The barrel length in comparison with the calibre is obviously an important factor in muzzle velocity. In cannon calibres, this is expressed as the "calibre length", which is simply the length of the barrel divided by the calibre. For example, the current Bofors 40mm gun has a barrel 2.8m long, and therefore has a calibre length of 70, expressed as L/70. The WW2 Bofors had a less powerful cartridge and needed a calibre length of only L/56.

RECOIL

The recoil force generated by firing a gun has two components; the momentum of the projectile, and of the escaping gas. Projectile momentum is easy to calculate; just multiply the projectile weight by its muzzle velocity (so a cartridge firing a 10g bullet at 1,000 m/s should have the same bullet momentum as one firing a 20g bullet at 500 m/s). Note that this is a different calculation from muzzle energy, as bullet weight and muzzle velocity are of equal value. This explains why in different bullet-weight loadings of the same cartridge which generate the same muzzle energy, the heavy bullet loading will produce heavier recoil.

The recoil caused by the escaping gas - a kind of "rocket effect" - is much more difficult to calculate because it depends on the relationship between the burning characteristics of the propellant and the length of the barrel. If you assume two rifles firing the same cartridge, one with a barrel of optimum length and the other with a much shorter barrel, the optimum length one will produce the higher muzzle velocity and therefore the greater recoil through bullet momentum. However, in the short-barrel gun the gas will be at a higher pressure when the bullet leaves the muzzle, and will therefore expand more violently, causing more muzzle blast and flash and generating a stronger "rocket effect". In this case, a higher proportion of the recoil will be generated by the expanding gas than with the optimum barrel.

For this reason, there is no simple ratio which will tell you exactly what proportion of the recoil is generated by the escaping gas as opposed to the projectile. However, a good approximation can be made, based on the weight multiplied by the velocity of the propellant compared with the weight multiplied by the velocity of the projectile. In a large number of empirical tests, the velocity of the gas escaping from the muzzle of a rifle has been determined to be 1,200 m/s (4,000 fps) plus or minus 10%. In larger high-velocity military weapons, which can operate at very high pressures and velocities, the escaping gas velocity may be significantly higher.

It is therefore fairly simple to work out what proportion of the recoil impulse is generated by the escaping gas. Take for example the 7.62x51 NATO military rifle/MG cartridge in M80 ball loading. This uses 3.0g (46 grains) of propellant to fire a 9.5g (146 grain) bullet at a muzzle velocity of 840 m/s (2,750 fps). The calculation goes like this (the units of measurement don't matter as long as they are used consistently):

Bullet momentum: 9.5 x 840 = 8,000 (rounded). Propellant momentum: 3.0 x 1,200 = 3,600. So the total recoil momentum is 8,000 + 3,600 = 11,600, of which the gas produces 3,600 / 11,600 x 100 = 31%

This figure of around 30% is typical for a medium-velocity rifle cartridge. In a higher-velocity rifle like the 5.56mm NATO it is in the region of 35-40%. In handguns it is much lower, in the region of 10-15%, although in the big Magnums it can exceed 20%. In powerful military cannon it can be as high as 50%.

The only way of reducing the recoil force generated by a cartridge, while maintaining the muzzle energy, is to reduce the effect of the escaping gas by diverting some of it, either to one side or (preferably) to the rear. A device to achieve this is known as a muzzle brake. The extent to which a muzzle brake can reduce recoil obviously depends upon the proportion of the recoil impulse generated by the propellant gas - it gives the greatest benefit in very powerful, high-velocity weapons. One text on military cannon states that an efficient muzzle brake can reduce the recoil impulse by up to 30%. Higher figures are possible, but only by using brakes which are so large that they would be impractical. A disadvantage of a muzzle brake is that the rearwards-deflected gas greatly increases the muzzle blast and noise perceived by the firer, and may also kick up dust, revealing the weapon's position and affecting the user's visibility.

A type of muzzle brake is the compensator. This deflects some of the muzzle blast upwards in order to counteract the tendency for the gun barrel of a hand-held weapon to point upwards as a result of recoil. It is therefore mainly found in powerful handguns, or in automatic weapons like sub-machine guns.

