Abstract
Mammals have presumably evolved to adapt to a diverse range of ambient environmental conditions through the optimized heat and mass exchange. One of the crucial biological structures for survivability is the nose, which efficiently transports and thermally preconditions the external air before reaching the internal body. Nasal mucosa and cavity help warm and humidify the inhaled air quickly. Despite its crucial role, the morphological features of mammal noses and their effect in modulating the momentum of the inhaled air, heat transfer dynamics, and particulate trapping remain poorly understood. Tortuosity of the nasal cavity in high-olfactory mammalian species, such as pigs and opossum, facilitates the formation of complex airflow patterns inside the nasal cavity, which leads to the screening of particulates from the inhaled air. We explored basic nasal features in anatomically realistic nasal pathways, including tortuosity, radius of curvature, and gap thickness; they show strong power-law correlations with body weight. Complementary inspection of tortuosity with idealized conduits reveals that this quantity is central in particle capture efficiency. Mechanistic insights into such nuances can serve as a tipping point to transforming nature-based designs into practical applications. In-depth characterization of the fluid–particle interactions in nasal cavities is necessary to uncover nose mechanistic functionalities. It is instrumental in developing new devices and filters in a number of engineering processes.
Introduction
Animal evolution has allowed efficient adaptation to highly fluctuating and extreme external conditions. A central component that modulates the evolutionary process is the need for continuous heat and mass exchange with the environment (Schroter and Watkins 1989). The nose is a central player in the body and environment interaction; it efficiently transports and thermally preconditions the external air before reaching the internal organs. The inhaled air is rapidly warmed and humidified in the respiratory track. Strikingly enough, evidence shows that the air is heated to 35°C from 23°C outside, besides attaining 98% relative humidity, before reaching the nasopharynx (Basu 2021), the airway region situated at the back of the main nasal passage and superior to the larynx, which is crucial for maintaining and protecting the vulnerable internal tissues of the lungs (Wolf et al. 2004; Elad et al. 2006; Doorly et al. 2008). A remarkable example is the southernmost mammal, the Weddell seal, which breathes air between −3°C and −26°C and warms it using complex flow mixing induced by the labyrinthine structures lining the nasal cavity (Boyd 1975; Mellish et al. 2014).
The respiratory region in the nasal cavity contributes to the body’s first line of defense against diseases by filtering external particles and airborne pathogens. The mucus is secreted on the respiratory epithelium, and the mucous membrane traps particles (Sobiesk and Munakomi 2019). Immunoglobulin A and lysozyme are present in the mucous membrane to aid in the immunological defense (Mellert et al. 1992; Kurono et al. 1999). Winding nasal passages helps to improve the particle capturing by retention of the particles in the passage due to the secondary flows. Also, penetrating further into the main nasal cavity, we note that part of the olfactory epithelium would extend along the very upper recesses of the dorsal passage (Bamford 2012; Kuper et al. 2012; Asgharian et al. 2016;Jenkins et al. 2018; Amini et al. 2019). Overall, key roles in the respiratory region are warming, humidifying the air, and filtering out external particles, but also a minor role in olfaction.
Despite its crucial role in the interaction between body and the external environment, the underlying thermal and momentum transport controlled by an animal nose is far from being well understood. Here, we explore two critical physical features that may play a key role in thermal conditioning and inhaled particle trapping. The first feature is the volume variability of the nasal cavity along the main path of the inhaled flow; its cross-sectional area increases monotonically up to the middle of the nasal airway and then decreases as it approaches the nasopharynx (Goldberg et al. 1981; Wolf et al. 2004; Hörschler et al. 2006 ). This feature allows an increase of the air-residence time without varying the inlet and outlet cross-sections, and promotes vortical flow patterns that enhance air mixing in the nasal cavity (Jackson and Schmidt-Nielsen 1964; Hahn et al. 1993). The second feature constitutes the multiple intruded turbinates (Schroter and Watkins 1989; Van Valkenburgh et al. 1999; Pless et al. 2004; Yuk et al. 2022 ). These structures, lining the intranasal tissues, are elongated in the streamwise direction. The protrusion length and geometry can significantly alter the flow as it bifurcates the incoming air into several compartments. This promotes the formation of vortices and the instability of boundary layers. Additionally, the turbinates increase the contact area between a heat source, namely the nose tissue, to enhance heat exchange.
