ACUSIM Software has been acquired by Altair Engineering and the AcuSolve software is now part of Altair's HyperWorks product line. Click here to obtain latest information about AcuSolve.

Compact Heat Exchanger Elements

Among the number of passive heat transfer enhancement techniques, internally-finned tubes have been one of the most vigorously investigated geometries due to their potential impact on energy savings as well as environmental issues. Internally-finned tubes are ideal for testing various numerical modeling capabilities because they possess many interesting and complex hydrodynamics associated with various fin geometries, and are experimentally testable with reasonable accuracy in a lab over a wide range of geometric and operating conditions.

It would be computationally expensive to simulate an entire finned tube, as in the left-hand figure. Using periodic boundary conditions and a computational mesh created by extruding a surface mesh with boundary layers two layers in the streamwise direction was proven to be sufficient to model the periodic flow condition.

Figure 1. Geometry Schematics

Three turbulence models were found to be adequate for this work: Lam-Bremhorst k-epsilon (1981), Spallart-Allmaras (1992), and Goldberg k-epsilon (1998). In applying these models to fin tube problems, it was found that both Spalart-Allmaras and Lam-Bremhorst's k-epsilon models predicted significant laminarization in the inner fin region. The severity of the predicted laminarization varied depending on the fin tube geometry. (It became more severe with more than 30 fins and lower Reynolds numbers - especially on the leeward side of the fin.)

All three turbulence models work reasonably well with fewer than 30 fins or at high Reynolds number, where the effect of laminarization is expected to be minimal. The following figures show some results for a fin geometry where the three models perform well. The flow is into the screen and the fin is winding to the right.

Figure 2. Circumferentially-Averaged Friction Factor

Figure 3. Velocity (left) and k (right) Distribution at Re ~ 36,700

Experimental results indicate that depending on the fin shape or directionality (unsymmetric fins), the overall friction factor can vary as much as 10-15% over a range of Reynolds numbers. Our computations indicated that tubes with micro fins are more likely to be affected than tall fin tube cases. Also, depending on the fin shape, azimuthal recirculation (similar to a separation bubble) was observed. The figures that follow are for 54 fins with a 45° helix angle.

Figure 4. Rectangular Fin Shape.

Figure 5. Tilted Fin (Impinging) Shape.

This page was extracted from a Doctoral thesis by Je-Hoon Kim in Mechanical Engineering at Rensselaer Polytechnic Institute. Academic advisors are Dr. Michael K Jensen and Dr. Kenneth Jansen.