Cylinder Leak, Static O-Ring Seal 2

[mrsander7]

I'm trying to determine what is wrong with a double-acting hydraulic cylinder with an external cap (static female radial gland). We're getting occasional leakage through the threads without visible damage to the o-ring. We're using an #344 Parker o-ring with a single backup and all dimensions and tolerances are exactly to spec as found in the Parker O-ring Design Handbook. We recently changed to a size larger o-ring/backup due to this problem, and leaking is still an issue.

In a particular case, the alignment of the cap and cylinder chamfer results in a satisfactory clearance (dist between cylinder od and cap id) behind the gland (on the thread side), but a 3-thousandths additional clearance BEFORE the gland (pressurized side).

Would such a clearance on the pressurized size of the gland be cause for leakage? If so, I'm trying to wrap my head around the mechanism behind the seal failure (why leaking may occur in this case, how the o-ring becomes unseated). Has anyone run into this issue before or have any insight?

Thanks!

 

[hydtools]

Surface defect.
There may be a slight spiral groove, thread-like, left by the turning process on the cylinder OD surface. This 'defect' will prevent the o-ring from sealing. Better to use a turning tool with a large nose radius. Do not use the threading tool to cut the sealing diameter. Or finish the cylinder OD with a plunge cut.
Similar 'defect' may be in the cap gland ID.

Ted

 

[chicopee]

May be temperature could be a factor. Sounds familiar with one of NASA space shuttle?

 

[israelkk]

Calculate cap radial expansion due to the pressure. The thread tooth angle (60 def) contribute radial expansion due to the pressure that pushes the cap. More detail including actual parts interface dimensions and tolerances, hydraulic pressure, etc. can help.

 

[Eltron]

Could it be a chemical compatibility issue?

Dan

http://www.eltronresearch.com

Dan's Blog

 

[mrsander7]

hydtools, I'll check on that process/inspection and get back, although we're calling out a 32 rms finish on that surface.

Temperature and compatibility isn't an issue... it's a simple industrial application at room-temp with BUNA-N seals, which have excellent compatability with hydraulic fluid. We're not having any issues with our internally threaded caps, just this particular externally threaded one.

israelkk, my calculations suggest a radial expansion of only 8e-7" due to the combination of radial pressure and your mentioned thread consideration.

Attached is a dwg of the gland. Again, all gland dimensions, tolerances, and finishes are standard per the Parker O-Ring Handbook. Thanks again!

 

[mrsander7]

Correction for the drawing above: The 4.215 dim is for the gland diameter. That dimension should actually read 3.878 +.002/-.000. Apologies for the confusion.

 

[bcd]

Not knowing the details of how it operates (pressure, temperature, flow rates, etc.) or how much oil is leaking, I cannot confirm if these issues exist, but here are two thoughts I had:

1. Cyclic pressure, especially high pressure will cause the o-ring to heavily deform in one direction. When the pressure is released, the o-ring will relax close to its original position minus any compression set. This pumping action can allow oil to slip past the seal and push out everytime the o-ring is pressurized. This could be aggravated by compression set which will increase the o-rings propensity to leak at low pressure.

2. Any high volume flow rate past the tight, close tolearnce gap may be creating a vacuum, tending to pull the o-ring in the opposite direction. But again, don't know how fast the cylinder cycles so this possibility may be way off target.

 

[hydtools]

mrsander7, I have inspected surfaces that were supposed to be 32rms and found spiral tool lines. When the process was changed to produce circular tool lines, the o-ring sealed well.

Is the diamter to the left of the backup ring actually 4.097 +.004/-.000? Seems like too much clearance even with a backup ring. It should be no more than .010 over the 3.875/3.873 dia. with a backup ring. Parker recommends .003/.006 diametral clearcance.

Ted

 

[mrsander7]

hydtools, The diameter to the left is 3.878+.002/-.000. Gland diameter should be 4.215+.004/-.000 rather than 4.097. My mistake on the quick sketch-up; Apologies on that one.

I'm in the process of checking with manufacturing in our plant on the tooling issue. It's a very good possibility that this is the problem considering the failed o-ring changes and uniqueness of the gland machining compared to our usual cylinders. As a temporary solution a while back, they started welding the cap on... eliminating the o-ring and paying little attention to the tolerances... so it's a process getting the correct information from previous revs.

bcd, I'm definitely looking into the cyclic pressure/vacuum problem. It was my original thought along with eccentricty due to the 4-16 thread but it has been tough getting validation for that failure mode when we have to disassemble the cylinder to check for defects. Would a double back-up ring help to alleviate the vacuum issue if this were the case?

Thanks again for all of your insight.

 

[hydtools]

Vacuum should not be an issue, assuming the taper on the lead end of the male plug enters past the chamfered edge of the female so the o-ring cannot lose its seal. Adding an opposing backup ring will further limit movement of the o-ring and cover clearance that may exist if the plug does not full engage the female. You should check the gland width to be sure it will accommodate a second backup ring without preventing the o-ring enough room to deform due to gland squeeze.
Why would you expect a vacuum? Does the load over-speed the piston, run away from the hydraulic flow?

Ted

 

[mrsander7]

Some new information... according to QA Tech, screwing in the cap until it bottomed out did not stop the leak, but seemed to lessen the severity. During normal assembly, we bottom the cap out and thread it back up to 170deg for port alignment. That's a total of ~1/32 worst case backed off, which is shown in the image I posted earlier.

Hydtools-
Attached is an image of the cylinder barrel od finish. No threadlike grooves (though certainly there are circular ones), contamination, or other defects that I can find. Could you explain the "taper" you mentioned? We have chamfers on the leading edges of the male barrel and female cap (.06x20deg), and other than the thread lead-ins, we don't have any tapers on our parts. Do you mean the GLAND taper (0-5deg?). As I mentioned in the first post, we do have a case in which the male chamfer opens a gap before the o-ring, but not after (see previous pic/dims).

We're running at a max operating pressure of 2400 psi. We have check valves and other measures to relieve vacuum and prevent load-driven operation, but I seem to be turning to cyclic pressure and compression set for lack of any other viable explaination.

 

[hydtools]

I was referring to the .06x20deg chamfers.

The surface finish looks textured, it is not very shiny like the chamfer is. Is there tool chatter? It may be worth taking a light file finish to the barrel OD with a fine file to clean the turned surface. If you run your fingernail over the surface does it feel 'toothy', textured? Improve the finish to 16rms even though 32rms should work.

What does the gland OD look like in the cap? Is it cut with a grooving tool that left a cut stop edge across the surface? Have the cutter dwell for a couple revs or more to unload the chip and cutter and get a little burnish finish.

Ted