Error

[mprearo]

Good afternoon,

I cannot solve this error; i also see on this internet page 

http://wikihelp.autodesk.com/Autodesk_Simulation/enu/2012/Help/0220-Setting_220/0496-Set_Up_A496/050...

but i didn't find the solution.

Thank you in advance

Best regards

Manuel

[S.LI]

See the attached file.

Do you set these beam cross-section properties?

[mprearo]

No, i don't.

Can you explain me this concept???I don't know what is beam cross-section properties; i'm trying to learn to use this software by myself and several tutorial.I'm an Algor beginner.

This error is due to the algor bolt, isn'it?

Thank you in advance

Best regards

Manuel

[S.LI]

Beam element in FEA is a one-dimension element. All detailed cross-section information (round, I-shape, c-shape) does NOT exist in FEA formulation explicitly. But it's definitely necessary.

the attachment in my last post shows a simple example on how to set them. For fully details, you'd better go to search "beam/crosssection" in help file.

To simulate bolts, round cross section should be good enough.

[mprearo]

I follow your suggest

thanks a lot

[John_Holtz · Accepted solution]

Hi Manuel,

I can see that Part 3 has an element type of beam. This is what the error message is related to. (The solver does not know, nor does it care, that you want to think of part 3 as "bolts". It only sees the element type .) Edit the element definition (Definizione Dell Elemento) and make sure that the following cross-section properties are defined:

• A = area
• J1 = torsional constant (relates the angle of twist due to a torque load)
• I2, I3 = area moment of inertia about the 2 and 3 axes.

The other properties (S2, S3, Sa2, and Sa3) can be 0.

I think S.Li was trying to demonstrate that if you click in the row for the beam cross-section properties (probably Layer 1), you can then click the "Cross-Section Libraries". From there, you can choose "User-Defined" and pick "Round". By entering the radius, the software will calculate the cross sectional values. (Just to complete the discussion, the library also has many of the AISC cross-sections, such as W beams, channels, and so on.)

[mprearo]

Thanks a lot John. I don't know what will I do without you.

If i undestrand well the solver recognize only the type of element that, in the bolt case, is a beam; so the software created a beam in a FEA environment but it doesn't set the feature/properties. isn't it?

I'm used to generate bolt in Inventor and that software make each features by itself.

Thanks again

Have a good weekend

[mprearo]

Last question.

I do this simulation to see the effects on the plat by the applied force on the shaft.

The result, in my opinion, is wrong because it's impossible that the strain is concetrated only in the application point of nodal force.

What's your opinion???

[S.LI · Accepted solution]

It's possible that high stress appears at specific points, such as loading points, constrain points.

If the load in your case is not a point-force, you can try to split it into several, and apply them on multiple nodes. Or you can choose surface force instead of nodal force.

This really depends on your case.

[mprearo]

Thanks for your help.

I follow your suggest by applying a " surface force " and the software return me an expected result.

Is there a reason that Algor produce the previous nodal result??? is better that everytime i'll use a surface force than a nodal force???

Thnak you in advance

Have a good weekend

Manuel

[S.LI]

Compared with a surface load, a nodal load could generate kind of concentrated stress response. This sounds pretty reasonable and fair, since a surface load is a kind of distributed load.

Basically, both loads should generate similar results at places FAR away from the loading position.

In another word, the effect of how to apply a load should reduce dramatically with distance.