hard points

[volaff]

the previous thesisist has meshato section per section and I nn wanted to do nothing but follow the "sue footsteps".

If the section gives me a mistake, you should do it on the whole moedello.


cmq also gives me the mistake.
Don't you come out?

ps: I believe that meshare section per section is also better to make any subsequent changes: I wanted to create a text file for each section so as not to repeat the "hand" operation later.

 

[eih64]

the previous thesisist has meshato section per section and I nn wanted to do nothing but follow the "sue footsteps".

If the section gives me a mistake, you should do it on the whole moedello.


cmq also gives me the mistake.
Don't you come out?

ps: I believe that meshare section per section is also better to make any subsequent changes: I wanted to create a text file for each section so as not to repeat the "hand" operation later.

I'm sorry, but I have to ask what you mean meshare section x section... you mean area x area because otherwise you think wrong....
serve the final db you produced and the area on which the problem occurs....
Otherwise you do not get out of the impasse.

 

[volaff]

practically the model "pannello" per section, a sort of "components" of which the panel is constituted (see attached image taken from the photo you posted).

each section is constituted by a number of areas, and I wanted to meshare one section at a time (in practice meshare more areas that constitute the section).

the area of the section that gives the problem is the one that contains the fact hardpoints.

the "final" db is basically what you also have from which I started to make meshare.

 

[stefano.garbin]

eye that "section" is a misleading term in how you use it.
ansys uses the term "section" to define sections of beams, shells, joints etc.

 

[volaff]

I ask vein....in this case are section inherent to shell elements (the panel is made of composite material made of fact).

 

[volaff]

Hello, everyone. I'm finding considerable difficulties in realizing the mesh of the areas that contain hardpoints and I can't come to the top.

I thought if it was possible to erase all hardpoints and sift them with simple keypoints.
the problem, however, is to realize the mesh in such a way that at each keypoint corresponds a knot.

sec you this thing is feasible fairly easily or should you necessarily use hardpoints? ?

ps: the nodo-keypoint(hardpoint) coincidence is necessary because, subsequently, a beam4 element will be inserted between each pair of knots that simulates a rivet.

Thank you very much

 

[stefano.garbin]

to each keypoint associated with lines-aree-volumi meshati always corresponds a knot.
then, for the rest, I repeat my advice: do not remain "blocked" at the work of the thesis before you. if possible rethink your problem, if necessary from scratch, as many have recommended on xansys.org
remove from the boxes the hardpoints (in so many have told you that they often create problems), and break your geometry so you only use keypoints for your rivets.
If you want to explore the option of workbench export an iges file and import it with design modeler. control geometry well, if necessary repair it, and then go with mesh.

 

[volaff]

Thanks for the answer.
fundamentally I would do that but I have import repobles in when from the classic anesys I can not generate the iges file as it gives me errors connected to hardpoints.

The problem is twofold: on the one hand there is my prof who says "do not make your own mind", on the other hand the program that tells me "error error".
from the end of the exams has passed more than a month and are always at the same point.

I have now noticed that there are hp repeating and this could be the cause of mesh failure.

I swear I'm really demoralizing. and only this is missing as a step.
Thanks for the seizures.
Have a good day!

 

[volaff]

cmq to think about all the tools what I know to use a little better is just ansys classic apdl therefore, sec me, it is better to continue working in this environment.

You'd better explain to me what you mean by "thick your geometry so you can only use keypoints for your rivets. ".

Thank you very much

 

[eih64]

cmq to think about all the tools what I know to use a little better is just ansys classic apdl therefore, sec me, it is better to continue working in this environment.

You'd better explain to me what you mean by "thick your geometry so you can only use keypoints for your rivets. ".

Thank you very much

It was said and laughed... and there are also files sent... when something is posted.... give us an eye...

 

[volaff]

Thank you very much for the answer but as not said.

the prof tells me about continaure to use hardpoints.
In practice he told me that:
"The problem could be due to the fact that hardkeypoints are very close to the area.
try to create sub-areas with smaller esize. "

I have to take the areas that give problems (the ones with the hp) and divide them further with a small "cut".
Hopefully.

thanks to all for courtesy and availability.
tomorrow we see what I can pull out by dividing the areas.

