(a) When an apparatus is capable of performing a function, it doesn’t necessarily mean (and typically doesn't mean) that the apparatus autonomously performs the function [unless you define it to be such], it typically means that the apparatus is capable of performing the function in response to the action of an agent. The agent can be a user (typically a person, but could be an animal such as a dog), but doesn’t have to be.
(b) Consider a container with a removable lid. The container doesn’t autonomously remove its own lid; a person can remove the lid. Similarly, consider a chair with a foldable back and foldable legs. The chair doesn’t autonomously fold its back and legs; a person can fold the back and legs.
(c) Now consider your instance of a door. First assume a door mounted to a frame with simple hinges. The door is capable of movement when sufficient torque is applied to it. The sufficient torque could be applied intentionally by a person or an animal; or, it could be applied unintentionally by a gust of wind or or unintentionally by a runaway shopping cart or unintentionally by shock and vibration from passing traffic (especially if the hinges are not spring loaded).
The next step is to add a latch operated by a door knob. The door is capable of movement if the door knob is intentionally turned and sufficient torque is applied to the door. This prevents the door from being opened by most animals and unintentional agents.
The next step is to add a lock. The door is capable of movement if the door is unlocked by a user with a key. This prevents the door from being opened by unauthorized users.
The next step is to seal the door permanently. For example, if we have a wooden door and a wooden frame, we can epoxy the door to the frame. If we have a steel door and a steel frame, we can weld the door shut. The door now cannot be opened without some destructive action: kick the door down, bash the door open, cut the door open, blow it up ....
(d) So the spec needs to spell out what enables the function to be performed. I don’t remember the cite, but there is case law that states that the function must be able to be performed without destroying the apparatus [unless that is the design intent]. I believe the claim under consideration was directed to an apparatus with detachable components. The Examiner cited prior art with fixed components, but stated that the components could be detached with a saw. The Examiner was wisely overturned.
(e) But the limits of “detachable” need to be clear. Consider two components nailed together. Are they detachable because you can remove the nails? Is that a destructive act? How does that differ from two components screwed together, if I remove the screws? What if the claim specifies that the components are detachable without tools? Well, what if the prior art discloses two components screwed together, and thumbscrews (which do not require tools) are well known?
(f) So, back to your hypo. You do need to distinguish your sealing element from the prior-art lock and clarify what you mean by “the door is incapable of movement relative to the door frame”. It is reasonable to say that the lock does render the door incapable of movement relative to the door frame when the door is in the locked state, if you exclude destructive actions such as bashing the door down. Does your sealing element permit only a single state [sealed only], as opposed to two states [locked, unlocked]? How does your sealing element differ from nailing, screwing, bolting, gluing, or welding the door shut?