AN EXPERIMENTAL STUDY OF 

 GRID TURBULENCE IN DILUTE 

 HIGH-POLYMER SOLUTIONS 



A. G. Fabula 



Naval Ordnance Test Station 



(after July 1967 renamed the Naval Undersea Warfare Center) 



Pasadena, California 



ABSTRACT 



Longitudinal, one -dimensional turbulence energy spectra were meas- 

 ured behind a towed grid in a tank filled with dilute aqueous high- 

 polymer solutions. Several types of hot-film flowmeters with constant- 

 tennperature bridges were used, with emphasis on quartz-coated, 

 platinum-film sensors of conical shape. Spectral nneasurements in 

 water were consistent with wind-tunnel measurements. Cylindrical 

 sensors were found unsuitable because of evidence of coating of the 

 hot-film leading edge. The additives -were poly(ethylene oxides) (Polyox 

 WSR-301 and coagulant), polyacrylamides (Separan AP-30 and JlOO) 

 and guar gum. Concentrations of up to 137 ppm were used. The guar 

 gum produced temporary suspensions which were not suitable for tur- 

 bulence measurements because of high noise level. With the other 

 additives, the sensor signal near the grid included a high-frequency 

 connponent (called "raggedness") which was attributed to temporary (or 

 temporarily altered) inhomogeneity caused by the action of the grid 

 bars. With extreme raggedness, the interpretation of the signal as a 

 turbulent velocity fluctuation would have implied impossible viscous 

 dissipation rates. 



The raggedness was insensitive to the range of solution preparation 

 techniques used, which included demineralization of the water and 40 

 hours of gentle stirring of concentrated "master" solutions. When 

 there was a decrease of raggedness (at a fixed distance from the towed 

 grid) with solution age, there was also a significant decrease of spe- 

 cific viscosity. 



Three types of spectral effects were the spectral distortion due to rag- 

 gedness, a greater experimental scatter of spectral levels near the 

 grid in newly prepared solutions, and the predictable spectral shift due 

 to viscosity increase. No non-Newtonian spectral effects were detected 

 under conditions of nonragged signals or at frequencies below the spec- 

 tral distortion due to raggedness. 



Note: This paper is based on a thesis listed as Ref. 1, 



39 



