مطالعه تجربي و عددي اختلاط آب يونيزه شده و پلاسماي خون در يك ميكروميكسر غيرفعال و بهينه¬سازي با استفاده از شبكه عصبي

Microfluidic devices , passive micromixers , ionized water , plasma , mixing efficiency , pressure dro‎p , optimization

Experimental an‎d numerical study of mixing of ionized water an‎d blood plasma in a passive micromixer an‎d optimization using neural networks

مهندسي مكانيك

Microfluidic devices are essential due to their ability to precisely control small volumes of liquids across various scientific an‎d industrial fields. These devices can be categorized into two main types: active an‎d passive. This thesis explores the mixing process of two liquids with different thermophysical properties—ionized water an‎d blood plasma—using two proposed passive micromixers that utilize fluid re-entry. The first micromixer features a novel T-Y shape aimed at enhancing mixing quality while maintaining a low pressure dro‎p. To validate the numerical simulation results, experiments were conducted an‎d five machine learning algorithms were employed to optimize micromixer performance. The study examined several factors, including the distance between inlets, the angle of the Y-shaped inlet, the velocity ratio of the two inlets, an‎d the aspect ratio. The findings revealed a significant increase in mixing efficiency with the addition of a Y-shaped inlet compared to a T-shaped micromixer. As the distance between the two inlets decreased, mixing efficiency improved, with optimal values identified for the Y-shaped inlet angle, inlet velocity ratio, an‎d aspect ratio. The optimization study indicated that the MPR algorithm was the most suitable for predicting the performance of the proposed micromixer. Since practical application of this micromixer requires two syringe pumps, a recirculating micromixer was proposed for mixing ionized water an‎d blood plasma, capable of functioning at high Reynolds numbers. The study also investigated how geometric parameters an‎d the number of mixing chambers influenced mixing efficiency an‎d pressure dro‎p. Results showed that channel depth an‎d inlet flow rate could create recirculating flows within feedback channels, generating vortices in mixing chambers that lead to chaotic flow formation. Moreover, mixing efficiency improved with increasing channel depth (aspect ratio) an‎d number of mixing chambers. Simulations demonstrated that channel depth significantly impacts both mixing efficiency an‎d pressure dro‎p. For instance, at a Reynolds number of 400 (based on ionized water properties), increasing channel depth from 50 to 1000 microns raised mixing efficiency from 23% to 73.255%, while pressure dro‎p decreased from 171.75 kPa to 85.12 kPa. At Reynolds numbers close to 50, the proposed micromixer exhibited a markedly higher figure of metric compared to most ordered an‎d disordered flow pattern micromixers. These two micromixers, characterized by short response times an‎d high practical functionality, are well-suited for applications in biochemical an‎d biological fields.

6