Velocity change with seating depth?

[Shinbone]

I started turning necks for the first time, and I experienced a dramatic improvement in my velocity extreme spread. With the newly improved velocity numbers, I am now seeing what appears to be a change in velocity with a change in seating depth. I think the trend has always been there, I just didn't see it due to the noise in my velocity numbers. Is such a change normal?

The velocity data appears to show a 0.6% change in velocity for a .035" change in seating depth. I know 0.6% is not a big change, but it is well within the resolution of the chronograph. Also, the trend in the data looks pretty clear.



Load details in the graph title. Velocity was measured with a Lab Radar. The Lab Radar was spot checked against a Magnetospeed during the shooting session, and found to be stable to within 0.1%, which, BTW, is the manufacturer's claimed tolerance. Powder charge weights were double checked on a second, more accurate scale and held to 0.03% variation. No wind. Air temp changed by about 10F while shooting.

Each velocity data point is the average of two shots. I know a two shot "average" is not much of an average. But, if the underlying inputs are good, then that average will also be good. And, again, the trend looks unmistakable, to me.

 

[Ned Ludd]

Two-shot velocity averages simply cannot be used to support a statistical argument that any trend you observed is real. It is even questionable whether 5-shot averages would be sufficient; however, the question can easily be answered in that scenario by comparing the difference between average velocity and the standard deviation. If the SD values are larger than the difference between average velocity values, a significant difference is not supported statistically.

Unfortunately, it generally takes a much larger sample set than most shooters are willing to generate to statistically support such trends, no matter how compelling they may seem. Burning out a barrel to provide solid statistical significance for such trends is not worth the any benefit that might come from it.

 

[Shinbone]

Ned, thanks for your input, and are you aware whether anyone has ever reported a variation of velocity with seating depth?

 

[Ned Ludd]

I've observed seating depth-dependent velocity changes on many occasions. I absolutely believe it happens, but not necessarily the way your graph would indicate. Certainly as the bullet is seated deeper and deeper into the case, the effective case volume is decreased, leading to increased pressure and velocity for a given charge weight. It is unquestionable that this phenomena occurs. However, a change in seating depth of at least .020" to .030" is usually required before you can even begin to reliably detect such a small difference with any statistical significance.

I've occasionally observed what appeared to be very interesting trends in velocity over much smaller seating depth ranges; for example, when testing a given load/bullet at -.003", -.006", -.009", -.012", and -.015" (off the lands). Unfortunately, even with 5-shots groups, the SDs were as large or larger than the difference between average velocities, suggesting there really wasn't any "trend" across that fairly narrow range of seating depth. Further, the perceived "trend" was typically not reproducible when testing the exact same loaded rounds a few days later. Of course, our eyes and brain would like to tell us there is a pattern; pattern recognition is something humans can be very good at. But if it can't be supported statistically/mathematically, it's most likely an artifact of sub-optimal sample size.

Nonetheless, if you believe something is going on that might be important, it is usually a testable hypothesis - simply repeat the exact same seating depth test a few days later and see whether the graphical results look anything at all like those you posted above. If they don't, it's probably error due to the small sample size. If they do look the same, you ***might*** be onto something. In that event, the real question becomes how many more times are you willing to repeat the same test to statistically support some perceived trend? In other words, if there really IS a trend, what benefit or advantage would knowing that information bring you? Would the knowledge gained be worth all the extra effort, reloading components, and barrel life? If having that knowledge provided some substantial benefit or gain, maybe it would be worth it. If it's just another interesting, but only marginally useful, piece of information, maybe it's not worth the effort. Everyone has to decide for themselves what margin of return is necessary for a given effort.

 

[rhovee]

I haven’t seen much from 5-40 off. But that’s with my big hunting rifles.

