U(PCRG), and U(SFG) > U(RG). It is renowned that the implementation of turbulator allows the acceleration of air flow due to the sudden contraction of flow passage. The introduction of a vertically aligned separation in the partially-covered grids promotes mid-plane jet formations, owing to the principle of mass conservation. Such separation effectively reduces the undesirable fluid bypass around the plate-fins, and forcefully regulates the working fluid to penetrate inter-fin regions. In addition, the accelerated airflow in between the fin array enforces greater wall shear stress along fin surfaces, which limits the growth of viscous sub-layer, thus effectively enhancing forced convection./p> Nu(PCRG)./p> 381.1 m2s−2. The pronounced v’ and w’ give rise to greater development of A, suggesting a larger area of flow boundary is being agitated at the localized inter-fin region, thus poses a unique advantage in supporting potent thermal dissipation. In addition, Fig. 5d, e showed that the fractal designs (SFG and PCSFG) generally produces greater variety of flow fluctuation as compared to the regular designs (RG and PCRG). The finding is crucial as it further supports the concept of PCSFG induced flows are capable to inherit turbulence features similar to that of SFG, whereby multilength-scale eddies are generated from the multitude fractal bar thicknesses and filtered within the fins. With the amalgamation of such feature and the added benefit of accelerated centreline airflow, PCSFG would induce highly effective turbulence structures in enhancing forced convective heat transfer, which is exceptionally suitable for localize cooling applications./p> the ensemble average. The L generated from the different 2D-planar grids are rationalized with δ, i.e. L/δ and is shown in Fig. 5c./p> Nu(RG), and could be enforced through the realization of smaller L(x)/δ. By considering vortices rotating in the X–Z plane, the dwindling L(x)/δ increases the vortices’ angular velocity around y-direction ωy, but in turns extending the vortex line laterally due to the conservation of angular momentum, i.e. vortex stretching. The elongation of vortex structures thus effectively interacts and disrupts the fins’ boundary layer. Similar findings were reported in Hoi et al.27, and it warrants the benefits of smaller L(x)/δ in enhancing heat transfer, as it encourages larger L(y)/δ formation to interrupt fins’ flow boundary layer more effectively. The implication concurred closely with RG induced L, yet poses an opposite effect of streamwise elongation, which directs the flow energy to undesired intermediate regime that lessened boundary layers’ agitation probabilities./p> 0, K > 3) are recorded for the former, and (|S|> 0, K < 3) for the latter duo. The ± S achieved by SFG implies extreme decelerated turbulent events documented in the (x, y) directions, along with turbulent accelerations in z-direction. These accelerations are considered to be rare and intensive, as indicated with the high positive K > 3. Interestingly, similar S developments are recorded for the partially-covered grids, but vastly disparate in K as evident from Fig. 8b. The realization of K ≈ 2 for PCRG and PCSFG depicts an increase in likelihood for the extreme decelerated (ax, ay) turbulent events, which could very well imply the high occurrences of alternating flow directions that escalates to the formation of copious vortices./p> Nu(SFG), even though the numeral polarity of S registered similarity. Conversely, NG and RG demonstrated moderate flow circulations, hence the low Nu. In general, the extreme decelerated flow events are capable of forming intense flow vortices, which is beneficial in disruption of fins’ boundary layer. Further research is still required to uncover the S and K profiles at different inter-fin locality, in order to uncover the overall flow structures that are preferable for maximising forced convection of plate-fin heat sink./p> PCRG. The phenomenon may imply that partially-covered grids are capable of generating broad array of high energy multilength scale eddies through (i) first stage grid-separation induced turbulence and (ii) second stage plate-fin eddies filtration for an intense vortex shedding process. As PCSFG comprised of different fractal bar thicknesses, there would be greater variations in eddies length scale, hence ampler variety of frequencies. Contrariwise, utilization of fully-covered grids masked the powerful vortex shedding effect, and is further subdued under NG configuration. The usage of SFG generates substantial turbulence intensity at centreline of x/Dh = 0.125, and one would infer that the energy level would be greater than RG, as I(SFG) > I(RG). However, the presence of largest grid bar on SFG produces sizeable wake with substantial flow recirculating in it. This causes the breaking of vortical structures which creates a less pronounce vortex shedding formation28. Even so, SFG prominent heat transfer, viz. Nu(SFG) > Nu(RG) is due to the effective distribution of flow kinetic energy in agitating fins’ boundary layer, as indicated by the wider spread of particle trajectory, especially in the spanwise direction. As for NG, the flow momentum is greatly sabotaged, causing the vortex shedding and energy level to subside. Surprisingly, the energy profile demonstrated by NG and RG are very much identical, revealing the impact of 2D planar grids in raising the preferable flow energy levels for forced convective heat transfer./p>