SAP2000: Resultant moment increasing with larger members

[IlliniEI]

I have a 3D model of a storage shed that I'm designing in SAP2000. I'm having trouble wrapping my head around the results I'm getting. The typical section view is a standard frame with a gabled roof (one stud on each side, pitched truss connecting them.) There is a significant wind load acting on these members. The top track that connects all of these sections together along the length of the building is experiencing a very large moment near the endwall of the structure. To counteract this moment, I have tried several changes. I've built up the top track to make it larger, I've doubled the amount of studs on each side, and I've even considered two trusses acting together. My question is, the more I stiffen up this connection with larger members, the greater the resultant moment becomes in that top track. I'm not understanding why this is the case. I assume it's calculating the resultant moment from F=Ku. I don't see why F should ever change if the loading is not changing. If K increases, should u not just decrease proportionally? It's been a few years since I've taken FEM and may be overlooking how SAP2000 determines its results. Any help would be greatly appreciated.

 

[Blackstar123]

This seems logical to me because, the higher the stiffness of a member, the more load it will attract. thus, if there are different load paths, the stiffer path will attract the more load.

This phenomenon of load attraction is more evident in indeterminate structure where all members should show consistent deformation otherwise there will be compatibility issues. This is also easily understandable from the definition of stiffness i.e K = F/u (force required to produce unit displacement). Lets say you have two members with different stiffness meeting at a joint which is displaced up to 2 in due to applied force. What do you think which member will require more force to deform up to the same units?

 

[IlliniEI]

Thank you for your response! I agree with you completely that it makes sense from a theoretical standpoint, but I can't make sense of it from a mathematical standpoint. I am simply considering if I look at a 2D structure that no matter the stiffness change, each member will experience the same resultant forces. Increasing the stiffness would only serve to decrease the deflection. If you're saying that a localized difference in stiffness from a 3D point of view restricts this change in deformation and therefore increases the resultant forces, I can understand that, but how do you design for this? I build up a member and get punished for doing so. That is the part that doesn't make sense to me. The smaller member fails strength capacity by 4%, but when I build the member up, it fails by 7%.

 

[Blackstar123]

Increasing the stiffness would only serve to decrease the deflection.

You are looking at only one part of the solution, that is, since deflections are reduced, F = K u should result in lower forces, right?
Well, not exactly.
The main question you should ask yourself is, how exactly the displacements are reduced? Answer will most probably be, because the stiffness is increased. (Off course provided that the external forces remain the same.)
In other words u is decreased because you increased K somewhere in the load path. Thus, it isn't necessary that you will get lower internal forces if u is decreased and it's simply not possible that you will get lower force in those member which are working the most in reducing that displacement.

how do you design for this? I build up a member and get punished for doing so. That is the part that doesn't make sense to me. The smaller member fails strength capacity by 4%, but when I build the member up, it fails by 7%.]

Since you have not shared enough details about your structure, the one possible solution I could think of your problem is, instead of increasing the size of the member which you are trying to pass, increase the stiffness of some other member (lets say a column or a brace) in the load path, so that it provides a more stiffer path for load to follow and take off some of the load from that member.

 

[IlliniEI]

I believe the variable I'm overlooking that you've touched on is that, in a 3d model, there are additional lateral restraints in and out of the plane. I believe I'm right in saying that this is what's causing the internal force increase. The stiffness increases and is therefore the more prevalent load path, but the displacement only slightly decreases due to these restraints. Therefore, the internal forces must proportionally increase. This has all made me realize that 3D modeling picks up variables that may be overlooked in 2D analysis.

[the one possible solution I could think of your problem is, instead of increasing the size of the member which you are trying to pass, increase the stiffness of some other member (lets say a column or a brace) in the load path, so that it provides a more stiffer path for load to follow and take off some of the load from that member.]

I'm not sure if I'm quoting correctly on here haha I tried this to no success. It is a storage shed, and the top track running along the top of the wall studs is what's seeing such a high moment. I attached a picture for you to see it. It's very localized to near the end wall. I tried to stiffen members around it, but no luck because the load path still goes back to that top track. I lowered the bracing/chord about a foot, and the top track passes now. The idea was that the bracing/chord would aid in reducing deflection along the span of the wall and therefore reduce the internal forces. Thank you for all your help!

 

[Blackstar123]

Yeah, something does look fishy here..
However, it's really hard to say what's causing the problem here since there are so many components at play.

First try to isolate the root of the issue here and then try working on to eliminate it.

for example

Delete the side wall area elements to make sure if the wind load that's causing the problem is coming from the pitched roof. Or vise versa.

I see a sort of horizontal platform at the start of the frame. It most probably is very thin and made of steel. But I couldn't tell if it has any intermediate suporting members. May be the joint is seeing moment in minor direction because of this slab.

Usually area slab is meshed at top/roof and side wall purlins. But I don't see this type of load path in your model for gravity or wind loads.

As I said, hard to tell a root cause of the problem when there are so many components at play.

 

[Blackstar123]

I'm not sure if I'm quoting correctly on here haha I tried this to no success.

Select the quote option in the tool bar for quoting someone. Or you can write the syntex
[Quote"] text you want to write ["/quote]
Don't use the inverted commas though. I only wrote it so that you can see the syntex.

 

[IlliniEI]

The front portion you are seeing is a porch area with three front posts as support. The structure is all wood with plywood sheathing.

The controlling load case is actually the roof live load case, which surprised me because I'm designing this for 160 mph winds. My best guess at the localized moment is that the roof is uniformly loaded and is therefore trying to uniformly deflect the walls outward. However, the intermediate end wall is keeping this from happening, and results in the large moment shown.

Lowering the bottom brace/chord throughout the length of the building prevents some of this outward deflection and led to a reduced moment at the end, which somewhat confirms that idea in my opinion.

 

[Blackstar123]

My best guess at the localized moment is that the roof is uniformly loaded and is therefore trying to uniformly deflect the walls outward. However, the intermediate end wall porch slab is keeping this from happening, and results in the large moment shown.

I think if you delete the shell element on porch slab, these moments will vanish.

I'm curious, how did you define the material model for wood in sap?

Also, I am not familiar with wood framing or its design, so I can't comment on that front.

 

[steelnz2003]

If I am reading the output correctly.
From what I see the eves member are providing lateral support to the truss. i.e the truss horizontal reaction carried by them
You may need to realise them at end end