On a lighter note, a musical reminder of the importance of production quantity https://www.youtube.com/watch?v=jIZq-alEsWM 

I also cannot resist bringing up the problems the US had with the M60 even with working FG42 and MG34/42s right in front of them.

The story of the 30-06 M2 is a pitiful one, at least as recounted in Chapter III of Hatcher's Notebook (which can be downloaded for free).  It is not that difficult to design a cartridge with better long-range performance.  The bullet originally used in the 1906 version of the 30-06 did not have a particularly high BC and was found to be significantly outranged by the .303 MkVII and the French Balle D during WWI.  In the mid 1920's, the improved 174gr .30 caliber M1 cartridge was developed to make up the deficit.   It had more recoil than the 1906 cartridge but a much longer range.  In fact, it flew so far that it could not safely be used at the many military rifle ranges that had been designed around the earlier ammo.  Once this was discovered in about 1936, the Army decided to develop a "short range" version of the cartridge with a trajectory that matched the earlier one.  That became the M2 ball that was used in WWII and Korea.  In other words, the US had a cartridge that outperformed the M2, especially at long range, but decided not to use it.

The 7.62x51 developed after WWII essentially matches the old 1906 round, just in a shorter and lighter package.  Part of the reason is again that it was to be used on the same practice ranges.  

OK, so might have to eat some of my words...  To look at retained energy with a focus on long range (subsonic) MG use I used the JBM ballistics calculator with Emeric's G7 BCs for .303 Mk VII and VIII and .276 Pedersen. For the Pedersen I also used Emeric's G2 BC and both Emeric's 820 mps and Tony's 2,520 fps muzzle velocities.

In addition I tried 0.467 G1 and 0.274 G6 BCs for .303 Mk VII that I found here and in an old spreadsheet of Nathaniel's respectively. (The latter also had a .248 G7 BC for the .276 but that was so close to Emeric's value I didn't bother.)

I did it to 4km - the longest possible range - but reduced to 3.5km where that gave an error.

Unsurprisingly the Mk VIII is on top for energy. But the Mk VII G1 model catches up with it by 4km. That seems strange to me. I didn't think Mk VII even shot to 4km.

G7 just seems like a bad model for the flat-base Mk VII (especially subsonic) so I ignored that. But the G6 model (which seems a plausible shape) predicts lower energies than G7 past 1200m. 

For the Pedersen, the G2 models lose energy a lot more quickly than G7 when they go subsonic. But G2 is for a conical ogive so the G7 seems more plausible.

Both the high and low velocity Pedersen G7 beat the G6 .303 Mk VII for energy past 1.5km (and are basically equal to each other). So that is why I might have to eat my words - although if I'm trying to salvage some of my original "Pedersen too weak for GPMG" position I can still say that is a lot further out than Tony's GPCs beat 7.62 NATO, and the Mk VIII is still indisputably on top. Plus bullet volume for AP, tracer & incendiary yadda yadda yadda...

Workbook attached if anyone is interested. I'm probably pushing the calculator beyond what it's designed for, but it was interesting to play with.

Yes, I always found that story funny, also that the Balle D and sS Patrone had much better BCs and form factors than any standard bullets around now, especially 7.62 and 5.56 NATO. I know there are various reasons for that - higher muzzle velocity, less recoil, fit in shorter action, better performance in extreme environments, volume for AP & other fillings - and it makes relatively little difference at most combat ranges, but still...

Thanks for the heads up about Hatcher's.

BTW, It is helpful to download a text-only version of the book as well as the one made from photocopies.  The photocopied version did not deal well with pages that were intended to be read sideways (landscape format), so some images and text have been cropped out. 

nincomp said:

...in about 1936, the Army decided to develop a "short range" version of the cartridge with a trajectory that matched the earlier one.  That became the M2 ball that was used in WWII and Korea.

M2 Ball was replaced by M2 AP as the standard combat load about midway through WWII.

US could make problems out of thin air even when presented with a complete and well-established design, just google up T-24 machine gun, or a more recent Century Arms AK-47 ;)

hobbes154 said:

OK, so might have to eat some of my words...  To look at retained energy with a focus on long range (subsonic) MG use I used the JBM ballistics calculator with Emeric's G7 BCs for .303 Mk VII and VIII and .276 Pedersen. For the Pedersen I also used Emeric's G2 BC and both Emeric's 820 mps and Tony's 2,520 fps muzzle velocities.

Don't forget that in my revised TFW the .276 replaces the .303 only in infantry squad weapons: rifles, ARs and LMGs. The .303 (in Mk VIII form) remains in use in legacy Vickers and Browning (for AFVs) MGs.

I have been looking into the matter of bullet air drag for some years now. Therefore, I think I know why you dropped the G7 drag model. At long ranges it yields  depressingly low bullet velocities.

The sad fact is that even these low velocities are really on the optimistic(!) side. Because G7 does not take into account the tendency of many bullets (particularly bottailed) to lose stability at subsonic velocities. 

G1 gives much better results on paper, but it is simply the wrong drag model for modern pointed bullets. If you start a calculation with a G1 ballistic coefficient obtained at high velocity, you get much too optimistic velocities at lomng ranges. For modern bullets, G1 simply creates an illusion that is very different from sobering real range results.

It is also quite natural that you compare an image of the G6 model projectile with bullet shapes like Mk VII or M1906 and come to the conclusion that it should fit these bullets better than G7. Alas, this is not the case, when you try to fit published velocity measurements, as I did. As a matter of fact, the U.S. firing table FT30-J-1 of 1944 for the .30 AP M2 bullet is based on the G5 (yes, G five) drag model with a totally different projectile shape.  

For pointed bullets, for which no actual Radar drag models  are available, the British origin G7 remains to be the best allround drag model. This is in my experience totally indepedent of the bullet base shape. The drag models derived from pointed straight base projectile shape, like G6 or G8 (which is G7 without the boattail) are according to my results not superior to G7 for describing drag of bullets like Mk VII.

Some other observations: 

Do not trust long range velocities from before Radar, because of the practical impossibility to measure reliably beyond about 300 m. These figures are always computed. 

Do not trust long range times of flight. They were measured by stopwatch (average of several observers), looking for the impact on water or a sandy/muddy shore. In rare cases, bullet impact on a target was measured using electrical devises. Anyway, the usual resolution of 0.01 seconds in the tables is too coarse to infer drag data. 

The relatively most reliable column in firing tables (range tables) is in my view the elevation. But this is also often rounded too much and in any case only useful for analysis starting beyond 500 m. 

The German sS bullet might have a good reputation for its aerodynamics, but frankly, in my (German) opinion the published military data is overly optimistic. It is not at all supported by Radar measurements of this shape.  

Last not least, do not overlook that before ICAO adoption by NATO in 1969, every military body used its own air density and speed of sound. In particular, Spanish (CETME bullet!) and Swiss atmospheres are valid for thin air "high in the mountains", while Sweden is the opposite with a colder, more dense atmosphere (5 instead of 15 degrees).                     

I hope the above helps to avoid some of the not so obvious pitfalls of comparing ballistic data. 

JPeelen said...

For pointed bullets, for which no actual Radar drag models  are available, the British origin G7 remains to be the best allround drag model. This is in my experience totally indepedent of the bullet base shape. The drag models derived from pointed straight base projectile shape, like G6 or G8 (which is G7 without the boattail) are according to my results not superior to G7 for describing drag of bullets like Mk VII.

In that case the G7 model of Mk VII gives more energy at long range than the G6 model and the Pedersen G7 models (the G7 is between the G1 and G6 above about 1km, as can be seen in the workbook I posted).

JPeelen said...

G1 gives much better results on paper, but it is simply the wrong drag model for modern pointed bullets. If you start a calculation with a G1 ballistic coefficient obtained at high velocity, you get much too optimistic velocities at lomng ranges. For modern bullets, G1 simply creates an illusion that is very different from sobering real range results.

Thanks, I think I understand this now you've spelled it out!

• The relation between G1 and G7 is pretty stable around Mach 2 with G1 having about twice as much drag (hence the rule of thumb that you can just divide G1 by 2 to get G7). This is where most of the measurement is done (and all that is relevant to most hunters and short-medium range target shooters, hence why G1 is still used so much).
• However below about Mach 1.5 the relative drag of the G1 model drops sharply and stays lower than 2 (but still above 1 i.e. the G1 reference projectile is absolutely draggier than G7 at all velocities) except for a brief spike just under Mach 1 (and under Mach 0.2 but that is hardly relevant). So if you fit a G1 curve to a G7 shaped bullet using data from the "flat" zone around Mach 2, the G1 model will then under predict the drag at most slower velocities and thus give overly optimistic results.
• By contrast the G6 is pretty similar to G7 at all supersonic velocities but much draggier at subsonic velocities (due to the lack of a boat tail?). So it will give worse results than G7 at long range when fitted to short range data.

Some good resources here http://www.jbmballistics.com/ballistics/downloads/downloads.shtml for anyone else trying to learn this stuff. (Sorry I was trying to upload some graphs but the forum wasn't liking it.)