error in running analysis

[Mahdisadeghi84]

Dear all
I modeled a truss structure and when I run my analysis I got error below:
zero diagonal terms encountered at a following nodes: 59,199,202,...
please help.

[huffte]

I wonder if you might have all members entering said node released for moment at that node? That would create an unstable structure. If so, then one member end needs to have the release removed.

[Stelios_Antoniou]

This message is an indication that at some particular DOFs the stiffness is zero (e.g. a node when only truss elements are connected, hence there is no stiffness in the rotational DOFs).
It is not necessarily wrong, but as Tim mentioned it might mean that the model is unstable. I suggest that you carry out an eigenvalue analysis to see if the deformed shapes are expectable

Stelios Antoniou
SeismoSoft Developer

[Mahdisadeghi84]

Dear Stelios and Tim
Thank you for your response.
I changed some members and I tried again but now I have the pervious error on some other nodes!
could you please take look at the link below (its a screen shot of my model):
http://s5.picofile.com/file/8103774042/truss.jpg
kindly note that I want to perform push-over analysis on my self-support tower.

[huffte]

Mahdisadeghi84,

In looking over your structure plot, it does appear that you have many nodes where the only members connected to the node are truss members. I base this on the fact that you have added the comment on your jpg file that "all others are truss".

Your tower is made up of what we will call (1) columns, nearly vertical members (2) beams, horizontal members connected at each end to a column (3) braces, truss members in the plane defined by two columns or a horizontal plane.

The concern and a possible source of the instability is those locations where only beams and braces are connected to a node. There are many of these. As noted in the SeismoStruct Help System, the only contribution to the stiffness matrix for a truss element is EA, all others are zero.

Currently, only the (nearly) vertical column members are inelastic frame elements. I wonder if a structure such as this, when actually built, would have continuous members where the "beams" are? Perhaps you could change the "beams" to either inelastic frame elements or, if you wish to ignore any inelastic behavior in them for some reason, to elastic frame members?

So if you had inelastic frame members for beams and columns and truss members for the braces only, it could be a positive step towards resolving the instability and might more accurately model the tower. You may still be ably to apply releases to the appropriate ends of the beams - where they are connected to columns.

Of course, you are the one who must decide if my suggestions are appropriate or not. You are the one who know how the tower will be built.

It is an interesting model Mahdisadeghi84. I trust that you will persevere and get a successful analysis. Best of luck.

[seismosoft]

Have you tried running the model with the truss members (eigenvalue, pushover or both)? It is reasonable that you get a warning message about the zero diagonal terms, since in many nodes only trusses are corrented and the rotational DOFs are free, however this is not worng. If the model is correctly structured (as it seems in the screenshot), it should run without problems. Note that the message that you get is a warning rather than an error message.

SeismoSoft Support

[Mahdisadeghi84]

Dear all,
Thank you for your explanation.
I changed all horizontal element (beams) to infrmFBPH and remain all brace ones to truss elements (notice that all columns are remained infrmFBPH).
Now when I run the pushover analysis I didn’t get the previous error but my analysis terminated and it says: Unable to apply the entire permanent load, while as I put the screen shots of my model, I defined incremental load!
http://s5.picofile.com/file/8104034234/1.jpg
http://s5.picofile.com/file/8104034342/2.jpg
http://s5.picofile.com/file/8104034518/3.jpg
http://s5.picofile.com/file/8104034568/4.jpg
http://s5.picofile.com/file/8104034576/5.jpg

Thank you all for you attention and help.
All the Best,
Mehdi

[seismosoft]

It seems that some parts of the model are not correctly connected. Try running an eigenvalue analysis to see if you get any strange deformed sheapes (e.g. the vibration of a unique member) that might indicate a lack of connection

SeismoSoft Support Team

[amir_nayebe65@yahoo.com]

hi
i modeled a 2D moment frame in X-Z direction(3bay + 3 story).i use linke element in all conections.but whan when i run the program it stope by this massege: "unable to apply the entire permanent load" and in front of itration number it has this massage "sol_prp".
befor this frame, i modeled same frame ( 3bay and 2 story) and i runed it withouat any problem.
could you please help me to find the problem?

[huffte]

Hi Mahdisadeghi84.

I think a forum search would have been in order here. If you search on "sol_prb" you'll get a hit that encourages you to closely examine the overall stability of your model.

The following is an excerpt from the SeismoStruct Help system in the section on Convergence/Numerical Instability"

"On occasions, very unstable models lead to the sudden development of out-of-balance forces that are several orders of magnitude larger than the maximum tolerance value. This in turn creates a so-called Solution Problem (i.e. the analysis crashes, albeit in a "clean manner"), and iterations within the current increment are interrupted, with a log flag message equal to Sol_Prb."

Since you have evidently used the same strategy for a similar model but with fewer stories, I would start with a check of things like connectivity, link properties, link local axes, etc. . .

Best of luck Mahdisadeghi84.