Of course, a recoilless gun deflects most of the gas directly behind the weapon, so in this case the "rocket effect" more or less balances the projectile momentum. However, this requires the use of several times as much propellant as with a conventional gun of the same muzzle energy, so the ammunition is bulky and expensive.

So far I have only discussed the "recoil force" as opposed to the recoil experienced. This is because the recoil experienced depends on the weapon, and on the mounting. In the case of a rifle or handgun, the weight of the weapon has a significant effect. Momentum works both ways, equally (you can't defeat Newton's law of equal and opposite reactions!), so the rearwards momentum of the gun matches the forwards momentum of the bullet plus the expanding gas. It therefore follows that the heavier the weapon, the more slowly it will move backwards under recoil, giving a smooth push rather than the sharper kick of a lighter weapon firing the same ammunition. The recoil momentum experienced by the firer is the same, but it is delivered in a different way; the lighter weapon, recoiling more quickly, has the same momentum but higher energy and is perceived to "kick harder".

For instance, let's take our 7.62x51 NATO cartridge and work out the recoil energy it would generate in different rifles. As we have seen, the cartridge generates a total recoil impulse of 11,600 (grams per metre per second). If a rifle weighs 4.0 kg (8.8.lbs) or 4,000g, it will therefore be pushed back at 11,600 / 4,000 = 2.9 metres per second. 4.0 kg at 2.9 m/s = 17 joules (whereas the cartridge develops 3,350 joules - which shows the importance of velocity in calculating energy). If a lightweight rifle of only 3.0 kg (6.6 lbs) is used to fire the same cartridge, it will be pushed back at 11,600 / 3,000 = 3.9 m/s, producing 23 joules muzzle energy - an increase in recoil energy of 35%, even though the recoil momentum is the same.

To translate this into practical consequences, the 7.62x51 NATO cartridge generates only about double the recoil momentum of the 5.62x45 NATO, but in rifles of the same weight this translates to double the rifle recoil speed, therefore four times the recoil energy. An intermediate military rifle cartridge like the Russian 7.62x39 used in the famous Kalashnikov AK 47 rifle fits about half-way between the 7.62x51 and 5.56x45 in recoil generated; regarded as around the maximum for (just about) controllable automatic fire in a military rifle.

Self-loading weapons, whether recoil or gas-operated, tend to reduce the perceived recoil because some of the energy is used to drive the reloading mechanism. The shape of the weapon can also affect the perceived recoil; in rifles, a "straight" stock which directs the recoil impulse into the shoulder causes less barrel jump than a dropped stock, which has the thrust line over the shoulder. In handguns, a revolver has a higher thrust line than other types and this may also cause more perceived recoil (although in the old-fashioned type of "western" revolver, the grip tends to rotate in the hand, absorbing some of the kick at the cost of considerable muzzle jump).

In heavier military guns which are fitted to mountings, the nature of the mounting can make quite a difference. In all but the lightest weapons, the mounting allows the gun to recoil backwards between shots (there is commonly some kind of buffer). Note that this doesn't reduce the recoil force - only a muzzle brake can do that - it merely reduces the peak recoil blow by spreading out the recoil force over a longer period and thereby puts less strain on the mounting (or to put it another way, allows a lighter mounting to be used). Detail design can make a big difference; the US Edgewater mounting used on aircraft HMG and cannon mountings in WW2 substantially reduced the peak recoil blow.

An even more effective way of doing this is with a differential recoil or floating firing mounting, in which the gun is held back in the full recoil position before firing. On firing, the gun runs forwards "into battery" and the weapon is fired just before it gets there. The recoil force therefore has to overcome the forwards momentum of the gun before it can start pushing it back again. This is a highly effective way of smoothing out the recoil pulses and is commonly used in modern automatic AA cannon.

Some automatic gun mechanisms have a greater recoil-smoothing effect than others. One of them is the long-recoil type, in which the barrel recoils a considerable distance between shots. This makes for a slow-firing gun and is mainly used in large-calibre weapons. It was commonly used in large aircraft cannon in WW2, such as the US 37mm and the British 40mm and 57mm guns - the big 57mm Molins had a peak recoil blow similar to that of the 20mm Hispano (although the total recoil thrust was obviously much greater). The other is the advanced primer ignition blowback type, as pioneered by Becker in WW1 and much used by Oerlikon and similar weapons in WW2. In this, the gun fires from an open bolt, which means that when the firing button is pressed, the bolt moves forwards and chambers a cartridge before firing. The key point of the API blowback is that the cartridge is fired while it is still moving forwards, so there is a kind of internal differential recoil effect.

