TestATunnocks

In Materials Science and Engineering, we look at how materials behave in different situations, and how environmental factors can influence this behaviour. This can be seen in something as simple as a chocolate bar, but the same properties and behaviour can be seen in composites materials for aerospace, building materials for nuclear reactors and nano-materials for electronic systems. Take a look at this video to see what we mean:

We went into the Materials Lab in The Diamond with a handful of chocolate bars to do some real industrial level experiments.The video gives a bit of info about some of these tests, but if you want to learn more about what we were doing, read on.

The material

While you might see just a simple caramel wafer chocolate bar, usually seconds before you eat it, Material scientists and Engineers will actually see something different - a sandwich structure composite. These are a special class of composite materials made by adding two thin, but stiff, ‘skins’ (also known as face sheets) to a lightweight, thicker ‘core’. 

While the core material normally has a lower strength, its greater thickness provides the sandwich structure composite with a high bending stiffness and overall low density. This makes it ideal for use in lightweight, stiff structures such as the fuselage of an aircraft. Infact, this very same caramel wafer style composite was used for production of The de Havilland Mosquito in 1940, but instead of confectionery, it used a balsa wood core with plywood skins.

Here, our Tunnock's Caramel Wafer bar is a multiple layer sandwich structure, consisting of five layers of wafer (the skins), separated by four layers of caramel (the cores). The bar is coated in chocolate -- in materials science and engineering terms, this can be thought of as a (yummy) surface finish.

One key thing to keep note of is the interface between the skin and core. In real composites, high stresses would build up at the interface between skin and core as these layers have quite different mechanical properties. In our chocolate bar version, the wafer and caramel layers have more similar properties, although the interface between the two is still where stresses will arise. If the bond between the two layers is too weak, the composite structures will most probably delaminate, that is to say the core and the skin will peel away from each other. Certainly not something that you’d want to happen if you were flying a de Havilland Mosquito! Keep an eye out for this happening when you test your own composites.

Bend testing

Bend tests are used to evaluate properties like the ductility and bend strength of a material. When we tested the chocolate bars in the lab, we used a three-point test, two on the bottom and one on the top. When you test them at home, you will probably have carried out a four-point bend test (unless you put your thumbs together in the middle of the bar) with two top, wider points of contact and two narrow lower points.

In the former, the maximum bend stress occurs at the central point opposite the centre contact point. In the latter, the maximum stress is distributed over the section between the two central narrower loading points.

A three-point bend is more suitable for materials that are homogeneous, while the four-point is particularly suitable for brittle materials where the presence of microscopic flaws has a big effect on the strength.

The results can be used to determine whether a material will fail under stress and are especially important in any construction process involving ductile materials that would be loaded with bending forces, such as the wing of an aircraft or a steel support in a building. If a material begins to fracture or completely fractures during a three or four point bend test, it is valid to assume that the material will fail under a similar stress state in any application, which may lead to catastrophic failure of the material and component.

Results of bend tests

Orientation of stress

We found that when the stress was applied on the surface of the wafers (perpendicular orientation), the chocolate bar deformed before it broke. The soft, pliable caramel (core material) was able to distribute the forces and stresses throughout the body of the bar without causing a crack (as it is more ductile), reducing the stress on the wafer skins.

However, when the stress was applied to the edge of the wafers (parallel orientation), a crack formed, leading to failure of the material, along with very little deformation. The crack initiated in the wafer and ran through it without being dissipated by the caramel material properties. This process was relatively slow due to the increased ductility and toughness of the material.

Results

	Perpendicular Load	Perpendicular Load	Parallel Load	Parallel Load
	Maximum Load (N)	Displacement at max load (mm)	Maximum Load (N)	Displacement at max load (mm)
-196°C	35.0	0.44	83.7	0.38
Ambient	42.2	1.53	80.9	1.70
30°C	15.6	2.73	43.4	2.35

Ambient temperature

How did the chocolate bar respond at normal room temperature? As we found above, when the force was applied perpendicular to the wafer orientation, the bar deformed before breaking. This bar withstood a force of approximately 42N, and deformed by 1.5mm before it broke. When the force was applied parallel to the wafer orientation, the bar withstood a much greater force (81N), and deformed by a similar amount (1.7mm) before breaking. Orientation of the stress and the composite is very important in structures.

Liquid Nitrogen (-196°C)

The effect of cooling a material to low temperatures is that the atoms are less mobile, and it takes a lower amount of energy to break the atomic bonds. You could test this out by putting your wafer bar in the freezer (although it will only get to around -20°C). We find that when a crack forms in a material, it travels easily through the structure - the material has become more brittle and has undergone a ductile-to-brittle transition. This could lead to catastrophic failure which is very dangerous in engineering!

