here's the second screenshot

hi Richard,

Thank you for your reaction.
In an earlier stage i've tried both of your suggestions but the results weren't satisfying.

In my opinion the best way is using the sliding/ no seperation option between the roller and the baseplate. When i do that, i get a similar messagebox.

On the other hand Autodesk prefers working with simplified assy or parts so in that case deleting the baseplate (and rollers), working with pin constraints would be better.
Question is, how do i use the pin contraints in the correct way so that it simulates the behaviour of a roller (with freedom in the xy plane)?

I also thought about deleting the rollers and introduce reaction forces but i don't know exactly how strong they have to be.

Hello Richard,

I didn't provide you of enough information in my last message, sorry for that.
The goal of this assignment is both learning and checking the prototype.

The 1/4 assumption is not correct. Main goal is to check the stiffness of the frame in case of a (vertical) peak force on one of the two Y-axis rollers. So it's a bit difficult to specify the reaction forces at each small roller.
In normal use it would be 1/4, i already checked that and it sustains.

Niels

You could apply the pin constraint to the outermost pin holes, the ones
without the added flange. There is no way that the loading will ever be
this extreme. If the design satisfies your needs under this scenario it
will under a more realistic loading.

Otherwise you'll need to model a pin that realistically mimics the roller's
actions.

wrote in message news:6283336@discussion.autodesk.com...
When i simplify it, it would be like in the attached image.
Actually the question is: How do i configure the pinconstraints of the four
rollers so that they simulate the behaviour of the rollers?

Neils:

I think I know what you mean. You're trying to have one roller remain
stationary in the X, Y and Z directions. You'd like a second one (co-axial
with the first) fixed in the Y and Z directions so that the frame doesn't
rotate. You'd like the remaining two fixed in only the Z direction.

This way, if the structural members bend, the first roller stays put, the
second can slide in the X direction and other two can slide in the X
direction or roll in the Y.

The pin constraint doesn't appear to give us that level of freedom. If you
remove the radial fix, you loose the constraint in the Y direction as
desired, but also in the Z direction which is no good.

As the frame flexes, these pins exert an unwanted tension or compression.
The question is whether this effect is significant on a structure such as
yours.

I don't know the answer yet. I suppose one could make a cross toggle (pins
in X direction and in the Y direction). Releasing the axial fixes
appropriately might give you the freedom you want.

Richard
wrote in message news:6283501@discussion.autodesk.com
I don't know what you mean exactly.

As far as i know i can select both holes for one pin constraint and it will
behave as a pin or axis like you say. But what i don't know is: how do i
configure those pin constraints? Fix them radial/axial.

When i fix them radial the behaviour is not as a roller but as if you put a
bolt in al the four rollerpostions, not correct if i would say. But when i
leave the radial fix (at 3 of the 4 roller holes) behind, to create the
freedom a roller would have, i get the message that there are not enough
constraints.

Hopefully you understand what i mean.

Neils:

I've just experimented some more.

I know this isn't the analysis you'll want but I think it will demonstrate a
path to your solution.

Forget the pin constraints. Use fixed instead.

The fixed constraint can be applied to a surface, an edge or a corner (and
possibly a workpoint). Apply separate fixed constraints to the corners
circled in red (and their counterparts at the other end.

On the one on the left, check all of x, y and z.
On the one at the bottom, check y and z.
For the other 2, check z.

I think this will adequately constrain the model but permit the freedom you
want. If this works, in the wheel pocket place a dummy member which gives
you a corner to select.

Richard

wrote in message news:6283534@discussion.autodesk.com...
i have already released the axial fixes.
don't know exactly what you mean by "cross toggle"

The results are promising!

The system behaves as expected, only thing that concerns me is the warning i get: "model seperates into 5 independent components" (see attachment). I don't know what that means. I guess it's about the welded frame and the four pins, but i can't figure out what is wrong with them.

my workflow:
- introduced 4 pins with flat side on it (aligned to XY-plane)
- 3 of the 4 got a Frictionless Constraint (so they are fixed in Z)
- 1 pin has got a Fixed Constraint (fixed in X, Y and Z)
- all the Contacts between the pins and frame are made Sliding/ No seperation. I've made a flange on the pins to get a Contact with the frame in axial direction. Otherwise the system applies "soft springs" to keep te pins in place.

A very good approach I think. My only doubt is the message that i get about the "independent components"

Neils:

The fact that the message is merely a warning, and one that makes sense, is
encouraging. I haven't (yet) had enough experience with this particular FEA
system to know if this is as expected or not.

I would think this gives a very good solution.

The down side I find to this is approach is that the pins are going to see
all of the high stresses and might distract you from the serious issues. In
your case the maximum stress appears pretty low so it's no issue but I'd bet
all of the stresses showing red are in the pins, which are of no interest to
you. Meanwhile you're getting no sense for what's happening in the frame.

Select Color Bar, unselect Maximum and replace the value with, say 15 MPa.

Richard

wrote in message news:6283984@discussion.autodesk.com...
The results are promising!

The system behaves as expected, only thing that concerns me is the warning i
get: "model seperates into 5 independent components" (see attachment). I
don't know what that means. I guess it's about the welded frame and the four
pins, but i can't figure out what is wrong with them.

my workflow:
- introduced 4 pins with flat side on it (aligned to XY-plane)
- 3 of the 4 got a Frictionless Constraint (so they are fixed in Z)
- 1 pin has got a Fixed Constraint (fixed in X, Y and Z)
- all the Contacts between the pins and frame are made Sliding/ No
seperation. I've made a flange on the pins to get a Contact with the frame
in axial direction. Otherwise the system applies "soft springs" to keep te
pins in place.

A very good approach I think. My only doubt is the message that i get about
the "independent components"