Optimal Combination of Mixing Units Using the Design of Experiments Method

• Published in: Micromachines (Basel) Vol. 12; no. 8; p. 985
• Main Authors: Juraeva, Makhsuda, Kang, Dong-Jin
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 19.08.2021; MDPI
• Subjects: combination scheme; cross-channel SAR (CC-SAR); degree of mixing; Design of experiments; Energy costs; Energy resources; Finite volume method; Fluid flow; mixing cell with baffles (MC-B); Modules; Optimization; Reynolds number; Statistical significance; Taguchi methods; Variance analysis; combination scheme; degree of mixing; mixing cell with baffles (MC-B); statistical significance; cross-channel SAR (CC-SAR)
• ISSN: 2072-666X; 2072-666X
• DOI: 10.3390/mi12080985

A passive micromixer was designed by combining two mixing units: the cross-channel split and recombined (CC-SAR) and a mixing cell with baffles (MC-B). The passive micromixer was comprised of eight mixing slots that corresponded to four combination units; two mixing slots were grouped as one combination unit. The combination of the two mixing units was based on four combination schemes: (A) first mixing unit, (B) first combination unit, (C) first combination module, and (D) second combination module. The statistical significance of the four combination schemes was analyzed using analysis of variance (ANOVA) in terms of the degree of mixing (DOM) and mixing energy cost (MEC). The DOM and MEC were simulated numerically for three Reynolds numbers (Re = 0.5, 2, and 50), representing three mixing regimes. The combination scheme (B), using different mixing units in the first two mixing slots, was significant for Re = 2 and 50. The four combination schemes had little effect on the mixing performance of a passive micromixer operating in the mixing regime of molecular dominance. The combination scheme (B) was generalized to arbitrary mixing slots, and its significance was analyzed for Re = 2 and 50. The general combination scheme meant two different mixing units in two consecutive mixing slots. The numerical simulation results showed that the general combination scheme was statistically significant in the first three combination units for Re = 2, and significant in the first two combination units for Re = 50. The combined micromixer based on the general combination scheme throughout the entire micromixer showed the best mixing performance over a wide range of Reynolds numbers, compared to other micromixers that did not adopt completely the general combination scheme. The most significant enhancement due to the general combination scheme was observed in the transition mixing scheme and was negligible in the molecular dominance scheme. The combination order was less significant after three combination units.