Abstract
It is suggested that the housing of regenerators may have a significant impact when experimentally determining Nusselt numbers at low Reynolds and large Prandtl numbers. In this paper, a numerical model that takes the regenerator housing into account as a domain that is thermally coupled to the regenerator fluid is developed. The model is applied to a range of cases and it is shown that at low Reynolds numbers (well below 100) and at Prandtl numbers appropriate to liquids (7 for water) the regenerator housing may influence the experimental determination of Nusselt numbers significantly. The impact of the housing on the performance during cyclic steady-state regenerator operation is quantified by comparing the regenerator effectiveness for cases where the wall is ignored and with cases where it is included. It is shown that the effectiveness may be decreased by as much as 18% for the cases considered here. A reduced number of transfer units (NTUeff) is proposed based on the calculated regenerator effectiveness that accounts for the effect of the housing heat capacity.
| Original language | English |
|---|---|
| Journal | International Journal of Heat and Mass Transfer |
| Volume | 65 |
| Pages (from-to) | 552-560 |
| ISSN | 0017-9310 |
| DOIs | |
| Publication status | Published - 2013 |
Keywords
- Wall effects
- Numerical modeling
- Nusselt number
- Low Reynolds number
- Aqueous heat transfer fluid
- Regenerator