Gusset Plate Design 1

[chad73]

I am designing connections of a Special Truss Moment Frame. The top and bottom chords are continuous while the web members are connected by welding to the gusset plates. The joint is treated as rigid. As a result, the web members will carry axial force, shear force and moment. I can't find any book or anything on the net that tells me how to design the gusset plate of this type. Help me please.

 

[ToadJones]

Is the moment resulting from eccentricities in the connections? Bending in the web members?

 

[chad73]

The moment comes from treating all the joints as rigid. Yes, bending moment in the web members. I look at UFM but those web members only carry axial loads.

 

[ToadJones]

I ask because usually when I do truss designs similar to yours I treat the top chords as continuous but model the web members as axial load members only.

ConnectEngr, where you been?
This seems to be your area of expertise.

 

[chad73]

TJ: So have you dealt with bending moment due to eccentricity of the load in web members?

 

[ToadJones]

In your exact case, no, I try to make sure there are no eccentricities which I believe is the same approach that the UFM uses.
With trusses I have dealt with (mill building type trusses and the like) the flexural stiffness of the web members and connections is so low that I never thought modeling them as rigid was appropriate.

 

[ToadJones]

...I would think (could be wrong here) that any moment in web members of a truss properly laid out to work as a classical truss, even with fixed connections, would only experience bending in the web members as a secondary force.
There has been papers written on this subject going all the way back to the early 1900's.
see attached....this is very old....probably outdated

 

[slickdeals]

There is a more recent one on secondary stresses written by Shankar Nair.

We are Virginia Tech
Go HOKIES

 

[connectegr]

Typically designing truss joints as rigid is not economical in terms of fabrication and erection. The design will produce a lighter truss, but the force combination will require additional connection material and welding. And in most cases reinforcement and stiffener plates. The overall cost to the owner is typically more for the rigid design. This is the reason for the lack of connection design examples.

http://www.FerrellEngineering.com

 

[connectegr]

The connections can be designed for the force combination you stated. But, more info is needed. What member types are used for chords and web members? HSS? Wide Flange? Web vertical or horizontal?

But, I do recommend releasing shear and moment in the panel points.

http://www.FerrellEngineering.com

 

[ToadJones]

Now that I think about it, I have modeled very heavy trusses for large, long span industrial conveyors with fixed joints at all locations. But, these trusses were VERY heavy and made entirely of heavy wide flanges (W14x193, for example) and had no gusset plates.
In this case I felt it made more sense given the high stiffness of the web members and their overall flexural rigidity.
For web members life small HSS or small pairs of angles I feel those are generally best suited as axially loaded members.

 

[connectegr]

Modeling truss panel points as rigid is a phenomena of the software programs now available. From the designer's standpoint the truss is lighter, due to reduced chord and web member sizes. "Less weight = less cost". This is a case where "sharpening our pencil" does a dis-service to our customer. This does not account for connection design, detailing, fabrication, and erection. As rigid members the web members are often sized to perform above 90% of their allowable capacity. As TJ mentioned this is usually a wide flange truss with webs horizontal. But, the model does not account for shear lag. So if the web is not extended into the chord, the member is immediately inadequate 85% < 90%+. If the web is extended, you have one or two diagonals and a vertical sharing the same workpoint. If the chord webs are relatively thin, the intersecting webs will overlap and so with their welds. The flanges of the diagonals must also be mitered and also clipped to avoid weld intersections. If these flanges are large, welds must be sequenced to avoid deforming the chord or vertical flanges. In order to control the welded framing the panel points are generally shop weldments with bolted splices in the chord diagonals. With the force combinations this will require reinforcement of the net section. If the designer has the foresight to consider these issues, the savings in truss weight are lost and the resulting member sizes may be larger than if selected for axial load only. My comments come from experience with connection design of "rigid" truss joints. Similar issues can occur in welded trusses with only axial force. But, many of these can be avoided by using only 75-80% of the members axial capacity.

http://www.FerrellEngineering.com

 

[chad73]

The truss we are making is for research purpose so we try to treat it as they are done in practice which is treating joints as rigid. Normally, they use double angles as chords, diagonal and vertical web. In our research, we are gonna be use 2C12x25 for all of the members. Cost and weight are not of our concerns. One panel of the truss is about 5.5" long and 4" high. Please help me figure out how to design the connection.

Thank you.

 

[connectegr]

"treated as they are done in practice", probably not rigid

With back to back channels I don't have any suggestions for the force combinations in the connections. Typically the channels are bolted through the web and gusset plate, or welded to the gusset plate. These connections are for axial force only.

http://www.FerrellEngineering.com

 

[ToadJones]

I agree with ConnectEngr here.

Connect-
Would you agree that with connections where the work points offer no eccentricity that there is no reason to design for moment?
My intuition tells me that the only moments on these connections will be secondary or the result of deflections?

 

[connectegr]

TS
I don't think the moment in Chad73's panel point is due to workpoint eccentricity. He is modeling the truss as a rigid frame, with the diagonals and verticals fixed at the truss chords. In the model, this works fine. And in your example of a welded wide-flange truss, it can work. But in a double channel truss the forces must transfer through a single gusset plate at the centerline of the cross section. Some moment can be taken in the channel web to the gusset. But the majority of the moment is in the flanges. The model sees lines with given properties and nodes. The actual shapes and load local stresses are not considered.

http://www.FerrellEngineering.com

 

[ToadJones]

Connect-
I see.
I have done a fair amount of trusses made up of WT top and bottom chords with double angle webs. I always modeled the top and bottom chord continuous and the web members as "truss" or axial members.
Usually I did not need gussets at all as the top and bottom WT stems were large enough to make the connections. When I did need a gusset to extend the web I had the fabricator full pen weld a plate to the stem and grind it flush. I only considered axial force on this plate and I only make the plate as big as necessary to accommodate the welding or bolting of the web members.

 

[chad73]

Well, in my case, for the Special Truss Moment Frame, the company that has used this system treats all the joints as rigid. They use double angles welded to the gusset plate. I'm sure that using double angles will induce less moment in the web members. In my case, we simply just change from double angles to double channels and I have no clue as how to design the gusset plates to take this moment.

 

[slickdeals]

I haven't designed a STMF, but I am not really understanding the logic behind forcing to design the joints for moment. It is a truss and does not need a moment connection (whether you get a partial moment connection due to connection fixity is another question).

the company that has used this system treats all the joints as rigid

Is it really required?

We are Virginia Tech
Go HOKIES

 

[ToadJones]

Are you by any chance using a Vierendeel Section in the middle span of your STMF?