Wednesday, 12 February 2020

Tears in Slump

An enquiry was received about a tear in the slump of an intended bowl. It was reported that it was tack fused, giving a variation in thickness between 6mm and 9mm. The blank was tack fused with the following schedule: 
139°C to 560°C  30
222°C to  621°C  30
139°C to 786°C  15
9999  to  515°C  120
60°C   to 427°C  10
115°C to  350°C  10

This is a bit odd, as it had a slow rate of advance to top temperature and no bubble squeeze. But it should not have been a problem for this piece.

It was then slumped. When the kiln was opened after the completion of the slump, this split was revealed in the centre of the bowl, rather than a complete break.

The schedule was again a bit odd and high:

83°C to  148°C  15
167°C to  590°C  10
83°C   to 720°C  10
222°C  to 410°C  120
83°C    to 427°C  10

The very slow start would protect against any heat shock, but the schedule doubled the rate of advance while the glass was still in the brittle phase.  This possibly induced some stress into the glass.  There is no reason why the rate of 167°C per hour could not have been continued to about 630°C to 640°C.  There is no need to slow the rate of advance once the glass is out of the brittle phase (ca. 540C).  With this slow rate of advance, the bowl may have been slumped into this simple ball mould at 620°C or less.  

This indicates that observation is needed when trying new layups and moulds to find the appropriate temperature.  It reveals at what stage any problem occurs.

There are several possibilities to consider in diagnosing the failure of this piece.

Rate of advance 
It might have been just too quick for the thickness of the piece. The piece is reported to have varied from 6mm to 9mm. From the picture this might have meant a single layer base, or more likely, a two-layer base and then the two-layer top pieces, making it 6mm to 12mm.  The firing schedule would need be as for an 18mm piece in the first instance, and 24mm in the second.

The initial rate of advance is more than slow enough for the calculated size of the thickest. The doubling of the rate of advance in the early stage of the brittle phase of the glass is more problematic. It may have had the effect of inducing a weakness (stress) that became apparent in the next segment with a slower rise to the top temperature. It probably heated the top much more than the bottom during this second segment, leading to the tear.

Was it adequately annealed?  
If you look at the fusing schedule, the anneal soak and anneal cools were adequate for a 12mm piece, so if compatible glass was used, there should be no annealing stress in the piece.  It is important to consider this, as a stressed piece can often break during an otherwise adequate slump schedule.

The placing of the piece may have had a significant effect on the outcome.  It is placed in the corner of the kiln.  A single element shows around the side of the kiln.  This would not be enough to heat the whole kiln, so we must assume it is mainly top fired. If the kiln has significant differences in temperature (only as much as 5°C), the location in the corner may have had enough of an effect to cause the stress.

When did the split occur
It is possible that the split occurred on the way up and then re-attached during the slow rise to the high working (lamination) temperatures.  This would give the appearance of an incomplete split, commonly called a tear.  We don’t have any information about the state of the edges. At 720°C an early split could have re-attached. The reason for considering this possibility, is the change of curve on the left end of the split. However, this does not seem likely as the split clearly shows on the edge of one of the squares, but does not go all the way across.  Lamination temperatures are not high enough to seamlessly heal a crack.

It is most likely the split occurred during the rise in temperature.  The reason for speculating this, is because of the distortions of the squares.  If the split occurred before any significant slumping the distoritions in the squares would be explained.

Range of Considerations

This discussion shows that there is much more than just the schedule to consider in diagnosing failures. 

Yes, the schedules are odd, but not impossible to get a good result with.  The fuse and slump schedules vary to a large extent which is not ideal.  

The fuse schedule to the bubble squeeze temperature is sensible, but the slow rate of advance is continued to the top temperature, which is not usual.  This means there was no consideration of a bubble squeeze soak, although this did not prove to be a problem. The annealing is suitable for up to 12mm.

The slump schedule starts very slowly and then doubles early in the schedule.  This rapid change of rate early in the brittle phase of the glass may have induced stress that could not be relieved later in the firing. A simple single rate of 167°C to the slumping temperature would have been adequate. 

The slump temperature used is in the lamination/sharp tack range, rather than the more usual 620°C to 677°C range.  This can bring several problems with excessive marking of the bottom of the piece, shrinkage, and even uprisings at the bottom of the piece.

Are any of these criticisms of the schedules adequate to point to as causes of the break? Without handling the piece, it is difficult to tell with the information available.

General Diagnosis

You need to consider the state of the break in diagnosis.  We do not know whether the edges were sharp or slightly rounded. It is clear the split piece fits the mould as it is, so the break probably occurred on the way up in temperature.

Was the high slumping temperature an element in the break? The third rate of advance was slow, so there should be no further stress introduced in the firing.  The annealing of the fused piece seems adequate, so a stress fracture does not seem likely.

The high slumping temperature may be disguising a problem, as the glass may have re-attached at the upper temperature.  But this does not seem likely as the picture seems to show the crack enters the red piece a short distance, but not all the way through.

Was the placing of the mould problematic? It is in a corner where there may be enough difference in temperature across the glass to induce this kind of tear.

These considerations show that multiple images of a problem piece need to be taken.  Various angles are required to eliminate problems with reflections.  Pictures of the whole setup in the kiln and the piece are needed.  Multiple pictures of the top and bottom are needed. And of course, close ups of the area(s) concerned.  These are all substitutes for handling the piece itself.  

If you have a community of kilnformers or a store you can take the piece to, you are likely to be able to give responses to questions and to get the information required about the possible problems and solutions.

No comments:

Post a Comment