Announcements
Attention for Customers without Multi-Factor Authentication or Single Sign-On - OTP Verification rolls out April 2025. Read all about it here.

Actions and Reactions not balanced

Anonymous

Actions and Reactions not balanced

Anonymous
Not applicable

Hi All,

 

I only have one instability warning left in my model: "The instability (3 type) in the Rz direction in the node 191420". But when I zoom in to node 19420 I do not see anything wrong there. See Sketch 5 - Calculation Message in the attached PDF. Model is also attached.

 

So far I applied only selfweight load on the structure. Load case DL1.
When I  checked the reactions, I noticed that Actions and Reactions are not balanced. See attached Sketch 6 - Action -Reaction balance. The difference is about 10600 kN but it should be zero..

 

I also put few more sketches in the attached PDF:

- Sketch 1 - Deformed Shape
- Sketch 7 - Foundation stress
- Sketch 8 - Foundation settlement.

 

It seems that the building is tilting towards one side. I now have excessive stresses and settlements in foundations. Note that this is only due to selfweight force that is DL1, at the moment. Things will get worse when I apply wind and seismic horizontal forces.

 

Could it be that the warning message is due to global instability of the structure, rather than just a simple local modelling issue?     


What causes the action reaction imbalance in my model?

Many thanks,
Aleksandar

0 Likes
Reply
Accepted solutions (1)
2,201 Views
17 Replies
Replies (17)

Stephane.kapetanovic
Mentor
Mentor

hi aleksandar1234,

 

After a short reading, some panels mismatch with nodes and bars.

 

Look at nodes 92, 24509, 34, 32, 43, 52, 54, ... and panels 347, 156, 10107, 1658, 1659, 1660, 1666, 2942, 2943, 2946, 8885, 8886, 9068, 10152, 10154, 10244, 10247, 10272.


Edges 1660[1], 10272[2] seems too short.

 

Bars  1 56 66 77 78 179 272 303 312 344 346 348 396 397 400 409A412 437 440 453 456 497 554 625 627 660 667 679 682 690 698A701 714 716A720 730A735 747 750A752 768A772 774 785A791P2 804A809 821A827P2 are misaligned with global UCS.

 

These are small differences but greater than what the model's consolidation tolerance can compensate.

 

I don't know if that fixes all the instabilities but you will hardly be able to avoid a check.

 

Kinds Regards

 

image.pngimage.png

image.png

https://forums.autodesk.com/t5/robot-structural-analysis-forum/useful-addins-for-robot-api/td-p/3899...

 

Anonymous
Not applicable

Hi @Stephane.kapetanovic ,

Thanks for the quick response and for the link to the useful add-ins.

May I ask, if Robot reports only one warning in Calculation Messages, how did you find out all those other mistakes? Do you have some special diagnostic tool?

I believe I fixed all the mistakes you mentioned except for those related to "[bars...] ... are misaligned with global UCS".
What do you mean by "misalignment with global UCS"? I checked one bar, I think No. 77, and its z axis is pointing upwards i.e it is not rotated about its axis. Why would the bars have to be aligned with global coordinates?

As I said, I believe I fixed the mistakes you mentioned but when I run the calcs:

 

1) Actions and Reactions are still not balanced. 4750 kN difference. See attached screenshot "Not balanced.jpg."

2) I get a calculation message "The instability (3 type) in the Uz direction in the node 191327". When I zoom in to node 191327 it looks like everything is perfectly connected.

 

Help, please.

Thanks,
Aleksandar

0 Likes

Stephane.kapetanovic
Mentor
Mentor

in fact robot only reports instabilities and not the way in which you obtained them, that which remains your responsibility.
Yes, I have developed tools for expertise and model diagnosis, but this is not the only reason.
With the link I sent you, the rounding tool will allow you to edit quickly the coordinates of the panels.

 

For the bars which are not aligned, it should be understood that the inclined members bring horizontal forces and twists of columns. This imperfection participates in the non-symmetry of the reactions to the supports and peaks of contraints on slabs.
Bar 77 has its node 188557 in Y = -86.462 and its node 111 in Y = -86.464. The difference is minimal for member 77 but I wanted to let you know because it happens that we make point-to-point copies or even that we make outlines using the peripheral bars. To close the subject of alignments, look instead at member 698 and all the similar little bars on their differences along X. Members are short and differences more importants.
For instabilities, the study requires more time than what I have actually and I am in training until January 25.

 

On the other hand, if the instability does not come directly from node 191327, it could come from your releases at origin of columns at node 191325 because you have placed a small bar in torsion console on the beam.
This is the case everywhere in your model. Review these links.
If you are releasing the torsion of a bar at one end, make sure that at the other end you have at least one beam that blocks the bar at its other end.
This is common in case using bbblll releases for all your bars which is your case.

