View Single Post
Old 04-04-2005, 03:41 AM
John Croucher's Avatar
John Croucher John Croucher is offline
Registered User
Join Date: Jul 2003
Posts: 62
John Croucher is on a distinguished road
Hugh, the trident seal test takes no longer than a ring test .In the case of a paper weight created for the saw test much more. Both need to be annealed, but if the samples are thin -ie around 2 mm or 1/10th of an inch or less, its all over in under 90 minutes. The trident seal advantage is that it is easier for a neophyte to get more acccurate results. Getting the ring test to have equal thicknesses of both glasses is quite difficult and equal thicknesses of both glasses are critical for an accurate outcome. The fusers all have a reasonably equal sheet glass thickness to work with.

The trident seal test allows a quiet and unhurried measurement to take place with the accuracy of a micrometer before fusing the respective glasses, whilst overlays circling around a glory hole don't. Ed Skeels or Lino or Peter V may be able to do overlay ring tests with the respective glasses thicknesses within 0.1mm of each other, but I would wager most glass blowers can't. Pete V rails against the trident seal test as being too sensitive, but in our experience it can accommodate a difference of 200nm/cm, which for most glasses is around a 4-5 point difference. That's a wide enough gap to test for and make adjustments. Some glass formulations are really peculiar. For instance, a fluorine opal glass with around a 20% PbO content is unbelievably unresponsive to react to an alkaline addition or subtraction re nm/cm. What we thought would require a chemical addition of 1-2 points LEC required an 8 point addition on our chemical spreadsheets. Some glass families adjust really rapidly to minor alkaline additions or subtractions, others require major alterations. Actually, outside of our measured clear glass parent families, I don't care what the measured LEC'c are of our 9 glass matrices are. All we care about is whether all the glasses fit each other. within an annealing range of 470oC -5202oC.

The other thing I forgot to mention in my earlier post is how critical the temperature range is when LEC measurements are concerned. Theoretical comparisons are useless when comparing English and Turners factors with Appen or when real world actual measurements are concerned. Most of the C&R Loo published results are either theoretical or voodoo nonsense. E&T worked everything out over the range 25-90oC. Appen calculated all his factors over 20-400oC. In Woolley's invaluable monograph, SP87 measured 94.9 over 0-300oC and 97.8 over 25-300oC. The E&T numbers come up at somewhere around 88-99. Quite a big difference. Corning uses a US$50,000 Theta dilatometer that has no better than +/- 1 LEC point accuracy, so even there the numbers are vague. That's why we say AROUND 96 x 10-7 over 0-300oC with our standard with an annealing temp of AROUND about 490oC . It's in the ball park but no more.
Reply With Quote