fem almond cart

[ch.br]

Bye.
I'm about to quote in my first fem analysis: an algear cart for drift as in fugura. the structure is all made of rectangular or square sections welded. I'd like to use hypermesh.
the boat could:
- on a central longheron point
- on a rope tied between one head and the other of the two vertical sections above the wheels. The rope is not drawn in the figure.
so baricentro and mass of the boat, then avoided to draw boat and rope and break in 3 forces: on the longherone and on the two vertical sections.

I have some doubts:
- do I consider it as if they made the cart one or do I have to shape them somehow?
-the wheel hub would not be welded down to the lateral section as now, but would be like the attached one. the parallelepiped should be inserted into the section and screwed. this because it is good that the hub is steel and not aluminum. How do I consider this piece? as resistance is one, but they are different materials.
- The wheels are full, plastic.

Thank you so much in advance to anyone who wants to help me. :smile:

 

[ch.br]

error!
I didn't realize that the second photo is a collage of many photos. this is the image of the hub and its support:

 

[SOLID]

- do I consider it as if they made the cart one or do I have to shape them somehow?

intraprende puoi entrambe the strade.

-the wheel hub would not be welded down to the lateral section as now, but would be like the attached one. the parallelepiped should be inserted into the section and screwed. this because it is good that the hub is steel and not aluminum. How do I consider this piece? as resistance is one, but they are different materials.

1. The photo is small tropo
2. you have given the answer

However it depends on the level of detail you want to achieve.
for example if you hypothesize a whole one you can do the equivalence of knots otherwise you can do a contact analysis.

- The wheels are full, plastic.

and the doubt what is it?

 

[ch.br]

intraprende puoi entrambe the strade.

I was hoping for a little more precise advice. Of course you can do everything and nothing. But what could come to me in this case?

 

[Onda]

begins to make a beam model, without welds and considering perfectly rigid sprays and values stress and deformations. then if it is the case you make a more precise model. depends a lot on what you're looking for, which you didn't specify.
If you want to see global stress and deformations, go great with a beam model.
If you try to determine what happens in detail you need to model it.

 

[SOLID]

I was hoping for a little more precise advice. Of course you can do everything and nothing. But what could come to me in this case?

shape bolted joints, welded and other by what I learned from my modest experience, it is quite complex because you should always correlate the mathematical model with experimental results, to have a certain degree of confidence.
so before thinking about what should be done, you should first understand what you want to evaluate.

for the rest quoto wave

 

[ch.br]

grazie mille. :smile:

 

[ch.br]

I have problems understanding how to assign thickness.

practically my parts are all profiled with a certain thickness, except the plastic wheel and a part of the hub, which are full.

I used pshells, but for solids?
I assumed the command was psolid, but in the component table I get a red box (I attach the image). Is that normal?
Besides: of my profiles I left the outer surface, so if from the analysis comes out that the structure is too weak I can increase the thickness with a click.
but how do I tell him that my thickness is inside the surface I have? I don't know how he came for now.. .

Thank you so much and apologize for the newborn skewers.

 

[Onda]

thickness is in the direction of the normal element.
if not from offsets this is by default place half of the shell.
then you have to give a half offset of the expected thickness.
you have to verify that all the elements have the normal turning towards the inside.
for solids you don't have to give thickness because the element doesn't need solid.
I advise you to make a model with shell elements and solid elements because the joining to the interface is not trivial. it should be done with rbe3 elements but it is quite complex. solid knots only have translations like dof while shell knots also have rotations. so to make a transition you have to tie at least two knots of solids to a shell knot. morals don't!
wave

 

[ch.br]

the full hub part I can see to make it thick sheet if it is a problem, but the wheel is full plastic, there is nothing to do.
What do I do then, do I do everything solid? I get too complicated the model (to say if I had to retouch the mesh) or is it feasible?

 

[Onda]

simply do not model the wheels!
What are they for?
What are you looking for from your model? If you want to see stress and deformations in the tubes that you make wheels?
And then what do you want to do? mold rubber?

make the basic structure model and constraints for example to the hub, then if you want to make a detail model by inserting the loads from the general model.
But I don't think you have to make a wheel model! Just go to the store and buy it!

your model from my point of view is not even to shell but to beam.
so if you need to change tube diameter only change parameters.
if you just like to see the colors on the surfaces you do it shell, at cost of 3 times so much time to model and complications to put the constraints.
at the ends of the tubes where the rods put rigid rbe2s representing the hub and on the master node (independent) put the wheel constraint.
ties on a xyz wheel on the other in yz and to the third support (at the bow) only in z, so it is also isostatic. If you're beaming it in 20 minutes, you did the model.
Hi.

wave

 

[ch.br]

the wheel is not rubber, it is full plastic.
I would have just replaced it with a spring element.

 

[ch.br]

How do I project a knot on a line or surface?
If I go to geometries and choose nodes I have various choices to create them and I no longer find that to make a knot projecting an existing knot on a line or on a surface.

 

[Onda]

even full plastic wheel you can replace it with a spring, just know the stiffness of the object. and however if you restrain it in an isostatic way you can delete it and nothing changes to the model. unless you want to study wheel deformation
I do not remember the command in hm to project on a line, exists, searches on the guide.

 

[SOLID]

If I go to geometries and choose nodes I have various choices to create them and I no longer find that to make a knot projecting an existing knot on a line or on a surface.

go to the tool->project menu
selections to surface or to line and in the menu along vector you must select surface normal.

 

[ch.br]

go to the tool->project menu
selections to surface or to line and in the menu along vector you must select surface normal.

Thank you. They showed me the other week, but I didn't remember.

 

[ch.br]

that you have to do if... surfaces of different parts that seem attached are not?
I meshato everything seemingly good. I have more components. I meshato so that the knots of the joint parties coincide. But apparently it is not so because with simulation the various pieces instead of deforming are detached.
I don't know if anyone
I think there's something I don't understand. How do I solve this?
I attach a figure of one of these points of contact.

 

[SOLID]

You must perform the equivalence of knots.
I don't have the software at hand, but if you're looking for help you'll find everything you need

 

[SOLID]

another command that might come in handy in these cases is find-->attached
allows you to view all the elements attached to the selected ones.