Found the same thing regarding locks and keys from Dutch sites.
In the (supposed) Vicus area with the adjacent grave sites you find remains of boxes, keys and (pad)lock parts.
Have a theory that a grave site is generally about 500meters away from the center of a settlement.
Have not read yet that these box/locks remains have been found in a Dutch Castra setting.
It could be that the living quarters of the soldiers had close fitting wooden floors instead of a normally assumed sand/loam floor.
A wooden floor is easier to maintain and warmer in the Northern regions, and a fireplace can be on sand level without burning the planks.
Only if you really want to hide something under the floor, a plank could be taken out and a hole could be dug to put a box in.
For narrowing it down in a time frame, first of all I am looking for 1st century AD evidence.
This is for our museum where we want to built a display box(footlocker) which holds the gear of a soldier.
The building where this box will stand is reconstructed after a 1st century AD locally excavated floorplan.
The reconstructed house has a wooden floor made of planks and wooden roof shingles.
The evidence of floor planks and shingles come from other regional excavations but were used in this reconstruction as part of Experimental Archaeology in our museum.
So in fact this could go two ways:
1
The storing box used by the active soldier when in barracks.
2
The storing box used by the retired soldier living in the Vicus.
Could be the same box though.
And for those who reenact the ancient Roman soldier a historical correct alternative for the plastic and aluminum transport boxes.