There is a popular fallacy that firing a large-calibre cannon in an aircraft (such as the 75mm M4 in the B-25) had a drastic effect on aircraft speed - or even briefly brought it to a stop! A simple comparison of speeds and weights between the shell and the aircraft will show that the aircraft had about 200 times the momentum of the shell. Firing several shots in quick succession would slow the plane a little, but not by more than about 5%.

EXTERNAL BALLISTICS

Just two key factors determine the external ballistics of a projectile; the muzzle velocity and the ballistic coefficient. The ballistic coefficient is significant because it determines the rate at which the projectile slows down, and in conjunction with the muzzle velocity this decides the maximum range (at any given elevation) and the time of flight to any particular distance. The time of flight in turn decides the amount by which the projectile drops downwards as this happens at a constant rate due to gravity. The curved path of the projectile which results from the muzzle velocity, the ballistic coefficient and gravity drop is called the trajectory.

In most types of long-range shooting (whether by rifles or large cannon) a short time of flight is considered desirable because it maximizes the hit probability by reducing the time of flight and flattening the trajectory. It also results in the projectile striking the target at a high velocity and therefore with greater effect. The main exception is when artillery fires in the "upper register" (above 45 degrees elevation) to achieve plunging fire.

The advantages of a high muzzle velocity in reducing the time of flight are self-evident. So are the disadvantages: more propellant is required, the barrel will need to be longer, the gun will be heavier and (in the case of a mounted weapon) so will be the mounting to cope with the greater recoil. In an automatic weapon, the rate of fire is also usually lower. As we have seen, there is also a practical limit to how high the velocity of any given projectile can be pushed. To make the most of the muzzle velocity, we need to achieve a high ballistic coefficient.

There are two elements which decide the ballistic coefficient (BC); the sectional density (SD) and the form factor (FF). The SD is a simple calculation as it is the ratio between calibre and projectile weight. The formula is:

For metric measurements: multiply the projectile weight in grams by 1.422, then divide the result by the square of the calibre in millimetres. So for a 12.7mm bullet weighing 40 grams: (40x1.422)/(12.7x12.7) = an SD of 0.353

For Imperial measurements: divide the projectile weight in pounds by the square of the calibre in inches (if bullet weights are in grains, divide the result by 7,000).

The higher the SD figure, the better the velocity retention (assuming equal form factors).

What the SD measures is the weight (or momentum, when moving) behind every square millimetre of the projectile calibre (i.e. the cross-sectional area of the projectile). If projectiles were solid cylinders then for a given SD figure they would all be the same length regardless of their calibre. In practice, of course, the length varies with the calibre; a 40mm projectile will be about twice the length of a 20mm, and will therefore have about double the SD figure. This explains why artillery shells travel much further than rifle bullets, no matter how fast or streamlined. Other things being equal, the bigger the calibre, the longer the range and the shorter the flight time to any given range.

Other things are of course far from equal, which is where the form factor comes in. The FF measures the aerodynamic efficiency of the projectile's shape, and is much more complicated to calculate; without access to manufacturers' data, only approximate estimates can be made. It is obvious that a projectile with a pointed nose will have much less air resistance than a simple cylinder, and it will therefore have a better FF, but problems arise when you try to become more specific.

The first problem is that the FF is different at subsonic and supersonic velocities, because shapes which work best at subsonic speeds are not the best at supersonic velocities. At subsonic speeds, the drag caused by the low-pressure area created at the back or base of the projectile is significant, and major reductions in drag can be made by tapering this to some extent (boat-tailing). At supersonic speeds, it is the nose shape that is critical; finely pointed noses are needed, but the back end doesn't matter so much. Some taper towards the base is useful, but the optimum taper angle is different from that at subsonic velocities. The benefit of boat-tailing at very long range can be demonstrated by two .30-06 bullets, both weighing 180 grains (11.7g) and fired at 2,700 fps (823 m/s). At sea level, the flat-based bullet will travel a maximum of 3,800m, the boat-tail 5,200m.

