Is there a way to join panels together in Robot?

[Anonymous]

Hi,

Newbie here. I'm looking to model a one-way slab in robot.

The slab spans 3.5m, with line supports at 0m and 3.5m.

Between 0m and 0.5m I want a depth of 0.15m. Between 0.5m and 3m I want a depth of .25m. And between 3m and 3.5m I want a depth of 0.15m again. The top of all three panels should be flush. Poory drawn diagram attached.

I have tried to model this by drawing 3 rectangles in top view, and turning them into panels of different depths. I quickly discovered that to get the panels flush at the top I would need to draw the central section slightly lower than the two at the edge. 

Once I had all 3 panels flush on top I put the line supports in. When I hit go however, the middle part of the span just dropped out, and then 2 at the edge just folded in on the linear pin supports. I got lots of error messages.

How do I join the panels up so that this doesn't happen?

Or am I going about modelling my slab in completely the wrong way? If so, what other ways are there to model my slab?

Many thanks,

Will

[Rafacascudo · Accepted solution]

Link the nodes of the 2 panels using full rigid links

[Anonymous]

Thank you. I will try that this morning and let you know how it went.

[Anonymous]

Do I need to mesh the panel first?

[Anonymous]

Okay I went away, meshed my panels, and then rigid linked the new nodes. Unfortunately I'm getting an instability (type 2) in direction Ry detected at the corner nodes of the 150mm panels where I have rigid links to the 250mm panel.

The line supports supporting my panel from beneath the middle of the 150mm slab (as shown in poorly drawn diagram) have been left free to rotate about direction Ry because this is how the support will work in real life.

Is this what is causing the error, and what can be done about it?

[Artur.Kosakowski]

Could you attach your model?

[Anonymous]

Hi Artur,

I've attached my model. If you could have a look and tell me where I've gone wrong that would be great.

Cheers,

Will

[Artur.Kosakowski]

As suggested by Rafacascudo the rigid links should have all d.o.fs blocked

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Rafacascudo]

By the way, your panel mesh is pretty bad. For a 2,5mx1,0 m panel you should have at maximum 0,25m as mesh size on meshing options.

[Anonymous]

Thanks for the tip. I thought I had a 0.1 squared grid but it must have changed back. Is there anywhere I can find rules of thumb on meshing like the one you just stated?

Also for some reason I'm getting super lower deflections. All my nodes have zero deflection when I hit results which is strange.

[Rafacascudo]

Set the displacement units (Tolls/job preferences/units and formats/other) to mm and increase the number of decimals. If still zero , send the model

For the mesh size , depending on the importance the panel has on the analysis ,I always try to divide the smaller panel side in at least 5 segments.

The more important the panel , the smaller will be the mesh size.

[Anonymous]

Hi

Rafacascudo

Increasing the number of decimals does indeed reveal small, albeit non-zero deflections. At first I wondered if I had messed up materials or section properties somehow, but a quick hand calculation indicates that 0.16mm deflection at centre is reasonable, even if it doesn't sound like it. L/270 for a 3m span gives a max deflection of 11mm, so I'm miles under this, and almost certainly over designed for serviceability.

Interestingly I've had a look at reactions at my line supports, and they are not symmetrical at the edges (Fz=12.05 vs Fz=11.92 as shown in screenshot) I will have a go at redrawing to see if I somehow messed up a coordinate somewhere. I wouldn't be concerned if I got this on a moment distributor that did a limited no of passes over the structure, but FEA should solve all points simultaneously so the fault probably lies with me. Other than errors in drawing my panels or line supports, is there anything else that might have caused this?

[Rafacascudo]

On the Reactions tab set it to show linear values. There are 2 options that can be shown graphically, average or total value.