In the past, I have used high speed video to analyze various aspects of AR-15 operation. Today, I did some testing at the range with several action springs. This is intended to be an ongoing test, so all three action springs were brand new. They are:
The test weapon was a 16" carbine gas AR-15 with a .063" gas port, H buffer, 5.56mm NATO chamber, M16/Auto bolt carrier, and AR-15 fire control group. The bolt carrier group was cleaned and wiped down with a dry cloth before the test. Two types of ammunition were used, Prvi Partizan M855 and Federal XM193. One magazine, a Lancer L5 20rd, was used for the test, and it was always loaded to 20 rounds at the beginning of each string of fire, which consisted of five rounds.
Rate of fire calculation is based on the time it takes for the bolt carrier group to fire, unlock, extract, eject, cock, feed, chamber, and lock, plus the time it would take for a standard M16A1 fire control group to allow the hammer to fall once more. These calculations are therefore theoretical, but this method has always been verifiable when compared to actual M16A1 high speed video testing.
The rate of fire was highest with the Springco Blue action spring (726rpm M855/680rpm XM193), median rates of fire were achieved with the Brownells Chrome Silicon spring (685rpm M855/667rpm XM193), and the lowest rates of fire occurred with the Brownells AR-15/M4 recoil spring (656rpm M855/626rpm XM193). The differences in cycle time and bolt carrier velocity between the three springs were always noted during the feeding, chambering, and locking portions of the cycle of operations; put simply, the lower rates of fire came as a result of reduced forward bolt carrier velocity. As forward velocity fell, so did consistency.
Relative to the other action springs, the Sprinco Blue action spring was very consistent overall, with a total cycle time standard deviation of .55 for M855 and 3.03 with XM193. The Brownells Chrome Silicon spring was not far behind, with M855 and XM193 standard deviations of 1.87 and 3.81, respectively. The Brownells AR-15/M4 recoil spring could not be considered as consistent, with standard deviations of 5.59 and 12.87.
Further testing will be conducted.