutely not complyant with those expected for the waterlogged wood preservation. In fact it was found that the contact between the saturated steam water and the wood (generally already impregnated with starch) are connected by a leaching of the impregnation products. Other quality vectors 

The absence of cracking and the maintenance of colour are two others quality parameters, estimated or quantified, to compare the effects of the different variants. Over the dimensional stability, the best results on all these aspects, like the maintenance of the colour and the reduction or absence of cracking for the main part of the wood samples, have been obtained applying the thermal treatment, and particularly the variants 2 and 3; moreover we note that the variant 3 is distinguish by the best colour preservation. 

Table 2 

Other quality parameters (resistance, mechanical properties) will be discussed in the next chapter; we can, for the moment, indicate that the thermal treatment seems to improve the waterlogged wood quality, at different levels depending on the variant, with a preference for variant 3. 

The best results (the dimensional stabilization, the colour maintenance, the reduction or absence of cracking for the main part of the wood samples have been obtained with the application of this thermal treatment.

Figure 3: Tests comparing the thermal treatment on wood ECBL-1.
(two series of impregnated samples ECBL - imp 1 et ECBL - imp 2 and a series of not impregnated samples ECBL - 1). 

Finally we note that the higher colour degradation determined with variant 2, although the good values obtained for the dimensional stabilization (ASE), can be eventually explained by a probable oxidizing reaction because the treatments are performed within a dry atmospheric air between 130 and 133°C in 20-30 minutes. Excluding the colour alteration the variant 2 doesn’t represent a bad treatment method. It’s possible to say that it can be used where we want to perform some changes in the colour, for example for some objects that have lost their original colour during their long immersion or under acid waters. 

The results of the thermal treatment depend on the variations that change the temperature in the environment in, but most of all on the kind of atmosphere in the treatment chamber. The extraordinary improvement induced in the structure depend, moreover, on a very specific physical parameter: the speed of decompression and the passage to the vacuum through a cooling phase and, on the other hand, the steam generation. This is an aspect that we want to study in the following paragraph. 

3.5.3 Conclusion 

This intermediate thermal treatment allowed, after its optimisation, to reach some high levels of finished products. It avoids the modification of the dimensions of the different samples without causing wood degradation because the used conditions are in a specific range of time and temperature. Its intervention, clearly more efficacious after a starch impregnation, is relatively positive also in the case of non impregnated wood.