Abstract
This paper describes the measurement of convective heat transfer coefficients and friction factors for sCO 2 flowing in additively manufactured tubes with internal pin-fins at the US DoE’s National Energy Technology Laboratory (NETL) in Morgantown, WV. The measurement procedures were validated by conducting benchmark tests with smooth stainless-steel tube and comparing the results with published correlations for Nusselt number ( Nu ) and friction factor. Over Reynolds numbers ( Re D ) ranging from 5×10 4 to 2.5×10 5 , measured Nu was within 5% of the Dittus-Boelter correlation and measured friction factors were within 5% of the McAdams correlation for smooth tube flow.
The candidate pin fin patterned pipes were additively manufactured (AM) at the Oak Ridge National Laboratory. The pins were circular or elliptical in cross-section and were printed at a 30° angle relative to the inner wall (to meet AM constraints). The pin arrangement was helical to promote enhanced heat transfer due to swirl flow. Pin length to diameter aspect ratio was 1.33, 2, and 8, while the pin diameter to tube diameter ratio was 0.188, 0.125, and 0.063. Tests were performed for Re D varying from 6.9×10 4 to 2.2×10 5 and at conditions equivalent to the low pressure (LP) outlet (8.69 MPa, 361 K) and the high pressure (HP) inlet (20.7 MPa, 350 K) of the low temperature recuperator (LTR) in an indirect sCO 2 cycle. The Wilson plot technique was utilized to measure the bulk heat transfer coefficients.
For the best performing design (tube A, pin length to tube diameter ratio: 1.33, pin diameter to tube diameter ratio: 0.19), the local heat transfer coefficient increased by 112% relative to the Dittus-Boelter correlation at the LTR LP outlet and 106% at the LTR HP inlet. These correspond to a 282% and a 271% increase in the product of the heat transfer coefficient and surface area product, respectively. Large pressure drops across the test articles were observed. For tube design A, the average friction factor, across the range of Re D considered, was significantly larger than the McAdams correlation at both the LTR LP outlet and the LTR HP inlet. A thermal performance factor was utilized to express the ratio of material required to build a finned heat exchanger relative to a finless heat exchanger with the same heat duty and pumping power. Tube design A was estimated to decrease the required heat exchanger material by 10%.
