| Summary: | 碩士 === 國立清華大學 === 工業工程與工程管理學系 === 104 === This research is aimed to provide a method for designing a half-facial respiratory mask based on real 3-D craniofacial data. In addition to the filtration capability of the filtration material, two of the most important factors that affect the performance of a respiratory mask are the airtightness and dead space. The airtightness is the degree of seal between the mask and the face of the mask wearer. The dead space is the space enclosed between the mask and the face, which could detains the tidal volume of the wearer and lower the effective permeability of the filtration material. To enhance the airtightness and lower the dead space, this research proposes a method for designing a half-facial respiratory mask based on 3-D craniofacial data. Furthermore, it is inclined to design a respiratory mask with a limited number of sizes (e.g. 3 to 5 sizes) for all of the users (e.g. 8 million labors). This research proposes a method for sorting the sizes of face based on the size and depth of the face.
Based on this airtight design concept proposed by 楊宜學, this research further proposes an improved two-step method for sorting the facial sizes into groups. This method first sort the parent group into groups based on the facial sizes (e.g. group M), and uses the external contour line of the mask and the real 2-D curve (which is different from two 1-D sizes such as the facial length size and the facial width size) as the basis for sorting. Next, the 2D external contour line is projected to the 3D face and thus derives the fitting interface curve between the 3D mask and the face. Afterwards, the degree of depth distribution of this 3D fitting interface curve is used as the basis for sorting the facial depth. The facial size groups are further sorted into N groups. This two-step sorting method is used for designing a set of airtight and low dead space respiratory masks by sorting manner according to a limited number of facial sizes, so as to be applicable to all of the wearers.
The results reveal that the external contour line of the 2-D mask based on the facial size is distributed between 90.04mm to 120.08mm in face length, the full pitch is 30.04mm, and the allowance value of the facial size is 20mm (±10mm). Therefore, the facial sizes are sorted to two groups, that is, size S group and size L group. Next, these two groups are sorted based on the facial depth according to the 3D fitting interface curve. It can be known from the spread curve of the size S group has a depth distributed between 4.6mm and 23.59mmm, the full pitch is 18.99mm, and the allowance value of the facial depth is about 14mm (±7mm). So the facial depth Size S group is further sorted into two groups. ”shallow size”:4.6mm~14.095mm and”deep size”:14.095mm~23.59mm. Likewise, the spread curve of size L group has a depth ranged between 4.6mm and 23.59mm, the full pitch is 18.99mm, and the allowance value of the facial depth is 14mm (±7mm). So the facial depth Size L group is further sorted into two groups. ”shallow size”:4.6mm~14.095mm and”deep size”:14.095mm~23.59mm. Next, we find the person whose facial size is proximate to the middle value as a standard person for all of the sizes. Based on the craniofacial data of this standard person and the airtight design concept of 楊宜學, we capture the 3D fitting curve of the standard person and construct the mask to meet the requirements of airtightness and low dead space. In accordance with the computer simulation, the requirements on airtightness are met for four size groups. The maximum dead space volume according to the computer simulation is 192ml, and the minimum dead space volume of the marketable products measured by real water syringe is 237ml which means that the dead space volume is lowered by 45ml. Certainly, this result must be verified with real products.
Methodologically, the design of the respiratory mask of this research is actually a design method for a real 3D mask. This design is different from the conventional mask which is achieved with reference to simplified 1D size such as the 1D facial length size and the 1D facial width size, such that the conventional mask lost the complex 3D information. The design method of this research is achieved by directly capturing 3D fitting curve from the curved surface in the 3D craniofacial data, while the 3D information is substantially reserved. This method is applicable to other designs, for example, helmets, spectacles, and shoes.
|
|---|
