O-ring Seal Friction - Pressure equalized and Differential Pressure 1

[johnchrc]

My question is really regarding the difference between seal friction on a dynamic o-ring in several conditions. It is axial Piston seal on rod. The o-ring is 300 series HNBR because it is robust. It will only cycle 3 times then lock open and the seals will become static. The conditions are listed below. I believe I'll have over 300# friction on a single o-ring with pressure equalized downhole at 5,000 psi.

1)Used in downhole tool and friction will add to spring force. There will by hydrostatic pressure at the tool up to 5,000 psi.

2)Differential pressure is applied across a piston area created by o-ring seals.

3)I have a nomograph that gives me the approximate friction with differential pressure and it is magnitudes higher than static friction and is large enough it must be considered to determine the hydraulic force to move the part down with the spring.

4) The spring is a 90 Belleville stack with hysteresis that reduces the return force.

5. I have a concern when differential or surface pressure is bled to 0 psi that the pressure equalized seals with hydrostatic pressure will generate almost as much friction created by differential pressure at the O-rings. With pressure acting on both sides, it will create a high contact force with gland ID.

Is this hydrostatic friction a real concern? My colleague thinks it will force the o-ring into square shape as pressure is applied on both sides of the seal and the contact area with the seal surfaces will be large. The normal force would be slightly higher than the hydrostatic pressure created force because of initial squeeze.

I have never considered O-rings with no differential having such high breaking friction once surface pressure was bled off. We will test to see. However, I never recall having a problem due to this condition before.

Consider that a 348 HNBR Rod O-Ring would have over 300# friction at 5,000 psi. With another seal and static friction it could around 600#. Then you have the spring force...

Is this the proper method to calculate friction. Comments?

- CJ

 

[Compositepro]

A picture might help, but if you have a high hydrostatic pressure on all sides of the o-ring, that pressure will not cause any contact force or friction.

 

[johnchrc]

Compositepro -

See attached and see if that explains why equalized hydrostatic pressure will create high friction just as differential pressure. I'm not claiming to be correct but I can't see how it wouldn't be almost as high as differential pressure friction.

- CJ

 

[Compositepro]

Your free body diagram for case two is just wrong. Case two should look just like case one. If you submerge an o-ring in the ocean to a depth where it is under 5000 psi it will not have changed shape at all from when it was on the surface. There might be some subtle changes due to the compressibility of rubber at such a high pressure, but it will still be round. In your case two the 5000 psi does not affect the contact pressure of the rubber with steel in any way. The 5000psi fluid is between the rubber and the steel, and contact pressure is due only to the elasticity of the rubber.

 

[johnchrc]

I think the diagram is correct. If there is squeeze and pressure applied from both direction the round seal becomes square and the contact forces increase with hydrostatic. If there was no squeeze , then I would agree but you have no seal.

How can the contact pressure not increase? Just explain because I can't see how it does. Maybe you can draw what you are trying to explain to make it clear?

- CJ

 

[johnchrc]

I would like a seal expert to put this to bed. Case 3 would be more like 7000 on one size and 5,000 psi on the other. Bleed the 2,000 psi off to Case 2 and you do not have 5,000 psi acting on o-ring from all directions in nice o-ring circle without the loss of squeeze and any future sealing compromised. I did not think about this until last week.

- CJ

 

[Compositepro]

Your drawings are accurate except you show a round o-ring becoming sqare for no reason. Fluid does not have a shape so fluid pressure cannot change the shape. Your drawing shows arrows dipicting pressure squeezing on the o-ring. It is not the backing rings squeezing on the o-ring, because they do not seal. The only difference between case one and case two in your drawing is that the fluid in the space around the o-ring is different. The net forces are the same. As I said, contact pressure between the o-ring and the steel is due only to the elasticity of the o-ring (i.e., the diameter is slightly larger than the gap). Think about it. There is really nothing more to say.

 

[johnchrc]

The drawing is not exact representation. You can even remove the backups from the example.

I have thought about it. I just haven't heard an explanation that explains why initial squeeze won't increase because of the way O-rings behave.
You are saying the initial squeeze remains the same as long as pressure is the same on each side of o-ring. The net forces are the same across the o-ring at 5,000 psi and at 0 psi. But the force at 5,000 psi would be much larger but equal in both directions compressing the "round" o-ring into a different or more rectangular shape with larger contact length and force at ID of seal and OD of gland. If the initial squeeze isn't lost, I don't see how it doesn't increase with pressure acting on the o-ring.
I have never experienced a problem with friction except with DP so I may be overthinking.


- CJ

 

[johnchrc]

UPDATE:
Testing performed at the largest Oilfield Service company in the world proved my case. Differential pressure friction = Hydrostatic pressure friction.

- CJ

 

[rickfischer51]

Johnchrc, I believe you are right. We used to treat o-rings as an incompressible fluid when doing gland calculations. Elastomers have Poissons ratio very near 0.5, which is compressible. Loading the o-ring axially deforms it and increases the pressure on the OD. Doesnt matter if the load is a pressure or a reaction force. The net axial forces are equal magnitude and sum to zero. Pressure distribution varies because of the different boundary conditions.

JOHN CHRane packing Co?

Rick Fischer
Principal Engineer
Argonne National Laboratory

 

[rickfischer51]

Correction: "which is incompressible"

Rick Fischer
Principal Engineer
Argonne National Laboratory

 

[johnchrc]

No... Try 4-3-1 and a name including last, first and middle initial.

My colleagues have had had recent experience with this. Small hydraulic area, high pressure, large seal dia's. This resulted in much higher breaking friction than planned for. There were many variables but the friction from high static pressure made the load unmanageable (pressure would not move piston) until a PEEK debris barrier was removed in floating j-slot. It did not create a seal but put a bind that pressure could not overcome but hammering with a hammer and brass bar would.

The static friction was irrelevant from a functional sense. but with dynamic or hydrostatic pressure growing the peek ring maintained a load Ff > P x A(.785 sq. in)and up to 20,000 psi. I don't have the details. I do know Peek can behave erratically when it comes to moving force but this...

Maybe I can get a pic. It is a simple device but I'm too tired to sketch and scan.

Good evening...
CJ

- CJ