• Registered: 06-09-2010
  • Posts: 21

Topic: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Here is my input file!

#analysis using 3d model
LinearStatic nsteps 1
domain 3dShell
OutputManager tstep_all dofman_all element_all
ndofman 3 nelem 1 ncrosssect 1 nmat 1 nbc 2 nic 0 nltf 1
#All node
node      1 coords 3     0     0     1200     bc 6 0 0 0 0 0 0
node      2 coords 3     0     0     0     bc 6 1 1 1 1 1 1
node      3 coords 3     2000     0     0     bc 6 1 1 1 1 1 1
#beam element
Beam3d 1 nodes 2 2 1 mat 1 crosssect 1 refnode 3 boundaryLoads 1 2
SimpleCS 1 thick 1 width 1 beamShearCoeff 0.833333 Iz 1.2e+006 Iy 1.2e+006 Ik 2.4e+006 Area 1000
IsoLE 1 d 1 E 70000 n 0.3 tAlpha 0.83333
BoundaryCondition 1 loadTimeFunction 1 d 0.0
ConstantEdgeLoad 2 loadTimeFunction 1 Components 6 1.0 1.0 1.0 1.0 1.0 1.0 loadType 3 ndofs 6
PeakFunction 1 t 1.0 f(t) 1.

Below is the result

Domain type: 3dshell, default ndofs per node is 6, per side is 0



==============================================================
Output for time  1.00000000e+000
==============================================================
Output for domain   1


DofManager output:
------------------
Node           1 (       1):
  dof 1   d  0.00000000e+000
  dof 2   d  0.00000000e+000
  dof 3   d  0.00000000e+000
  dof 4   d  0.00000000e+000
  dof 5   d  0.00000000e+000
  dof 6   d  0.00000000e+000
Node           2 (       2):
  dof 1   d  0.00000000e+000
  dof 2   d  0.00000000e+000
  dof 3   d  0.00000000e+000
  dof 4   d  0.00000000e+000
  dof 5   d  0.00000000e+000
  dof 6   d  0.00000000e+000
Node           3 (       3):
  dof 1   d  0.00000000e+000
  dof 2   d  0.00000000e+000
  dof 3   d  0.00000000e+000
  dof 4   d  0.00000000e+000
  dof 5   d  0.00000000e+000
  dof 6   d  0.00000000e+000




Element output:
---------------
beam element 1 :
  local displacements   0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000
  local end forces      0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000  0.0000e+000




    R E A C T I O N S  O U T P U T:
    _______________________________


    Node        2 iDof  1 reaction  0.0000e+000    [bc-id: 1]
    Node        2 iDof  2 reaction  0.0000e+000    [bc-id: 1]
    Node        2 iDof  3 reaction  0.0000e+000    [bc-id: 1]
    Node        2 iDof  4 reaction  0.0000e+000    [bc-id: 1]
    Node        2 iDof  5 reaction  0.0000e+000    [bc-id: 1]
    Node        2 iDof  6 reaction  0.0000e+000    [bc-id: 1]
    Node        3 iDof  1 reaction  0.0000e+000    [bc-id: 1]
    Node        3 iDof  2 reaction  0.0000e+000    [bc-id: 1]
    Node        3 iDof  3 reaction  0.0000e+000    [bc-id: 1]
    Node        3 iDof  4 reaction  0.0000e+000    [bc-id: 1]
    Node        3 iDof  5 reaction  0.0000e+000    [bc-id: 1]
    Node        3 iDof  6 reaction  0.0000e+000    [bc-id: 1]

User time consumed by solution step 1: 0.000 [s]


Finishing analysis on: Thu Dec 22 19:19:24 2011

Real time consumed: 000h:00m:00s
User time consumed: 000h:00m:00s

What's wrong? Help !

2 Reply by bp

  • bp
  • Administrator
  • Offline
  • From: Prague, Czech Rep.
  • Registered: 02-03-2007
  • Posts: 1,047

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Hi, attached you will find corrected input. If you will have any questions, consult input manual or post a reply to this topic.
Borek

Post's attachments

ppp01.in 760 b, 11 downloads since 2011-12-22 

You don't have the permssions to download the attachments of this post.

bp's Website

  • Registered: 06-09-2010
  • Posts: 21

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

thanks !
I Get it!
The format of this array is
2  size lnum(1) id(1) : : : lnum(size) id(size);
where size is total number of loadings applied to element, lnum(i) is the applied
load number, and id(i) is the corresponding entity number, to which the load
is applied (for example a side or a surface number).

