[scc2052]

brycelarson, on Dec 29 2008, 06:11 PM, said:

scc2052, on Dec 29 2008, 04:47 PM, said:

In the results I'm confused and don't really understand what the numbers on the graphs that are on they left mean can someone help me out? thanks

the two numbers that we're looking at for this test are:

mean velocity

and

SD (standard deviation)

mean velocity is a number that tells you how fast the ball was leaving the barrel. A higher mean velocity indicates that with the gun set the same - that barrel is going to shoot faster. This indicates that you could get more shots at a specific velocity from the same tank out of a barrel with a higher mean velocity.

Standard Deviation (SD) is an indication of how consistent that barrel is. the lower the number the more consistent. A low SD will directly correlate to having a lower +/- fps when chronoing a gun. Additionally, it's my assumption that a more consistent gun will be a more accurate one.

does that help?

yes thank you very much great job with your test very well put together thanks again

[Poe]

brycelarson, on Dec 29 2008, 06:11 PM, said:

...
Standard Deviation (SD) is an indication of how consistent that barrel is. the lower the number the more consistent. A low SD will directly correlate to having a lower +/- fps when chronoing a gun. Additionally, it's my assumption that a more consistent gun will be a more accurate one.
...

I realize it's a lot to ask, but will the individual velocity and impact location of each ball be charted during accuracy tests? Would be interesting to see how much accuracy is effected by velocity fluctuations. Maybe that's one for another test?

[Lord Odin]

Since the gun is going to be clamped down, I don't see why it couldn't be chrono'd as well. To see if there is a direct correlation, it would be suggested to record the velocity and its hit location together. I wouldn't separate those two values through sorting or exclusion as they may be linked.

[Snipez4664]

Lord Odin, on Dec 31 2008, 10:41 AM, said:

Since the gun is going to be clamped down, I don't see why it couldn't be chrono'd as well. To see if there is a direct correlation, it would be suggested to record the velocity and its hit location together. I wouldn't separate those two values through sorting or exclusion as they may be linked.

Especially if you have an estimate for what the drop should be based on velocity - use the excellent paintball trajectory calculator for that. That way, each shot can be normed to it's own zero - interesting when you're only investigating, say, the effect a barrel has, independent of gun.

It's a pain to send one man downrange while the other shoots, but damn...it would be interesting.

[cockerpunk]

we will be re-chonoing the gun in the XY grid test. i dont know if we will chrono each shot in that test, we have alot to do that day, so we might just make sure they are all shooting in a certain range before testing.

This post has been edited by cockerpunk: 31 December 2008 - 11:25 AM

[Snipez4664]

cockerpunk, on Dec 31 2008, 11:18 AM, said:

we will be re-chonoing the gun in the XY grid test. i dont know if we will chrono each shot in that test, we have alot to do that day, so we might just make sure they are all shooting in a certain range before testing.

Sounds good - a consistent enough gun is plenty enough control IMO.

[brycelarson]

last time with the backspin test we got fps, x, y data on all shots - don't see why we can't on this one as well

I intended to do that.

[cockerpunk]

id like to do that to, i think that will give us the final word on using chronograph readings to compare to actual XY accuracy.

id also like to do the accuracy VS fps test that day too.

This post has been edited by cockerpunk: 31 December 2008 - 01:05 PM

[brycelarson]

cockerpunk, on Dec 31 2008, 11:43 AM, said:

id also like to do the accuracy VS fps test that day too.

yup, I assumed we would use the best barrel from the accuracy at 275 test - then shoot at say 220, 240, 260, 280, 300 and 320 and see what happens.

[Poe]

Fantastic. Best of luck.

[woodwose]

This is very nice work fellas!

I think that it is important to note that the CP barrels have a significant amount of porting - just over 4". I happen to have a 14" .685" CP barrel that I bought because it was as close to underbored as I could get in a 1 piece, and it has roughly 10 inches of unported length. The Tom Kaye Barrel article suggests that the optimal barrel length for efficiency is 8-10 inches, and that the effective length of a barrel is the length only until the porting begins. It is very cool that this test verifies this article, since this has basically become the current conventional wisdom on barrel length vs. efficiency.

Also, where do you get Grain Belt beer? Is that midwest thing?

w0se

[woodwose]

Another question this helps to answer is whether it is better to size an entire barrel, or just part of it (like you do with a kit). If we look at the data for the 12/.685 one piece CP vs. the 12/.685 front/back kit, I think we might have an answer that makes a lot of sense.

The 1 piece barrel is more efficient (290.25 vs. 276.95) but less consistent (5.73 vs. 4.70). This seems to validate the theory that a 2 piece barrel does most of the acceleration of the ball in the first 5 inches (or whatever the length of the back is) instead of the full 12 inches of barrel length, so therefore there is less barrel length that can act on the ball (via friction that magnifies the ireegularity of the paint) and cause variation in velocity. Do you guys agree with this?

I can't wait to see the accuracy test. I would also like to see a Palmer barrel in there since that is the only other significantly different barrel design besides 1 piece, 2 piece, and rifled. That could be test 3 - where you test the effects of barrel porting

w0se

[woodwose]

Ok, one more thing and I promise to stop spamming this thread

Every time I read about someone shooting +-3fps or +-5fps I wonder what they are doing that I'm not. I am very happy to see that even with an entire scuba tank and a $750 emag, you guys are getting a standard deviation of around 5 fps. That means that if you shoot 20 shots at an average speed of 280fps, roughly 19 of them will be between 270 and 290 fps (2-sigma) and one will be outside that range.

