THE SPRING BREAK!
The mainspring - why it can fail, and how to prevent it
With one exception, spring breakages are very rare, so let’s get that exception out of the way first; it occurs when a spring coil is deliberately heated and collapsed – or the end cut and flattened – and either overheated or, sin of sins, quenched, either of which alters the spring’s metallurgy and makes it brittle and almost guaranteed to break at some point. Other than that, I cannot remember the last time I had a mainspring break, and that’s in 50 years of shooting airguns, until a few days ago.
Around eight years ago, Mike Wright and I commissioned a small batch of mainsprings intended primarily for our HW95s, and they were the polar opposite of the springs that have been fashionable among tuners since the 1980s, when leading rifles like the Feinwerkbau Sport and Weihrauch HW77 had their original long, thin springs swapped for short stiff springs by tuners. Although not as soft, and nowhere near as long as the springs fitted by Feinwerkbau and Weihrauch, our batch of 12 mainsprings were a sort of halfway house between the manufacturers’ and tuners’ springs.
This was not our first batch of springs; we had started by blindly following fashion and ordering springs wound from 3.25mm diameter wire, which produced a stiff spring and a ‘quick’ shot sensation, but when I took my first tentative steps towards recording the recoil cycle, we realised that they wasted a lot of energy, manifested as piston bounce and consequent recoil surge. The key to increasing energy efficiency and reducing piston bounce and surge was to go for thinner section wire, and more length of wire in the form of more coils.
Our second batch of springs were 19.1mm coil diameter, 2.9mm wire diameter, and had 36 coils, which was longer than we thought we needed, but which allowed us to cut the springs to a suitable length for each calibre. At the full 36 coils, the springs gave around 63mm of preload in the HW95, and within a
whisker of the legal limit in my .177 HW95 – and would have been over in larger calibres – so the springs were shortened by degree until we were safely legal at just over 11 ft.lbs., which was with 34 coils in .177, and 33 in .20 and .22.
My six springs went into a variety of rifles to assess their suitability, all bar two being further reduced in length until they became too short to be of use; they were the lucky ones, because the two 34 coil springs were put through spring hell, being subjected to dry-firing, detonation and dieseling, all deliberate, and all in the name of science. Following a couple of years of such abuse, one was retired to my .177 HW95, in which it powered at least 18,000 shots, probably many more, and was then re-deployed, along with a lightweight piston and a piston extension that took the stroke down to 77.5mm, in my HW98, which was then loaned to a friend for use in HFT.
Following the return of the HW98 a year or so later, the piston and spring found their way back into my .177 HW95. Muzzle energy was down from that in the HW98, thanks to the 95’s shorter barrel, but at 10.6 ft.lbs., it was not really worth bothering about, so the rifle was put into commission.
Following a zero session necessitated by a change of scope, the pellet POI dropped over the course of perhaps two dozen shots by an inch or so at 25 yards, so I put the rifle over the chrono and found 50fps had been wiped
“using a low-stress spring means few if any warranty repairs”
off the muzzle velocity, and the most likely reason for such a sudden big drop in velocity was a broken mainspring.
On stripping the rifle, I found that preload had reduced to just 34mm, so I knew there was a spring breakage which, unusually, turned out to be near the front end of the spring. Breakages are more usually found about an inch from the rear end, and are caused by repeated wave displacement collisions, in which the rear of the spring is still expanding as a wave of compressed coils hits from the front, greatly stressing the point at which the wave fronts collide. That obviously was not the cause of this front end breakage, so what was?
When fitting the spring in my HW95, I’d measured the preload at 53mm, yet the other 34 coil spring, which has seen rather less use, gave 55mm, so the former appears to have lost 2mm in length through creep, compared to even a well-used spring, which from extended compression tests with other springs from the same batch, suggests the fracture could be due to a combination of extreme old age and a very hard life.
Another possibility is that, as the spring lost length through creep, it reached a point at which the compression stroke altered the timing of wave displacement collisions to coincide with the point of breakage, possibly in concert with the spring’s age and health. Let’s look at the mechanism that causes the wire to fracture.
What I think happens is this: When a coil spring is compressed, the entire length of the wire comes under torsional shear stress, and when the piston bounces and the expanded front end of the spring starts to compress, it also is subjected to torsional stress, which means it’s effectively in torsion in the opposite direction to that of the still expanding rear of the spring. This is fine until a wave from the expanding rear end of the wire and a wave from the compressing front end meet, and where the two waves meet, the two opposing forces place huge strain on the wire, and it can fracture.