 

[volaff]

I tried a few combinations to make mesh but always gives the usual mistake.

to avoid mesh problems, according to you, I could create a mesh "mapped" (type grid to understand us) and not free: so I believe (and hope) that the problem is "less".

in the morning I get busy and we hope well (cross fingers)!

Good evening to all!

 

[volaff]

I noticed something strange.

there are some hp that have a double identity i.e. one thing like:

hptcreate, area , 235 , 10400 , coord , 298.00 , 294.90 , -74.4 ,

♪

hptcreate, area , 449 , 10400 , coord , 298.00 , 294.90 , -74.4 ,

in practice the hp has the same identification num, the same coordinates but num different area.

if plotted the two areas, however, are completely different.

Does anyone want to tell me if it's a mistake or can it be?

Thank you very much

 

[volaff]

ps: the prof told me to insist with the hp but, in the end, I decided to follow another road, that is to replace the hp with simple kp and then build a mesh "ad hoc".
In this way I can also go compare the results to see if I made mistakes in one model or another.

I'll keep you up to date because it's only because of you that I did something else... I saw it very hard.

I wish you a good Sunday to all!
And... thank you for everything!

pps: hp is quite a bit, so today, I think I will have to waste a lot of time erasing and recreating

 

[eih64]

In fact... to erase are less than 1 sec. hptdelete,all and to create them there was not a txt file where the coordinates of the hp were inserted? .
I see it really hard for you.

Good job

 

[volaff]

Bye.
you are right there was a text file where the hp coordinates were inserted.

I realized something yesterday.
If I go to view the hp coordinates by opening the model on "klist" and I go to compare the coordinates with the text files I have, I realize that they have a different z share.

in practice with klist ho:

no. x,y,z location kesize node elem mat real typ esys
6016 0.207e+04 897. 2.10 0.00 40101 0 0 0 0

while the text file says:

hptcreate, area , 299 , 6016 , coord , 2067 , 897.4 , 3.70 ,

as you can observe the coordianta z is different! !

the prof told me:

"probably the area on which the hkp are created is positioned at the share other than that of the command and therefore, compared to the command, the hkp are moved. alternatively check if the reference system is always the same"


even if I have no idea how to control this.

Thank you.
Have a good day!

 

[volaff]

I take a part of the speech I wrote in another topic because, in fact, a question follows part of the speeches already addressed in this discussion I read again from the beginning to the end.

Unfortunately I realized that working with hp is boring and annoying as you always fall in error by the software: In fact, however, "necessary" I find myself embracing this cross and moving forward ("cit of prof : he must work with the hp as we make everything semi automated").

I have to say that I'm right, but the hero in the hip is pretty tough.

as it patiently suggested to me stefano (which I always thank for the disability) (and then eih64) I identified the areas that contain the wrong hp and to solve the problem I followed the following strategy:


1.I have selected the wrong hp (hardkeypoints)
2.ho selezioanto nodes and elements (beam 4) attached to the hp and I deleted them
3.Clean the mesh of the areas that contained the wrong hp
4.I created the correct hp
5.re-mesh of the areas I had previously cleaned up (point 3)
6.created new elements (beam4) between correct hp pairs.


all this on the model "final" already loaded,vincolato and meshato.

if I go to launch static analysis (which on the original model gives about 25600 warnings) the following error appears:

_
the model data was checked and there were found errors.
please check output or errors file (
c:\users\desktop\ university\tesi\lanci\file.err ) for
messages.
"
this is the content of the file.err.
They suggested that I detach the mesh from the solid model but in the end I have to present a "classic" model so I don't know how much it is to do this (which I don't even know how to do between the others).

I apologize for the length.
perhaps this is not even a question/problem but only a small rash and a great thanks to those who accompanied me (helped) with ansys.

good Sunday to all