A further factor affecting military projectiles is the addition of tracer elements. These generate gas which helps to fill the low-pressure area at the base, reducing drag. This gives them a different trajectory by comparison with non-tracer rounds, not helped by the fact that as the tracer burns up the weight of the projectile reduces, thereby worsening its sectional density. Tracers can therefore never achieve a perfect match with other projectiles and can only ever be an approximate guide to their trajectory.

Putting all of this together, the most aerodynamically sophisticated projectiles in use today are the long-range artillery shells known as ERFBBB (extended-range full-bore base-bleed). These have a long, finely pointed nose to work well at their initial supersonic speeds, and a tapered base filled with a "base bleed" burning chemical which essentially does the same aerodynamic job as a tracer. Furthermore, the nose is so pointed that only the base of the shell is in contact with the barrel, so small streamlined stubs are fitted part way up the shell to keep it centred in the bore. It was discovered that these generate some aerodynamic lift, like tiny wings, and extend the range still further. The advantage of all of this can be seen in the range improvement over a conventional 155mm HE shell; in a 39 calibre barrel, the standard M107 shell has a range of 18,100m, the ERFB shell 25,500m and the ERFBBB 32,400m. Furthermore, unlike rocket-assisted or sub-calibre shells, there is no penalty in effectiveness as they carry at least as much HE (in fact, the South African 155mm M57 ERFB shell contains 30% more HE than the standard M107 shell).

It is possible to obtain some idea of typical FFs by comparing manufacturers' BC data with the calculated SDs for the same projectiles. In the case of small arms bullets, this provides the following approximate FFs (this figure should be multiplied by the SD to give the BC):

Comparing the BCs with ballistic tables for the ammunition gives the following results. These figures show the approximate percentage velocity loss over 100m for supersonic projectiles (900 m/s) with the following BCs:

Another type of Form Factor is traditionally used for artillery - and especially naval - shells. This is the "caliber radius head" (CRH) which measures how pointed the nose is. To give an example, if the curve of a shell nose is the same as that of a circle with a radius of 500mm in diameter, and the calibre is 100mm, then the shell has a CRH of 5. The higher the CRH, the better the FF.

In calculating SD and BC, it should be noted that the notional cartridge calibre is not necessarily the same as the actual projectile diameter, particularly with small arms. The bore diameter (ie the inside diameter of the barrel ignoring any rifling grooves) may be used instead, or some notional figure. The following list of bullet calibres is followed by some of the common cartridge designations of that calibre (there are many other calibres and cartridges�):

An important aspect of external ballistics is the stability of the projectile. If an ordinary bullet or cannon shell were fired from a smoothbored gun, it would probably start to tumble, ruining its aerodynamics and accuracy. Old-fashioned smoothbore muskets fired round balls, so tumbling wasn't a problem, but such a shape has a poor BC. Even round balls fired from smoothbores tend to drift off target as the range increases anyway, as they will not be entirely symmetrical. The answer was found to be to cause the bullet to spin rapidly by cutting spiral grooves into the barrel, called rifling. This evens out any asymmetries and keeps pointed bullets heading point-first.

Initially, bullets and shells were provided with studs to fit into the rifling but these were slow to load. An alternative approach was to make a polygonal-section barrel with shells manufactured to fit. Subsequently, rifles were provided with Mini� type bullets which had a hollow base, designed to expand under the pressure of firing and "take" the rifling. Modern rifle, pistol and heavy machine gun (HMG) bullets are given a metal jacket (usually cupro-nickel) which has a slightly larger diameter than the bore of the gun. It is therefore squeezed into the rifling grooves on firing, which leaves characteristic angled grooves engraved into the bullet. Modern cannon shells are usually made of steel which is too hard for this, so they are given a driving band near the base of the projectile, which is larger in diameter than the shell and is gripped by the rifling. The driving bands are traditionally copper but this is too soft for modern high-velocity cannon which normally use soft steel driving bands instead. The 30mm projectile for the US GAU-8/A cannon are unusual in using plastic driving bands.

With large-calibre, high-velocity cannon there is some risk of the shock of impact with the rifling "stripping" the driving band. To combat this, some weapons have progressive rifling, in which the rifling grooves start out parallel then gradually increase in twist down the barrel. In a return to earlier "studded" shell principles, the WW1 Paris Guns had pre-engraved rifling bands to minimise the friction and the risk of stripping the band.

There is also some concern with cannon that the driving bands, which stick out from the shell body and tend to be rather chewed up by the rifling, have a poor effect on aerodynamics. One solution to this problem is to include a smooth (unrifled) final section of the barrel at the same diameter as the shell, which squeezes the driving band flat against the projectile. This is known as a "Probertised" barrel (technically "RD Rifling") after Probert, the British inventor, and was used in the high-velocity 3.7" Mk VI AA gun in World War 2.

Rifling permits a high degree of accuracy over the maximum range of a weapon. There is a relationship between the rifling twist (the angle of the rifling to the barrel) and the length of the projectile. For a given calibre, the longer (ie heavier) the projectile, the steeper the twist has to be in order to stabilise it. Clearly, with a particular rifling twist some light projectiles will be very stable, some medium-weight ones marginally stable and some heavy ones not stabilised at all. This can have consequences for the terminal as well as the external ballistics, as we shall see. As projectiles become steadily longer so rifling can no longer cope, and long, thin projectiles such as APFSDS (armour-piercing fin-stabilised discarding-sabot) require fins, like arrows, at the back to keep them pointing, nose-first, in the right direction. Such projectiles are disturbed by rifling and work better from an (almost) smoothbore barrel (some degree of spin is considered useful in effecting clean sabot separation). HEAT (high-explosive anti-tank, also known as hollow-charge) projectiles also work best when not spun, so these two types of munitions have become associated with smoothbore barrels, almost exclusively fitted to AFVs. Recently, automatic cannon with rifled barrels have taken to using APFSDS, but the spinning causes the projectiles to yaw (fail to point straight ahead) for several hundred metres, so they only become fully effective at 400+m.

The optimum elevation for achieving the maximum range depends on the range capability. Large caliber, high-velocity artillery (e.g. the WW1 German Paris Guns) achieved their maximum range of around 75 miles at an elevation of 55 degrees, because aerodynamic drag reduces along with air pressure so the sooner the shell gets up into the thin upper air the further it will travel. Rifle bullets are restricted to the lower atmosphere and their optimum elevation is about 30-35 degrees. For the same reason, an aircraft gun will have a much longer effective range at high altitude than in the thick air at ground level.

TERMINAL BALLISTICS

There are two different aspects to this; the effect of projectile strike against soft targets (animals or people) and the effect against armour.

First, against soft targets (the squeamish have permission to duck this section!). A military (i.e. fully jacketed, pointed, non-expanding) rifle bullet will be destabilised when hitting a soft target and will tumble. This is because its shape means that the centre of gravity of the bullet is towards the rear so it naturally prefers to fly base-first. Spinning the bullet by means of the rifling keeps the bullet flying point-first through the air, but flesh is about 400 times denser than air so spinning is no longer enough; the bullet destabilises and turns over to travel base-first, a process known as tumbling.  In so doing it obviously inflicts a far more serious wound than if it carried on flying straight through the body. Incidentally, bullets designed for penetrating heavy game animals like elephant - which need to penetrate very deeply and must therefore not tumble - have long, parallel sides and blunt round noses, just like early military rifle bullets.

Not all bullets tumble at the same rate. Other things being equal, small bullets will tumble more quickly than large ones, but the design of the bullet is also important; some visibly identical bullets will tumble at different speeds, generally depending on the internal construction. For example, the Yugoslavian bullet for the 7.62x39 has a lead core and has been found in tests to tumble much more quickly than the Russian steel-cored bullet in the same cartridge. Readiness to tumble may also be affected by how well-stabilised the bullet has been by the rifling. A well-stabilised bullet may pass straight through the target without having time to tumble. The original US Army .223 (5.56mm) 55 grain (3.56g) M193 bullet was notorious for rapid tumbling, but the current NATO 62 grain (4.02g) SS109/M855 bullet is fired from rifles with a much steeper rifling twist (1 turn in 7 inches - 18cm - instead of 1 in 12 - 30cm) and is more stable, to the benefit of long-range accuracy and penetration but at the cost of a slightly slower rate of tumble on impact. Various tricks have been used to increase the probability of a bullet tumbling; the British .303 Mk VII bullet had a lightweight tip filler with the weight concentrated towards the rear of the bullet, and the current Russian 5.45mm rifle bullet has a hollow tip.