In our tests, the cooled bar tested perpendicular to the wafers withstood only 35N before the initial crack started (and had deformed only 0.44mm). However, this crack reduced the stress in the material and to such a level that it temporarily stopped moving through the sample. As the force continued to be applied, the portion of the bar that remained intact was able to withstand 31.7N. The overall deformation of the bar was 1.12mm, until the crack finally continued again to failure of the material.

The cooled bar tested parallel to the wafers withstood 84N load before breaking quickly and catastrophically, only deforming by 0.4mm.

Elevated temperature (30°C)

When we raise the temperature of a system, we introduce more energy into it. The atoms are then more mobile and the material is more ductile. Rather than bonds breaking, it is possible for the material to rearrange itself a little bit by ductile flow (think about a sticky fluid being squeezed). This can make the cracks less sharp and less likely to move through the material quickly, breaking it. 

In our tests we found that the bar tested perpendicular to the wafers started to break at a force of just 16N, but deformed by 2.7mm before a running crack formed which moved slowly through the material, giving a final displacement of approximately 5mm.

For the bar tested parallel to the wafers, the maximum force was 43N and the deflection before failure was 2.4mm.

Compression testing

We wanted to know what happened to our chocolate bars when we squashed them. This is important for any material that is bearing weight.  We placed one bar at a time between two parallel plates and exerted a downward force on them.

When we applied the force perpendicular to the wafer orientation, the bar squashed in a uniform manner, with caramel and wafer fragments being forced out of the side by equal amounts as this is the weakest, and most ductile (and so can flow more easily) part of the composite.

However, when we applied a force to the edge of the bar (parallel to the wafer orientation) the bar sheared as the wafers slid across each other, delaminating. Where this starts is caused by the bar not being perfectly square, and the sides not being parallel, meaning that the applied force isn’t evenly distributed. However, we have also activated large shearing stresses in compression, due to the different mechanical properties of the core and the face sheets.

Charpy impact fracture toughtness testing

The Charpy Impact measures how easy it is to move a crack through a material and how fast it moves once it’s started. This is really important in engineering applications as a low fracture tough material, commonly known as a brittle material, only requires a small amount of energy before it catastrophically fails. Ductile materials require a greater energy to break, and ductile fracture is always a preferred mechanism of failure. Many cases have occurred through history where catastrophic failures have occurred as a result of brittle fracture. The most infamous of these is the sinking of the Titanic.

To measure the fracture toughness of a material, it is first placed on a Charpy machine on an anvil. A pendulum of known mass is swung from a set height (so we know its potential energy) and allowed to strike the testpiece at its lowest point (and therefore at its maximum velocity, and highest kinetic energy). The amount of energy the testpiece absorbs gives the fracture toughness and is measured by the remaining energy in the pendulum and how far it continues to swing after it breaks the material. 

 

The fracture toughness of a material is not always fixed. At low temperatures some materials are classed as brittle are actually ductile at room temperature. The change between these two is where the fracture toughness changes, and is known as a ductile to brittle transition temperature.

We tested this ourselves using our chocolate bars at ambient temperature and at -196°C (having submerged them in liquid nitrogen). We would expect to find that the colder samples have a lower resistance to the impact, meaning that they have lower toughness and are brittle. We also tested the bars in different orientations, so that the impact was both perpendicular to the wafer orientation and parallel to it.

Our results showed that the average energy absorbed by the bars tested at room temperature was 0.11J when the impact was parallel to the wafer orientation and 0.12J when the impact was perpendicular to the wafer orientation.

For the bars tested at -196°C, the average energy absorbed was 0.02J when the impact was parallel to the wafer orientation and 0.015J when the impact was perpendicular to the wafer orientation.

These results show that the bars exhibited greater fracture toughness when tested parallel to the wafer orientation, and that they were tougher (more ductile) at room temperature.

Optical microscopy

We sectioned one of our chocolate bars with a sharp knife and examined the surface with an optical microscope. We looked at the different components of the bars at different magnifications from 5x to 50x. We observed that each component was made up of different phases. The photograph below shows the chocolate in the centre, with the wafer in the bottom right corner and the wafer in the top right.

Fourier Transform Infrared (FTIR) Spectroscopy

Fourier Transform Infrared Spectroscopy is a technique used to obtain an infrared spectrum of absorption or transmittance of a solid, liquid or gas. It is used to identify organic molecules, as the frequencies of absorption or emission correspond to particular vibrations between interatomic bonds.  The technique is also used to observe certain chemical properties.