The nasal cavities of animals exhibit distinct differences among species; they form different shaped turbinates and varying nasal gaps. Turbinates can be broadly categorized based on their number, e.g., 2 for humans and about 20 for elephant seals, gap width, ∼0.35 mm for humans, curvature, and length. Negus and Straatsma (1960) proposed a classification to categorize the nasal cavity based on its properties, including single scrolling and double scrolling, folding, and branching for turbinates. To examine the nasal cavity of animals, the complexity of the nose structure has been generally measured by changes in the cross-sectional area and perimeter of the coronal plane (Schreider and Raabe 1981; Gross et al. 1982; Patra et al. 1987). The nasal structure photographed using magnetic resonance imaging (MRI; Yeh et al. 1997; Ranslow et al. 2014) and microcomputed tomography (micro-CT; Wagner and Ruf 2019; Smith et al. 2021) has allowed for high-resolution examination of the turbinates. Related studies have uncovered the rich functionality of the nose by segmenting and staining nose tissue in a histological manner (Adams 1972; Larochelle and Baron 1989; Yee et al. 2016). That allowed characterizing the nose structure and the functional distinction between breathing and olfaction. Numerical simulations have explored airways in the nasal cavities, which are difficult to measure (Subramaniam et al. 1998; Craven et al. 2009; Smith et al. 2019). Known to have a high sense of smell, dog and mouse noses have been of particular interest (Mery et al. 1994; Barrios et al. 2014 ), but those of cats, donkeys, loris, and marmoset, among others, have also been characterized (El-Gendy and Alsafy 2010; Van Valkenburgh et al. 2014; Smith et al.2014, 2019; Pang et al. 2016).
In general, animals with high olfactory capabilities have more labyrinth-like air pathways. Characterization of the 3D shapes is needed to determine commonality and variation in the structural features of turbinates across different species (Rohlf and Slice 1990; Klingenberg 2010). Here, we present the physical structure of the nasal cavity, especially the respiratory region of animals and their external particle-capturing mechanism, primarily focusing on a pig nose as a basic model system. Detailed materials and methods are provided in the “Methods” section, results of characterizing nasal structure features are provided in the “Results” section, and a discussion is given in the “Discussion” section.
Methods
Turbinates induce a series of flow patterns that contribute to the performance of various functions, including odorant and particle capture and thermal conditioning. We have elaborated on the morphological details of a representative pig nasal pathway through analysis of CT imaging data. The various methods used are indicated as follows.
CT scanning
We used a Toshiba Aquilion 16-slice CT scanner with a 0.5-mm-slice-thickness resolution and 512 × 512 pixels in-plane resolution for a detailed view of the scanned nasal cavity of an averaged adult pig, as illustrated in Fig. 1(A). The pig’s head was placed on a platform with the nose facing down. Then, CT scan was done normal to the platform, which produced cross-sectional images perpendicular to the skin between the nostril and forehead. These CT-scan data provide a valuable basis for characterizing nasal airway structures and generating realistic computational models. Figure 1(B) shows a series of CT images of the pig nasal cavity mapped on the frontal plane at 2.4, 3.6, 7.3, 8.5, 10.9, and 14.6 cm from the nostril. Using a customized image processing and a photo editor, we have set the black region as representing an empty air passage and the white region representing the mixed bone, muscle, tissue, and blood inside the nose, as described in Fig. 1(C).
(A) A 3D pig model reconstructed from CT scan images of the front, top, and side views from left to right. The dark gray structure in the 3D model shows the nasal conduit. (B) Cross-sectional images of the pig’s nasal cavity. Black regions denote cavities. Airway cross-section images from close-to-the-nostril to the nasopharynx are shown from left to right. (C) Images extracted from only the cavity part in panel (B). (D) Process for specifying the nose structure. Gap thickness is measured from black and white images extracted from CT scan image. Then, the color is inverted and converted to an 8-bit grayscale image, which is skeletonized, allowing determining distance, LC, and displacement length, LS, of each branch and curvature.
