Surface load

[g.ceruti]

Hi, does anyone know how NiC translate a surface load in FORCE nastran cards? which is the function used?

Looking at the content of the nastran *.nas file, I found:

• with linear elements, an upward surface load is spreaded as upward nodal loads on element vertex nodes (Fig.1)

• with parabolic elements, the same upward surface load is spreaded as (Fig.2):
  • upward nodal loads on element midside nodes
  • downward nodel loads on element vertex nodes

Although, in both case the sum of all nodal loads provides the same total loads, locally each element undergoes to unexpected loads.

I used an other FEM software, to import and visualize the nastran file content.

Fig.1 Linear elementFig.2 parabolic element

[John_Holtz · Accepted solution]

Hi @g.ceruti

The load with parabolic elements is unexpected because of the mathematics of FEA and elements with midside nodes. I do not know the theory well enough to be able to provide equations, but I know that the load distribution can be "unexpected". (The load distribution is definitely not uniform with parabolic elements like it is with linear elements. The details depend on whether the load is applied to an edge or the entire element, etc.)

You might be able to see something similar if you can look at the reaction forces from a model and compare linear to parabolic elements. (I am not sure if InCAD provides enough details to get the reaction forces at the midside nodes or not.) For example, a bar fixed on one end with a tension (or compression) pressure on the opposite end. The reaction forces at the constraints will show what the "load distribution" is at the midside nodes that corresponds to the pressure at the other end. I will not be surprised if the reactions at the corner nodes and midside nodes are in opposite direction!

These questions come to my mind:

1. Are the results of your two models correct? In other words, does the load distribution "look wrong" because the results are wrong?
2. Are you trying to understand something by looking at how the load is converted to forces at the nodes? If this is the case, what are you trying to do?

[David_Kind]

@g.ceruti, what software did you use for the visualization of the .nas file?

[g.ceruti]

Hi John,

I am newbie in NiC, but I have more experience with other FEM softwares, including also MSC Nastran.

Generally, I look at the nastran file content just to understand what NIC do in translating "something" applied on geometric entities into "something" applied to FEM entities, beacuse Nastran cards are what Nastran solves.

Applying a surface load, I just expected to find pressure load cards in nastran file, or simply the same force load at each node.

I have checked the FEM theory in my university notes, and I confirm that the Nic load distribution on parabolic elements is the right one.

[Anonymous]

I've found a couple of good references on transfering distributed loads to mesh nodes.

This one:

https://www.colorado.edu/engineering/CAS/courses.d/IFEM.d/IFEM.Ch07.d/IFEM.Ch07.pdf

And this one: