Project Synopsis

Mr James Barclay MEng (Hons) ACGI AMIMechE, Rolls Royce

Download the project synopsis as a PDF [772 KB]

The texture of food is critical to how it is perceived by the consumer. Sensory analysis has defined terms such as ‘crispy’, ‘crunchy’ and ‘crumbly’ to describe the experience of eating a given food. When eating food products such as wafers and crisps, one expects a ‘crispy’ texture, and is disappointed if it is ‘soggy’ or ‘crumbly’. Food manufacturers have realised the importance of sensory perception to consumer satisfaction, but are currently only able to define terms such as ‘crispy’ in a subjective manner. An objective basis to measure sensory properties is highly desirable, and could be incorporated in the manufacturing process.

This project aims to define crispness in terms of engineering properties. A consumer will define a food as ‘crispy’ by the way that it fractures in the mouth, by the forces required to cause fracture, and by the sound emitted during fracture. The project therefore involved measurement of the fracture energy, Young’s modulus, and acoustic emission of a variety of wafers. The wafer biscuit was chosen as it is broadly described as a ‘crispy’ food in sensory analysis. Clearly if a particular mechanical / fracture / acoustic property – or combination of these - , were found that could be reliably used to define a food as being ‘crispy’, it could be used extremely effectively in quality control of mass produced crispy foods. A consultation with Nestlé emphasised the importance of moisture content on the sensory perceived crispness, and therefore Kit Kat wafers were tested both fresh baked, and at 12% moisture content (wafers were supplied in sheet form and were not ‘robed’ in chocolate). A low density, fresh baked ‘light’ wafer was also tested for comparison. In sensory terms, fresh baked Kit Kat wafer is ‘crispy’ and to a lesser degree ‘crunchy’, fresh baked ‘light’ wafer is ‘light’ and ‘crispy’ and Kit Kat wafer of 12% moisture content is ‘soggy’ and ‘chewy’.

To provide a firm foundation for the project, a thorough literature research was carried out on the mechanical, fracture and acoustic properties of wafers. A wafer has a cellular structure, made up of solid cell walls and air pockets. Extensive preliminary tests were performed using a range of methods to measure Young’s modulus E, fracture toughness K, fracture energy G and acoustic properties. Only the test methods that gave highly reproducible results were carried through to final testing.

Final tests included:

• Uniaxial compression, spherical indentation, and 3 point bending to measure Young’s modulus.
• Single edge notch tension (S.E.N.T) to measure fracture toughness (K) and fracture energy (G).
• Acoustic measurement during crushing using compression plates, spherical probe and artificial teeth to plot a force time trace synchronised with a sound time trace.
• A Fast Fourier Transform (FFT) used to plot the frequency spectrum from the sound time trace.
• Micrograph analysis performed on fractured specimens, to explain the mechanical and fracture
• properties.

Figure 1 – Tests for Young’s modulus

The shape of the stress strain curve shown in figure 2 from the uniaxial compression results gives an excellent insight into the process of fracture during crushing, and allows the concept of ‘jaggedness’ to be introduced.

Figure 2 - Stress-strain plot from a typical compression plate test for each type of wafer

Looking at the fresh baked Kit Kat curve, one can see that when compressed the wafer exhibits an initial linear region, followed by brittle fracture. The brittle fracture zone is characterised by the size and number of load drops, also called its ‘degree of jaggedness’. After significant brittle fracture, opposing cell walls begin to touch, giving a region of densification. The curve for the ‘soggy’ 12% moisture wafer shows no such jaggedness, indicating that a switch in failure mechanism from brittle to ductile has occurred, due to plasticizing effect of water. The shape of the brittle crushing zone for the ‘light’ wafer also appears different to that of the fresh baked Kit Kat wafer. It is demonstrated on several occasions throughout the project that the degree of jaggedness can be used to characterise the different types of wafer.

Figure 3 shows the Single Edge Notch Tension (S.E.N.T) test rig, along with a tool made especially for the project, used to cut a precrack of consistent length.

Figure 3 - The SENT testing setup, and the tool used to cut precracks

The S.E.N.T test setup gives highly reproducible results for fracture toughness and fracture energy, and shows that these properties correlate poorly with crispness.

Turning to sound testing, the aim was to produce a synchronised plot of load against time with sound pressure level (SPL) against time, and to produce a frequency spectrum (power spectrum) from the sound time trace. Sound was recorded during compression tests, which were performed using compression plates, a spherical probe, and a pair of artificial teeth shown in figure 4 below. The compression tests were carried out on an Instron machine. A tie clip style microphone was used to pick up the sound.

Figure 4 – Shows artificial teeth used in sound testing.

A device was made that would enable the sound – time trace measured using the microphone to be synchronised with the load – time trace from the Instron machine. The devise (shown in figure 5 below) works in a similar way to a clapper board used in the film industry, by making a sound the moment the cross head of the Instron machine starts to move.


Figure 5 - Light to sound converter used to synchronise load time from Instron with sound time trace.

Figure 6 – Synchronisation of load and sound pressure level (s.p.l) plotted against time

Figure 6 demonstrates well that a sound pulse is produced at the moment of a load drop, which represents the occurrence of a fracture event. A fraction of the stored energy in the wafer just prior to collapse is converted to sound at the instant of collapse. Looking at the results from acoustic testing in more detail reveals that the degree of jaggedness of the load drops during compression can be linked to a change in loudness of sound produced during crushing, which sets the crispy fresh baked wafer apart from the soggy 12% moisture wafer.

Drawing conclusions from the project overall, it is clear that mechanical properties can be used to characterise crispy products by the degree of jaggedness in the load – displacement plot from a compression test, and that the loudness of acoustic emission during crushing can also be used as a measure of crispness.

The jaggedness found in the load-deflection plot for crisp fresh baked Kit Kat wafer due to the brittle fracture mechanism is totally absent from the plots for the ‘soggy’ 12% moisture Kit Kat wafer. Microscope photos show that the moisture has the effect of changing the failure mechanism from brittle fracture, to gross plastic deformation followed by eventual fracture. Compared to the fresh baked Kit Kat, the loudness of acoustic emission is greatly reduced for the 12% moisture wafer, because the plastic deformation prior to fracture means that less energy is available to be released as sound when fracture finally occurs.

The Young’s modulus and fracture toughness of the light wafer was found to be less than that of the fresh baked Kit Kat wafer. This is due to the density of the light wafer being lower than that of the fresh baked Kit Kat wafer, which results in thinner wall thickness of the cellular structure. It is also possible that differences in ingredients (e.g. differing amounts of sugar) affect the mechanical properties, and therefore influence the sensory perception of the wafer.

The primary future challenge is to perform uniaxial compression tests and spherical probe indentation tests on both light and Kit Kat wafers and analyse the jaggedness of the load – displacement curve for characteristics such as the number of spatial ruptures (the ratio of the total number of peaks to the distance of puncturing) and the average specific force of structural ruptures (the ratio of the sum of force drops per peak to the number of peaks). It would also be of benefit to perform mechanical and fracture tests (compression, three point bend) on light wafer at a range of moisture contents, and on Kit Kat wafer at a range of moisture contents. This will allow the comparison of the effect of moisture on wafer containing sugar (Kit Kat) to wafer without sugar (light wafer).

Back to top