The filmmaker Sergei Eisenstein once conducted an interesting experiment. He cut together strips of film showing an old woman with various objects, then asked test subjects what they thought. When the image of a woman’s face cut to a loaf of bread, the subjects thought she was expressing hunger. When the image of her face cut to a picture of a dead child, the subjects said she was expressing grief. Her expression never shifted, but the interpretation changed with the accompanying image. In practical terms, this suggests that readers often interpret emotional reactions based on context. You don’t always need to show a character crying to suggest grief. If you show action, and that action is tragic, then the reader will assume that the character is in pain.
As a general rule, a scene is over once a character begins crying. Why? Because once a character cries, the scene can’t easily progress to a more intense emotional value. The writer has already played the peak emotion. The scene flounders because the fiftieth tear isn’t more meaningful than the first one. Remember, the task is to move your reader, and if the character is doing all the crying, it doesn’t leave much room for the reader to feel much except sorry.
Here’s another concept: you can keep scenes moving by encouraging your primary character to resist the pain, rather than by yielding to it. Why? Because resisting the pain creates tension within the character, and tension within the character creates tension in the scene. Yielding to the pain releases the tension in the character and in the scene.
If the character moves from tears to some specific action that's stronger than crying, then the scene doesn't have to be over at the first tear. For example, the character moves from crying to throwing things, or some other kind of violence. Rather than releasing the tension, the character’s tears build to the expression of a stronger action.
The same principle applies to sequences of scenes, though not as strictly. For example, if a character cries in the first scene of a sequence, what's left? If the character simply cries again, it won't turn the scene or sequence in a new direction. It simply repeats. And emotions that repeat without progression quickly become tiresome.
It's often more effective to build strong emotion in a character but not to release it at the first opportunity, because once you release it, you're faced with the law of diminishing returns; every time you return to play that same emotion, the effect on the reader will generally decrease if the level of emotion is the same. If you want your character to cry, then you must build to it, and realize that once that character cries, then either the character must progress to a stronger emotion, or the scene/sequence has effectively peaked.