COST ACTION FP1302 Short Term Scientific Mission Report Reference number: COST-ONLINE_STSM-FP1302-20382 Applicant: Doganis Efstathios, TEI of Athens Title: Evaluating the effect of restoration on wooden recorders by acoustic impedance Period: from 30 June 2014 to 4 July 2014 Host institute: Cité de la musique, Paris, France Chair of Action FP1302: DR SANDIE LE CONTE (sleconte@cite-musique.fr) STSM Manager: DAVID MANNES (david.mannes@psi.ch) Science Officer: DR FATIMA BOUCHAMA (fatima.bouchama@cost.eu) Administrative Officer: MS ARANZAZU SANCHEZ (Aranzazu.Sanchez@cost.eu)
Evaluating the effect of restoration on wooden recorders by acoustic impedance 1. Objective of the STSM This STSM that took place in the Cité de la musique aimed to study whether the repair of cracks in woodwind instruments could alter their amplitude or/and frequency by measuring the acoustic impedance of four wooden recorders before and after the conservation of the instruments. Four different adhesives/fillers were aimed to be tested in order to examine their ability to repair and preserve the acoustic propertiesof four wooden recordersmade of maple wood (figure 1). During the experimental part the recorders were examined in three different stages: a) in their undamaged condition, b) after developing an artificial crack and c) after repairing the crack. Figure 1: The four wooden recorders used in this STMS 2. Description of the work carried out during the visit During this STSM the researcher learn about the technique of acoustic impedance. The STSM lasted five days. During these five days, the STSM researcher familiarized and learned more about the acoustic impedance technique and he was trained using the host institution s equipment. During the first day the STSM researcher attended a small training demonstration on acoustic impedance in order to be familiarized with the method. He learned
how to calibrate the system and he was familiarized with the software used for the experimental part. Also, he organized all the necessary set up (figure 2) for recording the acoustic impedance of the four wooden recorders(baroque fingering) made by Gewa of maple wood that have been transported from Greece. Figure 2: The experimental set up A full documentation of the instruments was curried outbefore their recording, where they have been photographed and measured (figure 3). Figure 3: Recorder dimensions
Acoustic impedance In order to begin the experiment, the sensor was calibrated and it was tested by the formula F=2n-¼ * C/L. Then, the mouthpiece of the recorder was removed and the diameter of body was measured in order to add the information to the software. After, the instrument s body was placed to the sensor (figure 4, 5) and the acoustic impedance of the instrument was measured. The duration of the signal was 9 seconds. To test the reliability of the results, some instruments have been tested more than two times. All the recorders were recorded in 8 different fingering positions. Figure 4: The sensor Figure 5: Placement of the instruments to the sensor After measuring and documenting all the recorders in their initial condition, an artificial crack was developed onto the instruments. The artificial crack was made in the hole at the back (thumb) (figure 5) with a cutting tool (cutter). The crack was near 1mm width and 7,4 mm height in all recorders.
Figure 6: The artificial crack The third faze of the experiment was the restoration of the cracks with different adhesives/fillers for each recorder as it is shown in table 1. In most cases, instrument manufacturers in Greece repair the musical instruments with natural adhesives with no chemical processing such as natural resins and protein adhesives. For this reason, the adhesives selected for the filling the cracks were fish glue and a blend of beeswax and mastic called "keromasticho". For comparative purposes, two more adhesives were chosen; polyvinyl glue, used for wood welding and a methacrylate copolymer for its great durability and reversibility. All fillers were applied on the cracks with a paintbrush. After the restoration the acoustic impedance was recorded again in eight different fingering positions. Recorder Adhesive/filler Name 1 Mastic & Beeswax Keromasticho 2 Protein based glue Fish glue 3 Ethyl methacrylate copolymer Paraloid B72 40% w/w 4 Polyvinyl acetate glue Vinavil Table 1: Adhesive/Fillers used for the restoration of the recorders
3. Description of the results obtained The STSM demonstrate a number of key points that must be mentioned in this report. The impacts of the adhesive to the sound Comparing the spectra before and after the crack restoration,it is observed that in the majority of the recorders the sound spectra have been preserved with low various degrees of distortion. Comparing the initial sound spectrum of each recorder (before the crack) with the final sound spectrum, (after the restoration) the effect of the adhesive to the sound can be evaluated. In spectrum 2 and 4 it is observed that the sound spectrum after the restoration has been shifted right or left to a degree higher than 3 Hz. That means that the recorders have eventually lost their tonality. Also, the figures show that the sound spectra acquire different peaks after the repair, indicating that their volume has been increased or decreased. Spectrum 1: First recorder restored with beeswax and mastic As shown in spectrum 1, the initial spectrum and the spectrum after the restoration, have the same picks as well as they follow the same frequency in contrast with the red spectrum (cracked). Spectrum 2: Second recorder restored with fish glue In the spectrum 2, it is observed that the flute repaired with the fish glue has approximately the same picks as the original; but its frequency has shifted right. In contrast, figure 3 shows that the spectrum of the repaired recorder with ParaloidB72, has shifted some hertz to the left but its peaks remained the same.
The 4 th spectrum illustrates the recorder restored with the polyvinyl acetate glue. As it is observed, the restored spectrum has not returned to its initial state after the restoration and both picks and hertz have been shifted. Spectrum 3: Third recorder restored with Paraloid B72 Spectrum 4: Fourth recorder restored with polyvinyl glue 4. Conclusions The most appropriate adhesive/filler based on the preservation of recorder s audio properties, appeared to be a traditional material used in Greece for repairing wooden wind instruments, which is a blend of beeswax and mastic "keromasticho". Fish glue was the second best material tested, followed by Paraloid B-72. Nevertheless, both natural fillers, keromasticho and fish glue, do not meet some conservation standards regarding biodegradability, durability and properties after curing unless they are improved with various additives. Paraloid B-72 could be a candidate material as it can compromise its good properties over time with acceptable acoustic properties of the recorders. 5. Future research This study aims to be the base for a future research project. After the examination of the conservation materials on the acoustic properties of wooden recorders, it would be interesting to examine and observe how aging affects these materials and their acoustic properties. This can be achieved through accelerated aging (UV and RH/T circles). Therefore, in a future experiment, the «acoustic impedance» will be measured also before and after the aging of at least four conservation materials.
6. Projected publications/presentations resulting from the STSM The results of this work are planed to be presented at the COST FP 1302 conference in Cremona, Italy on the 29th of September 2014. Furthermore, these results will also trigger the preparation of a paper on the effect of restoration in woodwind instruments in the near future.