Actually, magazine spring life is not as simple as “load and leave - they will last forever”, or “cycling is the only thing that wears out magazine springs”.
In 1966 the Army did a five year study on this very subject. A report entitled “Evaluation of Pretreatment Process and Long Term Storage on Magazine Springs” was published with the results of this test.
There were two objectives, 1) determine the effect of various pretreatments on spring life and 2) determine the effects of long term storage of magazines stored at normal, hot, and cold temperatures. The parts we are concerned with are the life of a “normal” spring, as none of the pretreatments yielded cost effective improvements, if any improvement was seen at all, and the results of extended storage of fully loaded magazines versus unloaded magazines over 5 years of storage.
(Data from “Evaluation of Pretreatment Processes and Long Term Storage on Magazine Springs”.)
LONG TERM STORAGE:
In order for the “spring to not take a permanent set in service” the spring must never be loaded beyond specifications, which is generally not more than 50% of the free length. Most magazine designs have the springs over-compressed when loaded to maximum capacity. M16 magazines are such springs, when fully loaded, they are compressed to 14% of their free length (18% if you only go to 28 rounds).
The magazines, some loaded and some unloaded, were stored for 5 years, at set intervals a portion of the springs were removed from storage and had the free length and load-at-length measured, the ones removed from storage did not return to storage after testing. After approximately, 12 weeks the loaded springs had free lengths reduced below the original value, and lost strength, around 84% specification value. After 24 weeks, the load-at-length values were 76% there original value.
From 24 week the loaded springs showed a very slight decrease in both length and strength, dropping to about 72% after 2 years and stabilizing at that value after that. After 1-1/2 years, in the loaded condition, spring exhibited an average set of 11% below the initial free length. This stabilized for the rest of the test period.
This indicates that the strain on the spring has an effect on the loss of strength and length, but if the strain is constant the reduction reaches a stabilization point, and how low that point is depend on the initial strain. Unloaded magazines also showed a decrease in free length and load-at-length values, but stabilized in 24 weeks just inside the acceptable drawing limits.
The functional testing of the magazine after storage was to fire all of the loaded rounds in them and see if there were any magazine attributable malfunctions, there were none.
So, you do loose a measurable portion of your spring strength after long term storage, but it will stabilize after approximately 2 years. A spring with a strength of 72% of its new specification strength does not adversely effect function, but it has some impact on total life, as we know cycling of the spring will also lead to loss of free length and strength.
SPRING LIFE:
Spring were gymnasticated at a rate of 116 strokes per minute between the assembled height (unloaded length) and the maximum design compression (fully loaded length), with the free length and load-at-length checked a set intervals. This cycle rate was considered a bit excessive, but as the in service cycle rates would be slower this was the “worse-case”. By 1655 cycles, about 1/3 of the springs’ load-at-length value had dropped below the new spring limit, and by 12,000 cycles, all of the springs had dropped below 50% load-at-length (which is probably insufficient for proper function), or had become so distorted cycling was impossible.
DISCUSSION:
The service limit is not necessarily the same as the production limit, and according to this study, a spring with 72% of the production limit is capable of providing adequate performance. However, it would be incorrect to say that long term storage of fully loaded magazines “has no effect on the spring”. On average it took about 700 to 1000 cycles to reduce the spring strength values from “new” to 72%, so that much spring life has been lost. So, it depends on how often you load and unload your magazine, as to which is better for the life of the spring, storing loaded, or loading and unloading.
The up-shot is: It doesn’t make any noticeable difference as to the question of “Which is better for a spring, loaded or unloaded?”
EDIT: If you want to know when your spring is near the end of its life, when unloaded, the force need to depress the second round in the magazine should not be less than 2.5 pounds, preferably closer to 3 pounds.