Edge Beam Twist due to overhang detail

[Contraflexure74]

Hi,

I have a job on site where I have an unrestrained roof level edge beam spanning 6m between 203UC's with an overhang detail, see attached sketch. The outriggers are at 1.2m centres and the edge beam is fully welded all round at each end at the UC's.

The steel fabricators are welding up the RHS on site and have told me that the end of the overhang deflects by 10mm out at the tip of the PFC when the RHS was only stitch welded in place. They are proposing to over jack the overhang upward and then fully weld the ends of the 6m beam before releasing the props, hoping this will counter act the twisting effect. Any slight twist of the RHS is obviously magnifying the deflection out at the tip of the overhang. Having drawn it to scale myself, a vertical drop due to twist of the outer wall of the 250x150x5 RHS by 0.9mm gives a 10mm deflection out at the tip.

I have modelled this on my software and torsion appears to pass by a country mile but I'm not getting anything like the deflections experienced on site on my software and this 10mm deflection experienced on site is only under its own self weight before any wind or live load hits it.

Anyone have any thoughts on this as I can't seem to find any guidance of allowable twist in the design codes and I'm a bit worried now I've missed something!!!!

John.

 

[hokie66]

The rotation could be due to yielding of the "stitch" welding, particularly if it is not on all sides.

Have you looked at the deflection due to the moment applied to the 200UC? I assume the UC is bent about its weak axis.

Not a bad idea to provide some camber to the outrigger arrangement. The problem is in determining how much. But residual camber is always preferable to sag in a case like this.

 

[Contraflexure74]

Thanks for reverting hokie66.

You could be right on the stitch weld yielding.

Yes columns at both ends are bending about their weak axis. I applied the moment at the top of the UC and I'm only getting a deflection of 0.2mm of deflection but having said that, that column was only stitch welded in the opposite direction also. So possibly they loaded the outriggers before everything was fully welded up.

John.

 

[BAretired]

Your detail suggests several factors which will contribute to deflection of the channel, namely:

* Bending of the column due to applied moment​

* Deflection of the RHS beam​

* Torsion of the RHS beam​

* Local bending of the 5mm RHS wall​

* Torsional rotation of the four bolt group due to eccentric connection to channel​

* Deflection of the outriggers​

* Torsional twist of the outriggers​

* Deformation of stitch welding​

Have you considered all of these?


BA

 

[hokie66]

Another to add to BA's list.

*Fabrication of the end plates out of square.

 

[dhengr]

JohnFitzgerald74:
I can’t think of any specific additions to add to BA’s and Hokie’s listings, but I do have a couple other thoughts. I would take a cut at each of those different contributors to the deflection problem out at the channel tip. Note that most of the deflections are a one-shot deal as a function of the detail causing the particular deflection. But, the bending and torsional deflection of the 250x150x5 RMS is zero at the columns and a max. at the center of the span, so these vary at each outrigger. Once I had a handle on (estimate of) each type of deflection I might install the 250x150 rotated counterclockwise a degree or so at the column connection to compensate for some of the anticipated deflection and torsional rotation. The bulging of the right face shell (plate, web shell?) of the 250x150 at each outrigger (BA’s item #4) might be larger than you think, due to welding and canti. pulling force on the face. You could camber the outer channel to help compensate for the deflections. You could shim at the bottom bolts on the moment connection pls. on the outriggers, zero shims near the columns and max. shims at center span.

 

[Contraflexure74]

Hi,

The summary of what my model shows are as follows for the dead load case:

The max torsional moment on the 6m main RHS under the dead load condition is 1.4kNm.
The deflection of the main RHS is 0.06mm for the dead load condition.
The deflection of the outrigger at the tip is 0.5mm for the dead load condition.
The max moment of the outrigger for the dead load condition is 0.48kNm.
The moment induced at the column due to dead is 2.8kNm inducing a column deflection of 0.23mm

The loads on the over hang are as follows:

Dead = 0.5kN/m2 + Self weight
Live = 0.6kN/m2
Wind = 1.2kN/m2

As you probably guessed I have little experience in torsional design and I'm not sure how to check the local bending of the RHS 5mm wall.

I guess I should check the outrigger for torsion ASAP due to eccentric connection. The frame is up and propped so I'm in a bit of a pickle.

My confidence is now officially shot and am worried of the stability of this detail.

Any help would be greatly appreciated.

John.

 

[hokie66]

I can assure you that stability is not an issue. But serviceability is, in terms of the visual impact of differential deflection of the outriggers. But determining exactly which or how many of the issues which we have listed as possibly contributing is very complex, and may in fact not be determinable. I think your best course at this stage is to adopt dhengr's shimming idea at the end plates of the outriggers.

 

[Contraflexure74]

Thanks hokie66. I guess I'll have to see how it turns out on site and make an assessment then.

 

[KootK]

1) I agree with Hokie, I see nothing here to make me worry about strength or stability.

2) In what respect is 10 mm deflection a big deal for your structure? What serviceability problem does it cause? With common cladding systems, I typically design my structures for about 20 mm of movement to occur after the installation of cladding. Also, as dhengr mentioned, a good chuck of your deflection might just be the taking up of slack in individual connections. The effect resulting from future loads may not be nearly so pronounced.

3) You might want to look a bit closer into the nature of the "stitch welding". If those stitches are not well distributed about the section perimeter, you can get some significant twist as a result of local deformation at the member ends. And installation crews are usually focused on vertical load resisting capacity rather than torsion. If you had, for example, only a single stitch weld at the center of each vertical HSS web, that could cause some problems.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[BAretired]

The max torsional moment on the 6m main RHS under the dead load condition is 1.4kNm.
The deflection of the main RHS is 0.06mm for the dead load condition.
The deflection of the outrigger at the tip is 0.5mm for the dead load condition.
The max moment of the outrigger for the dead load condition is 0.48kNm.
The moment induced at the column due to dead is 2.8kNm inducing a column deflection of 0.23mm

The loads on the over hang are as follows:

Dead = 0.5kN/m2 + Self weight
Live = 0.6kN/m2
Wind = 1.2kN/m2

I'm not sure how you arrived at 0.48kN-m for the moment of the outrigger. It seems high for the dead load condition. Neglecting self weight, M[sub]outrigger[/sub] = 0.5*1.2(0.8)[sup]2[/sup]/2 = 0.192kN-m.

With outriggers spaced at 1.2m o/c, each column supports 5 outriggers, so the column moment would be 5(0.192) = 0.96 kN-m (less than your 2.8 above).

Please let me know where we differ.

If the outrigger dead load moment really is 0.48kN-m, I would be concerned about the local effect on the wall of the 250 x 150 x 5 RHS. This could be rectified by adding a stiffener welded atop the 80 x 40 x 5 RHS and the main RHS wall.




BA

 

[bookowski]

Are you sure this is deflection they are seeing, not just the fit up tolerance? Deflection of 10mm would imply that they had it propped and then measured it dropping 10mm after pulling props, is that the case? Does it vary along the outriggers?

Hang a weight off of one (maybe an ironworker?) and measure how much it deflects.

 

[Contraflexure74]

Hi BA,

0.48 includes self weight of outrigger inclusive of edge channel out at the tip which is 0.28kN/m.

The columns at each end of the RHS support 6 outriggers in total, 3 each.

John.

 

[jwilki]

If this a fascia beam I think 10mm is getting up there for when considering a deflection for a line of sight perpendicular to the beam span. Also, what is the deflection & rotation happening from the fascia from the 2-M16 bolts, are they tensioned to friction grip or just snug? Have them laser/string-line exactly what is rotating and deflecting (column, 1st outrigger, mid span outrigger???) relative to the ground

 

[BAretired]

A yield line analysis of the outrigger connection to the 250 x 150 x 5 RHS indicates that the ultimate moment capacity is approximately 4.6 kN-m, so it does not appear to be a problem so far as strength is concerned but the connection tends to be a bit flexible, so a stiffener may be warranted.

BA

 

[dhengr]

JohnFitzgerald74:
Since I don’t work in the kN/m/mm/kg system every day, I have a bit of trouble developing a quick gut feeling for proportions and load and stress magnitudes, without some considerable effort in the conversion. Like Hokie, I don’t think the thing is going to fall down immediately, so I’m not so worried on that account. But, I’m not sure it was well designed in the first place, without any allowance for some adjustments, and to really take the various conditions listed above into account. The shims should look like an upside down W, with two slots open downward to fit over/around the two bottom bolts. These fit over the lower bolt and can’t fall out, and they might be slightly wedge shaped and of several different thicknesses. I’m not quite sure that I understand how you load that canti. structure in kN/sq.m units, it all looks like line loads and concentrated loads to me. How is this thing finished in terms of roofing, soffit and fascia? Can you hide part of the deflection with some adjustment of the final fascia height connection? Will ponding and snow drifting near the fascia be an issue? I would check the top weld on the 80x40 tube to the face of the 250x150 tube. That is a very short, difficult weld around corners, with potential start and stops, etc., for a primary tension weld; then you have the flexing of the 250x150 face plate under load too. You could remove the outer channel and have someone put some camber in it to help compensate a bit, with some adjustment at the two bolt shear connection. You might stretch some piano wire lines/strings, at the channel tips, and other locations to get some idea of the relative deflections and rotations at each of the outriggers. I would try to study and understand what each different part or detail is contributing to the overall deflection problem.

 

[Contraflexure74]

Dhengr.

It's finished with a layer of aluminium on top to a fall with a external render on the soffit which brings up another question do I have to limit deflection to L/360 due to soffit finish.

Ponding won't be an issue as the top has a fall, snow drift won't be an issue.

John.

 

[DETstru]

The max torsional moment on the 6m main RHS under the dead load condition is 1.4kNm.
The deflection of the main RHS is 0.06mm for the dead load condition.
The deflection of the outrigger at the tip is 0.5mm for the dead load condition.
The max moment of the outrigger for the dead load condition is 0.48kNm.
The moment induced at the column due to dead is 2.8kNm inducing a column deflection of 0.23mm

When you give these deflections is that the deflection at the tip of the outrigger due to the rotation of each component?
As in: The deflection at the tip of the outrigger due to rotation of the main RHS is 0.06mm for the dead load condition.

Because when I calculate the deflection at the tip of the outrigger due to beam deflection and rotation combined (I'm using an HSS10X6X3/16, is that close?) I get about 3mm of tip deflection just due to dead loads (1.7mm from bending and 1.2mm from rotation).

Then you have to add the tip deflection due to the bending of the outrigger and the rotation of the column. Add in some fit-up tolerance and I can see 10mm being realistic.

 

[Contraflexure74]

The 0.5mmm value is deflection at tip due to rotation. Going to site tomorrow to take some measurements to try and assess whats happening. Thanks everybody. Will let u know the outcome.