So why doesn’t this happen every shot, why is it such rare occurrence, and why are some springs in some rifles and tunes more likely to fracture than others.
The springs in most factory air rifles are specified not to be highly stressed during the shot cycle, and that’s intentional because, just as the manufacturer uses enough grease internally for the rifle to last many years without attention – which most tuners deem to be way too much – using a low-stress spring means few if any warranty repairs, which all manufacturers wish to avoid at all costs, because warranty repairs eat into profits. In all but very aged factory standard spring airguns, spring fractures are very rare, and on the very few occasions they do occur, it’s normally due to a spring manufacturing fault.
“If you notice any of the symptoms of a broken mainspring, don’t use the rifle”
There are other possible causes of spring fracture with factory springs, but they need something drastic to go wrong, such as a detonation that causes ferocious piston bounce, or perhaps one too many dry fires. Short of pouring a highly volatile liquid down the transfer port, or serial dry firing, spring fractures in factory standard springers are extremely rare.
The springs used in modified airguns are another matter, because tuners, both amateur and professional, seem more concerned with effecting a change in the shot cycle that’s perceived as beneficial, or simply raising the muzzle energy, than they are with whether or not the spring, or any other modified component, might fail two, three or more years down the line. It’s worth noting that the mantra amongst many tuners is that ‘less is more’ when it comes to lubrication; again, the focus is more on what the newly modified rifles does now, rather than in the future when the grease runs out.
There’s no hard evidence to support this, but my gut feeling is that the more extreme the tune, the greater the likelihood of spring fracture. Many tuners today seem obsessed with creating minimal sight picture disruption, so they reduce recoil by shortening the compression stroke and/or reducing cylinder diameter, in either case necessitating higher peak cylinder temperatures and pressures, and asking ever more from the mainspring – something has to give, and it seems that something is often the mainspring.
Another tuning modification that seems to increase the risk of spring breakage is to overdo lightening the piston, because the lost mass has to be replaced by spring force, and the mass of the piston helps to slow piston bounce acceleration, which otherwise creates a fierce rearward displacement wave in the spring.
A few highly modified reduced cylinder bore rifles have reportedly suffered multiple spring breakages, accompanied by rear spring guide breakages, incredibly, even parting two-piece steel guides’ tubes from their bases, and the consensus seems to be that too little preload is allowing the rear end of the mainspring to lift away from the guide base, then hammering it back on the back block and causing the breakages. If that reasoning is correct, it’s because reduced cylinder bores mean reduced piston bores, which means reduced diameter mainsprings, and reduced diameter mainsprings are stiffer mainsprings, so you cannot use much in the way of preload.
The first sign that a mainspring has fractured often comes when you break the barrel or unlatch the underlever and they drop further than normal before there is any spring pressure. If you try to cock the rifle, the chances are that it will feel incredibly rough as the fractured ends grate against the piston side or spring guide, though that does not always happen, if the break is in the sections of spring surrounding the spring guide or top hat, and the two ends screw into one another, as happened with my spring breakage.
If you load the rifle without noticing anything amiss because the cut ends are intertwined, the next symptoms are a drastic softening of the shot cycle, possibly accompanied by a lot of spring noise, and a large downward shift in pellet point of impact.
If you notice any of the symptoms of a broken mainspring, don’t use the rifle until it has been stripped and checked, because the ends of a broken spring can cause very nasty internal damage.
My HW95 now has the remaining 34-coil spring inside, and is shooting very nicely indeed at 10.6 ft.lbs. with H&N Field Target Trophy, with the option to increase that, whilst reducing piston bounce, by adding preload, if I feel inclined. For the present, I think I’ll do some accuracy testing to make sure all is well, and then just enjoy shooting it. I
My next move will be to assess accuracy from the bench.
The springs were shortened to 34 coils for .177 (top) and 33 coils for .20 and .22 (below).
I cannot ever remember a spring breaking at the front end.
The fractured spring was the one I’d left coil bound for 1,000 hours back in 2012. It’s had a hard life, all right!
The wire fractures at the point it is subject to torsional shear in opposing directions.
The shortened piston stroke might have contributed to the spring failure.
With a static spring force on the piston of just 132 lbf, the cocking effort is low.
The surge amounts to 20% of the recoil; more preload would reduce that and increase muzzle energy.
A deliberate detonation compressed the spring by 50mm in seven milliseconds, easily enough to fracture a spring.