Originally Posted by rich
Interesting data; a x-y scatter plot shows quite strong inverse correlation between pellet weight and velocity, which is of course what you might expect, but is useful in confirming how pellet weight really needs to be taken into account for every chrono reading.
Interesting data Waldo. But I have to disagree with rich.
Looking deeper into the data of all 60 shots shows that the "strong" correlation between weight and velocity is due to only THREE shots (no.'s 45, 49 and 50). And even then the correlation is not very strong: R squared of only 0.01. I'd say that is a very low correlation.
(a 100 percent correlation would give an R-squared of 1.00, at 0.60 it begins to mean something)
The correlation without those three shots (that may be called outliers) is even poorer: R squared of 0.0007.
And that is a strange thing. Because, as Rich says, you would certainly expect a correlation between pellet weight and velocity. And it is not apparent from these figures.
My guess is that either there is something wrong with this dataset (weighing scale or chrono being not accurate enough, or the data of weight and velocity is mixed up) OR this 0.20 grain difference in pellet weight has no effect on velocity. Maybe it is this last fact: I have seen larger variation in weight in a tin of JSB Exact pellets. Between 8.2 and 8.7 grain in a tin. The variation here was only 0.2 grain. Probalby sorted pellets were used for this test.
The HFT 500 performs very well: a spread of 27 fps over 60 shots. And from the chart it looks like there are many more shots in a charge: