How long does a 308 barrel need to be to reach max velocity

Nothing surprising here, riterates and reconfirms that the .308 will achieve 90-95% of its maximum velocity potential in a 20-24" barrel.

What a cool video. It was definitely over the top, but they finally got it cut down to where the really useful data were. Neat stuff.

Years ago, Clint Smith did a similar test (but without the extremes) for precision relative to barrel length, but with a standard barrel, cutting it down by an inch with each shot. His work substantiated that length is not a determining factor of precision and that shorter barrels were not necessarily less precise.

Nature Boy said:

Neat video. I learned a few things. I know they said POA shifted horizontally as they shortened the barrel but was there a change in precision?

Click to expand...

POA remained the same (if the shooter wasn't trying to influence the test). POI is what changed some.

mcb said:

So I tweak the powder charge to get 2600 fps from a 24 inch barrel. But where they got 2784 fps from the 69 inch barrel QL predicts 3057 fps this tread of the prediction being fast holds until the lines cross ~24 inches. The the QL prediction starting coming in lower than the real world data. MDT got ~2568 at 19-inches and QL predicts 2471 fps.

Click to expand...

QL can only predict what the velocity will be out of a gun because QL does not control for things like internal barrel dimensions that will influence a bullet's velocity.

Double Naught Spy said:

POA remained the same (if the shooter wasn't trying to influence the test). POI is what changed some.

Click to expand...

I meant POI, brain said POI, but my fingers typed POA

Double Naught Spy said:

What a cool video. It was definitely over the top, but they finally got it cut down to where the really useful data were. Neat stuff.

Years ago, Clint Smith did a similar test (but without the extremes) for precision relative to barrel length, but with a standard barrel, cutting it down by an inch with each shot. His work substantiated that length is not a determining factor of precision and that shorter barrels were not necessarily less precise.



POA remained the same (if the shooter wasn't trying to influence the test). POI is what changed some.



QL can only predict what the velocity will be out of a gun because QL does not control for things like internal barrel dimensions that will influence a bullet's velocity.

Click to expand...

Actually you can tweak internal bore dimension in QL. If you have an accurate measure of the cross sectional area of your particular barrel's bore there is a place to enter this as opposed to using the default cross section value that is usually taken from the SAAMI spec. It even has a little widget to help you calculate that cross section arear based on measurements of your rifled bore. It the friction model that I think is failing us here (assuming I actually had the right powder). There is very limited ability to adjust the dry friction and viscous losses in the system.

mcb said:

Actually you can tweak internal bore dimension in QL. If you have an accurate measure of the cross sectional area of your particular barrel's bore there is a place to enter this as opposed to using the default cross section value that is usually taken from the SAAMI spec. It even has a little widget to help you calculate that cross section arear based on measurements of your rifled bore. It the friction model that I think is failing us here (assuming I actually had the right powder). There is very limited ability to adjust the dry friction and viscous losses in the system.

Click to expand...

Either way, the program is failing to control for all the variables, hence coming up with predictions that were not a great match for reality.

Interesting. I wish that they had taken 10 shots per test length, and had cleaned the bore and de-burred the crown each time they cut it down...As it is, it makes a SWAG at an answer.

It sort of bugs me when some one takes the care to make an interesting test fixture and doesn't collect enough data...

illinoisburt said:

The missing variable is powder burn rate. Loading with faster or slower powders could change his results measurably. I'm sure someone playing with Quickloads could figure out an optimum length.

Click to expand...

that is what I say being some powders lose more velocity in shorter barrels and some powders lose speed a lot slower in shorter barrels

Double Naught Spy said:

Either way, the program is failing to control for all the variables, hence coming up with predictions that were not a great match for reality.

Click to expand...

That's what I said in my original QL related post in not so few words...

So this was bothering me and I came back to it this evening and kept playing and researching. I have been using QL for 7+ years now and just today I learned that it does indeed have a long barrel model that allows you to add a customizable barrel friction to the model. After turning on this feature and experimenting with the friction number I got a model using 175gr Sierra Match King bullet pushed by 43.5gr of IMR 4064 to replicated the data that MDT showed in their video to within 3.5% or better across the span of all their data points. I used a screen grab from the video and a online program that lets you take an image of graph in turn it into a table of data. I put their converted data and my QuickLoads results into Excel and calculated the error at each length they tested. At the shortest length 19-inch QL was under by 3.48% and at 37 inch QL was over by 2.66%. From 62-69 inches it was off by less than 1%.

The only changes I made to the default 308 Win QuickLoad entry, to make this work, was to increase the case volume slightly (2.6%) and turn on the friction model, setting barrel friction to 3000 psi. Given the default 30 cal bore cross sectional area of .073641 in^2 that equates to about 220lbs of force to push the bullet through the barrel due to friction. Having lodge a bullet or two in a barrel a few times in my day that does not seem like an unreasonable number.

Double Naught Spy said:

Either way, the program is failing to control for all the variables, hence coming up with predictions that were not a great match for reality.

Click to expand...

Quickload only works well within relatively narrow parameters. It is quite bad with straight wall cartridges, for example. It seems 69" barrels are another failing.

Goosey said:

Quickload only works well within relatively narrow parameters. It is quite bad with straight wall cartridges, for example. It seems 69" barrels are another failing.

Click to expand...

Did you not see my last post? I got to within 3.5%. For a fluid and thermo simulation that is pretty darn good, and that is guessing what powder they are using. In reality they are probably using a canister powder that is not even in Quickloads or avalable to hobby reloaders but similar to 4064.

As for straight walled cartridges I used it to work up loads for my 450 Bushmaster before there was much published reloading data and did so safely and when the data did come out my selected powders and charge weights where within published load data. I have also used it to work up low recoil loads for my father's 45/70 though there is lots of good data to reference for that.

I have also used Quickloads to work up new loads for Webley revolvers in both 38/200 and 455 Webley using bullets not in most published data sets. Most modern data is for lighter modern bullets not the original bullets for either cartridge. I found a source of correct replica bullets (Matt's Bullets) and used Quickloads to work up safe loads that replicated the original velocities pretty well using modern propellants.

I could go on, that's why I came back to the problem last night and did what I did in my previous post. Quickloads is not perfect but it is pretty good when you use it right.

I really don't care that a barrel has to be over 60" long to achieve maximum velocity. I just would like to know whether the barrel length/twist is compatible with the rifle's optimal useage, whether it's deer hunting, varmint hunting, or target shooting, owner's desires, etc.

Folks get all worked up about cartridge efficiency - well this activity offers interesting evidence to that point.

Sure, it’s compelling that a 20” barrel achieves ~92% the velocity of a 69” barrel. But that ~8% difference in velocity corresponds to a 16% difference in Kinetic Energy. Effectively, that means a 20” barrel is only converting ~84% of the Potential Energy available which is being converted in the 69” tube. On a 50grn load in 308win, that means effectively 8grns worth of powder aren’t doing anything for us downrange.

Effectively, that means we pay over twice as much inefficiency “taxes” on a 20” barreled 308win than we paid in sales tax on the powder… And this is data from what is typically considered to be a very “efficient” combination - a 20” 308win… think of how much tax we’re paying in extra powder when we use something like 243win or 300win mag… “combustion efficiency” shouldn’t mean anything…

Let’s not factor in the required barrel gas pressure of my PTR-91s; roller- delayed blowbacks.

The bolt carrier doesn’t begin moving until the bullet leaves the muzzle.

Well, having empirical data from a single sample data collection come within +/- 3.5% of a theoretical model probably isn't that bad given the number of uncontrolled variables in this exercise.

BTW, note the quote from Feller in my tag line regarding theoretical models. (And I was trained as a pure mathematician, but became an empiricist after an immersion in the real world.)

The only “criticism” I might see in the model vs. the empirical data is their intersection and relative inflection point locations. The MDT data has an asymptotic approach to an upward bound, while the theoretical model predicts an inflection maxima with a retreat. Being close in modeling is great, but in many cases, “going the wrong direction” and lacking proximal inflection is more indicative of misfit than the raw error.

The model does predict what most folks would tend to assume - that the bullet achieved some maximum velocity and in fact then begins slowing down somewhere within an “almost reasonable” barrel length, but the empirical data shows the bullet remaining to slowly increase in speed for nearly 20” beyond the inflection maxima of the model prediction. In a way of speaking, being off my a small error for relatively large numbers might not be as indicative as the “all up isn’t the same as “up then down”.

This could be coincidental to the single experiment, or it could be indicative that some boundary parameters exist which diverge a “long barrel” from the behavior of a short barrel (normal sub-40” barrels). For example - maybe bore friction decreases after having been drawn for 40” of barrel. Maybe expanding gases aren’t cooling as quickly as modeled, such the in-bore pressure isn’t reduced as much as expected. Maybe, maybe, maybe.

I’ve been saying for a long time that “no practical barrel length will ever achieve full velocity and energy potential for any centerfire cartridge.” This test asserts that. So all of the BS about maximum potential from rounds like 300blk, x39, 6.8 and 6.5 Grendel, 223 and 308, etc that gets kicked around online can look to these results and extrapolate the implications. We pick barrel lengths on practicality for the task, not true maximum potential or powder utilization.

Varminterror said:

The only “criticism” I might see in the model vs. the empirical data is their intersection and relative inflection point locations. The MDT data has an asymptotic approach to an upward bound, while the theoretical model predicts an inflection maxima with a retreat. Being close in modeling is great, but in many cases, “going the wrong direction” and lacking proximal inflection is more indicative of misfit than the raw error.

The model does predict what most folks would tend to assume - that the bullet achieved some maximum velocity and in fact then begins slowing down somewhere within an “almost reasonable” barrel length, but the empirical data shows the bullet remaining to slowly increase in speed for nearly 20” beyond the inflection maxima of the model prediction. In a way of speaking, being off my a small error for relatively large numbers might not be as indicative as the “all up isn’t the same as “up then down”.

This could be coincidental to the single experiment, or it could be indicative that some boundary parameters exist which diverge a “long barrel” from the behavior of a short barrel (normal sub-40” barrels). For example - maybe bore friction decreases after having been drawn for 40” of barrel. Maybe expanding gases aren’t cooling as quickly as modeled, such the in-bore pressure isn’t reduced as much as expected. Maybe, maybe, maybe.

I’ve been saying for a long time that “no practical barrel length will ever achieve full velocity and energy potential for any centerfire cartridge.” This test asserts that. So all of the BS about maximum potential from rounds like 300blk, x39, 6.8 and 6.5 Grendel, 223 and 308, etc that gets kicked around online can look to these results and extrapolate the implications. We pick barrel lengths on practicality for the task, not true maximum potential or powder utilization.

Click to expand...

For the minimal time invested and working with several unknows and educated guesses (propellent being the big one) I think the model it pretty good. I am confident more better data and a little more time invested I could move the inflection point out further beyond the end of the 69 inch barrel and match the experimental data even better. With the friction model off the inflection point moves to infinity so it simply a matter of tweaking the friction and a few other parameters to match the real world. If nothing else it taught me a bit more about my tool and I have another dial to turn to help increase my match to real world data collect for my own reloading.

Varminterror said:

The only “criticism” I might see in the model vs. the empirical data is their intersection and relative inflection point locations. The MDT data has an asymptotic approach to an upward bound, while the theoretical model predicts an inflection maxima with a retreat. Being close in modeling is great, but in many cases, “going the wrong direction” and lacking proximal inflection is more indicative of misfit than the raw error.

The model does predict what most folks would tend to assume - that the bullet achieved some maximum velocity and in fact then begins slowing down somewhere within an “almost reasonable” barrel length, but the empirical data shows the bullet remaining to slowly increase in speed for nearly 20” beyond the inflection maxima of the model prediction. In a way of speaking, being off my a small error for relatively large numbers might not be as indicative as the “all up isn’t the same as “up then down”.

This could be coincidental to the single experiment, or it could be indicative that some boundary parameters exist which diverge a “long barrel” from the behavior of a short barrel (normal sub-40” barrels). For example - maybe bore friction decreases after having been drawn for 40” of barrel. Maybe expanding gases aren’t cooling as quickly as modeled, such the in-bore pressure isn’t reduced as much as expected. Maybe, maybe, maybe.

I’ve been saying for a long time that “no practical barrel length will ever achieve full velocity and energy potential for any centerfire cartridge.” This test asserts that. So all of the BS about maximum potential from rounds like 300blk, x39, 6.8 and 6.5 Grendel, 223 and 308, etc that gets kicked around online can look to these results and extrapolate the implications. We pick barrel lengths on practicality for the task, not true maximum potential or powder utilization.

Click to expand...

In the empirical data, how many shots were fired the establish the "muzzle velocity"?

In the empirical data, did the fact that the barrel was two barrels screwed together make a difference in the velocity behavior?

Does the air mas in front of the bullet make a difference in velocity?

Efficiency be damned! I’m here for performance!

As such, I’ll still be limited by my puny pugil-sticks. Since Arnold I am not, looks like normal barrel lengths for me.

So, even with its alleged foibles, the test proves we have the better part of the performance of the cartridge still within the lengths of barrel that don’t require bench pressing a Volkswagen.

Hmm, turns out we’ve been making them right all along…

someguy2800 said:

The prediction isn’t bad at any particular data point, however looking at the data trend on either end it’s deviating very rapidly such that it appears it would be significantly off if trend continues. Can you run your calculation down to 6 inches? I want to compare it to known data from a cut down test on a 223 I saw awhile back. I know a 223 and 308 are apples and oranges but I want to compare since that’s what I have data for.

Click to expand...

Yes it can go as short as you want to go. Share the data for the load you have in mind and we can see how it does. The more info the better I can tune Quickloads.

lysanderxiii said:

In the empirical data, how many shots were fired the establish the "muzzle velocity"?

In the empirical data, did the fact that the barrel was two barrels screwed together make a difference in the velocity behavior?

Does the air mas in front of the bullet make a difference in velocity?

Click to expand...

The air in front of the bullets makes a measurable difference. IIRC Quickloads uses standard atmospheric conditions but that you can vary those to match your test conditions. The changes in temperature and pressure commonly seen for most real world situations has minimal effect on the losses due to the mass of air in the barrel. It likely effects the propellant more and Quickloads can temperature compensate for that too.

I am shooting an 18" right now, and it seems pretty good for velocity and hunting.