The Effect of Additives on Fluid Friction 



Furthermore, additional experiments have shown that the effect is not en- 

 hanced by increasing the viscosity of solution of guar by "complexing" with so- 

 dium borate [9]. Increasing the viscosity in this way resulted in lowering the 

 drag-reduction effect based upon the weight of guar in solution. In a typical test 

 the viscosity was increased by a factor of 22 over the guar solution alone, by 

 addition of sodium borate, and the drag reduction than obtained was only 70% of 

 that which would have occurred using guar only. 



However the friction reduction is produced, it seems clear that the action 

 involved is suppression of turbulence intensity. Figure 13 shows test data from 

 the 45.7 cm diameter disk for guar, correlated with disk Reynolds numbers 

 based on water. At concentrations of guar up to 311 wppm, the slopes of the test 

 curves are roughly parallel to, but lower than the turbulent flow water data. For 

 621 wppm and above, the slopes are roughly parallel to, but higher than, the 

 laminar water flow case. From Fig. 10 it can be seen that no significant changes 

 in Fig. 13 would result from use of Reynolds numbers based on the measured 

 viscosities of the solutions for under 500 wppm. 



PIPE FLOW EXPERIMENTS 



The friction reducing effect of polymer solutions can be easily studied by 

 measuring the pressure drop occurring in a given length of pipe in which the 

 polymer solution flows. Many experimental facilities of this type have been 

 constructed, and in general they are similar to that shown schematically in Fig. 

 14, except for the use of air-pressure pumping to minimize degradation of test 

 solutions [10]. Pre-mixed polymer solution contained in tanks is forced through 

 the pipe test section where the static-pressure gradient is measured. Flow 

 rates can be determined by weighing the amount of polymer solution discharged 

 in a given time. Discharged solution is discarded to minimize bias due to shear 

 degradation which occurs very rapidly for many of the solutions. By compari- 

 son of similar data taken using pure water as the flowing medium, drag reduc- 

 tion may be calculated. 



Typical data using poly(ethylene oxide) of 4 million molecular weight are 

 shown in Fig. 15. Drag reduction of well over 75% is easily obtained. Similar 

 data using the same polymer in sea water, but in a different apparatus,^ are 

 given in Fig. 16. 



Another pipe flow apparatus, which is essentially a turbulent flow rheome- 

 ter, has recently been constructed according to the sketch of Fig. 17. The pis- 

 ton of the cylinder is moved upward at 1.245 cm per second, forcing fluid through 

 the small diameter pipe. The entire apparatus is mounted vertically to allow 

 entrapped air to escape. 



Some representative data from this instrument, taken at a constant flow ve- 

 locity of 12.65 meters/sec (Reynolds number based upon water at 21.1°C of ap- 

 proximately 14,000) are given in Fig, 18. 



5 

 Data taken by the Western Co. under U. S. Navy contract. 



953 



