Overcoming the problems of laminating toughened glass, toughened laminates

Until recently, architects in Southern Africa were rarely offered toughened laminates as an option. They had to utilise either monolithic toughened glass or laminated annealed glass, both of which have their strengths and their weaknesses.

The main reason for the reluctance by Laminate manufacturers to offer these products was the unacceptably high reject rate.

The last three years have seen an increase in demand for toughened laminates with several high profile projects requiring both the structural strength and integrity that only a toughened laminate can provide.

This report chronicles two projects, namely the Nelson Mandela Bridge in Johannesburg and the Gorilla enclosure at the National Zoological Gardens of South Africa in Pretoria. Details are given on the lessons learned, advances made and the significant reduction in quality losses. Recommendations based on hands-on experience, for the best practices in edge-working, tempering glass for lamination and successful de-airing are all included.

Increased demand

Previously the use of toughened laminates was discouraged by inflating prices, which was necessary to protect the manufacturer from the potential costs of high reject rates.

It is now time for manufacturers to review their procedures and reduce their reject rates so that they can enjoy the rewards from the increased demand for toughened laminates.

Customers may also benefit as the reduced cost of manufacture may lead to price reductions. This paper discusses the thought processes in deciding whether a product can be made and if so by what manufacturing method. This is then illustrated by the decisions made for two projects and the lessons learned during manufacturing.

Degree of difficulty

The first step is to determine the degree of difficulty in manufacturing the required toughened laminate.

Each project should be assessed individually as many factors have an effect on the manufacturing risk.

The degree of difficulty can be defined as being the product of the negative factors divided by the positive factors. Or figuratively by the following formula:

Negative Factors

Where ‘S’ is equal to the Size of the toughened laminate panel - the larger the size, the greater the risk.

‘G’ is the Glass type and substance. Here, the thinner the glass the more likely that roller wave distortion could be a problem. Also as a general rule, clear float glass is easier to keep flat than tinted and/or coated float glass.

‘N’ is the Number of glass plies in the laminate construction; the degree of difficulty is greater for multi-laminate constructions.

‘A’ refers to whether the construction is Asymmetric (i.e. contains more than one type or substance of glass). This is the biggest factor as it affects the capability of nesting the glasses together.

Positive Factors

‘P’ refers to the PVB thickness. While a thicker vinyl layer has more capability to absorb flatness irregularities it is also advisable to use as few layers as possible to minimise the entrapment of air between layers.

‘C’ stands for Correct stacking, while this goes without saying, it is not that easy and will be discussed in more detail later in this paper.

‘F’ refers to the Flatness of the toughened glass – the better the quality of the toughened glass, the easier the task of laminating.

‘M’ is for lamination de-airing Method, There are two methods that can be used: calendaring (pre-pressing); or vacuum de-airing.

While pre-pressing is the lower cost method, it does carry a higher risk factor than vacuum de-airing. If the negative factors are too great then the additional costs of vacuum de-airing are justified.

Nelson Mandela Bridge

‘Difficulty’ decision

The toughened laminate panels required for the Nelson Mandela Bridge were made from two 1930 x 1120mm plies of 6mm clear float with a clear PVB interlayer. The size and make-up of the glass made it suitable for a pre-pressing method of lamination.

Cutting & Processing

As with all toughened laminate projects, it is essential that cutting, edge working and drilling is done to a tight specification. Glass was first cut on a Lisec cutting table then drilled and polished on a Bavelloni Alpa 102 S - CNC machine which gives us the accuracy we desire.

Tempering

The glass size allowed for three pieces to optimise the furnace bed load. The direction of the glass was indicated by the placement of the silk screened logo. All glass for laminating is toughened to a quality standard, which requires a higher degree of flatness.

Laminate Assembly

The glass was assembled for laminating so that the respective logos on the two glasses were positioned on top of each other. The PVB used was a single layer of 1.14mm Du Pont Butacite which has been found to give better results than 2 layers of 0.76mm PVB.

