One question that we are often asked by customers is “can we use the old sheet that we have in stock and not have problems?”
In general, the answer to this question is yes, as long as the sheet is still masked and has been stored in good conditions. The properties of the sheet will not deteriorate over time. The only thing that you may need to do, depending upon the application, is dry the sheet before laminating or thermoforming (and we would recommend that you dry even new sheets for these applications).
However, we recently came across a very specific problem that one of our customers was having while using old sheet. The customer laminates polycarbonate sheet to glass using transparent polyurethane. There was some old sheet that the customer wanted to use in this application, the sheet had been produced about three years ago, but was fully masked and had been stored carefully. When the polycarbonate was bond to the polyurethane, the bond strength was measured and was found to be about 10% lower than when using new sheet. In ballistics laminates, this lower bond strength can potentially lead to some longer term problems with ballistics performance and de-lamination. An extensive study was therefore carried out to identify the cause of this lower bond strength.
The conclusion of the study was that over time, plasticizers from the masking had migrated to the surface of the polycarbonate sheet. These plasticizers then weakened the bonding of the polycarbonate to the polyurethane. As a result of these studies we are now recommending that polycarbonate used for ballistics laminates is used within 12 months of production.
In theory, this recommendation sounds simple to achieve. The customer would just need to look at the production date on the packaging and confirm that it is less than 12 months old. However, the customer needs to understand that the production date on the packaging and/or masking is probably the date that the hard coat was put onto the sheet. In practice, the actual production of the original sheet could be many months or even years before this date. The plasticizers would have started migrating to the surface of the sheet ever since the original production date of the sheet and the coating process would not have removed them.
Fortunately, in a world where most produces have quality systems in place, traceability of material is nearly always available. We would now recommend that all laminators of ballistics material ask their polycarbonate supplier to send them the material lot number, the coating date of the sheet (if coated) and the production date of the base sheet. If the production date of the base sheet is over 12 months old, it is likely to have lower bonding strength which may decrease the product performance. If the sheet supplier is unable or unwilling to supply these dates, we would recommend that adhesion testing is carried out or material is purchased from another supplier.
De-lamination of transparent armor is an ongoing problem. This blog post aims to explore the subject using some technical theory, with the aim of presenting simple solutions to minimize the problem. The proper design of the laminate, manufacturing of the laminate and selection of materials can all lead to a significant increase in the life of the laminate. This post will explore some of the issues and provide recommendations.
To start looking at the problem of de-lamination we want to start with a mathematical analysis of the problem and we therefore used a simple formula to model the stresses that cause de-lamination. The simplified formula is taken from the paper “Thermal Stress in Bonded Joints” by W.T.Chen and C.W.Nelson. It examines the thermal stresses in a bonded joint between two materials using an adhesive interlayer. The paper also gives a more complex formula for three layers instead of two; for readers who would like to examine the formula for three layers, the paper is easy to find by entering the title of the paper in Google. For more complex structures involving more than three layers, the formulas can be derived using the same principles. The paper shows how the following formula is derived, but for this blog post, we will just take the formula as given. For those that prefer to see the derivation, the paper is available to read. It is recognized that equating a laminate to a bonded joint is somewhat simplistic, but it does give a good starting point to analyze the problem.
The formula presented in the paper for calculating stresses in a two-layer joint is:
Τ= (α1 –α2) T G sinh (β x)
β η cosh (β L)
β2 = G [ (1/(E1 t1) + (1/(E2 t2) ]
Τ = Shear Stress (Pa)
α1 = Thermal expansion coefficient of layer 1 (/C)
α2 = Thermal expansion coefficient of layer 2 (/C)
T = Temperature change (C)
x = Distance from center of joint (mm)
L = Distance from center of joint to end of joint (mm)
G = Shear modulus of interlayer (Pa)
η = Thickness of interlayer (mm)
E1 = Elastic modulus of layer 1 (mm)
E2 = Elastic modulus of layer 2 (mm)
t1 = Thickness of layer 1 (mm)
t2 = Thickness of layer 2 (mm)
The formula can be used to calculate the Shear stress at any point in the laminate from the center to the edge. When x = L at the edge of the laminate, the shear stress will be maximum, and:
Τmax = (α1 –α2)T G
This formula is somewhat intuitive. The stress will be greater if the difference in coefficient of thermal expansion of the two materials α1 –α2 is large. The stress will also be greater as the Temperature change T increases. Also if the interlayer is thicker (η), it allows the stresses caused by the expansion and contraction of the materials to be reduced.
