Physics of Cooking

Last week we began measurements on McVitie’s Digestive biscuits and carrots. We steamed the samples for different periods of time in order to modify their moisture content. We chose these two foods because of their very different structures. We want to probe how the moisture dependencies of acoustic and mechanical parameters are influenced by structure.

We steamed biscuits for 30, 45 and 60 seconds and carrots for 1,2,3,4 and 5 minutes in a steam oven. The acoustic data was consistently limited for different steam-time samples in both foods and failed to give any useful information. However, the force-deformation curves were more successful for both foods and we were able to identify parameters in each that showed a dependency on steam time.

Read The Relationship between Structure and Crispness in full

The article below was written for Imperial College London’s physics department and will be used for outreach purposes to show the range of projects undertaken by MSci students. It gives an introduction to our 2014/15 MSci project in the context of wider research on the physics of multi-sensory cuisine.

Read Introduction to Food Texture Research in full

Up until recently our project has focused on engineering an acoustic-mechanical texture analysing device. As Leon mentioned in his last post, we have now completed the construction of the device and, over the past couple of months, have been carrying out preliminary measurements. During the data acquisition and analysis process, we have drew heavily on existing research. The picture below links to my review of all the pre-existing literature concerning acoustic-mechanical measures of crispness.

In my first post I discussed how there was a sparsity in existing research with regards to quantifying crispness in crusted foods with high-moisture cores (e.g.

Read Literature Review on Existing Acoustic-Mechanical Research in full

Acoustic data was acquired using the program Audacity at the standard sampling frequency 44100Hz, producing signals showing sound pressure level (SPL) over time. Background noise was removed by subtracting the noise profile from the signal. Different noise reduction levels were trialled on an example signal (see below) and a 48dB reduction proved most effective.

A Fast Fourier Transform (FFT) was computed on the signal to translate it to the frequency space. A Hanning window function was used for the spectrum and different sizes of FFT (number of frequency bins) were trialled to see which gave the most suitable frequency resolution (see first diagram below).

Read Preliminary Acoustic Measurements – Pringles in full