Gap thickness measurements
We explored the features of different animal species to put the analysis of the pig nose in perspective to comparable airway geometries. Images from different species were collected from various sources; in particular, CT scan images, magnetic resonance images, and anatomical images were obtained from selected resources (Craven et al. 2009; Casteleyn et al. 2010; Macrini 2012; Rodgers 2012). CT data for Caluromys philander (Bare-tailed woolly opossum), Dasyurus hallucatus (Northern quoll), Potorous tridactylus (Long-nosed potoroo), and Petauroides volans (Greater glider) were obtained from Dr. Macrini (2012). A female Labrador retriever’s high-resolution MRI was acquired from Craven et al. (2009), a male human MRI image was acquired from Shi et al. (2006), a Rabbit’s histological section image in dorsal view was obtained from Casteleyn et al. (2010), and Cavia porcellus (Guinea pig)’s CT scan images aquired from Rodgers (2012) and Shi et al. (2006). These cross-sectional images are located in the middle of the entire nasal cavity of each animal and contain the maxilloturbinal structure. However, depending on the animal’s nasal characteristics, a portion of the olfactory region of the nasal cavity could be included in the nasal images.
The analysis of the noses first considers the size of the passage in nasal cavity. Here, we define the gap thickness as the width of the air passageway, not its length. The images were treated using ImageJTM and calibrated with a line tool. The thickness was measured manually by the side length of the air passage, that is, the width of the channel. The gap thickness was measured at a minimum of 15 different locations, including the thickest and thinnest parts of the air passage.
Tortuosity measurements
Skeletonized images were used to determine tortuosity. The skeletonization process involves converting images to 8-bit grayscale images, which are processed with the ImageJ software [see Fig. 1(D) to determine the shortest distance, LS, and the actual distance, LC]. In this case, the length measurements were estimated for all segments with two end-points, one junction and one end-point, or more than one junction. Using this information, tortuosity was calculated by dividing the actual distance (LC) by the shortest distance (LS), accordingly the tortuosity, τ = LC/LS.
Curvature measurements
Skeletonized images were used to determine the curvature, κ. We manually drew the lines along every branch and approximated curves using various points. This was used to estimate the average value and standard deviation of κ for each line.
Numerical simulations of transport in animal airway and bio-inspired cavities
We performed two series of numerical simulations on a finite volume solver; one focused on replicating realistic inhaled airflux and the associated heat transfer trends across the intra-airway tissues in a realistic CT-derived pig nasal cavity (see Fig. 2), and the other explored basic particle management phenomena in idealized conduits with different tortuosities, τ = 1.19 and 1.90 (see Fig. 5A). The τ = 1.90 case resembles typical tortuosities observable in the pig airway (see Fig. 3B). In contrast, the conduit with simpler geometric complexity, τ = 1.19, allowed examining the impact of tortuosity in particle management and capturing a phenomenon that has been postulated to elevate particle capturing along the upper respiratory airway, which also includes the olfactory epithelial region, in different mammalian species.
(A) CT-derived digital reconstruction of the nasal airway of a pig. The cross-sectional view depicts the turbinate structures on a representative section. (B) Reconstruction of a pig airway along the outline of the pig body. The segment bounded by the solid dark lines (and mapped via a dashed line) corresponds to the position of the panel (A) reconstruction. (C) and (D) depict two separate views of the segmented model.
Basic relationship between animal weight and (A) gap thickness, (B) tortuosity, and (C) radius of curvature. Dashed lines represent the trends for the animals, not including humans. (D) Radius of curvature versus gap thickness. The abbreviations denote opossum (OP), northern quoll (NQ), guinea pig (GP), great glider (GG), potoroo (PT), rabbit (RB), dog (DG), pig (PG), and human (HM).