If a bullet has a relatively weak jacket, the stresses of tumbling may cause it to break apart while it is travelling sideways through flesh - a process known as fragmentation - which further increases the wounding effect. Most 5.56x45 military bullets fragment, although they have to be travelling at high velocity to do so. This limits their maximum effectiveness to fairly short range, particularly from short-barrelled carbines which have a lower muzzle velocity. Most 7.62x51 NATO bullets do not fragment, although the German one does - by accident rather than design. Fragmentation is not an official requirement for any military bullets; if it were, there might be some legal challenge over the international prohibition on bullets designed to cause unnecessary suffering. The noses of hunting rifle bullets (and many commercial handgun bullets) are designed to expand on impact, which greatly increases the size of the wound channel.  Such bullets are illegal for military use.

It is often claimed by hunters that as the striking velocity of the bullet increases beyond about 700 m/s (2,300 fps), so hydrostatic shock begins to appear, with the effect that animals drop dead much more dramatically than if hit in the same place with a low-velocity bullet. However, this effect does not seem to be replicated in people; there are many cases of soldiers continuing to fight for some time despite receiving severe (and ultimately fatal) wounds from high-velocity rifle bullets. Furthermore, serious shock effects are only likely if the bullet exceeds the speed of sound in flesh, which is around 1,500 m/s (4,900 fps), but even this has been disputed.

This brings us onto the vexed question of stopping power, about which it is impossible to make any pronouncements without stimulating fierce arguments. Stopping power may be defined as the ability of a particular weapon to immediately disable an opponent so he can take no further part in the fighting. It is not the same as lethality; quite low-powered weapons can be lethal, but considerably more power is normally required to achieve reliable stopping power. Incidentally, this shows that the notion that modern military rifle bullets are meant to wound rather than kill is a myth; if it is powerful enough to disable, it is more than powerful enough to kill.

Clearly, bullet placement is vital to achieving effective stopping power; it is much more effective to hit an immediately vital area with a low-powered weapon than to inflict a minor wound with a high-powered one. Also, the psychological state of the target has a considerable effect. Someone who is relaxed, or frightened, may be put out of the fight by a minor wound, someone who is highly charged with aggression will require far more power to stop them, and yet another person high on drugs may continue fighting despite suffering the most appalling wounds.

Stopping power is simplest to define with pistols, which have too low a velocity for hydrostatic shock to be a factor. The classic formula, named after the American Julius Hatcher, is calculated by multiplying the bullet weight by the muzzle velocity and then by the square of the calibre. The result is then multiplied by a form factor, similar in principle to that used for calculating the BC, except that in this case, the blunter the bullet shape the more effective it is. It will immediately be seen that calibre is the most important factor, and indeed large calibre pistols such as the .45" have always had a good reputation for stopping power. It should be noted that even the most powerful handgun or rifle will not physically knock someone down; if they were that powerful, Newton's law would require the firer to be thrown backwards with equal force. The recent spread of body armour has changed the perceptions of desirable pistol ballistics to some extent, as a high-velocity small-calibre bullet will punch through body armour which will easily stop a large-calibre, low-velocity bullet.

This brings us onto the subject of penetration. This is not just to do with military armour, but also against tough animals like elephants or water buffalo. As already indicated, early big game hunters found that the most important characteristics of a bullet against such tough game were that it should be round-nosed, strongly built, and have a good SD. A pointed bullet would not follow a straight path through a mass of bone, and one with too high a velocity also often followed an erratic path. Amazingly, one of the most successful early elephant guns (albeit only in very skilled hands) was the little 6.5mm Mannlicher. Why? Because its very long, 160 grain (10.4g), round-nosed bullet and its moderate velocity allowed it to penetrate remarkable thicknesses of bone - but it was only effective with a precise aim.

The subject of the penetration of armour is highly technical and complex. Furthermore, different national definitions of penetration and different types and qualities of armour used to test projectiles against make comparisons difficult. However, certain broad principles still hold generally true. As with elephant guns, a high SD is desirable and so (more surprisingly) is a blunt nose, although this is often concealed by a pointed ballistic cap or windshield. However, one major difference is that the higher the striking velocity the better, at least until the velocities are so high that the projectile is more likely to shatter than penetrate. For hardened steel penetrators, this happens at velocities much over 1,000 m/s.