The first thing to note is that transparent armor is often exposed to environmental temperature changes in military applications. ATPD.2352 requires testing over a temperature range of -31 C to +60C or a 91 degree C temperature range. Although the laminate will not see this range in temperature every day, it is certainly possible that it could experience these conditions during its life.
It should be noted that if normal operating temperature is say 15 C, this is not the temperature that has zero stresses. The temperature that has zero stresses is much closer to the temperature during fabrication the polyurethane sets and bonds to the glass and polycarbonate. Depending upon the polyurethane, this temperature could be 80 C or higher. Selection of the polyurethane therefore has some impact on the maximum stresses that a laminate will see. A polyurethane that sets at 120 C will lead to much higher stresses than a polyurethane that sets up at 80C.
To illustrate this point, the maximum stress will occur in a laminate when the temperature of the laminate is the lowest, in the case of ATPD.2352 this will be -31C. Using a polyurethane that sets up at 120C rather than 80C will give about (120 – -31) / (80 – - 31) = 151/ 111 or about 36% more maximum stress in the laminate at the interface.
It should be remembered in the selection of polyurethane, that choosing a low melting polyurethane to minimize stresses should be done with careful consideration of the operating and storage environment that the laminates will see. It is extremely counterproductive to have solar heating leading to the melting of the polyurethane, as this will lead to melting de-lamination rather than thermal stress de-lamination.
This problem can be made even worse by poor laminating control. Polycarbonate expands or contracts a lot more than glass. If the laminate is not uniform in temperature throughout the entire thickness at the time the Polyurethane is setting up, it is possible that some of the polycarbonate could be at a higher temperature at its core at the time the surface is bonding to the polyurethane. This increased core temperature can cause increase stresses at the interface of the polyurethane. Proper manufacturing that allows the temperature of the laminate to stabilize throughout, just above the temperature where the polyurethane sets up can significantly reduce stresses.
One elegant solution to the problem is to use radio frequency lamination to lower the temperatures of the polycarbonate and glass at the time of lamination. This type of lamination heats only the polyurethane interlayer and can therefore reduce the zero stress temperature well below the temperature achieved by conventional autoclaves. We can provide laminators with information on this process if requested.
The other item to note from the formula is that the thickness of the polyurethane is important. Using a thicker polyurethane can allow the stresses to be significantly reduced. If we consider that case where 6mm glass is bonded to 6mm polycarbonate, using the above formula the stresses can be reduced from 13.7 MPa to 6.9 MPa if using 0.075mm polyurethane rather than 0.025mm polyurethane with a temperature swing of 111 degrees C.
Decreasing the amount of thermal stress generated will significantly affect the life of the laminate. Halving the stress, as in the above example, could be the difference between de-lamination and no de-lamination. The other factor that affects de-lamination is the adhesion between the polyurethane and the other materials – glass and polycarbonate. De-lamination will occur at the weakest of these joints, which is typically the polycarbonate, polyurethane interface. De-lamination will occur when the forces due to the thermal stresses are stronger than the adhesion of the polyurethane to the polycarbonate or glass.
One area where we have started to have some positive effects in reducing de-lamination in high-end laminates is increasing the bonding between the polyurethane and the polycarbonate. We have been tackling this area in two ways, firstly by correct selection of the polyurethane and secondly by modifying the chemistry of the polycarbonate. We have recently made available an enhanced grade of polycarbonate that has significantly higher bond strength to polyurethane.
The next area that should be considered is the area of laminate design. In some cases laminates are configured only to pass ballistics specifications and little consideration is give to how the configuration may affect stresses and de-lamination. To illustrate this point we will use the three-layer formula developed in the paper that we discussed earlier. Due to the formula’s length, we will not present it here, but again the paper can easily be found.