(A) Meshed model of the left nasal airway from the pig reconstruction in Fig. 2. (B) A zoomed-in visual of the exterior mesh profile. (C) Cross-section, along with zoomed-in visuals in (D) for the tetrahedral elements and the four prism layers extruded at the airway-tissue boundaries. (E) Heat flux map on the tissue surface for a simulated inhaled airflow of 40 L/min. The simulated data for heat transfer was postprocessed and visualized on FieldViewTM, under license provided by Intelligent Light (Rutherford, NJ, USA) through its University Partners Program.
Idealized conduits with tortuosity τ = (A) 1.19 and (B) 1.90. (C) Particle capturing trends for airflow rates of 15 and 30 L/min through each pathway, averaged on an array of 312 inlet entries (Yuk et al. 2022). The red and black curves represent the 1.19 and 1.90 tortuosity cases. The trend lines during flow rates of 15 and 30 L/min are shown by solid and dashed lines.
Logistics for inhaled airflow and heat transfer simulation in a truncated pig upper airway
Large eddy simulations (LES) were performed in the digitized pig model (Fig. 2). The geometry was rebuilt from scans imaged at 0.5-mm-depth increments. The geometry contains 1.1 million unstructured, graded tetrahedral elements and four prism layers of 0.1-mm-thickness extruded at the air–tissue boundary, with a height ratio of 1. The grid size conforms to our prior assessments (Basu et al. 2017, 2018) on the sensitivity that simulated data might bear to mesh refinement, especially in anatomically realistic and topologically comparable nasal geometries. The simulations used a pressure-based transient solver with the Kinetic Energy Transport Model (Baghernezhad and Abouali 2010; Ghahramani et al. 2017) as the subgrid-scale model; it was implemented for a total 1250 time-steps, with time-steps of 4 × 10−4 s. As a biophysical variable, heat flux correlates with the subjective sensation of nasal patency. The test simulations in the pig model imposed a thermal boundary condition of 20°C at the inlet and 32.6°C along the internal tissue walls (Lindemann et al. 2002; Zhao et al. 2011), see Fig. 4. In addition, note that the simulated airflow rate was 40.0 L/min, under an imposed inlet-to-outlet pressure gradient of 2.90 Pa in the truncated airway model (Fig. 2). Therein, the numeric scheme considered the inlet entry near the nostrils to be a pressure inlet, the posterior outlet was treated as a pressure outlet, and the enclosing intra-nasal tissues are considered as rigid walls with no-slip boundary condition. The characteristic values for the ambient airflow included a density of 1.204 kg/m3, dynamic viscosity of 1.825 × 10−5 kg/m/s, thermal conductivity of 0.0268 W/m/K, and specific heat of 1005.9 J/kg/K (Kimbell et al. 2013).
Logistics for tracking particle capture trends in bio-inspired channels
The simplified, bio-inspired tortuous channel geometries, with τ = 1.19 (Fig. 5A) and τ = 1.90 (Fig. 5B), contain 1.87 and 1.40 million unstructured, graded tetrahedral elements, respectively, with four prism layers of 0.1 mm thickness at the airway walls with a height ratio of 1. The computational scheme operated on a segregated solver with SIMPLEC pressure–velocity coupling and second-order upwind spatial discretization. Pressure-driven flows of 15 and 30 L/min were examined considering a section containing 312 channels defining an equivalent filter (Yuk et al. 2022). Each computation ran for 6–7 h to complete a 0.35 s interval of simulated flow interval, with a time-step of 2 × 10−4 s. We consider an air density of 1.204 kg/m3 and the dynamic viscosity of air was set at 1.825 × 10−5 kg/m/s. The simulations enforced the following boundary conditions: (1) zero velocity at the airway tissue interface, along with “trap” boundary condition for droplets, whereby the tracking of a droplet’s transport would cease once it reaches the mesh layer lining the cavity walls; (2) zero pressure at inlet, which were the pressure-inlet zones in the simulations, with “reflect” boundary condition; and (3) a negative pressure at the outlet plane, which was the pressure-outlet zone, with “escape” boundary condition for the tracked particles, i.e., allowing for the outgoing particle to leave the test domain. To drive the 15 L/min air flux, the average inlet-to-outlet pressure gradient was −11.35 Pa, and at 30 L/min, the value was −35.12 Pa. The particle trajectories were tracked by Lagrangian-based discrete phase inert particle transport simulations against the ambient airflow, and additionally the numerical scheme considered the effects of gravity and other body forces such as the Safman lift force exerted by a flow-shear field on small particulates moving transverse to the streamwise direction. In addition, a droplet size range with diameters 1–14 μm was considered large enough to neglect the effects of Brownian motion. For the numerical tracking, the initial mass flow rate of the inert droplets moving normal to the inlet planes into the nasal airspace was required to be non-zero, and was set at 10−20 kg/s. The total number of monodispersed particles tracked for each tested size was 1338 in the τ = 1.19 conduit and 1076 in the τ = 1.90 conduit. Finally, the particle density was assumed to be 1.3 g/mL, commensurate with the physical properties for environmentally dehydrated, often pathogenic, particulates (Stadnytskyi et al. 2020; Basu 2021).
