time history analysis

[dziub7]

Bonjour,

je veux faire une analyse temporelle d'une structure non lineaire (passerelle haubanée).

J'ai mon cas 1 poids propre, que j'ai transformé en masse

cas 2 analyse modale

cas 3 le cas de charge de piéton

cas 4 analyse temporelle, methode de Newmark, j'ai défini ma fonction et mon cas 3

je n'obtiens pas les bonnes valeurs d'accélération.

quand j'affiche les réactions, j'obtiens presque les memes valeurs dans le cas 1 et 3

est ce que le problème vient de là??

merci d'avance

[dziub7]

"As far as I undestood you \ we got very high accelerations in this model that's why we had to dense calculation steps to get maybe more accurate results.

UNfortunatelly such model did not converge."

if you read message #14, high accelerations are found in particular nodes, i don't know why!?

results in nodes 93, "seem" to be correct, but i didn't analyse long time enough.  

Im doing it during 50s right now but calculations take a very long time!!!

I hope you are calculating it with Pdelta.

Yes I did!

concerning your setting: time step of 0.002s may be too long to calculate, don't you think?

Do we have any choice? This is the setting which MAY lead us to convergence. We will see.

perhaps!

I already did time history analysis with Robot on a simple linear model, and I found the expected results, but on a simple example I analyse during 300s, the acceleration becoming permanent after 50s. Calculations were not so long. Damping was the same: 0.004. And there was no particular elements.

[Rafal.Gaweda]

Calculations were not so long. 

Because it was linear model.

[dziub7]

hi,

I still try to do my THA,

i found something strange, i want to send you a file but it is 33Mb

in filemail they ask me for a emaiil address, how can i do?

[Rafal.Gaweda]

send it to your address then post only the link here

[dziub7]

Ok,

https://www.filemail.com/dl.aspx?id=VXZFLAHGXSNFAIK

you will see, the analyse stop at step 548/1017

if you look at the mast at this step, he is buckling

what is weird: it has bending moment at foot, although it is a ball joint...

I don't understand!

[dziub7]

Hi,

anyone have an idea?

I tried to divide the mast, and the buckling appears on new little element of the mast.

thanks

[Rafal.Gaweda]

We got the same divergence...

[dziub7]

do you know the reason?

[dziub7]

also, at the very first step, the mast have bending moments in both direction at foot...!

[Rafal.Gaweda]

also, at the very first step, the mast have bending moments in both direction at foot...!

---

You may see sometimes such effect in 1st step wht all loads are started to be apaplied 

Results from step 2 till divergence are OK.

[dziub7]

why the divergence??

[Rafal.Gaweda]

do not know yet. I guess it must be checked by developers.