- Top of page
- Materials and Methods
Instruments produced by Antonio Stradivari during the late 17th and early 18th centuries are reputed to have superior tonal qualities than more recent instruments. Dendrochronological studies show that, during his later decades, Stradivari used Norway spruce wood that had grown mostly during the Maunder Minimum (Burckle & Grissino-Mayer, 2003; Topham & McCormick, 2000), a period of reduced solar activity when relatively low temperatures caused trees to lay down wood with narrow annual rings, a high modulus of elasticity and low density (Esper et al., 2002).
Traditionally, wood used in the manufacture of musical instruments is treated with primers, varnishes or minerals to stiffen it. Such treatment can strengthen the adhesion between cell layers, but increases the density and vibrating mass because the cell lumina become occluded by the substance (Barlow et al., 1988; Schleske, 1998, 2002a), which ultimately reduces the speed of sound.
The increase in density has an adverse affect on the radiation ratio (R= speed of sound (c)/density (ρ)), reducing the speed of sound and its resonance frequencies (Barlow et al., 1988; Schleske, 2002b). Tests of other chemical treatments have shown that they increase the dynamic modulus of elasticity (EL and ER) and decrease the damping factor (δL and δR) (Yano et al., 1994; Ono & Norimoto, 1984; Meyer, 1995). Such treatments do not alter wood density, but increase the crystallinity of the cell wall, which is considered disadvantageous for wood processing (Yano et al., 1994). Other authors suggest that the wood of violins made by Guarneri and Stradivari was chemically treated to kill woodworm and fungi (Nagyvary et al., 2006).
An alternative approach to improving the acoustic properties of wood is to reduce its density by fungal or bacterial degradation. Some degradation probably resulted from the practice during the 17th and 18th centuries of floating tree trunks in water (Gug, 1991), but there is no evidence that this caused any noticeable reduction in wood density. According to Nagyvary (1988), the microbial degradation of pit membranes that occurred during this treatment would have resulted in an increase in wood permeability, so that subsequent penetration of varnish was enhanced. Recently, a new thermal treatment has been used to improve the acoustic properties of resonance wood. Treatment at high temperatures results in a reduction in density, because of decomposition of hemicellulose and cellulose, but the modulus of elasticity is reduced (Wagenführ et al., 2005a,b). A negative side-effect of the treatment is that the material becomes brittle, causing problems during the manufacture of instruments.
Most of the described treatments alter the woody cell wall and adversely affect the properties of the compound middle lamellae, both of which have a pivotal role in determining the overall stiffness of wood.
In a homogeneous bulk material, ignoring surface effects, the speed of sound, c, is governed by two mechanical properties: the modulus of elasticity and the density. In wood, which is strongly anisotropic, c varies directionally and is increased by any discontinuities in the compound middle lamella, such as those resulting from microbial degradation. Using the formulae shown in Table 1, it can be deduced that such degradation, even if very slight, results in an abrupt reduction in the E modulus and speed of sound (Schwarze et al., 1995) and has a negative impact on the acoustic properties of the wood.
Table 1. Principal acoustic properties used for the assessment of tonal wood quality of axial (L) and radial (R) samples
|Density, ρ (kg m−3)||ρ for the specimens in L and R directions|
|Young's modulus of elasticity, E (MPa)||E for L and R directions|
|Speed of sound, c (m s−1)|| for L and R directions|
|Radiation ratio, R (m4 kg−1 s−1)|| for L and R directions|
|Damping factor, δL for L direction and δR for R direction|| where fr is the resonance frequency, Δf the associated damping and K is a coefficient which varies between and |
The compound middle lamella is penetrated or otherwise altered by most species of wood-decay fungi, except for members of the Xylariaceae (e.g. Kretzschmaria deusta and Xylaria longipes), which have little ability to degrade guaiacyl (Nilsson et al., 1989; Schwarze et al., 1995), which the compound middle lamella contains in very high concentrations. As a result, this layer remains as an intact skeleton, even at quite an advanced stage of decay (Nilsson et al., 1989; Schwarze et al., 1995; Schwarze, 2007), which explains why the speed of sound through the wood is little affected until that stage (Schwarze et al., 1995, Schwarze, 2007) and is the reason why decay caused by K. deusta is hard to detect in trees by means of acoustic devices (Schwarze et al., 1995, 2004; Schwarze, 2007).
The objective of this study was to investigate whether wood-decay fungi, such as the soft-rot fungus X. longipes, which lacks the ligninolytic ability to degrade the compound middle lamella, or the white-rot fungus Physisporinus vitreus, which does so only at an advanced stage of wood decay, can be used to improve the acoustic properties of resonance wood. For this purpose, wood specimens of Norway spruce and sycamore before and after incubation were assessed microscopically, mechanically and physically (Spycher et al., 2008).