 

What you need to do is review your releases, check your results for all the movements and rotations of suspect nodes and sort them in order and find the associated degrees of freedom. What you can also do is look at the user nodes with regard to the computation nodes because if Robot has created several nodes very close it may indicate bad junctions and short elements.

 

Kinds regards

In second model, beam 90 and 9720 are not paralel to edges of panels. 

Beam 51 and panel 17[14]

Beam 917 and panel 118(2)

Panel 2943 and Edges 10244[3], 10247[11], Node 11

Panel 10170(3) with panel 10174[16] 

Panel 10174[16] with panel 10170[4], node 2544

Panel 10244 with panel 2943[3], Node 11 and Beam 90

Pane 10247 with Beam 51, 90, 2943[3], Node 11, Beam 9720

 

Anonymous
Not applicable

Thanks @Stephane.kapetanovic for your effort.

I think I found out what was the problem:

 

1. At the connection of walls to foundation pads, one of the diagonal infinitely stiff "dummy bars", had a 620000 kN !! axial force. 

2. At upper floors, I placed  rigid links (flexible diaphragm) at both ends of those stiff dummy bars. So the dummy bar was trapped between rigid links and somehow it attracted about 120000 kN axial force.

 

I will now fix these issues and share more details later.

 

Many thanks,
Aleksandar 

0 Likes

Stephane.kapetanovic
Mentor
Mentor

Alex, 

 

I believe your problem is being solved and eventually you will adjust your model gradually without too much efforts.

So if your problem is solved, you can close this topic.

 

Best regards.

Anonymous
Not applicable

Hi Stephane,

 

I fixed the issues that I myself found (mentioned above in Message 6), but it turned out that it is not enough so I will proceed with fixing the model step by step as per your instructions above.

 

Thanks a lot for your help. I will let you know how I am coping with the model.

Many thanks,
Aleksandar 

0 Likes

Anonymous
Not applicable

Hi All,

 

Something is definitely wrong with Robot.

 

I did the following simplifications in my model:
- removed all spring supports
- removed all foundations (so there are no foundation elasticity parameters any more) and replaced them with fixed supports
- removed all beam releases
- removed all slab releases
- replaced all orthotropic slabs with regular 400THK RC slabs
- made all smaller beam sections to be 400x800mm

 

My max. node displacement is now about 9mm, which seems as expected.

There is still no balance between actions and reactions.

 

Attached are:
- original model
- model with simplifications
- screenshot of not balanced actions and reactions

 

Seems to me that Robot is not really the go to software for real life complex geometries.

 

Regards,
Aleksandar

0 Likes

Anonymous
Not applicable

Update.

 

I almost get balanced actions and reactions when I remove the flexible diaphragm that I created with rigid links.

So, how do I do the seismic analysis if I cannot create a diaphragm? I was going to do an equivalent lateral force analysis of my own (not the built in lat. force analysis in Robot)  and apply forces in the master nodes which I would place in the center of mass of each floor.

Attached is the Robot file with deleted diaphragms and a snapshot of action-reaction balance.

Thanks,
Aleksandar

0 Likes

Anonymous
Not applicable

 

Why do you use rigid links?, wouldn't it be better to use slabs with a flexible rigid diaphragm?

Did you solve the problem?
0 Likes

Anonymous
Not applicable

Hi @Anonymous ,

Thanks for your interest. I have not solved this problem yet. Still struggling with it.

Changing to rigid slabs has not helped. See attached file.

Regards,

Aleksandar

0 Likes

Anonymous
Not applicable

Hi @Anonymous.

I see that you did not add supports to the slabs that are supporting the "DUMMY BARS". This may lead to instabilities

0 Likes

Krzysztof_Wasik
Autodesk Support
Autodesk Support
Accepted solution

Hi @Anonymous

In 210110 Level 1.rtd reducing E and G of "Super rigid" material helps.

Please refer to  article Sum of reaction 

 



Krzysztof Wasik

Stephane.kapetanovic
Mentor
Mentor

little thing but:

Take care to panel 126 at point 5 and Node 33 of beam 104

Same for panel 211 (node 33 and 441)

 

 

 

0 Likes

Stephane.kapetanovic
Mentor
Mentor

A) 5845 7600 7602 7821 - I have not found these nodes in my model. maybe I have deleted them earlier.
These numbers are objects that are unknown but exist in your model. Just select them from text.

 

Node 8658

Probably no direct reasons

0 Likes

Anonymous
Not applicable

Hi @Anonymous ,

 

Those slabs have defined foundation parameters hence these act as supports themselves and need not any additional supports. 

0 Likes

Anonymous
Not applicable

Thanks @Krzysztof_Wasik !

Reducing E & G solved the problem 99.9%

 

I have now put E for "Super_rigid" material to be 9000 GPa, so about 45 times the stiffness of steel. The differences between actions and reactions are now very small. See attached screenshot. Do you think these differences are acceptable?

I know that I have some instability warnings to take care of.

Thanks,
Aleksandar
 

0 Likes