Results
On the animal nasal structures
We first characterized the reference pig data and the ones from literature (Boyd 1975; Van Valkenburgh et al. 2011; Mellish et al. 2014), which illustrate cross-sectional anatomical structures. The air passage in the selected cross-sectional views was evaluated based on the nasal structure’s width, tortuosity, and curvature. The gray translucent internal part of the 3D views of a pig head, illustrated Fig. 1(A), highlights the nasal structure.
Figure 1(B) shows geometric changes along the nasal cavity from the nostril to the nasopharynx. Figure 1(C) illustrates these images in black and white binary coloring to explore the nasal cavity in detail. Here, the black region represents the nasal cavity. The first panel shows two nostrils and the second panel evidences slightly wider nostrils. Up to the second panel, both the right and left nostrils consist of air passages with a simple shape. However, a junction that divides an air passage into two or more paths is formed starting in the third image. From the third panel, the air passage is twisted in a spiral-like shape; this feature gradually enlarges and curls deeper in the nose, as indicated in the fourth, fifth, and sixth panels. As a consequence, the area of the nasal cavity increases. Figure 1(D) illustrates an example of the skeletonization steps that we used for the analysis. The binary images prepared for skeletonization allowed measuring the gap thickness. The channel passage size of the BW image in Fig. 1(D), D, can also be obtained with the cross-sectional area, Ac, and the perimeter, P, as D = 4Ac/P (Craven et al. 2007), which gives a value of 6.2 mm. The variation of this quantity within the cross section is worth noting, which has a 3.6-mm average value.
The curvature and tortuosity were characterized based on the shape of the segments. As shown in Fig. 1(D), segment 1 is a thicker air pathway connected to the nostril, and connects to segments 2 and 3. Likewise, segment 2 connects to segments 4 and 5. Note that segments 3 and 4 have a spiral shape, whose tortuosity is 4.4 and 2.2, respectively. However, segments 1, 2, and 5 are weakly curved. As a result, the associated tortuosity resulted in 1.1, 1.2, and 1.1. The radius of curvature in segments 3 and 4 is approximately 3.9 and 4.5 mm; it was 9.7, 6.4, and 14.2 mm for segments 1, 2, and 5. These quantities, namely curvature and tortuosity, significantly varied along with the nose.
This process is repeated for the other animals; then, the tortuosity, gap thickness, and radius of curvature are linked to the animals’ typical body weights to explore basic allometric relations. The body weight, W, was either the average value of the same species (Atramentowicz 1995; Wallis et al. 1997;Oakwood 1997; Mitchell et al. 2010; Jensen et al. 2015; Daily Pork Reports: https://www.ams.usda.gov/market-news/daily-pork-reports; Fryar et al. 2021) or reported values ( Craven et al. 2009; Casteleyn et al. 2010). The various relationships are shown in Fig. 3(A–C); they exhibit subquadratic power-law trends of the form ψ∝CWα, where ψ represents the tortuosity, gap thickness, and radius of curvature, power α is illustrated in Fig. 3, and C is a characteristic constant. Relatively high R2 > 0.8 of the radius of curvature and gap thickness evidence a clear dependence with bodyweight. In contrast, an R2 ≈ 0.5 between tortuosity and W indicates a weaker linkage. Remarkably, gap thickness and radius of curvature exhibit a very strong relation of R2 ≈ 0.99 with a quasi-linear relation with α ≈ 1.1.
