(554a) Kinetics of Polyethyleneoxide U309 Degradation during Turbulent Drag Reduction in a Hydrodynamically Smooth 4.58 Mm ID x 180 L/D Pipe

This work presents experiments to infer the kinetics of polyethyleneoxide U309 degradation during turbulent drag reduction in a test pipe of internal diameter D = 4.58 mm with electropolished bore of rms roughness 0.0003 mm and nondimensional length 180 L/D divided into 6 identical square-edged sections. The assembled test pipe had a square cut-off entrance, set as axial origin x = 0, and was placed in a pump-driven single-pass flow system that provided turbulent flow for solvent Reynolds numbers 4000 < Re < 80000 and wall shear stresses from 3 < Tw Pa < 600. At high Re > 8000 sets of five friction factors f_j&k were derived from pressure drops between all adjacent tap pairs, from 1&2 upstream to 5&6 downstream, while at low Re < 8000 tap pairs 5&6 and 4&6 were used, the latter providing twice the normal signal. The Polyethyleneoxide additive, abbr U309, of molecular weight 12.8x10⁶, was studied at concentrations C from 0.02 to 10 weight parts per million in deionized water DW at constant temperature T = 23 + 2 C. Solvent friction factors at any fixed turbulent flowrate were constant to within 1.1% for all tap pairs, that is, solvent flows were fully developed, with "entrance length" Le/D < 42.3. Too, at each flowrate, all tap pairs adhered to the Newtonian Prandtl-Karman law (E1) 1/√f = 4.0 log Re√f - 0.4 to within 0.2 ± 0.1 units of 1/√f for 300 < Re√f < 6000, showing that the pipe was hydrodynamically smooth in that interval. Solutions of U309 exhibited turbulent drag reduction, with flow enhancement S' = [(1/√f )p - (1/√f )n ] Re√fs that ranged from near zero to the maximum possible for flow along the MDR asymptote (E2) 1/√f = 19.0 log Re√f - 32.4. A typical solution, say C = 0.5 wppm U309, had a trajectory that ascended linearly at low Re√f, with all tap pairs clustered cohesively as in a jet; then, for Re√f > 1000, 1/√f increased with progressively shallower slope and tap pairs separated into a spray, upstream pairs above downstream, that attained a maximum at Re√f ~ 1500 before sinking towards solvent for Re√f from 2000 to 5000. The cohesion of friction factors from all tap pairs at Re√f < 1000 and their separation into a spray at Re√f > 1000 is evidence of polymer degradation becoming more pronounced with increasing flow strength and also with increasing residence time in the pipe. A "degradation falloff point" could be quantitatively associated with the maximum observed flow enhancement by the 0.5 wppm solution at (Re√f^, S'^) = (1000, 4.0), marking a characteristic flow strength at which degradation becomes discernable. Similar trajectories were exhibited by all concentrations of U309, and in each case the observed ratios of S'/S'^ at any Re√f > Re√f^ could be interpreted as a fraction of polymer still undegraded (1-X), and therefore as a kinetic "severity" -ln(1-X) of the overall polymer degradation reaction. Plots of degradation severity vs residence time in the pipe for each polymer concentration then provided apparent first-order degradation rate constants kd 1/s as a function of turbulent flow strength Tw Pa. Data from all of the present experiments showed that, for any given solution of U309, kd increased roughly linearly with increasing Tw, and further, that the "degradation modulus" (kd/Tw) = 0.033 +/- 0.007 (1/Pa s), was roughly independent of solution concentration C, implying that the overall polymer degradation reaction was of first order.