Atlas Observer

Back to the Atlas Observer

HSS Connections: The Top Things You Should Know Part 2

on June 17, 2015

Okay. Moment connections is kind of like the big one when people start talking about HSS. I mentioned prequalified moment connections. There is this book, if you will, called “The Prequalified Connections for Special and Intermediate Steel Moment Frames for Seismic Applications,” or, as we call it, 358. That is basically a listing of connections that have been tested and prequalified, if you will, in accordance with the criteria that’s set out in Chapter K.

Unfortunately for HSS, there’s really only one prequalified connection in 358, and that is the ConXtech ConXL Connection. ConXtech is a company out in California, which has come up with this process of kind of creating these clamps that actually clamp on the outside.

You can kind of see it in this picture here. They actually weld onto the tube these little brackets. But then what that allows them to do is you actually bolt diagonally at 45 degrees, and you actually are creating a clamping force that then helps develop resistance to . . . And they’ve been fully tested for seismic loads.

It’s unfortunate that we don’t have more that are prequalified. That’s something that Atlas is doing now. We’re actually doing some thoughts on whether we want to get into the business of prequalifying connections. You know, looking at working with other people on prequalifying connections.

What it boils down to is there is research that’s being done on HSS-to-HSS connections at the University of Michigan. We’re a sponsor of that, and we’ve actually provided the tubes, and we’re actually providing the oversight to that. So, there’s a lot of stuff that’s being done, as far as HSS moment connections. But it’s not in the code yet. It’s not happening right now. But it’s there’s stuff coming in the future.

When we talk about moment connections and you know, they say, “Well gee there’s only this one prequalified connection for a moment connection, and it’s proprietary, and I’m not a big fan of having to pay somebody else to do my designs for me.” So how do you get around using HSS in their seismic force resisting systems?

I say, “Well I’ve done a little research, and once again, this is, some of this is a little, I don’t want to say ‘nebulous,’ but it’s a little fuzzy. But there are some things that you can do to use HSS in seismic systems.” And the first one is to avoid using the seismic provisions.

Sounds kind of sneaky, doesn’t it? You know, “Just don’t use them!” Well, the reality is, how do you not use them? Well, you use them by being in Performance Category A. If you’re in Performance Category A, you don’t need to evoke 341. If you’re in Seismic Performance Category B or C and you’re using R of less than 3, according to ASCE 7, you don’t need to be using the seismic provisions.

So if you’re in that world–which, I think, most of us probably are in that world here in Chicago- and you’re using R of less than 3, you do not need to use the seismic provisions. So therefore, you can use whatever connections you want.

That’s one thing, I think, that people get lost and they get a little fuzzy. They see the big world out in California, and they see all that goes on out there, and they say, “Oh, everything needs to be prequalified if I’m going to do it seismic.” But seismic is not a California problem. It’s a national problem. I mean, I don’t know how many people have dug projects in Philadelphia or New Jersey or the southern part of Illinois. I mean, you’ve got to deal with seismic, right? You’ve got to do the detailing.

But a lot of times, you are in this world of A, B, or C and an R of less than 3. And I mentioned this before. You know, from HSS-to-HSS moment connections the other option of using HSS is prequalify your connections. You know, if you’re working for a project that has a lot of spare money and the owner really wants to go out and do something zoomy and new and he wants to fund you to prequalify your connections, go for it. There is current research being done at University of Michigan on, as I said, HSS-to- HSS moment connections.

All right, so now we’re on the subject of moment connections. Let’s say we are working in the world of where we can get away with not using prequalified or where we’re only controlled by wind, we’re not in a seismic region. So what, even then, HSS connections moment connections are a little scary for people.

So, these next few slides are just kind of, like, some typical connections that I’ve come across that I’m exposing people to and saying, “These are, just, these are pretty typical connections that are out there, that just maybe you haven’t been exposed to, if you’re looking to develop continuity.”
And the first one is, well, run your beam continuous. For a single-story structure or at the top of a multi-story structure, if you need continuity in your roof beams or you’re going to do cantilever construction, run it continuous.