The pre-pressing (hot-rolling) was done so that the direction through the pre-press was the same as the direction through the toughening furnace.

Autoclaving

A modified autoclave cycle is used where the pressure is held at 3.5 Bar until a temperature of 120°C is reached, then the pressure and temperature is increased to 12 Bar and 140°C respectively. This modified cycle has significantly reduced the occurrence of edge bubbles which are more prevalent on toughened laminates due to the thicker interlayer and the glass flatness.

Results

Despite all of the above precautions the first batch of laminates had a high reject rate of more than 25%. It appeared as though the problem was as a result of improper nesting. While care was taken to ensure that all glasses were processed in the same direction, the roller wave varied so that correct nesting was not possible.

For the second batch of laminates the furnace was loaded with the front, centre and back glasses being marked A,B & C respectively and loaded in the exact same positions for each furnace cycle. Glasses from two cycles were then made into matched sets (i.e A:A, B:B & C:C).

The lamination of these matched sets showed a reduction in rejects to less than 3%.

Roller wave experiment

An experiment was carried out to show the effect position relative to furnace rollers has on the shape of the roller wave deformation.

Four glasses of 4.0mm float were cut to a size of 250 x 1500 and loaded onto the furnace so that the short edge was parallel to the rollers. The glasses were staggered so that each subsequent glass was approximately 75mm behind the preceding glass.

The furnace was set deliberately to give excessive roller wave, after toughening the roller wave was measured using a dial gauge on a flatness table.

The results show that the roller wave pattern is dependent on the position in the furnace relative to the rollers. Only glass that has been tempered from exactly the same starting position can be correctly nested together.

Gorilla Enclosure, Pretoria Zoo

‘Difficulty’ decision

The toughened laminate panels required for the Gorilla enclosure at Pretoria Zoo, consisted of five plies of 10mm glass with a maximum size of 2950 x 1190mm. The size and make-up of the glass made it an unsuitable proposition for the pre-pressing method of lamination.

Cutting & Processing

Edgework was especially critical as many of the long edges would not be framed.

Tempering

The glass size allowed for only one piece to be tempered per cycle it was therefore essential that furnace conditions were kept stable until all five glasses per set were completed. Again the glass was tempered to the highest quality standard suitable for laminating.

Laminate Assembly

The glass was supplied to Rene Turck and Associates who are laminate manufacturers specialising in vacuum de-airing by means of vacuum bagging. The glass was assembled for laminating so that all five logos were directly on top of each other.

While this may appear to be an excessive amount of logos to use, it does ensure that the glass is correctly nested. Again the single layer of 1.14mm Du Pont Butacite was selected as being the best interlayer to use for this construction.

Vacuum de-airing/Autoclaving

After assembly the laminate was prepared for de-airing by using release fabric followed by breather fabric.

The vacuum bag was then sealed and vacuum applied. The laminates are then loaded under vacuum into an autoclave where a cycle of cold then hot vacuum, followed by a pressure cycle is followed.

The vacuum is kept on throughout the cycle, while the silicone rubber tubes used to connect the vacuum bags to the vacuum ports will be self sealing under pressure. It is however beneficial to be drawing vacuum during the release of autoclave pressure to eliminate possible formation of edge bubbles.

Results

The gorilla enclosure contains 48 panels of 54.56mm laminates and no rejects were made in lamination.

Conclusion

While it is possible to laminate almost any type of toughened glass, it is advisable to assess the degree of difficulty before making any commitments to manufacture. Ensure that all glasses have identical roller waves to allow for good nesting. Where possible advise against constructions that will cause potential problems. Consider the use of vacuum de-airing for the more complex constructions. If in any doubt, manufacture one or more sample.

Text and pictures taken from John Gilleeny’s presentation ‘Toughened Laminates: How to prevent an architect’s dream becoming a manufacturer’s nightmare? A Process Case Study’ - delivered on Monday June 20th 2005 at Glass Processing Days.