 

[CubCouper]

I just posted an explanation of how seating depth can have a big impact on velocity and tuning on this thread... I've spent days building Shot Start (Initiation) Pressure charts for my PPC's by working backwards from known charges and velocities to calculate the impact of neck tension, seating depth, and even tuner position.

http://forum.accurateshooter.com/th...ent-powder-lots-that-shoot-different.3990202/

 

[barefooter56]

I've observed seating depth-dependent velocity changes on many occasions. I absolutely believe it happens, but not necessarily the way your graph would indicate. Certainly as the bullet is seated deeper and deeper into the case, the effective case volume is decreased, leading to increased pressure and velocity for a given charge weight. It is unquestionable that this phenomena occurs. However, a change in seating depth of at least .020" to .030" is usually required before you can even begin to reliably detect such a small difference with any statistical significance.

I've occasionally observed what appeared to be very interesting trends in velocity over much smaller seating depth ranges; for example, when testing a given load/bullet at -.003", -.006", -.009", -.012", and -.015" (off the lands). Unfortunately, even with 5-shots groups, the SDs were as large or larger than the difference between average velocities, suggesting there really wasn't any "trend" across that fairly narrow range of seating depth. Further, the perceived "trend" was typically not reproducible when testing the exact same loaded rounds a few days later. Of course, our eyes and brain would like to tell us there is a pattern; pattern recognition is something humans can be very good at. But if it can't be supported statistically/mathematically, it's most likely an artifact of sub-optimal sample size.

Nonetheless, if you believe something is going on that might be important, it is usually a testable hypothesis - simply repeat the exact same seating depth test a few days later and see whether the graphical results look anything at all like those you posted above. If they don't, it's probably error due to the small sample size. If they do look the same, you ***might*** be onto something. In that event, the real question becomes how many more times are you willing to repeat the same test to statistically support some perceived trend? In other words, if there really IS a trend, what benefit or advantage would knowing that information bring you? Would the knowledge gained be worth all the extra effort, reloading components, and barrel life? If having that knowledge provided some substantial benefit or gain, maybe it would be worth it. If it's just another interesting, but only marginally useful, piece of information, maybe it's not worth the effort. Everyone has to decide for themselves what margin of return is necessary for a given effort.

Great post Ned! I have to add one thing though. The main factor in repeatability MAY be the how the " lose nut behind the butt plate" is hard wired during later testing also. This is a point that is very overlooked when it comes to load development IMO. Comments?

 

[Shinbone]

The main factor in repeatability MAY be the how the " lose nut behind the butt plate" is hard wired during later testing also.

Interesting point. But please elaborate on the details. For example, how does the shooter's technique affect velocity repeatability?

 

[gman47564]

A light hold on the rifle will yield less velocity than a tight hold will.. meaning the more solid you hold the rifle the more velocity it will make..

 

[barefooter56]

Interesting point. But please elaborate on the details.

We , along with external conditions are constantly changing. One day a person may be " in the zone" so to speak with his/her concentration ETC. And gets great results during a test or match. Next time out .. not so much. Concentration is hard to keep as is "staying in the moment" so the old M1A1 brain doesn't start getting involved. Its the " oh man! the last two shots went in the same hole THOUGHT". Then as the crosshair goes over the two previous bullet strikes Mr. Brain turns on and gives the command , NOW! Resulting in a pulled shot.... The "ZONE" can be hard to find everyday.

Interesting point. But please elaborate on the details. For example, how does the shooter's technique affect velocity repeatability?

 

[Ned Ludd]

Great post Ned! I have to add one thing though. The main factor in repeatability MAY be the how the " lose nut behind the butt plate" is hard wired during later testing also. This is a point that is very overlooked when it comes to load development IMO. Comments?

I agree, Phil. However, that is also a difficult thing to quantify with any statistical accuracy. We're human and we definitely do things, both at the reloading bench and behind the rifle, that can affect pressure, velocity, precision, etc. The very fact that we are always trying to shoot the very smallest groups possible indicates a certain amount of bias in the testing process. Nonetheless, we attempt to quantify results and analyze those results statistically to shed some light on whether a change has actually occurred. Most of the time, small sample sizes are the norm, so we have to live with a certain amount of uncertainty, like it or not.

 

[barefooter56]

I agree, Phil. However, that is also a difficult thing to quantify with any statistical accuracy. We're human and we definitely do things, both at the reloading bench and behind the rifle, that can affect pressure, velocity, precision, etc. The very fact that we are always trying to shoot the very smallest groups possible indicates a certain amount of bias in the testing process. Nonetheless, we attempt to quantify results and analyze those results statistically to shed some light on whether a change has actually occurred. Most of the time, small sample sizes are the norm, so we have to live with a certain amount of uncertainty, like it or not.

