In this chapter, the authors describe fluid dynamics considerations regarding odor dispersal in real environments and their relationship with realistic odor presentation using an olfactory display. The authors propose the use of a Computational Fluid Dynamics (CFD) simulation in conjunction with the olfactory display. A CFD solver is employed to calculate the turbulent airflow field in a given environment and the dispersal of odor molecules from their source. The simulation result is used to reproduce realistic changes in the odor concentration with time and space at the nose. The results of sensory tests are presented as a demonstration of CFD-based odor presentation. The effect of body heat on odor dispersal in indoor environments and how it affects odor perception is also discussed.
TopIntroduction
An olfactory display is a device that generates odors in the form of a gaseous vapor of odorous chemical molecules and presents them to the user. It can be used to add special effects to computer games, for example, by creating a game with a “Press to Smell” button. Presenting a smell in response to a click on the button will provide the user of the game with a completely new experience that no traditional game machine can provide (Nakamoto, et al., 2008). The olfactory display should at least be able to generate a single type of odor associated with the button. However, most game machines can generate different sounds and images; therefore, it is natural to expect them to have multiple “Press to Smell” buttons. Thus, most research on olfactory displays so far has thus been devoted to the development of hardware devices capable of multiple odor generation and rapid on-off switching.
Practically any color can be generated by mixing three primary colors, i.e., red, green, and blue, because our eyes have only three types of color receptors. However, considering the complexity and diversity of olfactory receptor cells in mammalian noses (Buck & Axel, 1991), it is unlikely that an arbitrary smell can be generated simply by mixing a small set of primary odors. Therefore, in a typical olfactory display system, one ready-made mixture of odorous chemicals in liquid form is prepared for the generation of each specific smell to be presented to the user. The olfactory display system developed by Nakamoto and Yoshikawa (2006), for example, can hold up to eight bottles of solutions containing different mixtures of odorants. An odor vapor in the headspace of a selected bottle is delivered to the user’s nose through tubing. The odor can be switched from one to another using computer-controlled solenoid valves. The intensity of the odor presented to the user can also be changed using the solenoid valves to dilute the odor vapor with clean air. An olfactory display system developed by Sato et al. (2008) uses ink jet devices to attain more precise control of odor concentrations and rapid switching between odors.
These sophisticated olfactory displays also allow attempts at realistic odor presentation. The concentrations of odorants reaching the nose in real environments vary with time and position. Therefore, it is expected that olfactory displays can provide users with a realistic experience if they reproduce the change in the odor concentration reaching the nose. For example, if the odors are intended to enhance the reality of a specific scene in a movie, their concentrations should be carefully adjusted. The appropriate release rate of an odorant for a faint scent of flowers drifting in the air can be extremely different from that for a strong unpleasant odor sensed by the nose. If the scene is changed from a distant flower garden to a close-up view of roses, the release rate of the odorant should be increased accordingly.
To reproduce the change in the perceived odor intensity while walking through an environment, Yamada et al. (2006) assumed a mathematical model for odor distribution in a virtual world. They developed a wearable olfactory display system. In their experiments, each subject was asked to carry the wearable olfactory display and move around in the given space to find the position of a virtual odor source. The concentration of the odor released from the wearable olfactory display at the subject’s nose was adjusted assuming isotropic diffusion of a gaseous chemical substance into the air. However, this assumption holds only in environments with extremely weak airflow (Murlis, et al., 1992). Moreover, the odor distribution in a given environment is affected by the presence of obstacles. Consider the scenes shown in Figure 1, for example. In Figure 1(A), a woman is reading a menu in an Indian curry restaurant. When the curry is ready to be served from the kitchen, as shown in Figure 1(B), the woman sitting at the table would perceive a faint smell of curry. When she walks toward the kitchen, as shown in Figure 1(C), the smell of curry would gradually become stronger. In Figure 1(D), the woman is looking into the kitchen over the counter. She would then sense a much stronger smell because most of the odorants evaporating from the curry pot are confined to the kitchen. Only a small portion of the odorants emerges from the opening above the counter and spreads throughout the restaurant.