  • Registered: 10-02-2009
  • Posts: 79

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Hi Borek,

I am wondering whether 3d beam elements can be used with any nonlinear material models, like misesmat or j2mat. It seems only isole is working. Is that correct? Thank you.

Xuejian

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

As far as I can tell, Beam3D doesn't support anything but elastic behavior.

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

ppp01.in is no more available to download... same problem for me with this input deck:

Uno-Qk1.out
test model 5/28/2015 3:33:01 PM
linearstatic nsteps 1
domain 3dShell
OutputManager tstep_all dofman_all element_all
ndofman 3 nelem 2 ncrosssect 2 nmat 1 nbc 3 nic 0 nltf 2 nset 2
node 6 coords 3 5000 0 0  bc 6 1 1 1 1 1 1 
node 5 coords 3 2500 0 0 
node 4 coords 3 0 0 0  bc 6 1 1 1 1 1 1 
Beam3d 1 nodes 2 4 5 mat 1 crossSect 2 refAngle 0 
Beam3d 2 nodes 2 5 6 mat 1 crossSect 2 refAngle 0 
# fictitious section for LumpedMass elements
SimpleCS 1 thick 0 drillstiffness 0
SimpleCS 2 area 0.150000005960464 Iy 0.00112500041723251 Iz 0.00312499981373549 Ik 0.004250000230968 beamShearCoeff 0.833333333333334
IsoLE 1 d 2.55E-09 E 210000 n 0.3 tAlpha 0
BoundaryCondition 1 loadTimeFunction 1 prescribedvalue 0.0
ConstantEdgeLoad 2 loadTimeFunction 2 Components 1 -10 DOFs 1 3 set 1
ConstantEdgeLoad 3 loadTimeFunction 2 Components 1 10 DOFs 1 3 set 2
ConstantFunction 1 f(t) 1.0
PiecewiseLinFunction 2 npoints 2 t 2 0.0 1.0 f(t) 2 0.0 1.0
set 1 elements 1 2
set 2 elements 1 1

7 Reply by bp

  • bp
  • Administrator
  • Offline
  • From: Prague, Czech Rep.
  • Registered: 02-03-2007
  • Posts: 1,047

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

The beam3d is geometrically linear, but could be used in simulations with nonlinear material behaviour. This requires to use fibre cross section model, allowing to use different material modes for each fibre. The only limitation is that the material model should support fibre material model.
I am out often office til Monday, will try to upload missing file again then.

bp's Website

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Thanks, I'll wait your check. However, in my deck I used no fibre cs and the analysis is linear elastic.

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

I'm pretty sure this is due to the beam element not supporting the new method of applying boundary conditions (using sets), though I haven't had time to check.

10 Reply by bp

  • bp
  • Administrator
  • Offline
  • From: Prague, Czech Rep.
  • Registered: 02-03-2007
  • Posts: 1,047

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Hi, I attached the modified input.
Two main points: the beam3d now does not support loads applied using the sets syntax (as pointed out by Mikael). Secondly, the "refAngle" is not supported (at least with version on  oofem.org repository). So I have to modify it to use reference node instead.

Post's attachments

eee.in 1.1 kb, 3 downloads since 2015-06-03 

You don't have the permssions to download the attachments of this post.

bp's Website

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Thank you for your help. However, two more things:
1. it is simplier to write an input file using sets for nodes and elements. I hope the modifications done by Mikael will be extended to the "ConstantEdgeLoad" command;
2. the refAngle feature was written by me and was added by Mikael (see here). As far as I know, these modification should have merged with your git, together with the support for UEL.

Thanks,
Giovanni

12 Reply by bp

  • bp
  • Administrator
  • Offline
  • From: Prague, Czech Rep.
  • Registered: 02-03-2007
  • Posts: 1,047

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Hi Giovanni,
I am pretty sure I have used the latest repository version on oofem.org. The changes you mention have not been merged yet.
Borek

bp's Website

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Borek is correct. I haven't had time to look over the (big) merge with oofem.org repo yet. I'll try to get to it within a week or so.

The support for loads through sets should be fixed for the beam elements as well, but again, I haven't had time to look at that. (the reason is that structural elements, like beams and shells, need specialized implementations, and was low priority to add since the old syntax still made it possible to use them).

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

I had a look at this.

