My own personal opinion is to store the Onions in a cool dry place and to chuck them out if you start to see stalks growing.

j/k
Actually, steel can creep.
Steel will
not creep when subjected to repeated cyclic loading. It will fatigue. Fatigue is what happens when the micro-structure of the steel (atomic bonding and/or crystalline joints) breaks down after repeated cyclic loading. Fatigue can cause the steel to break even if there is no associated creep.
Creep occurs when the steel is subjected to a static load that is within a certain percentage of the maximum yield stress of that type of steel. This percentage is lowered when the component is subjected to cyclic loading or elevated temperatures.
Why do we associate creep with cyclic loading? Because, quite often, there is a certain amount of static load included in the cyclic load. This is almost unavoidable in many of the applications we subject steel to.
Why do we not see steel structures around us creeping? This is not a simple subject and I will only be touching the tip of the reasons.
The stress required is quite high for immediately visible creep, or the length of time required is quite great (decades). Sufficient stress to cause 10% strain will result in a permanent deformation for most steels, but you can see that steels will strain (elongate) under even low tension. However, this elongation (the amount of stress required to create it and the amount of resultant strain) is also highly dependent on the ductility of the metal. Most structures are designed to never go close to the 10% limit for safety reasons. Even then, the cyclic loading it is subjected to will promote creep. We still are unable to see it because it happens in small increments over long periods of time. Objective measurements over extended time frames are needed to expose this.
Not sure if all that helped any, but my own personal take on storage of spring-assisted-opening devices is to store them in the opened or semi-opened position, whichever places the least amount of force on the spring/torsion bar.