Materials and Methods
- Top of page
- Materials and Methods
We selected 80 specimens of the heartwood from Norway spruce (Picea abies L.) and 40 specimens from sycamore (Acer pseudoplatanus L.), free from visible defects or knots and with narrow annual rings according to the criteria for resonance wood. The density of the Norway spruce and sycamore wood specimens ranged from 360 to 490 kg m−3 and from 530 to 630 kg m−3, respectively.
To determine acoustic properties in the axial as well as the radial direction, 20 specimens in each sample were cut with their longest sides axially orientated (‘axial specimens’) and another 20 were cut with their longest sides radially orientated (‘radial specimens’). The dimensions of the axial specimens were 3 mm (tangential) × 25 mm (radial) × 150 mm (longitudinal), and those of the radial specimens were 3 mm (tangential) × 25 mm (longitudinal) × 100 mm (radial). Before every measurement, wood specimens were preconditioned at 23°C and 50% RH until a constant weight was reached (i.e. the moisture content (MC) of the specimens was 10.5 ± 0.5%). The mass losses with their corresponding standard deviations (SD) were also measured before and after incubation at 23°C and 50% RH. Additionally, 40 specimens of Norway spruce wood were impregnated with 1% malt solution, so that the MC above the fibre saturation point was reached (approx. 28%) before incubation with Physisporinus vitreus (Pers.: Fr.) P. Karst. (a basidiomycete) and Xylaria longipes Nitschke (an ascomycete). The incubation process was initiated according to European Standard EN 113 (European Committee for Standardization, 1997) with the aim of exposing the wood to a high inoculum of each fungus to facilitate colonization of the wood. The samples were incubated in the dark at 22°C and 70% RH.
Five physical properties were assessed before and after 6, 12 and 20 wk incubation with the fungi, using resonance frequency (Görlacher, 1984) measurements according to Spycher et al. (2008) (Table 1). Differences between values in percentage before and after incubation were estimated for each specimen, and the average of these variations and the corresponding SD were calculated from 10 or five specimens for Norway spruce and sycamore, respectively.
Bending strength (σr) was determined by three-point bending tests, whereby a central load was applied to specimens with a span (L) of 100 mm (German Standard DIN 52186; Deutsches Institut für Normung EV, 1978). The tests were carried out with a universal 100 kN bending test machine with a load rate of 2.5 mm min−1. The load was measured using a 1000 N force sensing device with a maximum error of 2% and a midspan deflection (w) with a maximal error of 1%. Two values were recorded: the maximum stress (σmax MPa) reached and the bending strength for each specimen (σr MPa), where the maximal deflection was wmax. Mean values and SD were calculated from 20 and 12 wood specimens of Norway spruce and sycamore, respectively. One-way analysis of variance (ANOVA) of the recorded values was performed, with respect to acoustic properties and bending strength, for all wood samples using SPSS software (Chicago, IL, USA) with the significance level set at P < 0.05.
For light microscopy, the incubated wood was cut into smaller blocks (10 × 5 × 5 mm), which were embedded, sectioned and stained according to the procedures of Schwarze & Fink (1998). The specimens, with transverse, radial and tangential faces exposed for examination, were fixed in 2% glutaraldehyde buffered at pH 7.2–7.4, dehydrated with acetone, embedded in a methacrylate medium and subsequently polymerized at 50°C. The embedded specimens were sectioned at approx. 2 and 3 µm using a rotary microtome (Leica® 2040 Supercut) fitted with a diamond knife. For general observation of wood anatomy, sections were stained for 12 h in safranine and then counterstained for 3 min in methylene blue and 30 min in auramine. To detect early stages of selective delignification, duplicate sections were also stained with safranine and astra blue (Srebotnik & Messner, 1994). Safranine stains lignin regardless of whether cellulose is present, whereas astra blue stains cellulose only in the absence of lignin.
Colour micrographs (Kodak EPY 64T) were taken with a Leitz Orthoplan microscope fitted with a Leitz-Vario-Orthomat camera system.
- Top of page
- Materials and Methods
Incubation of wood with two species of decay fungi caused marked density losses and cell wall thinning; that is, the partly degraded wood resembled superior resonance wood grown under cold climate conditions. This finding is in good agreement with other research that shows that the gradual decomposition and loss of hemicellulose with time lowers wood density without affecting its Young's modulus, subsequently increasing the radiation ratio (Bucur, 2006).
