Long Span Joist - Out of Plane Load transfer - Connection

[cgstrucg]

Hello,

I have a project in NYC for which I have a lot of long-span joists. I have a bunch of rooftop equipments on dunnage, posts of which rest on these joists. I have a few questions related to connections of these posts to the joists and if anyone can refer me in the right direction, it would be really helpful.

First is the post to joist connection. I was planning to just put a base plate at bottom of the post and drill bolts through the plate and joist flange to act as a pinned connection. Yesterday my senior engineer told me that he prefers weld. His reason was that he doesn't want to reduce the joist flange area. Now my question here is this that I always thought weld connection is always a rigid connection and moment will transfer through it as it resists rotation. But the senior engineer said that design welds to act so they don't transfer moment and I don't know how I can do that.

Second is regarding out of plane loads that equipment transfers to the bottom of the post. Vulcaraft doesn't design for these out of plane loads so the senior engineer asked me to look into sideways puddle weld to transfer the load to the roof diaphragm. For this, I really can't find any standard detail design and analysis. Has someone encountered these situations and please share with me details or any alternate concept.

The final thing is regarding those out of plane loads which we will transfer to the diaphragm. We have to figure out a way to transfer that diaphragm load to beams running perpendicular to the joists. Has anyone seen detail for this before? This is something for which I can't find any details which are simple and not warp the deck.

Any help would be really appreciated. I am designing roof with joists for the first time and although learning, it is quite challenging and I want to avoid designing small bad details that mess the roof.

Thanks a lot

 

[phamENG]

cgstrucg -

1) all connections will transfer some moment unless it is a "true pin" in which the members are able to freely rotate about a single point. If you have one bolt with a nut that's installed to snug tight, it will (in reality) transfer some moment through friction until the frictional force is overcome. If you have two bolts and the connected member deflects enough, they bolts will bear on the bolt holes and form a couple that will transfer moment. In a typical steel structure, there are zero "true pin" connections. We consider them as such for simplicity, but they really don't exist. When properly designed, the relative stiffness of the various members and the connections will allow the structure to perform globally as we expect, if not at the micro-scale. In your case, I would suggest trying a two sided weld parallel to the direction of the resisted shear force. That will reduce the weld's Ix and reduce its stiffness. Question, though - if the base is pinned, where is its stability coming from? Knee braces/moment connections at the top of the post? If so, size your members as though you have a true pin at the base. That way, very small rotations at the base will "activate" the stiffness of the upper members and prevent further rotation (and, therefore, further moment transfer) at the base. Also, make sure your base plate is as thin as you can get away with for vertical loads and shear. That will increase the flexibility of your joint.

2) Those out of plane loads are easier than you think. I know it's tempting to think that that corrugated steel deck will crumple like an aluminum can in the "weak" direction, but as long as you have an appropriate diaphragm connection pattern on a suitable joist system, it won't buckle under normal loading. Take a look at the Steel Deck Institute's Diaphragm Design Manual. If you're designing steel deck roofs, this should be sitting on your desk. The manufacturers provide some useful values for their products, but the basic equations they give you are explained and expanded on in the DDM. They also have design examples for concentrated lateral loading to be resisted by the diaphragm. The trick is to keep them out of the joists by running channels or angles between the joists and welding the deck to them. The joist is prevented from moving by the load path through the strut and into the diaphragm.

There's a lot to think about in a properly designed joist roof. The steel deck institute and the steel joist institute both have valuable design guides to walk you through it, and AISC has some guidance as well. Is there already a contractor or are you specifying Vulcraft joists? If you know who will be supplying the joists, don't be afraid to call their engineers. They'll be much happier to answer your questions up front than do deal with crap drawings that require a month of RFIs to clean up after the fact.

 

[cgstrucg]

1. Yes, I have a plate welded at toe of beams for stability. So while designing those base welds, do I assume no moment is transferring and just design for shear and axial?

2. Thank You for making me aware of DDM. It looks really informative.

3. I was planning to talk to Vulcraft about this point, just want to be aware of the normal industry standards as I am not yet able to find anything on this.

 

[phamENG]

A 2d sketch of the frame and that connection would help. I think I know what you're saying, but can't be sure.

 

[SteelPE]

I went to a seminar once where the speaker said something similar to "if you can provided a positive load path for load in your structure, you structure will find a way to get to the load to the ground". Meaning, as long as you have a positive way to get the load to the ground (through moment connections or pinned connections) the structure will deliver the load properly to the ground. So I wouldn't get hung up in the fixed vs pinned vs partially fixed scenario you outlined.

I have taken care of out of plane joist loads in the past by adding bridging to the joists where the load are being delivered to the joists. If the load is large, sometime I will add angles under the deck. I have also added a C6x8.2 above the deck at locations where the posts hit the deck. These C6x8.2 are toed down, and can actually help with increase fixity at the bottom of the post if they are supported by multiple joists. I also agree with your senior engineers assessment of not reducing the area of the chords.

In regards to transferring diaphragm loads perpendicular to joists, it appears you are using long span joists (LH joists) these should have a 5" deep seat. You can either specify a roll over capacity for the seat, or you can provide shear collectors in-between the joist seats. Shear collectors can be anything that matches the depth of the joist seat (channel, HSS tube). These shear collectors are typically 4'-5' long and are placed intermittently along the beams (if the shear loads are low).

One book you might want to have is "Designing With Steel Joists, Joist Girders, Steel Deck" by Fisher, West and Van De Pas. This book is incredible with regards to practical application of these type of roof systems.

 

[cgstrucg]

Thank You, everyone. These were really good suggestions.