Numerics FEA interpolation order vs element number

In summary, the level of interpolation order in Finite Element Analysis (FEA) has a significant impact on the accuracy and convergence of the results. A higher interpolation order can provide more precise results, but it also increases the computational cost. On the other hand, increasing the number of elements in an FEA model can also improve accuracy, but it may not have as significant of an impact as the interpolation order. Therefore, it is important to carefully balance the interpolation order and element number in order to achieve accurate and efficient FEA results.
  • #1
member 428835
Hi PF!

Can someone help me understand the difference between interpolation order (linear, quadratic, etc) vs element number?

Like, if we had a 1D beam (for simplicity) what's the difference between using 1 element and a quadratic interpolation vs 2 elements but a linear interpolation?

Not sure if this is posted in the right spot?
 
Engineering news on Phys.org
  • #2
In the end the choice between more low-order elements or fewer high-order elements will depend on your application, and the stress gradients in the area you're analyzing. Fewer high-order elements can be more efficient in terms of solving time and model size, but it may still be necessary to have a large number of them depending on your goals.

Here is some additional reading on the topic: http://www.padtinc.com/blog/the-focus/mid-side-nodes-do-they-really-help
 
  • #3
Mech_Engineer said:
In the end the choice between more low-order elements or fewer high-order elements will depend on your application, and the stress gradients in the area you're analyzing. Fewer high-order elements can be more efficient in terms of solving time and model size, but it may still be necessary to have a large number of them depending on your goals.

Here is some additional reading on the topic: http://www.padtinc.com/blog/the-focus/mid-side-nodes-do-they-really-help
Thanks!
 

Related to Numerics FEA interpolation order vs element number

1. What is Numerics FEA?

Numerics FEA (Finite Element Analysis) is a mathematical modeling technique used to solve complex engineering problems. It involves dividing a continuous system into smaller, simpler parts called elements in order to obtain an approximate solution.

2. What is the interpolation order in FEA?

The interpolation order in FEA refers to the degree of the polynomial function used to approximate the behavior of the system within each element. A higher interpolation order means a more accurate representation of the system, but also requires more computational resources.

3. What is the relationship between interpolation order and element number in FEA?

The element number in FEA refers to the number of elements used to discretize the system. The higher the element number, the more accurate the solution will be. However, increasing the element number without changing the interpolation order will not improve accuracy beyond a certain point.

4. How does the interpolation order affect the accuracy of FEA results?

The interpolation order directly affects the accuracy of FEA results. A higher interpolation order will result in a more accurate solution, but also requires more computational resources and can lead to unstable solutions if not chosen carefully.

5. How do I determine the optimal interpolation order and element number for my FEA analysis?

The optimal interpolation order and element number for an FEA analysis depends on the complexity of the system and the desired level of accuracy. It is usually determined through a process of trial and error, starting with a low interpolation order and gradually increasing it until the desired level of accuracy is achieved.

Similar threads

  • Mechanical Engineering
Replies
6
Views
2K
  • Mechanical Engineering
Replies
9
Views
1K
  • Precalculus Mathematics Homework Help
Replies
4
Views
659
Replies
6
Views
1K
  • Set Theory, Logic, Probability, Statistics
Replies
1
Views
775
  • Mechanical Engineering
Replies
2
Views
880
  • General Engineering
Replies
4
Views
2K
  • Differential Equations
Replies
0
Views
633
  • MATLAB, Maple, Mathematica, LaTeX
Replies
8
Views
1K
Replies
2
Views
1K
Back
Top