What's the most likely cause for this carbon seal crack?

[Fracture Toughness]

TL;DR Summary

Cracked carbon graphite seal that is used in an inline axial piston, variable displacement hydraulic pump.

We have a molded carbon graphite seal that is used in an inline axial piston, variable displacement hydraulic pump. One of our customers reported that, when using the “A” parts in the past, they only needed to replace them due to normal wear. However, after switching to our parts, the replacement cycle seems to be much shorter due to “broken” or “cracked” failures. This issue was identified after hydraulic fluid leakage was observed. According to their records, the same problem has occurred three times over the past few years. We have sold over 1,000 parts to a dozen customers, and this is the only customer that has reported cracking issues. It is unclear whether this is a quality escape on our side or something happening at the customer’s end, but my first goal is to determine the root cause of the cracks.

Based on the photos provided, two of the failed seals show cracks initiating from the root of the fillet radius. You can see the tang of the seal retainer engaging with the anti-rotation slots. The shaft seal subassembly is held in position by the seal retainer, which is attached to the mounting flange with fasteners. To me, it appears the cracks may have been caused by some type of impact. Both components are static relative to the shaft, so neither should be rotating. Could vibration be the cause?

We performed a material analysis on the cracked parts, and no deviations were found. Hardness, density, microstructure, and porosity were checked and compared against the A seal, and the results were equivalent. On the drawing, the ID, OD, slot-to-slot distances, slot width, and fillet radius also show no significant differences compared to the A seal. However, we only performed a 1.5 AQL inspection on this part, and slot-to-slot distance was not measured, so I cannot rule out a quality escape. That said, since this is a molded product, the process should be highly repeatable.

Another possibility is that something went wrong during installation on the customer’s side. But since the A seal did not experience the same issue, I am inclined to believe the root cause may still be on our end.

Pump General Characteristics:​

• Type: Inline axial piston pump
• Displacement: Variable
• Rotation Direction: Clockwise (viewed from the coupling shaft end)

Performance Ratings​

• Maximum Continuous Operating Pressure: 209.0 bar (3031 psig)
  
• Rated Speed: 3750 rpm
• Maximum Continuous Speed: 4000 rpm

Installation Data​

• Hydraulic Fluid:
  • Alkyl Phosphate Ester
• Fluid Temperature Range:
  • -45 °C to +121 °C (-50 °F to 250 °F)
• System Filtration:
  • 10 µm (400 µin) nominal
  • 25 µm (1000 µin) absolute

 

[Baluncore]

Welcome to PF.

Is the fracture always located at the (tensile stress concentration) corner of the retaining slot?
Does the seal always fracture in the same two places?
Could that retainer be distorted?
Is the retainer replaced at the same time as the seal?

 

[Fracture Toughness]

Thank you! The first failure was a fine crack (part still intact) after a couple of hundred hours of operation. The second failure was a complete crack (in 2 pieces) after a couple of thousand hours of operation. The third failure (similar to 2nd failure) should be the one shown in the photo.

I forgot to mention that they are likely still using the A seals. It is certainly possible for the retainer to have defects, especially if it has sharp edges. However, if the A seals are not affected by it, then the likelihood of the retainer having significant defects would be slim, I think. But that's a good point, I should probably confirm that with the customer so I can rule out the retainer 100%. From the photo, there appears to be some clearance between the tang of the retainer and the seal's slot.

They are typically replaced at different intervals based on scheduled maintenance inspections, but they can be replaced at the same time as well.

 

[Fracture Toughness]

the last two failures appear to develop crack on the same location, the 1st failure is not in the slot. these 3 failures are on 3 different parts and different years

 

[Lnewqban]

What is inserted into that female spline?

 

[Baluncore]

Is the drive from a short inline spline, with a compliant coupling?
Or is it from a gearbox with rolling bearings?
Or from an offset V-belt drive with support bearings?

What controls the pump flow and outlet pressure?
How long does it take to cut the pump output when pressure rises?

Hydraulic shocks can generate side forces on pump drives.
To what actuators does the pump supply fluid?
Can those actuators collide with something and cause a shock?

 

[Lnewqban]

Following @Baluncore line of thinking, it would be interesting to see how is that female shaft supported by bearings and the condition or age of those.
Any cut view of the pump-coupling-motor assembly would be helpful for us to see.

Normally, bearings and seal go together, and proper alignment of shaft-bearing-seal is not an issue, as all the surfaces are machined in one step and perfectly concentric.

This case should be the same in theory, but any misalignment, vibration, or undesired axial load on the seal lip may be happening as result of one or more causes (like improperly aligned motor-coupling-pump, worn bearings, pumping pistons cyclic loads, shaft surface finishing, etc.).

 

[Fracture Toughness]

I can only answer some of the questions:

1. Short inline spline + coupling shaft to the engine gearbox; not an offset V-belt drive.

2. It’s pressure-controlled variable displacement: a pressure compensator senses outlet pressure and adjusts the yoke/swashplate to vary displacement. There’s also an electrically depressurizing valve that can lower/limit system pressure. The pump is directly connected to the engine gearbox and the output fluid flow will change in proportion to engine rpm.

 

[Fracture Toughness]

found a similar pump online: