FEAP User Forum

FEAP => Programming => Topic started by: JStorm on June 09, 2022, 01:05:41 AM

Title: FEAP 8.6.1n: bug in QUADrature ORDEr
Post by: JStorm on June 09, 2022, 01:05:41 AM

It seems that the 3D solid element ignores QUADrature ORDEr for the number of integration points.
However, the number of history terms for the element considers the quadrature scheme.
As a result, out of bounce operations are caused for a reduced integration scheme because too many integration points are used and called.

Code: [Select]

QUADrature ORDEr 1

Code: [Select]

History terms/pt  =     8

Code: [Select]

     Material    Element Tag   Element Type  NH1/2 History    NH3 History
          1              1      Solid                  8              0

Title: Re: FEAP 8.6.1n: bug in QUADrature ORDEr
Post by: Prof. R.L. Taylor on June 09, 2022, 06:08:31 AM

There is a uniform integrated element that uses only 1 set of values.  Called with DISP UNIF stab
where stab is a stabilizing parameter. 
I will look to fix the set of quadrature to consider any input of pts.

Title: Re: FEAP 8.6.1n: bug in QUADrature ORDEr
Post by: Prof. R.L. Taylor on June 09, 2022, 07:40:36 AM

After thinking about your request more, it is not a good idea to allow 1-pt quadrature for general 3-d solid elements.  If the elements are distorted (your mesh seems regular) this would lead to errors as even the volume is no longer preserved.

Instead, the request will be tracked as an error by checking the value of "d(5)" in the quadr3d.f routine.

Also, when I forced 1-pt on the element the TANG command reported a singularity.

Title: Re: FEAP 8.6.1n: bug in QUADrature ORDEr
Post by: JStorm on June 12, 2022, 02:15:41 PM

That behaviour is also fine for me.
I had special reasons why to run it with one integration point.
It was never planed to perform qualitative simulations with this setup.