In the first case we will consider a two-layer laminate consisting of 6mm Polycarbonate bonded to 6mm Glass using a 0.025 mm polyurethane. The change in temperature that the laminate will be exposed to will be considered to be 100 degrees C. We have calculated that the maximum stress will be 12.3 MPa.
If we then change the laminate configuration, with the aim of keeping the total thickness the same, to 3mm Polycarbonate, 3mm Polycarbonate and 6mm Glass, the total amount of polycarbonate and glass will remain the same. In this configuration the maximum stress between the glass and the polycarbonate will be 11.70 MPa. Although the difference may not seem to be much, it is a 5% reduction in the stress. In a laminate that is close to the point of de-lamination, reducing the stresses by 5% could be enough to significantly increase the life of the laminate or even prevent de-lamination occurring. Reducing thermal stresses, particularly when done in conjunction with increasing the bond strength between the polyurethane and the polycarbonate, can be very effective in decreasing de-lamination and increasing laminate life.
Other factors do affect de-lamination including edge seals, chemical attack and edge finishing, but the aim of this article is mainly to look at some of the factors associated with de-lamination caused by thermal stresses.
The key points to minimize thermal stresses and reduce de-lamination are:
- Select the correct thickness of polyurethane to minimize thermal stresses
- Select the correct type of polyurethane to minimize stresses and increase bonding, while also considering environmental conditions that the laminate will be exposed to.
- Optimize autoclave conditions to reduce thermal stresses.
- Improve the bond strength between the polycarbonate and the polyurethane by using an enhanced polycarbonate designed to increase bond strength in transparent armor.
- Design the laminate configuration to minimize stresses in addition to achieve ballistics requirements.
In the final blog post of this trilogy we will discuss Iridescence and how it can cause a rainbow effect on abrasion resistant coated Polycarbonate sheet. We will also discuss how the rainbow effect can be minimized.
Iridescence is the rainbow or oil slick type pattern that often appears on the surface of a Polycarbonate sheet particularly under artificial lighting conditions.
Coated Polycarbonate sheet has a thin film of coating on the surface of the sheet in order to protect the sheet against abrasion damage. It is this thin film of coating material that causes the problem, in much the same way that a thin film of oil on the surface of a pool of water exhibits the rainbow patterns. The effect is due to the process known as interference.
As discussed in previous blog posts, whenever light travels from a material with one refractive index to another material with a different refractive index, some of the light is reflected. In the case of the coated Polycarbonate sheet, when light moves from the air into the coating some of the light is reflected. Then, when the light moves from the coating into the actual Polycarbonate, some more of the light is reflected. When the light that is reflected from the first surface comes into contact with the light that is reflected from the second surface the light waves recombine.
Depending on how thick the coating layer is, the light waves may be in sync when they recombine or may be out of sync when they recombine. If they are in sync the two waves will added together and will have constructive interference. If they are out of sync the two waves will start to cancel each other out and will have destructive interference.
Since visible light has wavelengths of 380nm(violet) to 750nm (red) and a typical hard coat has a thickness of 4 to 7 microns [4000 to 7000 nm], the coating thickness is an order of magnitude thicker than the wavelengths of visible light. A small percentage variation in the coating thickness can therefore change whether the constructive interference or destructive interference occurs. If there is variation of coating thickness over a small area of sheet, even if the variation is only tens or hundreds of nanometers, then there will be areas of constructive interference and areas of destructive interference. This variation in the interference patterns is part of the cause of the iridescence or rainbow effect.
The question then becomes, how do we eliminate the variation in coating thickness? Abrasion resistant coatings are often added to sheet by a process known as flow coating. The sheet is hung vertically and coating solution is allowed to run down the surface of the sheet from top to bottom under gravity. The solvents are then allowed to evaporate. If the sheet is allowed to move before most of the solvents have evaporated, the coating surface can become uneven. However, we need to remember that the coating surface is not the only surface that we need to be concerned about – there is also the sheet surface that is reflecting light. The sheet is extruded between large chrome rolls which are powered by motors. If there is any variation in the motor speed of these motors or the motors pulling the sheet, there can be variation in the thickness of the sheet. While the variation in the thickness will be small, it only requires very small variation to cause iridescence.
