ricka0
Well-Known Member
[ QUOTE ]
<font color="purple"> Ok If anyone can shoot a hole through this theory, I welcome that shot.
This is more or less an [sic] emperical process based on accurate barrel
life in several calibers as reported to me by lots of folks plus a
couple dozen barrels I've worn out. All I did was study the data and
determine what math would give a best-fit formula. And if someone has
a better method, I'd like to know what it is.
</font>
[/ QUOTE ]
I have a MS in Applied math. Here is my review:
In scientific parlance, you don't have a theory - perhaps a conjecture. In colloquial terms, you can use theory.
>>All I did was study the data and determine what math would give a best-fit formula.
How did you measure best fit? I can easily get an exact fit to your data using splines or high order polynomials. Would a high order poly be a better predictor?; no - they oscillate wildly to fit thru each data point. How about a spline? Nope, splines can't be used to extrapolate data.
You have answered the problem yourself.
[ QUOTE ]
<font color="purple">
But this empirical formula has limits. The old .22 rimfire long rifle
burning 1 grain of powder in a bore whose capacity is 24 grains would
calculate to give a barrel life of some 1,728,000 rounds.
</font>
[/ QUOTE ]
IE, you can't extrapolate.
<font color="purple"> Ok If anyone can shoot a hole through this theory, I welcome that shot.
This is more or less an [sic] emperical process based on accurate barrel
life in several calibers as reported to me by lots of folks plus a
couple dozen barrels I've worn out. All I did was study the data and
determine what math would give a best-fit formula. And if someone has
a better method, I'd like to know what it is.
</font>
[/ QUOTE ]
I have a MS in Applied math. Here is my review:
In scientific parlance, you don't have a theory - perhaps a conjecture. In colloquial terms, you can use theory.
>>All I did was study the data and determine what math would give a best-fit formula.
How did you measure best fit? I can easily get an exact fit to your data using splines or high order polynomials. Would a high order poly be a better predictor?; no - they oscillate wildly to fit thru each data point. How about a spline? Nope, splines can't be used to extrapolate data.
You have answered the problem yourself.
[ QUOTE ]
<font color="purple">
But this empirical formula has limits. The old .22 rimfire long rifle
burning 1 grain of powder in a bore whose capacity is 24 grains would
calculate to give a barrel life of some 1,728,000 rounds.
</font>
[/ QUOTE ]
IE, you can't extrapolate.