©Lin Yangchen




The sight of organ pipes, with their exquisitely sculpted mouths and soaring columns of polished metal and finely appointed wood, evokes sonic magnificence. But that is just the surface. Deep within the seemingly plain walls of organ pipes is an inner magnificence hidden from the naked eye.

Using his specially configured petrographic-metallographic microscope, the author reveals for the first time the spectacular complexity and self-organized emergent spatial structures in materials that constitute the pipes of the Auferstehungsorgel, from the metallurgy of tin-lead alloys to the histology of botanical sections.


Click to sections:

Tin-lead alloy
Bamboo

or scroll down.




Tin-lead alloy


Pieces of pipe metal left over from the shortening of an octave of Klais Principal organ pipes for the Auferstehungsorgel. These give the author a rare opportunity to study the metallurgy of organ pipes that are in actual use. The same day I got them back from organ builder Robert Navaratnam, I personally rushed them halfway across the country on train, bus and foot to the one metallurgy laboratory that was willing to take the job.


A sample must be polished to a mirror-smooth finish for metallography, or metallurgical observation under a microscope equipped with reflected light or epi-illumination. This is to prevent scattering of the incident light by an uneven surface, which would obscure the true underlying structure of the material. The sample was first mounted in acrylic. As tin and lead recrystallize at relatively low temperatures, cold cutting and polishing was performed using water at room temperature to avoid disturbing the original microstructure. The final polishing was done by hand with 1 μm diamond slurry. The sample was etched with 3% nital to more clearly show the internal microstructure of the pipe wall. Nital is a corrosive, flammable and toxic solution of nitric acid in alcohol. The nitric acid serves the dual function of oxidizer and corrosive agent, while the alcohol suppresses ionization. See ASM International (1985) and Voort (1999) for polishing procedures and etchants.

Tin-lead alloy is eutectic, having a limited solubility range due to differences in the atomic radii and crystal structures of its constituent elements. When the alloy contains 38.1% lead, it has a sharp melting point, the eutectic temperature, of 183° C. If the percentage of lead is higher than 38.1, the lead begins solidifying before the cooling mixture reaches the eutectic temperature, forming dark blobs as seen in this Klais sample. The remaining liquid mixture subsequently solidifies at the eutectic temperature, where atomic diffusion gives rise to a striking lamellate pattern with occasional branching. This is made up of the two phases of the tin-lead system, the lead-rich α phase (dark areas) and tin-rich β phase (bright areas). Very similar patterns occur widely in nature, from sand dunes to brain corals to fingerprints to the skin of a python to signal processing patterns in the visual cortex of a monkey (see Horton & Hocking 1997).

Some tin alloys are anisotropic (Voort 1999); the β phase in the Klais sample turns out to be so. This makes it easy to see its grain structure, which is difficult to discern in plane-polarized light (previous image), as different shades of indigo in cross-polarized light (above). The pipe metal appears to have cooled non-directionally on the macroscopic scale during casting, the lamellae having grown simultaneously from multiple nucleation sites and formed grain boundaries where they met.


The predominantly lamellar microstructure of the pipe wall (upper image) indicates that the mixture was cooled relatively slowly during casting, such as by air; fast cooling such as in chill casting or soldering would produce fine globules (ASM International 1985) as has happened in the thin solder joint on the pipe (lower image).

During further cooling from the eutectic temperature down to room temperature, the solubility of the tin continued to decrease, causing some of the tin in the large α blobs to precipitate out as (light-coloured) β granules (ASM International 1985).

The visual contrast between lead and tin is caused by both the polishing and the etching; polishing tarnishes the lead (ASM International 1985), while nital selectively attacks the α phase. Nital produces a range of grey shades due to its sensitivity to crystal orientation, which translates to differences in reflectance under epi-illumination (Voort 1999).

The microstructure shows that the pipe metal contains about 40% lead, on comparison with Buehler reference photomicrographs. This is actually the composition of ancient pewter (see Scott 1991). The α blobs would not have been created at the eutectic composition or proportions of lead lower than that. Proportions of lead from about 45% to 55% would have produced the large α blobs characteristic of spotted metal.


The solder joint at the back of the pipe has a considerably finer grain (middle of lower image) than the pipe wall (left and right), perhaps because the former cooled more rapidly. Everything looks fairly homogeneous in plane-polarized light (upper image), suggesting that the solder and pipe wall were made from the same molten stock.

A micrograph of the lamellar microstructure was prepared for computational spatial analysis by thresholding it in GIMP with 14.5 out of 255.0 levels, which satisfactorily segregated the near-white β phase from the grey shades of the α phase. The resulting binary image (above) was processed in the R Language for Statistical Computing. The α:β pixel ratio is 0.507 i.e. the α phase occupies about a third of the total area.


Co-occurrence chart showing the proportions of αα, αβ and ββ neighbouring pixel pairs in the image. I considered each pixel to have eight neighbours in the square grid. The proportion of αβ neighbouring pixels is 0.041. The higher the value, the finer the lamellar pattern. The concept is somewhat akin to the surface area:volume ratio.

Summary statistics like these could serve as additional metrics for comparing tin-lead alloy samples from different pipes, stops and builders. For more sophisticated computer vision workflows for metallography see Rusanovsky et al. (in press).

This analysis demonstrates the utility of microscopy in inferring pipemaking materials and methods for forensic or cultural conservation purposes when such conventional sources of information as physical facilities or written records are unavailable. See Scott (1991) for specialized procedures for historical metals.




Bamboo


A slice of bamboo was sawn from a culm supplied by Jerry Ng from his cured pipemaking stock. The sample was morphologically similar to but of greater diameter than the lengths used in the 笛 (Bamboo Flute) pipes specially made by Ng for the Auferstehungsorgel.

One side of the bamboo slice (not shown) was dry-polished using 200-grit silicon carbide paper followed by 500 and 1200-grit diamond lapping discs.

 
The sample was mounted unstained and unembedded, using Bostik White Tack and a two-way spirit level to ensure horizontality.

Examining the polished transverse section of bamboo under epi-illumination with a 50× metallurgical objective.

This is what the cross-section of the wall of a bamboo organ pipe would look like (see acoustical analysis). Bamboo has the scattered arrangement of vascular bundles typical of a monocotyledon. The frequency, density and strength of the bundles increase towards the circumference. Bamboo is essentially a fibre-reinforced composite. Although much research has been done on the anatomy and mechanical properties of bamboo (e.g. Wang et al. 2012, Dixon & Gibson 2014, Dixon 2017, Huang et al. 2017, Osorio et al. 2018), how the characteristics vary with species and age (e.g. Wahab et al. 2010, Zakikhani et al. 2017, Darwis & Iswanto 2018, Darwis et al. 2020) and how they can be exploited as construction materials and other commercial products, there has been virtually no research on causal effects on acoustic properties.

Cellular structure of a cured bamboo culm. The protoxylem (PX) and metaxylem (MX) conducted water while the phloem (PH), partitioned into sieve tubes (ST), carried sugar from photosynthesis. The sclerenchyma fibres (SF) started out with a large central bore or lumen (LM) that shrank in diameter as the cell matured into a strong polylamellate fibre of cellulose microfibrils oriented in different directions (see Mannan et al. 2016) in different concentric layers, sheafed in lignin. The surrounding groundmass of parenchyma tissue (PA) stored energy in the form of starch. Starch is removed during curing, otherwise the organ pipe would eventually be eaten up by insect borers (see Bhat et al. 2005).

The epidermis, having dessicated during curing, looks like a mediæval stone wall. The cells are filled with silicon dioxide (Ito et al. 2015) and give the bamboo a smooth skin. The fibre bundles (brown) near the surface are so dense that they don't even have lumens any more. The white debris is probably silicon carbide contaminant that got embedded from the sandpaper during polishing.




References


ASM International 1985. ASM Handbook Volume 9: Metallography and Microstructures. American Society for Metals, USA.

Bhat, K. V., Varma, R. V., Paduvil, R., Pandalai, R. C. & Santhoshkumar, R. 2005. Distribution of starch in the culms of Bambusa bambos (L.) Voss and its influence on borer damage. Bamboo Science and Culture: The Journal of the American Bamboo Society 19(1):1–4.

Darwis, A. & Iswanto, A. H. 2018. Morphological characteristics of Bambusa vulgaris and the distribution and shape of vascular bundles therein. Journal of the Korean Wood Science and Technology 46(4):315–322.

Darwis, A., Iswanto, A. H., Jeon, W.-S., Kim, N.-H., Wirjosentono, B., Susilowati, A. & Hartono, R. 2020. Variation of quantitative anatomical characteristics in the culm of belangke bamboo (Gigantochloa pruriens). BioResources 15(3):6617–6626.

Dixon, P. G. 2017. The Structure and Mechanical Behaviour of Bamboo and Bamboo Products. PhD dissertation, Massachusetts Institute of Technology.

Dixon, P. G. & Gibson, L. J. 2014. The structure and mechanics of Moso bamboo material. Journal of the Royal Society Interface 11:20140321.

Horton, J. C. & Hocking, D. R. 1997. Timing of the critical period for plasticity of ocular dominance columns in macaque striate cortex. The Journal of Neuroscience 17(10):3684–3709.

Huang, P., Chang, W.-S., Ansell, M. P., John, C. Y. M. & Shea, A. 2017. Porosity estimation of Phyllostachys edulis (Moso bamboo) by computed tomography and backscattered electron imaging. Wood Science and Technology 51:11–27.

Ito, R., Miyafuji, H. & Kasuya, N. 2015. Rhizome and root anatomy of moso bamboo (Phyllostachys pubescens) observed with scanning electron microscopy. Journal of Wood Science 61:431–437.

Mannan, S., Zaffar, M., Pradhan, A. & Basu, S. 2016. Measurement of microfibril angles in bamboo using Mueller matrix imaging. Applied Optics 55(32):8971–8978.

Osorio, L., Trujillo, E., Lens, F., Ivens, J., Verpoest, I. & van Vuure, A. W. 2018. In-depth study of the microstructure of bamboo fibres and their relation to the mechanical properties. Journal of Reinforced Plastics and Composites 37(17):1099–1113.

Rusanovsky, M., Beeri, O., Ifergane, S. & Oren, G. in press. An end-to-end computer vision methodology for quantitative metallography.

Scott, D. A. 1991. Metallography and Microstructure of Ancient and Historic Metals. The J. Paul Getty Trust, USA.

Voort, G. F. V. 1999. Metallography: Principles and Practice. American Society for Metals, USA.

Wahab, R., Mustapa, M. T., Sulaiman, O., Mohamed, A., Hassan, A. & Khalid, I. 2010. Anatomical and physical properties of cultivated two- and four-year-old Bambusa vulgaris. Sains Malaysiana 39(4):571–579.

Wang, X., Ren, H., Zhang, B., Fei, B. & Burgert, I. 2012. Cell wall structure and formation of maturing fibres of moso bamboo (Phyllostachys pubescens) increase buckling resistance. Journal of the Royal Society Interface 9:988–996.

Zakikhani, P., Zahari, R., Sultan, M. T. H. & Majid, D. L. 2017. Morphological, mechanical and physical properties of four bamboo species. BioResources 12(2):2479–2495.









Powered by SmugMug Owner Log In