1d modelisation for discrete beam

YOYO

Dear Mates,

I'm a beginner to Code-aster. I want to make a 1D mode analysis of a discrete beam
I leant how to make a mode analysis for a continuous beam but for such a discrete beam with rigid bar and hinges connected them, I don't know how to define the hinges and apply the mass and stiffness on it. I read the manual of Code-Aster and there are several kinds of discrete elements such as CARA='K_TR_D_L' but I want to define the hinges as 0D elements and the bars as 1D elements as well as the modelisation if I should use MODELISATION=('DIS_TR', ) or not. Could somebody help? Thank you very much!

jeanpierreaubry

YOYO
i suggest you read my first book:
chapter 2 to 7
how to add a point mass on a beam
how to add a end release on a beam
this should help for a beginning
chapter 15
modal analysis of a beam

YOYO

jeanpierreaubry

Hi,

Thank you Jean. I read your book and now I know how to add point mass and end release.
But there's another question. I want to make a lateral modal analysis of this system but if I use a bar element, the frequencies given are in the longititude direction. Meanwhile if I use a beam element with high Young's modulus, it seems like the result is not correct.
Could you tell me how to use a bar element to make a lateral modal analysis, maybe it's better to use liaison_solide?
Thank you.

jeanpierreaubry

YOYO
please share your files
as i do not understand what you mean

YOYO

jeanpierreaubry

Attached is my .comm file. This is a 1d modal analysis to get the eigenfrequencies and I tried to use 'DIS_TR' on a 0d element. Then I used M_TR_D_N to add a point mass and I tested it to prove that it works well. But when it comes to the K_TR_D_N to add a rotational stiffness, it doesn't work. No matter how I change the value of Ry, the eigenfrequencies doesn't change. I guess that the when we calculate the eigenfrequencies, the point rotational stiffness will not affect the stiffness matrix... Could you help with this? Thank you.

stage-1.txt

2kB

YOYO

Ah and I tried to use element Barre but it doesn't work well since it is not an axial analysis, so I tried to use a beam element with big or 0 Young modulus....

jeanpierreaubry

YOYO
the .comm file is valuable with a given associated mesh
it maybe ok with a given mesh and wrong for another one
without this mesh i cannot help

YOYO

jeanpierreaubry

Sry I don't know why I can't upload a .med file here even I compressed it to a .rar or .zip file.
Here is a photo. The mesh is very simple just a beam simply-supported both ends with a 0D element in the middle.

jeanpierreaubry

YOYO
post the mesh!!!!
maybe on the cloud
it maybe simple to your eye but what you have done may not be what you want to do

YOYO

jeanpierreaubry

Attached are the links of the mesh file .med and the .hdf file. Thank you.
https://drive.google.com/file/d/1mno5XRiqPSAFMSQL87Npj7S-ATtT7E5t/view?usp=drive_link
https://drive.google.com/file/d/15F6SUox7RBNiKACMkal0I48Y2ByZT2tm/view?usp=drive_link

jeanpierreaubry

YOYO
i cannot see anything that qualify an hinge around HINGE group
the beams on right and left of this point are perfectly rigid linked
read appendix 3.4 in my book

YOYO

jeanpierreaubry
I added a seg2 element and to keep the symetrical properties I seperated the mass into 2. For the seg2 element I used K_TR_D_L with 0 0 0 0 0.78 0 to simulate a rotational spring and I change the beam elements to bars. But code-aster refuse to give a eigenfrequency. It seems like code-aster cannot find any eigenfrequency of this system.
https://drive.google.com/file/d/12vzJ_kEmvvFC9d7v_l6BmcOfOdtlC_UD/view?usp=drive_link
https://drive.google.com/file/d/1DokTgUt_cCWE2Bc6nRDaBJ_t4qldPuZN/view?usp=drive_link
https://drive.google.com/file/d/1EDHOxB3RjRgBWyfICdFCSXXpZ82w9aVn/view?usp=drive_link

jeanpierreaubry

YOYO
we are running in circle i do not understand what you intend to do
there seems to be very strange numerical vales in your analysis
please provide here a sketch of the problem and of the expected mode shape with the analytical value you calculated by hand

and when you share a new version of a file please change its name otherwise we will get lost

jeanpierreaubry

with very first look
it will never work with BARRE i.e. rod, you have to use POU_D_E i.e; beam
several rods one after the other in a continuous line as in this mesh will behave like a folding ruler i.e; folds on itself

before going to the relatively high level of adding mass and hinges solves this problem as a continuous beam until you get solution that is ok with you analytical calculation

YOYO

jeanpierreaubry

Yes exacly like a folding ruler! In fact I want to find the eigenfrequency of a stucture with rigid bars connected with rotational spring as hinges. The problem is I want to set all the mass on the hinges and the stiffness is also all concentrated on the rotational spring as RxY. That's why I try to set the bar element with high stiffness and 0 mass....but as you mentioned, the bar elements are used for axial analysis but I'm now trying to find the eigenfrequencies in the lateral direction ...

jeanpierreaubry

YOYO
i am afraid you do not seem to really understand what you writing in the .comm file
for example with this

_F(CARA='K_TR_D_L',GROUP_MA=('HINGE', ),VALE=(0.0, 0.0, 0.0, 0.0, 0.7854, 0.0))),

you create a NO link (free floating) connection in DX, DY, DZ, DRX, DRZ at the ends of the element which means it is not connected to anything!!!!!!!

please provide the mode shape you expect and the value of the first natural frequency you calculated by hand

YOYO

jeanpierreaubry

I tried to use a POI1 element or 1 SEG2 element with DIS_TR as a spring in the middle and setting K_TR_D_N with dx dy dz as EA/L, DRX DRY DRZ as EI/L with E=10000, a cercle section with r=0.1(so I=pir⁴/4 and A=pir²), L=1 to simulate a rotational spring and set M_TR_D_N 3.14 0 0.... as a concentrated mass. For the beam element I used POU_D_E and I set big young's modulus E=1e12 and small pho=1e-12 to simulate a behaviour of massless rigid bar. I set DRX=0 DRZ=0 DX=0 DY=0 to simulate pure bending senario with the deflection in DZ and rotation in DRY and set DZ=0 at both end as simply supported boundary condition.

I expect that the eigenfrequency will be 0.08 to 0.2 for the pure bending and there will be some other high eigenfreuqncies for axial or lateral vibration.

But it doesn't work... I can find some small eigenfrequencies for axial vibration but cannot find the expected eigenfrequency for pure bending...I don't know where the error is. I upload my files that stage 1, mesh1 and study 1 are with SEG2 element and stage 2 mesh2 and study 2 are with POI1 element. Could you help? Thank you.
https://drive.google.com/file/d/1E9DYguI4nms23KBhWcxbyvTc6AtB2kPm/view?usp=drive_link
https://drive.google.com/file/d/1yfCnzsuxul-8MjMxLj4o_ywhqYysJYUj/view?usp=drive_link
https://drive.google.com/file/d/1XfhQ7-MP-hzITr9ZsrvVqxRv0144_VFP/view?usp=drive_link

YOYO

jeanpierreaubry

Oh with SEG2 I got the right answer. Thank you very much!

jeanpierreaubry

YOYO
why it does not work is usually quite well explained in the .mess file, did you read it
is that the behavior you expect,