Does Glass Actually Flow Over Time?

[Harut82]

didn't know where to post it so I put it here


do you guys agree with the following?

One common belief is that glass is a super-cooled liquid of practically infinite viscosity when at room temperature. Supporting evidence for this position is that old windows are often thicker at the bottom than at the top. It is then assumed that the glass was once uniform, but has flowed to its new shape.

One possible source of this belief is that when panes of glass were commonly made by glassblowers, the technique that was used was to spin molten glass so as to create a round, mostly flat and even plate (the Crown glass process, described above). This plate was then cut to fit a window. The pieces were not, however, absolutely flat; the edges of the disk would be thicker because of centripetal forces. When actually installed in a window frame, the glass would be placed thicker side down for the sake of stability. Also, the sparkle is greater and the visual effect stronger when the thicker side is down. There is anecdotal evidence that occasionally such glass has been found thinner side down, as would be caused by carelessness at the time of installation.

Writing in the American Journal of Physics, physicist Edgar D. Zanotto states "...the predicted relaxation time for GeO2 at room temperature is 1032 years. Hence, the relaxation period (characteristic flow time) of cathedral glasses would be even longer" (Am. J. Phys, 66(5):392-5, May 1998). In layperson's terms, he wrote that glass at room temperature is very strongly on the solid side of the spectrum from solids to liquids.
Evidence against glass flow
If medieval glass has flowed perceptibly, then ancient Roman and Egyptian objects should have flowed proportionately more—but this is not observed.
If glass flows at a rate that allows changes to be seen with the naked eye after centuries, then changes in optical telescope mirrors should be observable (by interferometry) in a matter of days—but this also is not observed. Similarly, it should not be possible to see Newton's rings between decade-old fragments of window glass—but this can in fact be quite easily done.
Likewise, precision optical lenses and mirrors used in microscopes and telescopes should gradually deform and lose focus. This is also not observed.

http://encyclopedia.thefreedictionary.com/glass

 

[Jelfish]

Perhaps windows are more prone to flow because they are in a harsher environment while telescope and other lenses are typically well protected. Perhaps it's the shape of windows and that they're typically much larger and stand up right while lenses are typically smaller, lighter, and are more rounded shaped.

Whether or not glass is a solid really depends on the definition of solid. If I remember correctly from chemistry, a solid has a definite melting point and breaks with blunt edges while semi-solids have a melting temperature range and break with sharp edges.

 

[TheStatutoryApe]

I believe that glass is what is referred to as an Amorphous Solid.

 

[russ_watters]

Yes, that is correct, Harut.

Jelfish, small telescopes may be thicker and well-supported, but large research telescopes are enormous and so heavy that they require special external structure to keep them from bending/breaking under their own weight (in fact, structural stability is a severe impediment to making large refractors). So the forces at work there are far larger than on windows and any change noticeable in a window in decades by the naked eye would indeed be noticeable in a telescope in days. And don't forget - many such telescopes are decades old too.

 

[zoobyshoe]

Glass does not flow. Whether or not it is technically a solid or liquid is a separate question:

https://www.physicsforums.com/showthread.php?t=75562&highlight=glass

 

[Astronuc]

Glass and in fact, most, if not all ceramics flow - but very, very slowly - i.e. it creeps. I can't necessarily vouch for specific cases, but my company specializes in developing models for concrete behavior, including creep.

 

[zoobyshoe]

Glass and in fact, most, if not all ceramics flow - but very, very slowly - i.e. it creeps. I can't necessarily vouch for specific cases, but my company specializes in developing models for concrete behavior, including creep.

I don't know about any other materials but common glass doesn't flow over time. The whole notion that it does was a misunderstanding of the way stained glass window were made hundreds of years ago, and as the OP mentioned no ancient glass that we have shows any evidence of flowing over time.

Writing in the American Journal of Physics, physicist Edgar D. Zanotto states "...the predicted relaxation time for GeO2 at room temperature is 1032 years. Hence, the relaxation period (characteristic flow time) of cathedral glasses would be even longer" (Am. J. Phys, 66(5):392-5, May 1998). In layperson's terms, he wrote that glass at room temperature is very strongly on the solid side of the spectrum from solids to liquids.
Evidence against glass flow
If medieval glass has flowed perceptibly, then ancient Roman and Egyptian objects should have flowed proportionately more—but this is not observed.
If glass flows at a rate that allows changes to be seen with the naked eye after centuries, then changes in optical telescope mirrors should be observable (by interferometry) in a matter of days—but this also is not observed. Similarly, it should not be possible to see Newton's rings between decade-old fragments of window glass—but this can in fact be quite easily done.
Likewise, precision optical lenses and mirrors used in microscopes and telescopes should gradually deform and lose focus. This is also not observed.

The notion that glass flows over time is just the mistaken thinking of someone who was unaware that medieval glass blowers had not developed a way of making very flat glass of even thickness. The method they had of making flat sheets of glass was very crude by comparison to later methods: you blew the best sphere you could, and then spun it by rolling the blow tube with the hot glass sphere at the end on an elevated rail letting the sphere flatten into a disc as you rolled. The only flat glass they could produce was in disc form, and there's nothing remotely resembling precision in the uniformity of the thickness of these discs. Any piece you cut from such a disc is going to be visibly thicker on one end than another.

