I recently removed some crystal glasses from a cupboard. Two of them left 1 centimetre of their rims behind on the shelf. The glasses are more than 30 years old and were upside down. The rings of glass are not an even width, and there appears to be a starting point for the 鈥渃ut鈥�. What caused this? (Continued)
Readers have given The Last Word a real pasting over the past few weeks after we repeated, in an answer to this question, the urban myth that glass flows. So we sent all the original answers to John Parker, reader in glass science and engineering at the University of Sheffield. 鈥� Ed
The contention that glass flows is certainly wrong. From what we are told, the glasses had stuck to the shelf on which they were placed, so when the owner tried to remove them this induced sufficient stresses to cause failure. Two questions need to be considered: why did the glasses stick, and why did that particular failure occur?
Advertisement
An examination of the marks left on the shelf after removing the glasses might provide some clues. While glass does not flow, it is very slowly attacked by liquid water and the reaction products might just produce a weak bond with the shelf over time. Such bonding can be a serious problem in hot, humid climates if glass sheets are stacked together and condensation and evaporation occur alternately. Even after a short period in storage the sheets can become bonded. It would be interesting to know whether the glasses were put away in the cupboard wet, or whether conditions were so humid that further attack occurred over a long period.
If the shelf itself has a plastic coating this may also have contributed. Plastics do change over time as a result of flow, reaction with sunlight and so on. It is even possible that an airtight seal formed between the glasses and the shelf so that a difference in air pressure contributed to the adhesion.
The cleanness of the break suggests that the crack travelled around the glass as a result of quite low applied forces. Under these conditions the fracture surfaces tend to be smooth and growing cracks do not fork. The crack will have started at one point, near the back of the glass, where the action of pulling on the stem induced a bending stress that was tensile on the outside glass surface. Examination with a hand lens may allow the origin of failure to be identified. Once a crack has started it will run round the whole rim very easily.
But why did the crack start? Either because the applied stresses were locally high or because a flaw in the glass surface at a critical point boosted low applied stresses. This could have been caused by a thin region in the glass or because poor annealing has introduced extra stress. However, a severe flaw 鈥� for example caused by a scratch from a diamond ring as mentioned by one of your correspondents 鈥� can boost stress levels at the tip of the crack by orders of magnitude in a brittle material such as glass (a brittle material being one that does not flow under stress). A defect like this is the most likely explanation.
As always with these kinds of problem the danger is that only part of the story comes out in the first analysis and another factor turns out to be the key.
John Parker, University of Sheffield, UK