The reality is that neither the sheet producers or the abrasion resistant coaters have the ability to control their processes to the level of 10-100nm thickness. If we look at sheet producers, many of them state that their thickness specification is plus or minus 10%. On a 0.118″ thick sheet that corresponds to 300,000 nanometers. While this is an overall thickness tolerance and not a measure of local variation of thickness, it does give some idea of the magnitude of the problem. Using this information we can determine that the sheet producers and coaters cannot prevent the problem. In many cases sheet producers often blame the coaters for the problem and coaters often blame the sheet producers. This then leaves us with the question of how do we solve the problem?
To answer the question we will briefly move to another topic – different types of lights. Traditional incandescent light bulbs have a relatively smooth light spectrum across the visible region and are similar to sunlight in this respect. When sunlight is split into its component wavelengths (such as in a rainbow) there is a smooth transition from violet through the various colors to red. There are no wavelengths missing. An incandescent light bulb behaves in the same way (as do some full spectrum LED bulbs).
Fluorescent bulbs, mercury bulbs, sodium bulbs and non full spectrum LEDs are different. When the light is split into its component parts, there are peaks at some wavelengths and gaps at other wavelengths. For example, a low sodium bulb emits an almost monochromatic light source at 589.3nm and a standard fluorescent bulb has 22 peaks with the main four being Mercury at 437nm, Terbium at 543nm, Mercury at 547nm and Europium at 611nm. These wavelengths of a fluorescent bulb combine to yield a light that looks like natural light but has discrete wavelengths rather than the continual spectrum of natural light.
Having a light source composed of discrete wavelengths rather than a continuous spectrum is a major problem for iridescence; when the light is reflected from the two surfaces the discrete wavelengths make the problem much larger as there are no intermediate colors to cancel out the iridescent effect. In short, the light source can make the problem of iridescence much greater.
The best way to reduce the effect of iridescence is to change the lighting source to a full spectrum light source such as incandescent bulbs or full spectrum LEDs. If the only option is to use fluorescent bulbs, it is better to use a bulb with more emission peaks to more closely resemble full spectrum light.
Another option to completely resolve the problem is to use what is known as an index matched abrasion resistant coating. The Polycarbonate sheet has a refractive index of 1.585 and most coatings have a refractive index of 1.49. If an abrasion resistant coating with a refractive index of 1.585 is used, the light will treat the coated Polycarbonate sheet as a single layer material and the effect of iridescence will be completely eliminated. While this process sounds great (and HighLine Polycarbonate can offer index matched abrasion resistant coated products) there is a significant downside – index matched coatings are very expensive. In most applications it is better to install full spectrum bulbs to reduce the problem.
Finally, to illustrate the effect of lighting on the visual appearance of iridescence we will recount a case study about the problem. A rail car manufacturer was experiencing oil slick like patterns on the Polycarbonate windows of their railcars. The manufacturer of the windows was inspecting the windows prior to sending them to the rail car manufacturer to try and identify the problem. They were unable to detect the issue as their factory was lit with incandescent lights. When the windows were installed in the railcars, the oil slick appearance was easily visible because the internal lights on the rail car were fluorescent bulbs.
The most practical solution would have been to change the bulb type on the railcar, but unfortunately the window manufacturer did not understand the problem. They told the rail car manufacturer that the problem was due to birefringence, which, as anyone who has read these three blog posts knows, was not the cause of the problem. By understanding the cause of the problem it is easier to recommend a solution to the customer.
Anti-glare coatings are different to anti-reflective coatings. Anti-glare coatings are generally produced using an abrasion resistant hard coat with small particles in the coating to give a matte surface. This matte surface stops light being reflected from the sheet surface back to the viewer so that the user’s view is not obscured by glare from lighting or the sun.
One down side to the matte surface is that the light transmission of the sheet is lowered and the view through the sheet is hazy. The more of the matte agent that is put into the sheet the more the glare is reduce, but also the sheet becomes more hazy and the view more obstructed.
