Column Building Sway Supporting a Gantry Crane 1

[NathanNZ]

Im designing a steel portal frame building to AS/NZS 1418.18 that supports an 2 10 tonne internal gantry cranes.
Link
Section 5.13 Serviceability, requires that the columns supporting the crane are not allowed to deflect laterally more than 10 mm at the support under serviceability loads. This seems ridiculous considering the crane supports are 8m above ground and the portal frames over the building span 20m. This equates to designing the building for H/800 from sway from wind. Compared to H/300 to H/250 for a normal structure.
So to achieve this, i'm going to need 200kg/m welded steel beams for the portal frames rather than say the largest 610UB which is 125kg/m.

What are your thoughts?

 

[Agent666]

Structures I have peer reviewed and those designed by colleagues have usually complied with these requirements. Sometimes the crane suppliers have looser/similar/tighter requirements, so worth checking if they have any additional requirements.

One effective way to address the issue is to brace the end walls in the direction of the portals and fully brace the roof to act as a diaphragm to transfer loads back to these end frames. This is done typically so your lateral SLS deflection is reduced (partially from the portals and the end bracing sharing the loads depending on the stiffness of the roof structure). Alternatively you can simply increase the section sizes for the frames, but this doesn't get your best bang for buck, but of course this might depend on the building length/aspect ratio.

Usually for larger cranes you have no choice as the sum of the horizontal thrusting/inertial loads are often higher than the total lateral loads from wind, and as such these can govern for the lateral design.

Note you also need to come up with a strategy for concurrent loading of the cranes if its at all possible that the thrust loads can occur at the same time on a frame (i.e. both cranes operating in the same bay or similar), tying multiple portal frames together via the previously mentioned roof diaphragm can effectively aid in spreading the loads over multiple frames even with no end wall bracing (model structure in 3D to determine the roof bracing effectiveness in spreading the loads to multiple frames). Its possible as I understand it, if the cranes are operating on the same rails that they can be programmed to not approach each other to avoid any adverse loading conditions, but if they are on a second set of adjacent rails, then you can be forced to count as 2 times the thrust/inertial forces on a single frame for design purposes. As I understand it AS1418 has no real guidance on the combination of effects from multiple cranes, so its left to engineering judgement, but it should reflect all possibilities of the actual loading situation that could occur.

I've also seen designs that add roof/wall bracing suitable strength/stiffness wise for the serviceability loads, but under higher seismic or wind loads these braces are intended to yield such that the portal response becomes more dominant at the ULS. This allows some economy in the design if its included/considered in an appropriate manner as part of the overall design philosophy (i.e. capacity design approach if it yields as part of the loading scenario). If you had an ULS event then its replaced or repaired to reinstate the functionality required for crane operation, otherwise it simply is intended to act to reduce SLS deflections until such time that it fails under higher than SLS loading.

 

[Agent666]

Don't forget also that at the SLS, you might be able to justify some degree of base fixity from your baseplate connection, especially if you are considering them as being effectively 'pinned' for the analysis. This can be quite effective in reducing the lateral SLS deflections.

 

[NathanNZ]

Thanks for the advice, it is very helpful. We have braced the entire roof and end walls but because of the aspect ratio portal frame action is still fairly dominant. We have also assumed 20% base fixity in our model. Concurrent crane loading has been allowed for as it is small compared with the wind loads. My conclusion is that we design for the 10 mm and give the client and cost guys a surprise now with the increase in steel size rather than down the line when it is peer reviewed or worst yet, problems did occur to the gantry crane operation.

 

[Agent666]

If you are going with welded sections you can come up with an efficient/economic arrangement with tapering the members and varying flange/web thicknesses and putting any additional steel where you get the best bang for buck (closer to the knee). Putting the same section throughout isn't the most efficient use of the additional material obviously. Obviously a deeper lighter/stiffer member is possibly better than using a heavier welded 600 deep member to match the original design proposal. It might even work with columns and first 5 meters of the rafters being welded with the original 610UB between.

Other alternatives I have seen to possibly give you some other ideas(you didn't mention how far out you currently were from the 10mm limit so no idea if these would work):-
- External props down to the ground outside of the building envelope.
- Props or haunches at the knee of the portal if there is sufficient height above the crane, a increase in stiffness in this region can improve deflections fairly cheaply. Small raise in roof height might permit a prop/haunch that stiffens the knee to the point where it works with the original section.
- Design/detailing for a higher level of fixity at the base.
- Stiffen the bracing system (add more braces or use axially stiffer members to demonstrate that you can effectively spread the load over more adjacent frames) - ask yourself how many do I need to mobilise, and what might that bracing system look like to achieve this.

Definitely find out if the supplier also has any limits as these add support for what might be perceived as a drastic change or possibly 'over engineering' by the cost guys (you know the old we've never seen anything like this before in any other buildings that have similar cranes).

I would say also that I have seen one designer we peer reviewed that completely ignore all the limits, mainly because they didn't know about them initially, followed by a severe reluctance to change their design. No idea if it ever came back to bite them or not! We simply refused to cover this aspect and made it clear to the client and local authority that the designer was operating outside of best practice recommendations. Saying that, its hard to know at what point you would definitely have an issue, cranes seem quite forgiving of all building movements apart from the relative distance between the rails.

 

[KootK]

I'm no expert on crane tracking problems but, at some point, I would think that absolute limits would sensibly give way to span:depth limitations solely. And, in that respect, I would think that it would be one or both of the following:

1) Rotation of the rail about its own axis as governed purely by fairly generous H/delta and;

2) Lateral curvature in the rail run governed by Run/delta.

This assumes that:

1) P-delta and all of the other, usual engineering stuff has been tended to and;

2) Relative lateral movement on opposite sides of the gantry can be kept to a minimum.

Long story short, I feel like a long building should be allowed to have more, absolute lateral movement at the tops of the rails. One would think that erection tolerances over the length of a very long building would be producing pretty much the same effect anyhow. But, alas, the rules are the rule and there's risk in deviating from what has been shown to work well in the past. If anyone can expand upon the specific tracking issues associated with the 10 mm limit, I know that I would be interested to hear about it.

I've one solution to add to Agent's well considered list. It's unlikely to suit your clients programmatic needs but you could replace your columns with short concrete shear walls (precast/cip) on one side of the run. I saw this done on a precast storage facility recently (double tee roof) where solid end walls were not an option. I was surprised at how little wall it actually took to get the job done. This was in a low seismic zone. In your case, I wonder if you could use walls for serviceability loads but then allow those same walls to rock under higher, seismic demand.







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.

 

[Agent666]

KootK, I believe its the cumulative effect of the 'normal' erection/fabrication/cross sectional tolerances and the addition of the building sway vs the inherent tolerance of the crane itself that can possibly cause issues with the crane operation and tracking.

Refer to the following excerpt from an Australian Steel Institute (ASI) document which attempts to explain this Link. Note that the deflection limits in the standard are more strict than the limits in this document. I believe if the manufacturer specifically states limits then these can be used as an alternative, the standard allows this I believe in clause 1.2 where it states other methods not noted in the standard are not prohibited provided the designer can demonstrate that generally accepted methods or procedures have been employed.

Here's another take on looser deflection limits (again from an ASI document), it specifically notes the differences between the standard and limits found to work in practice).Link