We used this technique to investigate what made up the different components of the chocolate bar: the chocolate, the caramel and the wafer.

The analysis showed clear peaks at specific wavenumbers, and that each component was very similar in composition. Batista et al (2016) analysed cocoa beans and chocolate using FTIR, and the spectrum we observed shows remarkable similarities to those reported in the 2016 paper: the test detected phenolic compounds present in cocoa and its derivatives.

We also compared the materials against a database of standard materials, which did not contain any specific food-related molecules, the closest match was for a chemical known to be used in cosmetics. Either our tester needs to wash their blusher off their hands before they run their next test, or we need a new database more suited to foods!

It is fair to say that the results were inconclusive, suggesting that sample preparation wasn’t adequately rigorous, this test possibly isn’t appropriate for this kind of material, or that that we didn’t have access to a suitable database against which to compare our results.

Differential Scanning Calorimetry (DSC)

DSC is widely used for examining polymer type materials allowing their thermal transitions to be found. Important thermal transitions include the glass transition temperature (Tg), crystallisation temperature (Tc), and melting temperature (Tm).

DSC works by measuring the difference in energy between a reference material and the sample needed to keep the sample and the reference at the same temperature during heating or cooling. When a sample is heated or cooled phase transitions (for instance, a solid to a liquid) can happen. The reference sample is selected so it doesn’t have any thermal transitions over the temperature range we are operating in. Depending on the phase transition, more heat could be absorbed by the sample in comparison to the reference material - an endothermic response. It could also be that the sample absorbs less heat than the reference material - an exothermic response.

Crystallisation temperatures of either an amorphous material on heating, or a liquid on cooling can be measured using DSC. This phase transformation is exothermic. Heating to higher temperatures may also allow us to see the melting of phases (an endothermic phase transition). Both of these are first order transitions so we see a peak in the DSC curve on heating or cooling.

DSC can also be used to look at other thermal transitions. For example, when heating a glassy material it may pass through something known as the glass transition (Tg). The structure of the glassy material goes from a rigid material, to a viscous (sticky fluid) or rubbery material. This is not a ‘phase transition’ but rather a structural reorganisation (a second order transition). In the DSC curve this will be a shift in the baseline, not a peak.

In these tests the sample is heated from -80°C to +50°C and cooled back down to -80°C again.

For the caramel on heating the glass transition occurs at approximately -25oC (seen by the shift in the baseline from -0.2 W/g to -0.35 W/g). There are two endothermic transitions after this at 5.57oC and 44.65oC. These could show the melting of ingredients in the caramel, such as the hydrogenated vegetable oil. See Hitachi (2007).

The DSC curve for the wafer has a similar shape to the caramel, showing a glass transition followed by two endothermic peaks on heating. This may be due to the contamination of the sample with caramel as they were difficult to separate.

The DSC curve for the chocolate shows a large endothermic peak at 16.61°C on heating and an exothermic crystallisation peak on cooling at 11.86°C. This is likely melting and crystallisation of the fats (vegetable oil and cocoa solids) in the chocolate.

In our experiment the crystallisation peak is at a lower temperature than the melting peak. Cocoa butter is known to crystallise very slowly which means the liquid chocolate can be undercooled (liquid exists on cooling below the melting temperature) before the crystallisation is observed in the DSC - a key process in the tempering of chocolate (NETZSCH (2020)).

References

1. Hitachi High-Tech Science Corporation (2007), 'Thermal Analysis of Candy', Hitachi High-Tech Science Corporation Application Brief 2007.9. Available at: https://www.hitachi-hightech.com/file/global/pdf/products/science/appli/ana/thermal/application_TA_080e.pdf [Accessed 20 March 2020]
2. Nádia Nara Batista, Dayana Pereira de Andrade, Cíntia Lacerda Ramos, Disney Ribeiro Dias, Rosane Freitas Schwan (2016). 'Antioxidant capacity of cocoa beans and chocolate assessed by FTIR', Food Research International, vol.90, pp. 313-319. Available at: https://www.sciencedirect.com/science/article/pii/S0963996916304537 [Accessed 20 March 2020]
3. NETZSCH Analyzing & Testing (2020), 'NL #3 Article: The Chocolate Side of Thermal Analysis' [online] Available at: https://www.netzsch-thermal-analysis.com/en/landing-pages/nl-3-article-chocolate/ [Accessed 26 March 2020].