To illustrate the effect of an anti-glare coating we have taken three pictures of an anti-glare sheet with a 40% gloss level. The 40% gloss is quite a high level of matte agent – we commonly supply product with gloss levels of 60% and as high as 80%. The 80% gloss level is much more transparent but does not reduce the glare as much as the 40% gloss level material.
We are often asked how much does the reduction in gloss level obscure the view through the sheet? The answer depends on what you are trying to view. If you are trying to view something that is a long way away through the sheet, the object is still able to be seen but the view is very blurred. To show this effect, we positioned a typed page only 15″ behind the anti-glare sheet. The page is visible but the details are not.
We then moved the page to 5″ behind the sheet. Again the page is visible and you can even start to make out the detail of some of the larger font. 48 Point font is clearly legible, even 28 Point font is just visible, while smaller font can be seen but not read.
We then moved the typed page to immediately behind the sheet and the page was even touching the sheet. Nearly all of the font, even the smallest can be clearly read.
When choosing an anti-glare gloss level it is important to test it in your application. The questions that need to be answered are how much do you need to reduce glare and how much haze can you accept. The answers to these questions depend on what environment you are you using the sheet in and what do you need to see through the sheet.
In the last post we discussed how stresses in Polycarbonate can cause the material to become Anisotropic and exhibit Birefringent properties. Light waves parallel to the stress direction will travel through the sheet at a different speed than the light waves perpendicular to the stress direction.
It is possible to visualize the stresses in the sheet due to the birefringent properties of the sheet. A technique known as Photoelasticity is often used. In this method light is first passed through a polarizing filter, in order to block all components of the light not vibrating in the direction of the plane. The light coming through the filter is then known as polarized light. The light is then allowed to pass through the Polycarbonate part being examined. The birefringent properties caused by the stresses cause the polarized light to be split into two perpendicular components each moving at different speeds, which are governed by the amount of stress in each direction. The components of the light waves recombine on leaving the Polycarbonate. When this light is then viewed through a second polarizing filter it is possible to see the effect of the retardation of the light in the form of “rainbow” like patterns. There is a lot of theory that can be explored on the method of Photoelasticity and this theory can easily be researched by carrying out a web search. In this blog we do not plan to go into advanced theory of how the light waves recombine, but rather discuss how the method of Photoelasticity can be practically used.
In the picture at the top of this blog post is a photograph taken of a piece of Polycarbonate with a hole drilled through it. The photograph was taken with a simple phone camera and two polarizing filters bought from a camera shop for $25 each. One filter was put behind the Polycarbonate part and one filter was put in front of the part. Although this cheap set up does not compare with advanced equipment for visualizing and measuring Photoelasticity, it does provide a simple practical tool for visualizing stresses in Polycarbonate parts.
In the Photo it can be seen that there are high levels of stresses on each side of the hole. We suspect that these stresses were caused by poor drilling technique using the wrong drill bit for Polycarbonate and operated at the wrong speed. It is also possible that the drill was started while in contact with the sheet. The technique of Photoelasticity allows us to visualize these stresses and therefore allows us to adjust fabricating methods to minimize stresses. This information is particularly important, as we know that areas of increased stress are prone to cracking and damage, especially when exposed to certain solvents.
We invite readers who are involved in fabricating Polycarbonate parts to try this test method themselves to see the stress areas on the parts. All you need to do is buy two Polarizing filters from a camera shop.
In this section of the trilogy of blog posts on the subject of the rainbow effect, we have seen how stresses in Polycarbonate sheet can lead to birefringence and that these stresses can be visualized through polarizing filters as a rainbow type pattern.
However, it should be understood that rainbow effect seen on some hard coated Polycarbonate sheet without the use of polarizing filters is not due to the birefringence of the material. These rainbow type patterns on hard coated sheet are often very easy to see with just the eye and can cause the visual appearance of the sheet to seem very poor. In the last post on this topic, we will discuss what causes the rainbow effect on coated sheet and how its effect can be minimized.