All very true Ned. We do need to be "tuned in" to ourselves better to do better. If you go to the range just to get it done. And are in a distracted mood. Go home.

 

[Brians356]

If I read the data correctly, the velocity decreased as seating depth increased (jump increased from .003" to ~.033"). And velocity dropped precipitously between jump .028" and .033". So decreasing the case capacity (squeezing the boiler) did not increase pressure until after jump .033" was reached. Do I need new glasses?
-

 

[CharlieNC]

Sorry I don't agree with most of the interpretations. There is clearly a very high statistical correlation between seating depth and velocity as can be shown using linear regression. The sample size is not so important because you are not trying to determine statistical attributes of a single seating depth or comparison of one to another; the regression evaluates the cause/effect across the entire range. Similarly I often find an effect if depth on vertical point of impact.

 

[Brians356]

“There are three degrees of falsehood: the first is a fib, the second is a lie, and then come statistics”. (Anonymous, 1891)
-

 

[dellet]

When developing sub-sonic loads for a gas gun, Function and velocity take precedent over accuracy. Seating depth plays a big role in marginal loads.

In general moving the bullet out, will drop velocity and increase muzzle pressure. This allows adjusting velocity up and down and also can increase/decrease reliability.

You effectively change the burn rate of the powder. Higher or lower peak will mean less or more gas pressure at the port and muzzle when it's time to cycle the action.

These are small things at 3-4000 fps. but when the desired operating window is less than 50-100 fps the advantage of being able to fine tune velocity with seating depth is very helpful.

Sample size really doesn't matter much when you can repeatably move a bullet +/- .030" and measure the results with function/non-function, or sonic crack/no-crack

 

[gman47564]

If I read the data correctly, the velocity decreased as seating depth increased (jump increased from .003" to ~.033"). And velocity dropped precipitously between jump .028" and .033". So decreasing the case capacity (squeezing the boiler) did not increase pressure until after jump .033" was reached. Do I need new glasses?
-

I have seen this as well a decrease in pressure from seating deeper in the case... and its my understanding that seating the bullet farther from the lands allows pressure to escape around the bullet before it engraves the rifling and seals up...thus a bullet seated out farther seals up quicker retaining the pressure and allowing it to build more pressure... course I could be wrong about all that...

 

[Brians356]

In general moving the bullet out, will drop velocity and increase muzzle pressure.

Which direction is your "out"? Out of the case, or out of the lands?
-

 

[dellet]

Which direction is your "out"? Out of the case, or out of the lands?
-

Out of the case. Making a low density load lower.

 

[Ned Ludd]

Sorry I don't agree with most of the interpretations. There is clearly a very high statistical correlation between seating depth and velocity as can be shown using linear regression. The sample size is not so important because you are not trying to determine statistical attributes of a single seating depth or comparison of one to another; the regression evaluates the cause/effect across the entire range. Similarly I often find an effect if depth on vertical point of impact.

In fact, sample size is very important when there is a high level of uncertainty in individual velocity measurements. While I agree with your suggested use of linear regression as a general approach, using that approach with numbers that might have error values of 10, 15, 20, or even 30 fps for single shot velocity determinations is not going to give a statistically valid conclusion because your talking about an effect of a very slight change in seating depth that is probably not linear, and is actually smaller than the individual measurement error (variance). Further, the effect of bullets seated into the lands and the resultant increase in start pressure throws an additional level of complexity (non-linearity) into the equation.

My main point is that when you're only recording single digit velocity changes between groups where seating depth varied by .003" or .005", and it is established that ES/SD values for the load are equal to or greater than the variance between seating depth velocities, and the effect is most likely not linear, no mathematical method known can turn that data into a statistically sold conclusion. We can certainly look at the numbers for the trend we might expect (i.e. as seating depth decreases, effective case volume increases and pressure decreases). That is simple physics and is easily interpreted when the measurements fit the anticipated trend. However, when the data don't fit the expected trend, it is merely guessing.