AnsweredAssumed Answered

Seal Leakage Simulation

Question asked by Alessandro Tornincasa on Jul 14, 2014
Latest reply on Jul 17, 2014 by Alessandro Tornincasa
  • Comment • 5

Hi,

I've perfomed a non linear simulation of a valve seal compression.

These are the results.

Before compression:

Start.png

after compression:

End.png

 

Suppose there's a fluid leak on the right side, as shown in this picture:

Leak.png

This is what will happen:

  1. fluid will enter the seat
  2. it will apply pressure to the seat and the rubber seal
  3. seal will deform and change its shape
  4. more zones of the seal (anche the seat) will be subject to pressure
  5. pressure will deform the seal
  6. new seal shape will determine new pressure distribution

 

Is there any way to simulate this with non linear analisys ?

the big problem here is that you don't know in advance what regions/faces of the seal will be pushed by fluid pressure and also seal deformation changes faces/regions that are subject to fluid pressure.

 

Thanks.

 

Alex

Shaun Densberger
Shaun Densberger Jul 14, 2014 10:04 PM
Correct Answer

Unfortunately, I don't think you'll be able to directly simulate this. All I can think of (realistically) is looking at the contact pressure between the o-ring and the valve and making sure that it is higher than the fluid pressure.

 

You could estimate the size of the area that the fluid pressure is acting over (per the results you've posted) and use that to divide up the o-ring's geometry such that you can apply a pressure load over the new sub-region. This pressure load's pseudo time function will need to be defined such that it comes into play after the o-ring has been compressed. Of course, this won't capture that additional area of the o-ring that will be exposed to the fluid pressure (which will then result in more o-ring deflection).

 

To account for this, you could do one of two things:

  1. Measure the new area exposed, modify the size of the sub-region, re-run the study, and repeat as needed.
  2. Incorperate an robust factor of safety on the amount of contact pressure you need.

Of these two ways, item #2 is the most realistic (#1 would be very time consuming, but I think it's technically possible).

See the reply in context
No one else had this question

Outcomes