Discussion Groups

Robot Structural Analysis

Reply
Active Member
s073714
Posts: 10
Registered: ‎01-31-2012

Problems with time-history analysis

997 Views, 8 Replies
06-06-2012 12:09 PM

Hi

 

I have tried to make a dynamic analysis of a stair-structure.

 

First I have made a Modal-analysis only applying mass from self-weight. I have obtained, a modal frequency which result in F = 6,81 Hz (DS/EN 1991 acceptable criteria fn > 8Hz).

 

1) Is it possible to run a video simulation of each modal deflecting behavior? Or is it only possible to see each mode as a picture?  

 

Secondly I have made a Time-history analysis. I have applied bar-properties of only tension of the stiffening bars in-between the load-bearing main elements. This result in a non-linear analysis.

 

 

A : non - linear model: I have made a time-history analysis based on a non-linear dynamic analysis ,Newmark method (acceleration). I think the results looks strange, since the displacement and velocity seems to appear with the same amplitude though time, but the acceleration amplitude seems to grow. Is there something I have forgotten to consider while creating the analysis? 

 

B: linear model: I have modeled the stiffening bars elements with releases (instead of advanced properties, tension). I have run a static analysis for the load-case, which I want to analyze. The stiffening bars, which result in compression, are deleted. The time analyze are thereby investigated with a linear analysis approach. Is it possible to use this approach as a substitute for the non-linear model?

 

How long time due I have to run the analysis to be sure that I achieve the maximum acceleration?

 

Results are attached and so is a zipped file.

 

Best regards

Birgitte

 

 

IN-data

2 different load-cases are considered

Point-load

The following dynamic loading is applied in node 4:

P(t)=p*sin(360*f*t)   

t=[0;10]      p=2KN     f=3Hz

uniformed-load

The following dynamic loading is applied in node 4:

P(t)=p*sin(360*f*t)   

t=[0;10]      p=2KN/m^2     f=3Hz

(Results are attached)

 

No extra damping is applied to the system, so I have only applied the structural damping for steel psi=0,05.

I have made 2 different approaches:

 

Comparison A and B – for point-load case

As can be seen from the results the displacement and velocity is similar. However the acceleration is differs very much between case A and B.

Please use plain text.
Valued Mentor
tony.ridley
Posts: 653
Registered: ‎09-07-2011

Re: Problems with time-history analysis

06-10-2012 02:32 AM in reply to: s073714

In answer to point 1)

 

When looking at the mode shape (results-------->diagrams for bars---------->deformation) click on "animate"

 

Other points I'll have a look later.

 

Tony

Please use plain text.
Active Contributor
MARIOS75
Posts: 30
Registered: ‎01-04-2010

Re: Problems with time-history analysis

06-11-2012 12:04 PM in reply to: s073714

Probably there is something wrong with the Rayleigh damping. Those diagrams make no sense.

You should expect  after a few seconds the response to be almost sinusoidal.

Try to assign the 5% viscous damping to the first and second frequency (ω) from the "damping" menu.

Please use plain text.
Product Support
Rafal.Gaweda
Posts: 5,581
Registered: ‎04-26-2010

Re: Problems with time-history analysis

06-13-2012 01:11 AM in reply to: s073714

1. Damping is not set (take a look below - red rectangles)

2. You are looking at results (calculate model) in the time range of acting forcing function so no influence of damping can be seen. Run calculations for longer time than function acts (take a look below - blue rectangles)

 

damp.jpg



Rafal Gaweda
Please use plain text.
Active Member
s073714
Posts: 10
Registered: ‎01-31-2012

Re: Problems with time-history analysis

06-13-2012 05:26 AM in reply to: s073714

Hi

 

Thank you very much for the answer.

 

I applied the damping and yield other results. However the damping applied of Zeta=0,05 is considered relatively high for a relatively flexible staircase structure.

 

I have applied a damping ratio of Zeta=0,01, due to the literature see attached file. The result in Robot seems to stabilize for the displacement (attached file). However the velocity seems to increase a little in the last part of the time interval. The acceleration seems to increase a lot though time. Can this be correct?

 

When damping is applied I assumed, that all motion parameters (displacement/velocity and acceleration) would decrease and stabilize. I have to compare the maximum acceleration of the structure to a limiting acceleration requirement.

 

Best regards

Birgitte

Please use plain text.
Active Member
s073714
Posts: 10
Registered: ‎01-31-2012

Re: Problems with time-history analysis

06-14-2012 02:54 AM in reply to: Rafal.Gaweda

Hi

 

Thank you very much for the answer.

 

I applied the damping and yield other results. However the damping applied of Zeta=0,05 is considered relatively high for a relatively flexible staircase structure.

 

I have applied a damping ratio of Zeta=0,01, due to the literature see attached file. The result in Robot seems to stabilize for the displacement (attached file). However the velocity seems to increase a little in the last part of the time interval. The acceleration seems to increase a lot though time. Can this be correct?

 

When damping is applied I assumed, that all motion parameters (displacement/velocity and acceleration) would decrease and stabilize. I have to compare the maximum acceleration of the structure to a limiting acceleration requirement.

 

Best regards

Birgitte

Please use plain text.
Product Support
Rafal.Gaweda
Posts: 5,581
Registered: ‎04-26-2010

Re: Problems with time-history analysis

06-14-2012 06:09 AM in reply to: s073714

I was checking linear file.

It takes some time to calculate nonlin model. Be patient.



Rafal Gaweda
Please use plain text.
Product Support
Rafal.Gaweda
Posts: 5,581
Registered: ‎04-26-2010

Re: Problems with time-history analysis

06-14-2012 08:10 AM in reply to: s073714

Enlarge Division number:

(Calculations last ~2,5h)

 

div.jpg



Rafal Gaweda
Please use plain text.
Active Member
s073714
Posts: 10
Registered: ‎01-31-2012

Re: Problems with time-history analysis

06-18-2012 02:23 AM in reply to: s073714

Hi Rafal

 

Thank you again for the help. I will try the suggested and see if I obtain better results.

 

I thought that I posted my previous mail incorrectly, and that is the reason for the double mail. It was not my intention to rush. The help I resieved from you have been very helpfull.

 

Thanks again very much

  Birgitte

Please use plain text.