SUB-CALIBRE PROJECTILES

This term is used to describe projectiles which are smaller than the calibre of the gun they are fired from. Nowadays this normally means APDS or APFSDS, but I will also deal with two related developments; APCR and squeezebore guns.

As we have seen, a large caliber will permit more energy to be generated than a small one. On the other hand, for a given projectile weight a smaller caliber will have a higher SD and therefore better long-range and AP performances. Designers have therefore tried different ways of combining the advantages of the two.

The simplest type was known to the British in WW2 as APCR (armour piercing, composite rigid - I have seen an early document which referred to this as "composite rigid armour piercing" but they presumably thought better of the acronym�), to the Americans as HVAP (high velocity armour piercing) and to the Germans as Hartkernmunition or Pzgr.40. However, it was probably the French who fielded it first, in the M1935 loading for the little 37x94R round still being used in some tank guns (there's a picture of one, plus sub-calibre projectiles, in the http://www.quarry.nildram.co.uk/tankammo.html - photo gallery on this website). It is nowadays commonly known as APHC, for armour piercing hard core, and is mainly used in MGs, HMGs and small-calibre cannon.

As the names suggest, this consists of a lightweight projectile (normally mainly aluminium) with a hard, small calibre core (normally tungsten alloy, which is heavier and harder than steel). The light projectile in a large-calibre gun gives a high muzzle velocity but when it strikes the target, only the hard core penetrates so it can go through much more armour than a full-calibre projectile of the same weight. The only disadvantage is that the light projectile has a low SD and therefore slows down more quickly than a normal projectile, steadily losing its penetration advantage as the range increases. To overcome this problem, later versions tended to be little if any lighter than a standard shell, thereby trading some of their short-range penetration for better long-range effectiveness. A modern example of this is the 30mm API used in the GAU-8/A cannon fitted to the A-10 aircraft; this is also unusual in having a depleted uranium core.

Another approach to achieving the best of both worlds was the squeezebore gun, of which there were two basic types; the Gerlich and the Littlejohn. In both, a projectile fitted with flanges to fit a large caliber barrel was squeezed down to a smaller caliber before it left the muzzle. The difference between them was that the Gerlich guns had tapered barrels whereas the Littlejohns had normal barrels with a tapered attachment fitted to the muzzle, in principle not unlike a shotgun choke. These worked very well and both saw limited service in WW2, the Gerlich in some German AT guns and the Littlejohn (named after the Czech designer, Janecek, which translates as little John) in some Allied armoured car and light tank guns. Their main problem, apart from the cost of the tungsten-cored ammo (and in the case of the Gerlich, the expensive barrel manufacturing) was that they could only fire this type of ammunition; they could not fire full-calibre HE shells. For this reason, they lost favour as soon as a better solution emerged.

The better solution was APDS, for armour piercing discarding sabot. This was like the APCR shell, except that the light alloy sabot (French for shoe) was designed to fall away from the small-calibre penetrator as soon as the projectile left the muzzle. This therefore combined the advantages of a large caliber for maximum energy with a small caliber for best flight and penetration performance, and allowed conventional ammunition to be fired from the same gun. It was initially designed in France before WW2, but was then developed in Canada and the UK, being issued for British 6pdr and 17pdr guns from mid-1944 onwards.

Apart from the cost and availability of the tungsten (always an issue in WW2) the only problem was that early version were very inaccurate because the flight of the projectile was disturbed by sabot separation. The British carried on using conventional AP tank ammunition into the 1950s, and APDS only really became supreme with the British 105mm tank gun of the late 1950s, which became the NATO standard for many years.

The replacement for APDS in tank guns (it is still used in small caliber cannon and HMGs) was APFSDS, which takes the design principles to their logical conclusion in producing the longest and thinnest practical projectile. The problem, as we have seen, is that achieving stability by spinning doesn't work with such long projectiles so they have to be fin stabilised. Modern manufacturing quality means that a high degree of accuracy can be achieved, and APFSDS seems likely to remain the supreme penetrator until conventional guns are replaced by different technologies.