As the loads is an edge-load, it should be applied to a set of edges. For the beam i consider the midline as the first "edge".
This consistency is required for the sets to work properly.
"elements" (body/volumes) should be used if the load is specified as per volume, "elementboundaries" (surfaces/areas) should be used when applying a surface load, and "edges" should be used when applying edge loads.

Beams could support edge and body loads (surface loads seem unlikely, but not theoretically impossible if we know what the cross-section looks like).
Right now, there is no warning, because the set contained elements (which are volumes, since the cross-section has an area), and since the edge loads is applied to all the edges in the set, there are currently nothing to apply it to (thus no force).
If one added edge sets, there would have been an error message that beam3d lacks support.


Now, the second problem is that the "dofs"  was ignored by the Neumann BCs. Now, this can be solved by just specifying all the zero values in the load, but it is misleading that this value is not utilized.
The problem is that some loads don't have a meaningful interpretation for "dofs". Pressure loads and gravity for example.
We're going to have to have a long think about how to have these types of general neumann b.c.s combined with problem specific loads (like pressure and gravity)

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Thanks Mikael for the reply.
Basically, the only way for now is to write the edge load in the same line as the beam3d elements; however, I would say that having the load commands in the same line makes more difficult to write automatic meshers.

For the second problem, I agree with you: the zero components must not be in list. For simplicity, I would keep in the program the commands for general Neumann b.c. completely separated (and in general not consistent) from the commands for specific loading (eg. gravity, that for a beam can be easily expresses as an edge load).

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

dear Mikael, I tested the modifications on beam3d to support the edgeload with the input in post #6, but I get:

____________________________________________________
           OOFEM - Finite Element Solver
        Copyright (C) 1994-2014 Borek Patzak
____________________________________________________
_______________________________________________________
Warning: (util.C:65)
In oofem::InstanciateProblem:
Failed to construct engineering model if type "domain".

_______________________________________________________
_______________________________________________________
Warning: (oofemtxtdatareader.C:111)
In oofem::OOFEMTXTDataReader::finish:
There are unread lines in the input file
The most common cause is missing cause are missing entries in the domain record,
 e.g. 'nset'
_______________________________________________________
_______________________________________________________
Error: (src\main\main.C:255)
In main:
Couldn't instanciate problem, exiting
_______________________________________________________
No backtrace available
  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

you must have made a typo or something in the input file, can you pleaes post it?

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

I checked again, there was a typo, thanks.
Anyway, I get all the results zero. See attached file.

Post's attachments

beam3dconstEdgeload.zip 1.35 kb, 2 downloads since 2015-06-15 

You don't have the permssions to download the attachments of this post.
  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

As i mentioned, the loads are applied to all the edges in the set. The current sets specifies only volumes (elements are always volumes).


That aside, it turns out there was some other bugs as well. I just fixed one, but the other was a bit more complicated (due to the element missing an interpolator object, which the assembly code assumes).

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Bug is now fixed, and edge loads should now work on edge-sets with beam3d and beam2d.

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Thanks for your help; however I pulled from yout git repo and re-compiled, but the results are still zeros... please find attached input deck with checked units of measure.

Post's attachments

OneDistrib-Qk1.zip 1.23 kb, 1 downloads since 2015-06-16 

You don't have the permssions to download the attachments of this post.
  • Registered: 25-06-2013
  • Posts: 242

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Try oofem.org/gitweb, last Mikael's commit is "Fixed bugs with edge loads on beam2d and beam3d"

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

It is exactly what I did, and I also cleaned the vs compilation files.

  • From: Gothenburg, Sweden
  • Registered: 08-09-2009
  • Posts: 1,419

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

You have still only applied volumes in your set

Set 1 elements 1 1

It needs to be applied to edges.

Set 1 elementedges 2 1 1

25 Reply by johnnyontheweb (edited by johnnyontheweb 16-06-2015 18:08:10)

  • Registered: 16-10-2010
  • Posts: 417

Re: Help me! Beam3d with ConstantEdgeLoad ! All result is zero!

Thanks, now it works fine. 2 more questions:
1.Can you please explain in detail the sintax of

Set 1 elementedges 2 1 1

Is it: set #numberOfSet elementedges #numberOfValues(always 2 for beam)? #eleNumber #edgeNumber?
I cannot find any other example, even for other elements than beams.

2. in the results I read a reaction in direction 2 (no load in direction 2) and a moment in dof 6 at node 1, which should not be present.

Thanks for your time and patience.

Post's attachments

OneDistrib-Qk1bis.zip 1.36 kb, 4 downloads since 2015-06-16 

You don't have the permssions to download the attachments of this post.