Hide Glue vs. Silicone for Rebuilding a Pianola
by Ralph Nielsen

I have hesitated so far to join the fray in the vigorous debate on glues. But I know a bit about PVC-E and PVA glues, as my formal training and former employment was as a Ph.D. polymer chemist, with a specialization in latex/emulsion materials and industrial gelatin (effectively hide glue).

Both PVA and PVC-E glues are emulsion/latex glues which are milky in appearance until they dry, because they are a water-based suspension of very fine polymer particles that scatter light just like the suspended fat/protein droplets in milk. Both can be thinned with water, and both can penetrate into wood pores and fibers. The particles in both glues coalesce irreversibly when dried into a layer of water-insoluble plastic. Neither PVA nor PVC-E glue is water-soluble after that point, and in my experience, neither can be easily cleaned off of wood surfaces.

The properties of the dried glues are very different. PVC-E (poly (vinyl chloride)) is softer and more rubbery, and more hydrophobic or water repellent, while PVA (poly (vinyl acetate)) is harder and stiffer, with a more wettable surface and a greater tendency to absorb a little water and become a little softer. Thick bulk layers of PVC-E can sometimes be peeled or scraped off, because it is softer. And thick bulk layers of PVA glue can be swollen and softened by soaking in water (PVC-E with long! soaking) to aid removal. But neither is "soluble" in the same way hide glues are, and a thin surface layer or the material impregnated in the pores of wood or fabric is effectively impossible to remove.

Most woodworkers who have used both PVA and hide glues have seen how a spilled or oozed PVA glue usually creates a later finish defect, even if it is wiped off the surface immediately with a wet cloth, or even if the surface was lightly sanded after drying. Hide glue drips that are wiped off are much less likely to create a defect. Because the PVA penetrates quickly and dries irreversibly, it remains in the wood pores and shows up during finishing, while hide glue is much more removable and more forgiving.

Hot hide glue will 'wet' cured PVA glue surfaces better than cured PVC-E surfaces, and will adhere to some degree, but in my experience not as well as to a wood surface that hasn't been contaminated with PVA glue.

Cured RTV (room temperature vulcanization) is a cross-linked silicone rubber usually based on the polymer PDMS (poly(dimethylsiloxane)). It is extremely hydrophobic (water repellent) both before and after curing, so it bonds better to polymer surfaces like PVA-sealed wood than to bare wood. And it is entirely insoluble after curing, so it is also almost impossible to remove entirely from a rough surface, and wherever a little remains, the contaminated surface will repel most water-based glues including PVA, PVC-E or hide glues.

It is probably good that different rebuilders have different styles and use different materials. Maybe 20 years ago, I used lots of PVA glue in general carpentry (never in stacks), but I have since stopped using it almost entirely except where I'm working with a large joint where repositioning and longer working time with the cold glue is important.

I have found that hot hide glue is almost always superior for wood joints, in ease of use, "grab", clamping and positioning, heat response in veneering, avoiding finish defects, better bond strength and properties, and reversibility or renewability either next week or in 80 years. Because it gels on cooling, hide glue is also far superior for joining fabric, felt, or leather to wood without saturating and hardening them. Unlike PVA glue joints, hide glue joints can usually be convinced to break along the joint with little or no damage to the wood, a tendency that increases somewhat when the joints are 40-80 years old.

I also use some PVC-E glue ( for cloth-to-cloth bonding in some pneumatics, and the resilience is nice with some plastic keytops), CA glue (as a binder for ivory powder in keytop chip repair, for example) and epoxies (where deep penetration and space filling are needed, like some bridge repair).

I also keep my hot glue in a cheap plastic flip-top container with a plastic toothbrush and plastic paddle in it as applicators, so I can microwave it to ready in 45 seconds when I need it quickly, with no risk of burning it. The plastic container can be set in a thermostatted pot of hot water during longer jobs, and can be frozen indefinitely to avoid fungal growth, if it isn't needed for a few weeks (all tricks from the industrial gelatin world). And despite propaganda to the contrary, neither momentary boiling in the microwave or long-term freezing affects its properties at all.

I also find hide glue superior in stacks for positive airtight sealing. To me it seems apparent that early manufacturers specifically did design their pneumatic elements for reversible disassembly and rebuilding. Use of hide glue, often along with leather, buckram, or cork gaskets that aid removal of pneumatics from decks all indicate that, along with service manuals from the 1920s that instruct the technician to recover striker or expression pneumatics if there is a hint of fabric stiffness or damage. While it is a secondary concern, it is interesting to note that gelatin or hide glue has the lowest gas permeability of almost any polymer, while silicones like RTV have almost the highest.

For my part, I prefer the superior performance, reversibility, and time-tested performance of hide glue and traditional materials for most wood-joining applications, and especially in stacks. And I see no advantage in using insoluble, non-removable and irreversible materials like PVA glue or RTV based on claims that they solve problems I simply don't have with hide glue.

Ralph Nielsen
http://www.historicpianos.com/