Column subject to jib crane loads - your view on this connection detail! 6

[valleyboy]

Hello all

I would welcome your view on the connection detail shown on the attached sketch. This thread was prompted by a discussion I had with a colleague regarding how this connection would behave.

The building is a 3 span portal framed building. The building has 5 number overhead cranes, and some jib cranes at lower level than the overhead cranes. This problem relates to the connection of the column mounted jib crane connection onto the main portal columns.

The jib cranes have an SWL of 2 tonnes, and a jib length of 4m. The jib acts as a cantilever, and has an upper tie member to restrain the jib arm. The jib can slew 180 degrees in plan. The detail on the top left of the attached sketch shows the plan arrangement.

The jib is attached to the column flanges with bolted angle cleats, as shown on Elevation A-A. When the jib is orientated parallel to the web, the jib effects can be included in the primary frame analysis as a major axis effect, and designed for. However, when the jib arm is parallel with the flanges, a minor axis bending moment is induced into the column, which must be dealt with in addition to the major axis effects from portal action. To compound matters, this minor axis effect is applied to one flange only.

In terms of section capacity, the column section is adequate for the combined major and minor axis effects, if the full minor axis section capacity is considered. This obviously relies on there being some mechanism in place to distribute the load applied to one flange to the other. I had thought to do this be adding stiffeners to the section over the depth of the connection as shown on elevation B-B. My thought was that in order for one flange where the load is applied to be mobilised to resist the load in isolation, that the stiffeners would be mobilised as a tie to distribute a proportion of the load to the other flange.

Any thoughts on whether this would be an effective method to mobilise the full minor axis moment capacity? Any other solutions gratefully received!

Valleyboy

 

[hokie66]

I don't think the stiffeners do much good. The jib applies a torsional load on the column, and wide flange sections are not much good for that. The approach I have always used is to reinforce the flange on which the jib is supported by adding a plate or channel to take that weak axis bending.

 

[MRH1335]

I would check localized stresses and evaluated the global stress increase. Globally, I do agree with you about stress increase in the minor axis since they design these buildings with no reserved capacity and you should inform the manufactures’ engineer about this modification to his original design. He might be concern about the additional moment and possible torque that the jib crane imposed on the mid-height of the column. I also would check the localized stresses on the angles, welds and bolts due to full load capacity and fatigue loading. You may be down to the fatigue Category “D” per table A-3.1 (section 2, Item 2.3) of the AISC 13 Ed. Also do not forget prying action on the connectors.
Proposed Stiffeners are good but use them in pairs at top and bottom. I would put a pair in between as well.

MRH

 

[hokie66]

MRH,
What do you mean by "they design these buildings with no reserved capacity and your should inform the manufaturer's engineer..."? I assumed that valleyboy is the designer.

 

[msquared48]

Any chance of installing two horizontal beams between the jib verticals at the point of attachment of the jib arm to act as horizontal stabilizers in the weak axis direction, decreasing the kl/r and providing a lever arm for torsional resistance?

Mike McCann
MMC Engineering

 

[valleyboy]

Just to clarify - I am the designer of the structure - this isn't a retro-fit.

Hokie66 - initially I came to the same conclusion as you - that this will apply torsion to the section.

However torsion is usually generated when there is a net horizontal force applied with an eccentricity with respect to the section centroid. For this arrangement, the horizontal forces due to the couple are equal and opposite - so the net horizontal resultant force is nil - the only effect is a moment applied to the flange.

In view of that, it isn't a huge leap to look to mobilise the other flange using the stiffener solution. The rationale is that for the one flange to which the jib is attached to be mobilised in bending, the welds to the stiffener must assist by transferring the force between flanges.

Or is this flawed in some way?

 

[valleyboy]

@mssquared

I've already explored that option. This is a workshop structure and the introduction of the restraints you describe would hinder access from one bay to the other.

VB

 

[hokie66]

OK, I stated it wrongly. "Torsion" is a poor description of what happens. But the other flange already has to help, as the load is applied outside the carrying flange. So the remote flange is just the back support for the propped cantilever, and sees forces in the opposite direction, albeit of much less magnitude than on the loaded flange. The stiffeners help in that respect, but the loaded flange still has to take the load. It is analogous to a pair of columns supporting a cantilevered beam with a concentrated load at the end.

 

[valleyboy]

Many thanks for your input. Time t call it a day - I will take a look in the morning with a fresh pair of eyes!

vB

 

[JAE]

I think there is torsion, but it is in the upper section of the column (above the upper connection) and in the lower section below the lower connection.

The overall column has two equal and opposite forces, F that are 1300 apart (4'-3"). The overall column height (let's call it L).
The forces F basically create a moment near the center of the column of 1300F.