The observed pattern of degradation by P. vitreus seems to be unique with regard to the selective delignification of the secondary wall without degradation of the middle lamellae, even at advanced stages of decay. The significant increase (P < 0.05) in the damping factor (340% in the radial direction) that was recorded after incubation of 20 wk can be attributed partly to selective degradation of pit membranes (Schwarze & Landmesser, 2000; Schwarze et al., 2006; Schwarze, 2007). The degradation of pit membranes by P. vitreus is an important aspect that could have significant benefits in wood protection processes, namely for improving the permeability of waterborne wood preservatives (Schwarze et al., 2006). Similarly, an increase in wood permeability facilitates penetration of varnish, which is traditionally used to increase the stiffness (i.e. Young's modulus of elasticity) of the wood used for making violins (Nagyvary, 1988). Thus, it is conceivable that the significant reduction in Young's modulus of elasticity recorded in Norway spruce wood incubated by P. vitreus could be mitigated by additional treatment with wood-stiffening varnishes (e.g. copaiba balsam), which can result in an increase in the speed of sound by 18.8% in treated compared with untreated wood (Schleske, 1998). Such treatment would ultimately further enhance the radiation ratio. Incubation with P. vitreus for > 20 wk will adversely affect the properties of resonance wood, rendering it unsuitable for violin manufacturing.
In sycamore wood incubated with X. longipes, degradation began preferentially within groups of libriform wood fibres containing intercellular spaces, leaving fibre regions without such spaces, vessels and xylem ray parenchyma undegraded and largely intact, even when decay had become advanced elsewhere. These differences in cellular decay resistance have been previously reported for Armillaria mellea on sycamore wood and correspond to the degree of lignification within the two types of fibre (Campbell, 1931, 1932; Nilsson et al., 1989; Schwarze et al., 2000; Schwarze, 2007). Even after 20 wk incubation, the compound middle lamella in sycamore wood showed little signs of degradation, which indicates that even longer incubation periods could be used without reducing the speed of sound.
Particularly in the case of the top plates for violins, a large R (Table 1) of the material is desirable to produce a big sound (Holz, 1966; Wegst, 2006; Spycher et al., 2008). A high radiation ratio in the axial direction is of utmost significance for first-quality resonance wood (Ono & Norimoto, 1983; Müller, 1986). For the manufacture of an excellent concert violin for use by a soloist, the violin maker requires at least ‘very good’ material quality for the two quarter cuts (top plate: Norway spruce wood; bottom plate: sycamore wood). In the present study, degradation of Norway spruce and sycamore wood by P. vitreus and X. longipes, respectively, was accompanied by significant increases (P < 0.05) in R after 6, 12 and 20 wk incubation (Table 2). In the wood of both species, improvement in the radiation ratio was achieved by a reduction in density of approx. 12%, coupled with relatively little alteration in the speed of sound (Table 2). In Norway spruce wood, R values of 10 and 16 have been measured in acoustically ‘poor’ and ‘excellent’ specimens, respectively, with corresponding values of 5.5 and 8 in sycamore wood (M. Schleske, unpublished). Thus, in our study, the acoustic quality of Norway spruce and sycamore wood was increased from ‘poor’ to ‘good’.
The axial bending strength of incubated Norway spruce and sycamore wood specimens was not significantly reduced, in comparison with the controls, after 20 wk incubation, whereas a significant reduction (P < 0.05; Fig. 5) in the radial bending strength of both wood species was recorded. These results are in good agreement with those of previous studies that showed that, in comparison with controls, the impact-bending strength of Norway spruce wood was not significantly reduced after 12 wk incubation with P. vitreus (Schwarze et al., 2006).
The reduction in radial bending strength is important for Norway spruce, but may not be relevant for the use of sycamore wood in violin-making (M. Schleske, unpublished); that is, the mechanical impact on the sycamore bottom plate is mostly a dynamic effect, while the static forces exerted in compression are compensated by the geometry of the violin (Bond, 1976; Spycher, 2008). Furthermore, the top plate made of Norway spruce wood is responsible for the global sound emission of the violin, but not for its strength and stability. The potentially disadvantageous radial strength losses in Norway spruce and sycamore wood after incubation could be mitigated simply by using thicker top and bottom plates (Wegst, 2006; Spycher, 2008).
The quality of the resonance wood is very important for the acoustic quality of the violin. The procedure described here for modifying wood is intended primarily to enable the manufacture of better solo instruments. A solo violinist prefers an instrument that can play ‘against’ the orchestra. Its tonal properties include high projection, high volume and dynamic range, together with a sensitive modulation of tonal colours, and these depend directly on the material quality of the resonance plates of the violin, which in turn is correlated positively with the velocity of the longitudinal sound waves (both across and along the grain) and negatively with wood density. A material with a high ratio of the speed of sound to density increases the sound emission of the instrument, which means that the plate amplitudes are high in relation to the force that excites the strings. This enhancement of resonance makes the difference between a violin of average quality and one that is suitable for a top soloist. Because of the enormous size of modern concert halls, acoustic instruments made from wood modified by fungi will be desirable for meeting the needs of soloists in the future.
In further studies, we will be attempting to optimize the uniformity of colonization and decay processes, particularly identifying the critical incubation time above which the radiation ratio is adversely affected. The exact influence of the bending strength on the violin will also be determined, using a prototype violin made from fungal-treated wood plates. Additionally, the influence of the damping factor on the acoustic quality of resonance wood and the effects of its modification on the final properties of the violin will be investigated.