As the OP quote points out, for stability and aesthetic reasons the thicker side was most often oriented toward the bottom. It's easy to find examples of pieces where they didn't follow this rule, though, which are thicker at the top. How could that be explained if glass authentically flowed?

This, I think, ought to stand as the best evidence that glass is not flowing:

If glass flows at a rate that allows changes to be seen with the naked eye after centuries, then changes in optical telescope mirrors should be observable (by interferometry) in a matter of days—but this also is not observed.

I'm not sure how anyone can read that and still lend any credence to the obvious mistake made by the person who first happened to notice that cathedral stained window glass seemed to be thicker at the bottom than the top. That person obviously jumped to an erroneous conclusion.

 

[russ_watters]

Creep: http://www.engineersedge.com/material_science/creep.htm (won't let me copy a quote)

Some materials creep, some don't, and it also depends on temperature and load - some creep at one temperature and load, but not another.

 

[Kouros Khamoushi]

Is glass liquide or Solid?

At the page 1 of Introductory Solid State Physics Second by H.P. Myers states that: From the present point of view, solid state physics is mainly the physics of crystalline solids. Most of the inorganic solids we encounter in our daily lives are crystalline, the notable exceptions being glass, which is supercooled liquid, and soot which is amorphous...
Therefore glass is counted as Solid materials.
Regards Kouros

 

[Gokul43201]

I don't know about any other materials but common glass doesn't flow over time. The whole notion that it does was a misunderstanding of the way stained glass window were made hundreds of years ago, and as the OP mentioned no ancient glass that we have shows any evidence of flowing over time.

Zoob, that's not entirely correct. What is correct is that (silica based) glass does not flow noticeably over a time-scale of centuries. However, independent of the "observation" of uneven windows, physicists have studied the properties of glasses theoretically and arrived at the roughly same conclusion - that glasses flow (and most certainly creep). The only difference between the physical theory and the myth is the time-constant for flow. Theory predicts that this is of the order of about 10¹⁰ years or thereabouts for silical glass - not anything as small as a few centuries.

 

[Kouros Khamoushi]

Is the glass solid or liquids?

It depends in which way one classifying the materials. For example one can say that there are crystalline, amorphous inorganic and organic materials in this case we are concerning with crystal structure of the materials. In crystalline solids the atoms are arranged in a regular manner which is significant for their behaviour.
Glass are counted as solid material when we are looking at its structure and behaviour because most of inorganic materials which we are encounter in our daily lives are crystalline, the notable exception being glass which is super cooled liquid.
Myers stated that Solids are not necessary crystalline. They may also be obtained as glasses (page 58)1 and amorphous aggregates that are not crystalline on any significant scale; i.e there is no long-range correlation between atom positions and no translational symmetry.
Short range order over distances corresponding to one or two atomic spacing may arise, the degree of ordering depend on the type of bond that is formed. Amorphous metals are well described as random close-packed structure.
Nevertheless we can imagine the time average of a liquid structure, which may be approximated by radial distribution function (RDF).
Taking the origin as the momentary position of a given atom we define the probability of finding another atom in an element of volume d3 R at distance R from our choosen atom as
N/Vg2(R) d3 R

N/V is the average particle density and g2(R) is call the pair distribution function. Now when R is large let's say > 20 angstrom we do not expect any correlation between position of two atoms and probability can depends only on the size of d3 R and the average density, so g2(R) is unity for large R. At very small distances, g2(R) is zero because two atoms can not occupy the same position. However, for a certain range of values of R, entered around the average particle separation, g2(R) varies in a pronounced manner. For an isotropic liquid, the radial distribution function is defined as RDF=4*piR2 g2(R)
Integration of the RDF between the time limits 0 and R tell us how many atoms are contained within a sphere of radius R described around our chosen atom. One can define higher distribution functions, e.g. the three particles distribution g3(R12,R13) that are important in the liquid, but since no way of obtaining experimental knowledge about them, discussion of liquid and other disordered structure is usually confined to g2(R) and RDF.

1.
This literature about the glass written by H.P.Myers and specifically Second Edition of Introductory Solid State Physics are quite known and are teaching in so many institutions around the world. Printed 2002.

 

[zoobyshoe]

Zoob, that's not entirely correct. What is correct is that (silica based) glass does not flow noticeably over a time-scale of centuries. However, independent of the "observation" of uneven windows, physicists have studied the properties of glasses theoretically and arrived at the roughly same conclusion - that glasses flow (and most certainly creep). The only difference between the physical theory and the myth is the time-constant for flow. Theory predicts that this is of the order of about 10¹⁰ years or thereabouts for silical glass - not anything as small as a few centuries.

OK, I stand corrected: technically glass flows. Anyone engineering anything they expect to last ten billion years should take this into consideration.

 

[Kouros Khamoushi]

Yes, it is right as you mention glass flow, and different type of glasses have their own time limit, but the point is are we taking into the consideration if really glass is solid or not . The answer to this as Myer stated it is counted as solid, but it is still is under question and as I feel there is some uncertainty about that. What do you think about that?

 

[FredGarvin]

OK, I stand corrected: technically glass flows. Anyone engineering anything they expect to last ten billion years should take this into consideration.

Oh great. One more thing we have to put on our requirements documents..."must remain operational for 1 billion years."