Some time ago I recorded a simulation video which explains what happens when the piston-seal on a double-acting cylinder leaks, and it becomes a displacement cylinder. After watching the video, Larry from Canada sent me this in response:
"I enjoyed your video about cylinder drift not being caused by oil leaking past the seals. But I have to ask the same question as your previous sceptic: why do they have seals? If the seals are bad the oil is going to pass from the side under load to the other side. Doesn't matter if the pressure is the same or not. The load on the cylinder rod will move it and the oil will bypass the seals. If not, why do cylinder rods drift lower under load?"
I replied as follows:
You're not ready for the answer to the second question until you understand the answer to the first.
Fill a bucket with water to the brim.
Now plunge your arm into the bucket until your fist hits the bottom.
Your arm displaces water from the bucket.
But what if that displaced water had NO way of escaping from the bucket?
Do you still think you could get your fist to the bottom of the bucket?
Think about it.
And if necessary, watch the video again.'
This 'flicked the switch' for Larry almost immediately, and he responded:
"Thanks for the reply Brendan; I now understand. As my arm enters the bucket or the rod enters the cylinder they have to displace the water or oil which they cannot due to them being in an enclosed container. So is it safe to assume the oil is leaking back through the valve to allow the cylinder and load to drift down?"
That's pretty much it in a nutshell... when the load is acting to retract the piston-rod (positive load).
And if you haven't seen it yet, you can watch the simulation video here.