Discussion Groups

## Simulation Mechanical and Multiphysics

Member
Posts: 4
Registered: ‎07-29-2013

# Variable pressure - Shell elements with MES

259 Views, 4 Replies
08-06-2013 12:57 AM

I have an hyperelastic structure and I would like to apply a variable pressure in order to inflate it.

The problem is the pressure doesn't aply perpendicular to the surface althought I tick the box "normal to surface"...

Is anybody have an explication?

Please use plain text.
Valued Mentor
Posts: 498
Registered: ‎08-30-2012

# Re: Variable pressure - Shell elements with MES

08-06-2013 05:38 AM in reply to: sego31

The problem is that "variable pressure" is not analyzed as a pressure. The input pressure is decomposed into nodal forces and analyzed as forces. (Hence the blue colored arrows in your images and that the arrows are at the nodes and not at the center of the elements like a regular pressure.) Nodal forces do not change orientation throughout the analysis because it is not clear what a "normal direction" is at a node.

One option is to split the N surfaces you have now into more surfaces and apply a constant pressure on each surface based on the equation at some point on the surface. This option would apply the same pressure throughout the analysis but it would remain normal to the surface.

Since you have a linear equation, I suspect another option is to use a hydrostatic pressure. This would give you the precise linear pressure distribution and the pressure values would change based on the displacement.

Try a test model to see if either of these options will work for you.

John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

16 years experience with Simulation Mechanical
Please use plain text.
Member
Posts: 4
Registered: ‎07-29-2013

# Re: Variable pressure - Shell elements with MES

08-06-2013 06:23 AM in reply to: AstroJohnPE

I'm agree with you, hydrostatic pressure is more suitable but I can't use it because I have already hydrostatic pressure applied all around the structure (actually my structure is submerged under water and inflated with water) and I work on shell elements. In fact I try to simulate a mobile dam inflated with water (the membrane's material is assimilated to rubber) .

I can only use "variable pressure" to simulate the water pressure into the membrane, that's why I entered this linear equation in order to have an approximative hydrostatic pressure.

The solution to aplly constant pressure and split more surfaces doesn't resolve my problem, does it?

Please use plain text.
Valued Mentor
Posts: 498
Registered: ‎08-30-2012

# Re: Variable pressure - Shell elements with MES

08-06-2013 01:26 PM in reply to: sego31

Here's a thought. Doesn't hydrostatic pressure 1 + hydrostatic pressure 2 equal another hydrostatic pressure? That is, your inside and outside hydrostatic pressures can be represented with just 1 hydrostatic pressure. (As far as the analysis goes, I think this can only be simulated after the two pressures become equal. I doubt that you can have the hydrostatic pressure change direction during the analysis -- unless you do a restart analysis.)

Also, depending on the pressure inside the bag, it may be accurate enough to say that the pressure is uniform on the inside. So a hydrostatic pressure + or - a constant pressure is just another hydrostatic pressure, so you only need to apply 1 hydrostatic pressure. See the documentation for hydrostatic pressure to better understand how to do this.

Another common trick is to make a duplicate of the part and give the duplicate "nearly zero" strength. One pressure load is applied to the real part, and the other pressure load is applied to the fake part. Since the fake part has no strength, all that happens is the pressure on the fake part is transferred to the real part with then takes the load.

John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

16 years experience with Simulation Mechanical
Please use plain text.
Member
Posts: 4
Registered: ‎07-29-2013

# Re: Variable pressure - Shell elements with MES

08-12-2013 04:47 AM in reply to: sego31

If I decide that the pressure is uniform on the inside, what value am I suppose to take?

The max value, or I do an average?

Thanks.

Please use plain text.