Nickel Sulphide Inclusions
Nickel Sulphide Inclusions is not a new phenomenon, indeed it has been known of and studied by glass processors for more than 50 years, and yet even today it can still be met with surprise when it happens, why?
Let us start by looking at the text taken from a project specification, and please consider that the text is not an isolated event, this is almost word for word the same as many other projects I have worked on:
"Fully Tempered Glass (FT): Comply with BS EN 12150 or ASTM C1048, Kind FT.
Unless otherwise specified, all toughened glass shall be Heat Soaked' tested for a minimum of 2 hours at not less than 280ºC in order to minimize the occurrence of nickel sulphide crystals. This process shall be strictly controlled and carried out to BS EN 14179 “Glass in Building - Heat Soaked Thermally Toughened Soda Lime Silicate Safety Glass” paying particular attention to temperature and duration of treatment. The Trade Contractor shall state in his Tender submission his proposals in this respect, together with a written statement with regard to the avoidance of spontaneous glass failure. The Trade Contractor shall show to the satisfaction of the Owner, by statistical analysis of tested data, that the probability of failure in service is not greater than one occurrence in 130 tonnes. Fully toughened glass shall have a surface compression of not less than 100 MPa, or an edge compression of not less than 100 MPa, or both."
So at first glance do you see anything that concerns you? To many the answer will be no, but to those who have experienced NiS, the answer is most certainly YES!
'Statistical analysis of tested data, that the probability of failure in service is not greater then one occurrence in 130 tonnes'
Seems an innocent enough statement, but to all the mathematicians who understand that for the word probability to exist, the average breakage would need to be far less than one in 130 tonnes in order to limit the occurrence to the same. But is this possible?
Statements made by the glass industry, from where a probability of breakage was defined as one in 400 tonnes, was calculated using years of research and data from a glass processing plant. It has been noted by the industry experts, and already understood by the mathematicians, that for probability to exist, you will be likely to have one bad project which exceeds the one in 400 tonnes, combined with many projects which have little or no failure. But then surely this makes sense, after all the risk of NiS is generally linked to batch quality of raw materials and processing.
'Surface compression of not less than 100MPa'
Whats the problem here? Well for starters under ASTM standards, the minimum allowable surface compression is 69MPa as such the specification is calling for a far greater stress to be introduced into the glass.
Second, would you interpret the specified minimum allowable compression as 'pre' or 'post' heat soak testing? Indeed, if it is post, then the surface compression being applied to the glass during tempering needs to be in excess of 110MPa to make a due allowance for residual stress relief which will happen during the heat soak testing process.
The problem is that with the greater stress comes the greater risk of spontaneous failure, whether it is NiS or indeed some other foreign object within the glass, or even a minor defect to the edge.
Conclusion
1. Question and quantify the impact of specifications written with the intent of trying to exceed the defining limitations set within the international standards.
2. If you do not want to gamble with the risk of NiS, eliminate the need for fully tempered glass.
Let us start by looking at the text taken from a project specification, and please consider that the text is not an isolated event, this is almost word for word the same as many other projects I have worked on:
"Fully Tempered Glass (FT): Comply with BS EN 12150 or ASTM C1048, Kind FT.
Unless otherwise specified, all toughened glass shall be Heat Soaked' tested for a minimum of 2 hours at not less than 280ºC in order to minimize the occurrence of nickel sulphide crystals. This process shall be strictly controlled and carried out to BS EN 14179 “Glass in Building - Heat Soaked Thermally Toughened Soda Lime Silicate Safety Glass” paying particular attention to temperature and duration of treatment. The Trade Contractor shall state in his Tender submission his proposals in this respect, together with a written statement with regard to the avoidance of spontaneous glass failure. The Trade Contractor shall show to the satisfaction of the Owner, by statistical analysis of tested data, that the probability of failure in service is not greater than one occurrence in 130 tonnes. Fully toughened glass shall have a surface compression of not less than 100 MPa, or an edge compression of not less than 100 MPa, or both."
So at first glance do you see anything that concerns you? To many the answer will be no, but to those who have experienced NiS, the answer is most certainly YES!
'Statistical analysis of tested data, that the probability of failure in service is not greater then one occurrence in 130 tonnes'
Seems an innocent enough statement, but to all the mathematicians who understand that for the word probability to exist, the average breakage would need to be far less than one in 130 tonnes in order to limit the occurrence to the same. But is this possible?
Statements made by the glass industry, from where a probability of breakage was defined as one in 400 tonnes, was calculated using years of research and data from a glass processing plant. It has been noted by the industry experts, and already understood by the mathematicians, that for probability to exist, you will be likely to have one bad project which exceeds the one in 400 tonnes, combined with many projects which have little or no failure. But then surely this makes sense, after all the risk of NiS is generally linked to batch quality of raw materials and processing.
'Surface compression of not less than 100MPa'
Whats the problem here? Well for starters under ASTM standards, the minimum allowable surface compression is 69MPa as such the specification is calling for a far greater stress to be introduced into the glass.
Second, would you interpret the specified minimum allowable compression as 'pre' or 'post' heat soak testing? Indeed, if it is post, then the surface compression being applied to the glass during tempering needs to be in excess of 110MPa to make a due allowance for residual stress relief which will happen during the heat soak testing process.
The problem is that with the greater stress comes the greater risk of spontaneous failure, whether it is NiS or indeed some other foreign object within the glass, or even a minor defect to the edge.
Conclusion
1. Question and quantify the impact of specifications written with the intent of trying to exceed the defining limitations set within the international standards.
2. If you do not want to gamble with the risk of NiS, eliminate the need for fully tempered glass.