There's no reason to empty and reload the magazines. Springs will weaken from cycles of compression and decompression. You can leave springs compressed for a very long time before experiencing any significant weakening.
I'll agree that unloading and reloading a spring doesn't make "damaged" metal heal, but keeping SOME mags fully loaded can be harder on the spring than cycling the springs.
Spring life depends on with
how deeply the springs are compressed when routinely used, and
how long they're kept at that deeply compressed state. With some mag designs -- given good quality springs -- spring life may never be an issue, even if the springs is cycled many, many times or the mag is stored with the springs fully compressed (i.e., fully loaded). With others designs, the springs can degrade very quickly. If you never load a 15-round mag beyond 7-8 rounds, it'll probably last longer than the gun. With some mag designs, however, keeping the mag fully loaded for long periods could cause the spring to degrade. It depends on the mag's design and how the spring itself is routinely used.
- CZ, years ago, used the same springs for the 9mm 10-round mags as with the 15 or 16 round mags -- and spring life was noticeably different in those two types of mags.) I don't know if this is the case now. I've never had a 10-round mag spring fail, but saw a number of higher-cap mag springs degrade over time.
- Another example: the Rohrbagh R9, the smallest 9mm available a few years back, had a recoil spring with a suggested service life of 300 rounds/cycles,. That was later reduced to 250 cycles -- but the springs still worked for more than 250 rycles.. Those springs weren't kept compressed, but each shot compressed the spring to its design limit. Why the short life? Because those springs were intended to do the most work with the least material, to fit in a very small weapon, and the springs were an inexpensive and easily changed parts -- and were viewed as renewable resources. (I think Rohrbagh has retired sold the firm.)
Springs typically won't degrade from being kept depressed UNLESS the springs are in sub-compact or hi-cap mags, and if -- when fully loaded -- the springs are near or at their design limits.
Wolff Springs recommends downloading a round or two for long-term storage for those mags, but springs in standard 10-round mags (or 7-round 1911 mags) might outlast the gun. Those (7-round) 1911 springs are never pushed anywhere near their elastic limit, and tales of such mags being stored loaded for 50 years can be found -- still working like new. Early 8-round 1911 mags didn't last long at all, until they redesigned followers, etc.
Cycling doesn't really degrade a spring significantly UNLESS, when compressed (fully loaded and left that way), the spring is near it's design (aka "elastic") limit. And even then, the springs might outlast the shooter. But if you've got some 17 or 18 round mags, you might want to download a round or two if you store the mags loaded.
Note: If cycling alone degraded springs, many of the older cars and trucks on the highway would have quit running rights years ago -- as valve springs will typically cycle many tens or hundreds of millions of times but never fail. Why? Because those springs were designed so that when working never are compressed to or near their elastic limit.
On other forums, you'll find input and test results that show that cycling alone isn't always an issue. Discussions on this topic on those forums have included input from a Metallurgist or two, as well as engineers involved in the aerospace industry.
Here's a long-term test (still ongoing) by
JohnKSa on
The Firing Line that directly addresses the topic. John, an engineer, is a very knowledgeable handgun enthusiast, and also a long-time air gun shooter. The best air guns used in competition are frequently spring powered. The link:
https://thefiringline.com/forums/showthread.php?p=6005156#post6005156 You'll see that the springs tested immediately degrade, but continue to work despite some significant degradation in a few cases. A couple of years from now (i.e., the equivalent of long-term storage), the test results may show different results.