The caution here is that the bottom flange is not supported, or the top of the column is not supported, so you do need to brace that. You do need to provide some stiffeners. You need to extend your joist bottom chord in order to brace [inaudible 00:04:49]. Otherwise, because most people say there are unbraced lengths that are columnous to the bottom of the beam. The reality is that unbraced length is actually to the support point, and so you actually do need to take it to where the support point is.
But I don’t think you want to try to develop any kind of continuity from your bolts to beam there, so I would encourage you to brace the top of that column if you’re going to do this type of connection.

Now, that doesn’t say you have to do it at the top of a building. You can do it at column splices. You can do it at every floor. It gets a little your fabricator may freak out a little bit, and the erection guys may freak out a little bit at this. But it is an option. You can interrupt at each end of your columns and run your beam continuous. For more lightly loaded columns, you can just run stiffener plates in there to transfer your axial loads. And if you are using the full cross-sectional area of your HSS for your column loads, then you might want to think about splitting an HSS on either side and welding it to carry the load.

You know, these aren’t the prettiest connections. They’re not the most cost- effective. But it is a solution. And it’s something to think about. Some detailing issues to think about if your beam, you need to have your beam flange wide enough to support the base plate. You may have to use a rectangular HSS in there to fit on the top of your beam. And then the moment transfer that’s associated with this is dependent on the pieces and parts that you use for this. So you have to make sure to check all that stuff.

The next one is a through plate diaphragm. And actually, I’ll call it, I’m going to lump things together as a diaphragm altogether. But one of the versions is a through plate. And this is where you would actually cut your column, and then you actually run a plate continuous, and you have to weld your column back together. But it gives you a bolted connection. You know, you can do them with column splices. You can kind of see in the one picture there, you can actually do it at a column splice. The moment transfer is dependent on the welds you use.

The alternative to that is, we’ll call it a diaphragm plate that goes around the columns. So you don’t want to cut your column. It seems kind of . . . So you can have your diaphragm plate go around the column. It’s a very similar situation, except now you’re slipping your plates around the column and welding them in the plate.

I haven’t seen these used a lot, but they’re still kind of an option out there, where you can actually have either a plate welded to the face of the column and you could bolt to it, or you could do angles welded to the side of each HSS and then bolt to that. It is wider than the column. It’s going to interfere with finishes. It’s going to drive your architect nuts. But it is it is an option. It’s a good field-bolt type connection.

And then there’s the old-fashioned, just welded up. You know, if you put, if you change that HSS picture to a wide flange, everyone would be really comfortable with that, right? That’s what we do a lot of the times. We just, we have a full full-moment connection, and we weld it to the flange of the wide flange.

But now if we substitute an HSS here, everyone kind of freaks out a little bit, and they don’t know what to do. But the reality is you can do this with an HSS column just as easily as you can with a wide flange. But you have to realize some things. You’re not going to be able to develop the full moment capacity, the full Mp, of your wide flange. So if you’re looking at getting all the gas out of your beam, you’re not going to get there. So it’s not the right connection.

You can develop the full flexural capacity of the HSS. To maximize the efficiency of that, you want to make sure your beam flange is as wide as possible. You don’t want to use a real narrow flange on a wide HSS. You want to maximize that, spread that load out, make sure you’re using all of the flat dimension of the HSS.

The next one I’d like to, the next few subjects are all related to truss connections. And, because to be honest with you, this is really the most common connection out there that people use HSS for. They use a lot of HSS planar trusses.

The connections in Chapter K are related to single-plane, or planar trusses that are typically designed for pinned. You know the pin members, the branch members are pinned. We’re talking about tension and compression members. So that’s really the kind of connections we’re talking about.

And from a nomenclature point of view, this is what we’re talking about. We’ve got, the few things I want to identify here is the G here. That’s what we refer to as our gap. So if we’re talking about a gapped connection.

And then E is our eccentricity. Now, a lot of times people think there is no eccentricity of these, but actually, depending on the dimensionality of your chord and your branch members, there will be some eccentricity of your center lines. And that does need to be accounted for. It’s a secondary moment that does need to be accounted in the design of the chord.

And then the overlap. If you’re not we talk about gap connections, and then we talk about overlap connections. And an overlap can either be a hundred percent overlapped or partial overlapped. And that partiality is you can see there, it’s a percentage of how much overlap there is.

You know, when we’re talking about overlap, this is what we’re kind of talking about. A hundred percent overlap is, as illustrated there, two branch members completely overlap. And then, the partial overlap is where they partially overlap. And then you can see that there’s definitely more different cuts and different profiling that’s involved with a partial overlap. But there are some advantages to doing that.

Now, we’ll touch on that in a little bit. Now, when it comes to analysis of these things, as soon as you pin the branch members, you can use the equations in Chapter K for tension and compression. So, that’s what I encourage and there’s basically three ways you can analyze this truss. You know, everything’s pinned. Or you can do kind of the hybrid, where you pin just the web members, just the branch members, and your chord is continuous. That’s probably the better way of doing it, because that reflects reality for the most part.

If you really want to model the eccentricity, the nodal eccentricity, as we talked about the E factor, the E value, we talked about the nomenclature there, you can actually put in little stiff members and model your eccentricity that way. It’s up to you whether, how fancy you want to get with that.

And then, of course, the third version is what this engineer was doing, was with our original analysis where everything’s fixed. Now you may think that that’s a better solution. But the reality is, it makes it harder to design your connections. So I think either doing pinning everything or actually looking at a little more sophisticated model like the one in the middle there might be the better way of doing things.

So, when we talk about joint types, these are the different types of joints that are out there and are defined. We’ve got your T-joints, which has a subset of Y. We have your cross joints, using the nomenclature that’s in AIC. In the European vernacular, they’re called X-joints. We have gap K- joints and overlap K-joints, and then within a subset of the K-joint is something called N.

Now, the thing I want to point out is, these are, while it may look like these are all dictated by the geometry, they’re not. Okay? The geometry is really, or the classification of these joints is really based on the method of force transfer. Okay, and that’s important to realize, because if you look at the geometry of the connection, you think, “Okay, that’s a K- connection,” or, “That’s a T-connection.” But the reality is, it may be a combination of connections.

So even though the geometry looks like a K-connection, there may be a cross connection. There may be an N-connection. And you have to be aware of that, because it’s really more about the force transfer.

And what happens is, there are a lot of times, like for instance, this case I’ve got illustrated here, this doesn’t apply. If you look at this, it looks like a K-connection. Actually, it looks like an N-connection–that’s what an N-connection looks like, which is a subset of K. But if you look at it and you look through the formulas, this doesn’t really, there is no formula that actually you can do that with.

So what you need to do is, you actually need to break it into the force transfer. In this case, 50% of the load is being transferred as a K- connection, and then the other 50% of the load is being transferred as an X- connection, or a cross-connection. And you actually check each one of those situations separately, and then add them linearly. And then linearly, together the interaction needs to be less than 1. So it’s just linear interpolation.

That’s how you get around the limits of applicability within Chapter K, is you have to look into your force transfer method and realize that it may be multiple methods of force transfer. So you break it up into these different pieces and parts.

So we’re talking about trusses. We’re talking about fabrication. I’ve talked about gap joints, partial overlap. You can see from the diagram that maybe partial overlap connections have, that’s a lot of cutting and my fabricator probably would freak out on that.

So, let’s talk a little bit about fabrication costs and the impact that that will have on your decisions on what kind of connections you do. You know, not only for truss design, but all steel design, minimum weight is not least cost. And I think we’ve had that message hammered into us a lot of times, but it’s true. Sometimes by upsizing a few members, we actually lower our costs. It seems kind of counterintuitive, but it does happen.

Obviously for a truss, if we keep the number of different sizes to a minimum, we’re going to be better off. So having every single branch member be a different size as you go across your moment diagram for your truss, maybe it makes sense to group things a little bit or make them all the same size, depending on how adventurous you want to be.

And then the clear thing here that I’d like everyone to walk away with is to understand the joint configuration and the connection design criteria before you analyze the truss, before you select your members. You really have to do that. We live in a part of the country, a part of the world where we tend to push off the connection design onto our fabricators. It’s a system that works. It’s good.

But a lot of times we tend to ignore the connections because, “I’m not doing them. Why should I worry about it, right?” So what happens is, we design our truss. We come up with least weight. We have all the forces in there. We hand them off to the, we put them on a diagram in the drawings, and maybe we amplify them by 10% just to cover ourselves, and then we hand it off to the fabricator, and he can’t get the connections to work. “I need to put reinforcing plates in it. I need to displace, I need to add this, that, and the other thing to get these connections to work.”

A little forethought by the engineer on the front end of this, understanding that if you upsize the chord size a little bit, you can avoid, you can get your connections to work, you can avoid a lot of reinforcing issues that may have to happen at the fabrication level. So this is something to be aware of, that you really, in these type of situations where you’re looking at a welded truss, you really need to understand the connection configuration and the joint, and the capacity of those joints.

So along those lines, what this chart will show you is for different configurations you have different costs. So, while there is a cost impact for each one of these, there’s also a strength impact that goes, that’s opposite of this. So as you get up in strength, there is cost associated with it. So it’s something that you have to find that balance for your project.

Just jumping back here a second, okay, we talked about what the chord is, but I didn’t really talk about what the branch member is. Well, if you look through the limits of applicability in Design Guide 24 in Chapter K, you’ll see that it’s really, they only deal with like-to-like. It’s only round-to- round or square-and-rectangular-to-square-and-rectangular. That’s what everything is.

So what happens if you have a situation where you want to have round branch members and square chords? And especially if you do like a gapped connection, that still seems like a pretty simple connection. You have one planar cut welded up and maybe I like the aesthetics of having the round branch members.

But the problem is, this doesn’t really, isn’t really covered by Chapter K. It’s outside the limits of applicability. But, of course, if you look in the commentary, there is language in there that says you can use other verified design guidance. And so, therefore, we need to look. What other verified design guidance is out there?

There has been research that’s been done by once again, Jeff Packer’s name is up there. Professor Packer, he’s involved in a lot of stuff when it comes to HSS. But he wrote a paper a few years ago about the static and fatigue design of these types of connections using what we call a conversion method.

And what this basically does is, for calculation purposes, you convert the round sections into square sections, or equivalent square sections. And so, therefore, once you do that, you can then use the Chapter K requirements. And the formula that they’ve come up with there is, if you take the branch diameter of D and you replace it with a width of pi over 4 times D and the same wall thickness, you can then use the Chapter K requirements.

I’ve made a lot of references to a lot of different things. I thought I’d just quickly show them, a collection of these resources out there, because a lot of people, they’ve heard of them around before, but they really don’t know what they are. Well, clearly, Chapter K, that’s in the specification. That’s one of the greatest resources that we can have out there about HSS connections. Because right now we have something in the spec that tells us how to do connections for HSS, you know. Prior to 2005, they didn’t exist.

Design Guide 24 is the AIC design guide that’s based on Chapter K, of the 2005 version of Chapter K. It’s a great resource. I highly encourage you to, if you haven’t seen it, to check it out.
Professor Packer, once again, our friend from the North in Canada has designed, he has a design guide from back in ’97, which was really good about this issue, especially truss connections and HSS connections. It’s a bit out of date from a code perspective, but it’s an excellent resource.

And then of course the CIDECT design guide. CIDECT is a European organization that produced, does a lot of education and research about tubular structures. It’s actually a collection of tubular manufacturers that promote education and research, and they do a lot of research programs. And they publish nine design guides, and those are available for free on AIC’s website.

You don’t have to go to CIDECT to order them. You can actually go and download the PDFs right off of AIC’s website. You end up with a pretty good- sized PDF, but it’s still, it’s a free document.
And while they’re Eurocode-centric, they still have some, all the information that’s in there is based on, is the core information that was then put into Chapter K. So while Chapter K is a little limited sometimes, the CIDECT is an expanded version of that. So they’re a great resource if you want to go beyond what’s in our codes.

And then we’re trying to, on our websites, we’re trying to help you guys out by giving you resources as well. And we’re constantly going to be improving that, so I encourage you to check out our websites.

Threw a lot at you in that hour. Hopefully it was meaningful to some of you. Hopefully you had some . . . There’s a lot of things in there that I’m sure people will have debates about. And I’m sure I’ll get pummeled with lots of little email questions after this. But a lot of it’s food for thought. So it’s just things to think about. It’s things that maybe you haven’t thought about before when it comes to HSS connections. So, and I appreciate your feedback.

[iframe class=’sproutvideo-player’ src=’//videos.sproutvideo.com/embed/1c9bd8bc1418e4c894/f46e9867406ef795?type=sd®ularColorTop=ff2b00®ularColorBottom=bd1000′ width=’630′ height=’354′ frameborder=’0′ allowfullscreen ]