Gabryel F. Soares 1,†, Gilberto Fernandes Jr. 2,† , Otacílio M. Almeida 3,†, Gildario D. Lima 4,† and Joel J. P. C. Rodrigues 5,† *

1-Federal University of Piauí (UFPI), Brazil; Este endereço de email está sendo protegido de spambots. Você precisa do JavaScript ativado para vê-lo.

2-State University of Londrina (UEL), Brazil; Este endereço de email está sendo protegido de spambots. Você precisa do JavaScript ativado para vê-lo.

Abstract: Respiratory diseases are among the leading causes of death globally, with the COVID-19 1 pandemic serving as a prominent example. Issues such as infections affect a large population and, 2 depending on the mode of transmission, can rapidly spread worldwide, impacting thousands of 3 individuals. These diseases manifest in mild and severe forms, with severely affected patients requir- 4 ing ventilatory support. The air-oxygen blender is a critical component of mechanical ventilators, 5 responsible for mixing air and oxygen in precise proportions to ensure a constant supply. The most 6 commonly used version of this equipment is the analog model, which faces several challenges. These 7 include a lack of precision in adjustments and the inspiratory fraction of oxygen, as well as gas 8 wastage from cylinders as pressure decreases. The research proposes a blender model utilizing only 9 dynamic pressure sensors to calculate oxygen saturation, based on Bernoulli’s equation. The model 10 underwent validation through simulation, revealing a linear relationship between pressures and 11 oxygen saturation up to a mixture outlet pressure of 500 cmH2O. Beyond this value, the relationship 12 begins to exhibit non-linearities. However, these non-linearities can be mitigated through a calibration 13 algorithm that adjusts the mathematical model. The research represents a relevant advancement in 14 the field, addressing the scarcity of work focused on this essential equipment crucial for saving lives. 15

Keywords: Blender, FIO2, Mechanical ventilator, Simulation 16