The relations associated with a representative human nasal cavity are included to aid insight. Interestingly, the human nasal cavity does not follow the trends of animals. This may indicate distinct stressors modulating the evolution of functionality needs indicated here.
On the heat management and particle screening
In support of the ansatz about the importance of anatomic contortions in air warm-up, the total surface heat flux at the entry, marked by the left-anterior gray segment in Fig. 2C and D is 7.59 watt/m2, which decayed to 0.034 watt/m2 along the posterior airway walls (see Fig. 4E). The conditioning of the air, as it moves into the posterior parts of the animal’s nasal passage, is also quantifiable from the surface heat transfer coefficients, which are reduced from 0.431 watt/m2/K at the entry to 0.002 watt/m2 averaged over the rest of the airway. The preliminary estimates on heat transfer are in agreement with recent examples of multispecies investigations in literature (Ito et al. 2017).
The numerical inspection of particle screening trends in the idealized, bio-inspired channel geometries with τ = 1.19 and 1.90 shows lower capturing efficiency, as hypothesized (see Fig. 5). The capturing efficiency for the two geometries is the same for comparatively small particles, namely, viz. ≲3 μm for 30 L/min ambient air flux through the entire filtration unit, the same threshold being ≲4 μm for the 15 L/min flux. We note a 100% capturing efficiency for particle sizes ≳6 μm during the 30 L/min flux and ≳14 μm during the 15 L/min, for the τ = 1.19 channel. The same thresholds for the τ = 1.90 channel are ≳4 and ≳7 μm, respectively.
Discussion
The curvature and tortuosity of the nasal conduit may play a central role in the flow mixing, particularly at comparatively high breathing rates. The characterization of gap thickness, radius of curvature, and tortuosity through cross-sectional images of the various animals’ nasal cavities provides a primary metric for quantifying the geometry and the linkage to the effects on the flow topologically. Indeed, follow-up work is needed to incorporate three-dimensionality and establish a factor relating planar characterization with spatial geometrical features.
Regarding the geometric idealization and truncation of the anatomic airspaces, it should be pointed out that the CT-derived geometric model for the pig airway (e.g., see Figs. 2 and 4) is a truncated representation of its upper respiratory airspace. The demarcated cavity, however, does include part of the olfactory epithelial region (Amini et al. 2019) to ensure the relevance of the isolated anatomic geometry in context to olfactory mechanisms and the transport features involved therein. Nonetheless, we note that the olfactory bulb is located further back in the nose (Kuper et al. 2012), well past the truncated airway domain identified in panels (C) and (D) of Fig. 2. The pig upper respiratory space in our work comprises two main flow pathways in the main nasal cavity, namely the dorsal and ventral pathways. Part of the olfactory epithelium extends along the very upper recesses of the dorsal passage (Kuper et al. 2012; Asgharian et al. 2016; Amini et al. 2019). In addition, the reader should also note that excluding the posterior parts of the upper airway (going into the pharyngeal tract) in our test geometry does not bear any appreciable effect on the upwind flow physics tracked along the anterior airspace. Note that the streamwise length scale in a typical animal upper airway is orders of magnitude greater than the cross-sectional dimensions transverse to the mean incoming airflux. This unique geometric feature prompts the inhaled airflow to attain a full development profile within a short distance, and as such, it is reasonable to consider the modeling to be agnostic to the flow features outside of the nasal passage [see our earlier work (Yuk et al. 2022)] while estimating the regional deposition of inhaled particles in the cavity. However, in this context, we also direct the reader to an alternate approach, which albeit would warrant a longer computing time over a wider numerical domain. The process includes a reconstruction of the external nose and then would apply the far-field boundary conditions on, say, an external spatial box and subsequently allow the air to flow naturally into the intra-airway space through the nostril openings (Craven et al. 2009; Taylor et al. 2010). It is worth highlighting that the boundary condition at the end of the region explored considers negligible lateral pressure gradient. It is undoubtedly an approximation since topographical features at the region downwind of the explored domain may trigger a transverse pressure gradient. As a result, local flow distribution, particularly in the vicinity of the downstream end, can be affected.
Although it is challenging to measure airflux and pressure in a live pig airway, the present study aims to assess expected mean conditions in terms of typical thermal properties in the inhaled air and along the intra-airway tissue surfaces (Lindemann et al. 2002,, Wolf et al. 2004; Elad et al. 2006) as well as the breathing rates expected in an adult pig, based on studies using allometric scaling of mammalian breathing (Bide et al. 2000; Glazier 2005). It is worth noting that the assumption of a uniform and constant temperature along the airway wall is an approximation. The localized airway wall temperatures may indeed vary along the flow path through the nose. Also, evaporative cooling may play a significant role in nasal heat exchange. We expect to quantify and incorporate these effects in the modeling framework in our future work.
In the numerical tests, the length scale of the tracked particles is on the order of micrometers (Fig. 5C), which is typical (Zhi et al. 2021). However, odorants that eventually stimulate the perception of smell are at molecular scale (Kermen et al. 2011); our modeling approach considers that tiny particles enter the animal nose in the larger droplets and aerosols. Our focus was on exploring the trapping characteristics, such as a pig’s upper respiratory pathway, owing to the complex airflow features affected by the incoming streamlines inside the topologically evolved and highly tortuous conduit.
Finally, it is worth noting that despite the flow being rather laminar, the variability of the wall topography and the tortuous pathways may induce local flow instability and potentially local recirculation and vortical motions that are not captured when considering laminar flow in the simulations. LES allows for capturing those phenomena and a laminar solution when appropriate and ensures assessment of small-scale transport and the associated effect on local particle dynamics.
Conclusions
Species have evolved to optimize a number of critical functions, which led to improved, more efficient performance that has increased survival chances. As central components bridging the inner body with environmental conditions, nasal cavities may have undergone a series of adaptations to address several critical needs associated with heat transfer conditioning, maximization of particle mixing, and minimization of pressure loss and bulk dimensions of the structure. The weighting and balance between these needs are directly liked to the local environmental conditions. Indeed, the nose of animals in cold regions like arctic seals may emphasize heat transfer and biomass at the expense of increased pressure loss. A fundamental phenomenon yet to be uncovered is a quantitative assessment of topological features of turbinates and linkage with functioning. Our systematic characterization of turbinate morphology of various species and relation to pressure loss, heat transfer rate, and flow pattern allow the development of functional relationships that inform a new generation of air filtering, conditioning, and management in disparate environments. Our basic exploration of selected turbinates and idealized conduits illustrates the importance of functional trade-offs between the contributing factors.
The heat transfer simulations revealed the role of the pig airway complexity in a rapid warming up of the inhaled air. Comparative inspection of idealized geometries with reduced tortuousity of 1.19 and 1.90 at two flow rates illustrates distinct effects. The particle-capturing efficiency is a primary factor determining whether or not the channels can function effectively as filtration unit pathways. The results show that for both ambient flow rates, the particle capture rate is consistently higher for the more complex τ = 1.90 pathway. Also, a higher ambient airflow significantly elevates the particle-capturing efficiency by imparting a sweeping effect on the particle trajectories toward the channel walls.
ACKNOWLEDGEMENTS
The authors thank Dr Jeffry D. Schroeter, at Applied Research Associates, Inc., Raleigh, NC, USA, for the insight into the regional epithelial characterization of the pig nasal cavity.
Funding
This work was supported by the National Science Foundation [grant numbers CBET-2028075 to S.J., CBET-2028090 to L.P.C., and CBET-2028069 to S.B.].
Conflict of interest
Authors declare no competing interests.
Data availability
All datasets are available in the manuscript.