So an emag with a scuba tank - arguably an insanely consistent setup - only shoots +-10fps! Next time someone with a pump on 12ies claims +-3fps I'm going to create an animated gif of me kicking them square in the nuts! The only way you can get +- 3fps with most guns is if you only shoot 1 shot (joke).

w0se

footnote: http://en.wikipedia....ndard_deviation

This post has been edited by woodwose: 01 January 2009 - 03:23 PM

[brycelarson]

woodwose, on Jan 1 2009, 01:36 PM, said:

The Tom Kaye Barrel article suggests that the optimal barrel length for efficiency is 8-10 inches, and that the effective length of a barrel is the length only until the porting begins. It is very cool that this test verifies this article, since this has basically become the current conventional wisdom on barrel length vs. efficiency.

Also, where do you get Grain Belt beer? Is that midwest thing?

we've actually disproven this idea now. The CP .683 back without a front was significantly slower than with the 10" tip - and that tip has porting for nearly all of the length. This indicates that the ball is still accelerating through at least some of the ported portion of the barrel.

And yes, grainbelt is a Minnesota thing. It's a beer that was brewed in Minneapolis right across the river from downtown starting in the 1890's. In the 40's Grainbelt Premium was introduced - it changed hands a few times and went away in the 70's - but a brewery called Schells in southern Minnesota started producing it from the same recipie in 2002. This sign is a classic landmark on the river in downtown Minneapolis.

woodwose, on Jan 1 2009, 01:44 PM, said:

The 1 piece barrel is more efficient (290.25 vs. 276.95) but less consistent (5.73 vs. 4.70). This seems to validate the theory that a 2 piece barrel does most of the acceleration of the ball in the first 5 inches (or whatever the length of the back is) instead of the full 12 inches of barrel length, so therefore there is less barrel length that can act on the ball (via friction that magnifies the ireegularity of the paint) and cause variation in velocity. Do you guys agree with this?

I can't wait to see the accuracy test. I would also like to see a Palmer barrel in there since that is the only other significantly different barrel design besides 1 piece, 2 piece, and rifled. That could be test 3 - where you test the effects of barrel porting

yes, we agree that a majority - but not all - of the acceleration seems to happen in the first part of the barrel. We looked at the two piece with an underbored back this way: it gained the consistency of an overbore via the tip - but kept much of the efficiency of an underbore with the back. Sort of a best-of-both situation. Once we have the accuracy data then the decision between one and two piece may become more clear.

As to the Palmers - yeah, that would be cool to add, if anyone has one to lend for a test - we would love to give it a try. The nice thing about this test is that once we've got good baseline testing procedures set up we can get a good comparative test with a pretty small test. In other words - once we know how the barrels we have rank - we can then setup a new test, include some known barrels and add some new ones - and get a pretty good idea about how the new ones fit into the old data.

[cockerpunk]

woodwose, on Jan 1 2009, 03:22 PM, said:

Ok, one more thing and I promise to stop spamming this thread

Every time I read about someone shooting +-3fps or +-5fps I wonder what they are doing that I'm not. I am very happy to see that even with an entire scuba tank and a $750 emag, you guys are getting a standard deviation of around 5 fps. That means that if you shoot 20 shots at an average speed of 280fps, roughly 19 of them will be between 270 and 290 fps (2-sigma) and one will be outside that range.

So an emag with a scuba tank - arguably an insanely consistent setup - only shoots +-10fps! Next time someone with a pump on 12ies claims +-3fps I'm going to create an animated gif of me kicking them square in the nuts! The only way you can get +- 3fps with most guns is if you only shoot 1 shot (joke).

w0se

footnote: http://en.wikipedia....ndard_deviation

let it be said that we have seen numbers like an SD of 1.8 over 20 shots out of that gun before, but, as i claimed before, this gun is one of the best guns i have ever seen in my 8 or 9 years of playing over the chrono.

it was a bit cold in bryces basement, and the paint was not as good as when we saw that 1.8 SD.

regardless the gun is about as consistent as i have ever seen guns.

[brycelarson]

yup, it was about 50 - 55 degrees in my basement (great temperature for brewing beer, a little cold for scientific paintball testing ) and the paint was fine, but less than perfect. Last summer with the .679 freak insert and paint only a few days out of the factory we did record a 20 shot string with a SD of 1.68

bore size test

This post has been edited by brycelarson: 01 January 2009 - 05:38 PM

[Capt. Kirk]

i feel like an idiot.... but if this wasn't testing accuracy what was it testing?

[Lord Odin]

Capt. Kirk, on Jan 1 2009, 05:42 PM, said:

i feel like an idiot.... but if this wasn't testing accuracy what was it testing?

Consistency and efficiency and the different barrel attributes such as length, 1 piece vs 2 piece, etc. The accuracy test is coming up soon.

[brycelarson]

Lord Odin, on Jan 1 2009, 05:46 PM, said:

Capt. Kirk, on Jan 1 2009, 05:42 PM, said:

i feel like an idiot.... but if this wasn't testing accuracy what was it testing?

Consistency and efficiency and the different barrel attributes such as length, 1 piece vs 2 piece, etc. The accuracy test is coming up soon.

yup, we need distance to do the accuracy test - we like 50 feet. It's enough to get some spread, but not so long that you miss the target.

[Capt. Kirk]

so the first test was for efficiency? then why was t run off a scuba tank?