Heat Treating Glass – craft or science?
Share your story:

Books can and have been written on this topic but hopefully this extremely summarised account will enlighten some.

To increase the strength of float glass we can heat treat it.  There are 2 levels of treatment covered under the current standards. Heat Strengthened and Toughened (also known as Fully Tempered).  Strengthened glass increases the strength of the glass by 2 and Toughened increases it by 4 to 5 times compared to annealed glass of the same thickness.  Of the 2 only Toughened glass is classified as a class A safety glass.

There are factors which can be used in calculations to fine tune these values such as polishing the glass edges to reduce the risk of breakage.  Heat treatment would never be applied to glass with a raw cut edge which contains many small shells and cracks which weaken the integrity of the lite and can be the propagation point for failure.

It should be noted that heat treating glass to make it stronger does not make it stiffer, it will still bend to the same degree under loading just be more resistant to stress and breakage.

Broken glasses can be identified by their break pattern.  Toughened glass will break into small crystal type pieces as defined in BS EN 12600 while Strengthened retains the large shard pattern associated with annealed glass, but break in a manner specified by BS EN 1863-1 which limits the number and size of any “islands” created in the broken lite.  The break pattern for each of the glass types is part of the QC procedure laid down in the standards for the processing of the glass.

A significant feature of heat treated glass is that apart from assembly into an IGU the heat treatment is the last value adding process in the production of the glass.  Nothing can be done to work it in terms of edge prep, cutting or drilling when the glass is heat treated so if it is wrong in any way it is scrapped.

Toughened glass (not heat strengthened) is required to be Heat Soak Tested which is a destructive test to determine the presence of NiS (Nickel Sulphide) inclusions.  These are microscopic inclusions which occur naturally in the glass at a rate of approximately 1 in 13 Tonnes.  HST to BS EN 14179-1:2016 Annexe A will reduce this to approximately 1 in 400 Tonnes but as this is a check for a random occurrence the glass is required to be 100% checked.  Briefly NiS particles are converted to their smaller Alpha state during the heating process and frozen in that condition when the glass is toughened.  Gradually over time the inclusion reverts to the Beta state increasing in size and causing stress to build up in the glass to the point where spontaneous failure occurs and the glass shatters.  Without HST which accelerates the conversion by heating the glass to between a minimum of 250 and maximum of 290⁰c and holding there for a minimum of 2 hours before controlled cooling this process can take several years to happen.  At the end of the HST process any NiS inclusions should have converted and the affected glass shattered in the oven.

Despite all the automated machines, engineering and standards involved the production of heat treated glass is very much about craft as well as science.  An experienced processor with understanding of the fine tuning and adjustment of their equipment is vital to control the variables noted above and can make an enormous difference to the quality of the finished product.

Both strengthened and toughened glass are subject to deviations from flatness because of the heat treatment processes.  These deviations are defined as.

  • Roller wave
  • Overall bow
  • Edge dip
  • Local Distortion (or local bow)
  • Roller pick up (roller pluck)
  • Anisotropy

Roller wave is the result of the glass in the furnace being in constant movement on a bed of ceramic rollers typically 300mm apart.  During this process the glass is in an almost plastic state and thinner glass can sag slightly between the rollers. When the glass is cooled to its solid condition the shape is retained resulting in peaks and troughs.  This causes a lensing effect in the glass with the power of the lens defined by the radius of the curve.  The standards and most specifications restrict the acceptability of this to a trough to peak depth of 0.15mm.  Roller wave is measured no closer than 150mm from the edge of the glass. Glass processors will in most cases be able to better this standard but will charge a premium to guarantee a lower value because to do so requires increased QA measures and inevitably scrapped glass as it cannot be rectified afterwards.

Overall bow relates to the deviation from flatness across the entire pane from edge to edge.  The standard defines how this is measured and the tolerance values allowed.  It should be noted that the measurement of overall bow can only be done to a single lite as it comes from the furnace it cannot be done to laminated sheets or those assembled into IGUs.

Edge dip (or lift) is the condition where the glass can be seen or measured to lift at the edges relative to the main flat plane of the sheet.  This is a function of the contraction of the edges which cool faster than the centre of the sheet when quenched and the effect of the leading and training edges of the glass cantilevering off the end of the rollers before reaching the next when quenching occurs.

Local distortion is measured at 25mm from the edge of the glass and can be a result of the combination of roller wave and edge dip (lift).

Roller Pick up can occur with glass thicker than 8mm and is the result of the self-weight of the glass causing it to slightly stick to the ceramic rollers when in it’s semi plastic state and appears as slight dimples in the glass surface.

All the above conditions have an impact on the quality of any lamination to be done to heat treated glasses as out of sync roller wave where peaks are aligned with peaks can cause thinning and thickening of the PVB resulting in further lensing than that seen with roller wave alone.

It should be noted that BS EN 14179 states that the measurement and tolerance values for the above only apply to thermally toughened soda lime silicate safety glass without holes and / or notches and / or cut-outs.  For these glasses the standard does not apply, and special or specific values should be agreed before processing.

Anisotropy (leopard spots etc) is an optical effect caused by stress patterns within the glass during quenching which result in polarising light.  Particularly apparent when viewing the reflection of the sky or when wearing polarising glasses this phenomenon is recognised in Hadamar and is not considered to be a defect.

In recent years there has been a move away from toughened glass on the outside face of buildings – particularly following the case at 125 Old Broad Street where a total of 17 lites on the outside suffered spontaneous failure due to NiS inclusions leading to a £15m settlement for the builder and subcontractor.  This has led to more laminated glass being used which has its own set of features and flaws which will be looked at in another posting.