There is then two opposite reactions at the top and bottom of the column resisting the moment and equal to F(1300)/L. Let's call this reaction R.
So the lower segment has a uniform shear of R = 1300F/L as does the upper segment.

The shear, R, along these lower segments is located at the centroid at the bottom and top of the columns.
However, at the connection points, there is an offset of the shear located at the carrying flange. The eccentricity is d/2.

This means that there is a torsion on the section of R times half the depth in the upper and lower segments but no eccentricity in the center.

So your stiffeners don't help in between the forces - but would beyond the connections.

I would consider adding plates to form a tube full length of the column.

 

[dhengr]

Valleyboy:
Your second para. @ 17:24 is exactly what gives Hokie’s approach much merit. If you design the front flange to take that moment applied mid height on the column, I (we?) think you can rationalize the torsional problem away, or at least it is more manageable, but doesn’t completely go away. The torsion is less of an issue because the rear flange does not have to be brought into full play to make things work out. Furthermore, that torsional effect does not stop where your stiffener pls. terminate t&b, even though the two horiz. forces on the front flange appear to cancel out. Put those two loads “H” on the front flange and treat it like a beam spanning from the floor to the roof structure, and you will need horiz. reactions on that beam at the floor and roof and those forces are eccentric to your UB’s centroid the full height of the column, and in opposite directions. The torsion is in one direction above the top jib connection and in the other direction below the bottom jib connection, and transitions btwn. the two btwn. the jib connections. Draw the shear and moment diagrams caused by the two “H” loads on the front flange, as a beam. What you might want for a column at the jib cranes is a channel or another light UB for the front flange and the regular UB column for the web and back flange of the portal frame. Do this problem by superposition: design the front flange as a beam-column for the “H” loads near mid height when the jib is at 0̊ and 180̊ and add this to the portal frame column when the jib is at 90̊. And if you really want to confuse the issue, now look at the jib at 45̊. Undoubtedly, with some of today’s software you can do this as a combined section, on one model, in several different loading steps. I think I’m saying essentially the same thing as JAE is, in slightly different words.

 

[valleyboy]

Thanks for your valued input gents - it all becomes clear.

I will take a look at the options noted. At first glance I think it would be more cost effective to install a pfc vertically attached to the flange, and design this to span vertically from foundation to the next restraint above the top jib.

The top restraint exists at the column centre line, so the connection here will need to be designed to resist the applied torsion.

Thanks for adding some much needed clarity!

 

[BAretired]

If the top and bottom of column is hinged, there can be no torsion in the column. Gravity load P acts eccentric to the column, producing a moment P.x and P.y about the minor and major column axes respectively.

The load is transferred to the column centroid by the jib beam plus a torsional member parallel to the minor axis of the column.

The stiffener plates shown on the sketch may not provide much torsional stiffness by themselves, but additional horizontal plate stiffeners at the upper and lower jib connections would provide a closed shape between column flanges which may provide sufficient torsional resistance. In this way, the full minor axis moment capacity can be mobilized.

BA

 

[MikeHalloran]

I personally would prefer a 'mast type' jib crane, where the reactions are transferred to the floor and ceiling, or at least to the ends of the column, as opposed to the proposed detail, where the reactions attempt to wrinkle the column.


Mike Halloran
Pembroke Pines, FL, USA

 

[JAE]

BA - I don't think any building column is ever "hinged" at the base torsionally.
Nor really at the top either but possibly quite flexible (in a torsional sense) at the top.

 

[BAretired]

JAE, I agree that real columns are not hinged torsionally, but it is a conservative design assumption which illustrates the fact that torsion in the column is not required to satisfy statics. Torsional restraint at the ends of the column does not alter the conclusion that the full minor axis moment capacity of the column can be mobilized without forming a tube for the full length of the column, although this may be prudent in order to reduce deflection.

BA

 

[valleyboy]

Thank you all once again for views.

In the event I have decided to go with a 260PFC stitch welded over the loaded flange, spanning vertically between foundation and the next highest point of minor axis column restraint.

This arrangement does induce a torque into the bracing member, but the longitudinal member was originally governed by slenderness limits rather than structural capacity, and is capable of resisting the applied moment, together with axial loads present.

BA - I take your point - but the cross section analysis becomes rather complex, with the major axis effects due to portal action (which also carries a gantry crane!), minor axis effects due to jib crane loading, and the potential for twist at SLS in the columns due to the jib crane action. I would be more comfortable with the PFC option, which provides a dedicated load path.

@Mike - I also would prefer a foundation mounted jib but this sadly wasn't an option due to space constraints.

VB

 

[hokie66]

valleyboy,
Good choice. That's the way it is normally done in a heavy industrial environment where a slewing jib has to be supported on the building column.

 

[BAretired]

I agree that the channel reinforcement is a good choice.

BA