The recent letter by Skinner' draws attention to two features of capillary rheometry which are not often appreciated in routine viscometric meauurements. One of thebe effects has received considerable attention in these labor5 tories in connection with the flow of molten polypropylene and the establishing of a melt flow index test method for that material.2 Skinner shows that with the standard melt indexer there may be a 50% increase in flow rate during the complete extrusion of a full charge of polythene, assuming a power law of flow, and making no capillary end connection. It is, of course, a prerequisite that for a standardized test method, the measured quantity shall respond to the test conditions in the same way for all materials tested-in this case that the flow rate shall be constant at all points during the extrusion. It is clear from Skinner's example, that this is not true in the case quoted: whether or not this has been experimentally observed durin? routine measurements of melt index in the United States or Canada is not known but as the ASTM test method requires only one sample of extrudate to be cut, it is unlikely that changes in rate would be noticed: on the other hand, the B.S.I. method inrludes, in its proccdure, a check on the constancy of flow rate.With polypropylene, the position becomes much worse, for the index n may be as much as 2.5, and the consequent increase in flow rate is by a factor of 3.6 in the course of a run. It was the direct observation of such an increase that initiated work on this problem, which showed that the pressure gradient in the barrel, or "reservoir effect," was responsible.Oakes and Peover in 1946 and Clegg in 19578 convidered the Melt Indexer as a system of two extruders in series, and showed that the barrel pressure gradient was only of the order of a few percent for most polythenes at grading stress (1.7 X 106 dynes/cm.*). Clegg showed, however, that for materials with n > 2 (e.g., P.V.C.) the effect was very great, and could not be ignored in the measurement of melt indices.His treatment gives an expression for the output rate:where L = length of polymer column between die and piston, 1 = length of die, r = radius of die, R = radius of barrel, and n = flow power law exponent. This result can also be derived from the equation given by Skinner.In the Melt Indexer, with n = 2, Q increases by 100% between the beginnin% and end of the run, and attains the value equivalent to that due to the whole pressure on the piston only at the end of the run.Measurements of the pressure in the melt near the die entry were made (a) by the deflection of a diaphragm in the barrel wall2 and later ( b ) by a pressure transducer in a similar location (Lamb and Benbow*). A parallel register of the rate of flow was also made and both showed an increase as the piston descended. Values of apparent viscosity calcylated from the rate alone and assuming full pressure to be operative at the die entry showed a decrease by a factor of 2 in the course of the run, whereas those using the pressure...

No abstract