Parking Roof Design

[KonstantinBG]

Hi,

I have been trying to design my parking roof utilizing STAAD.PRO. I would like to ask some questions about stuff that amazed me, perhaps someone may help.

Let me tell you some words about the roof - basically the roof is a cantilever type (maybe some people call it trellis) having only one column to the ground foundation.

This is a picture of the roof structure in 3D.

And a picture of the roof structure in 2D.

The next 3 pictures illustrate the sections that I chose. The sections marked in red are corresponding the highlighted section on the right side.

The next picture is the load definitions.

This is the Steel Deisgn Tab and the commands I defined.

Now my questions and the weird thing that amazed me.

In the output file the only beam that fails is the beam which is furthest away from the column or in other words the beam which is at the highest point of the roof. This makes sense and STAAD.PRO offers me another beam which is 120x60x3 mm compared with the prevoius - 60x60x3 mm.

Now, for the columnds I chose 140x140x6 mm beam. It does not fail but STAAD.PRO offers me a ridiculous beam - 40x40x2 mm. This is really weird for me.

What do you think does that make sense and should I trust STAAD.PRO or just use the beams I defined?

And perhaps you would like to see the Maximum Resultant Displacement diagram.

Maximum Displacement for DEAD Load.

Maximum Displacement for SNOW Load.

Maximum Displacement for Combination.

And another thing is that STAAD.PRO does not connect for some reason the roof beams(120x60x3 mm lie down on the column) to the beam at the highest point of the roof. I tried to divide the beam at the highest point but it just divides the beam into smaller beams in between the roof beams.

In conclusion I would like to know your opinion.

Am I doing the design correctly and is it going to be safe to build the roof?

 

[canwesteng]

Don't know that STAAD will properly design the curved beams, I would just leave those out of the model as they are likely just architectural features and you should be stable without.

 

[KonstantinBG]

Actually the idea of the curved beams was to add additional support to the whole structure but I think that prorably STAAD.PRO does not consider them at all since they do not deform.

 

[KonstantinBG]

Any other opinions?

 

[JAE]

I gave up on STAADPro in 1990.

I would suggest that you get help from those who really know that program, understand all the parameters of your design, and can work with the model directly to see how you input your data and what assumptions you used.

The only people that I can think of that would fit that description would be your in-office peers, or supervisor, or send it to STAAD's technical support for them to look at.

Or post in the STAAD forum here: forum750



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[Canuck65]

I don't use STAAD, however, in other similar programs I have had success modeling the curved beam as series of connected straight sections. Perhaps break your curved beam up into 300mm long straight segments and at least see if the model will work and assign forces to the members. Then you will likely have to look at the forces in the segments and design the curved beam by hand from the data.

 

[TLHS]

You should be starting with new software by doing analysis that you know the answer to, and do a proper verification. Not so much to see if the program is wrong, but to figure out how it actually defines everything.

Stuff I immediately see:

- It looks like you're using a service load level load combination for strength code checks.
- You don't have any deflection criteria set, so the select command will just pick sizes based on strength criteria. Personally, I think the select command is more trouble than it's worth and I just pick my own sizes.
- With the way you've got multiple analysis and check commands one after another, it's reasonably possible that you aren't getting the results in the GUI that you think you are.
- I haven't tried to use curved members in years, but it's definitely limited. You really need to read the guides to figure out what it's doing. I suspect it isn't doing design calcs on them (at least fully) and it may not be doing deflection related things correctly, in which case the general structure's stiffness matrix will be wrong.
- Why don't you know if the columns are being designed wrong? You have 'Track 2' on, so go into the output file and review the design values against what you think they should be. Then you can figure out what it's thinking
- Do an animated deflection diagram and review it to check for continuity and to ensure that things are all deflecting the way they should be. The animated deflection tool is the most useful check in the program
- It doesn't look like you have any releases defined. Is every connection fully rigid? You're likely getting a different response than you want with three members rigidly connecting to the fixed base foundation on the left side.

Also, read your output file.


As far as items not being connected, every member has to have a node at the point you want to connect. If you just have lines running through each other with no node you can intersect those lines in the geometry drop down to create a connected node.

Non-typical members are not a great way to learn new software.

 

[TLHS]

Also:
-how is the dead load only the weight of the members? There must be some sort of decking, secondary structure, or similar.
-even if your structure has fixity problems, the snow load deflection doesn't make sense when compared to the dead load deflection. I suspect your snow load isn't being applied right. It looks like it might just be being applied to the perimeter beams or something.
-The deflection in those views will just be at the nodes, so you don't have a real feel for how the beam being supported by the curved member is deflecting, because there's no nodes in the area of max deflection
-Your columns are likely seeing close to pure vertical loads, to get realistic sizes you need to have lateral load cases to get bending moments into them