Grating connectors in high seismic areas

[C. Lodge]

Good afternoon,

For those who have experience in high seismic zones- what do you normally spec out for grating connectors to fasten the grating to the structural steel?

Do you weld, use saddle clips or some other type of clip product? If you use saddle clips, how do you justify the resistance to lateral forces under seismic conditions?

Thanks for your time.

Colleen

 

[SlideRuleEra]

Saddle clips. No justification needed, heavy load are not permanently fastened to to grating. They are supported directly by the structural steel. Loads that may be "sitting" on the grating during a seismic event are not restrained by the grating... they would just "slide around" during an earthquake. The only seismic loads the saddle clips resist is the weight of the grating and any (small) horizontal friction force from loads "sliding" around on top of it. We do use non-serrated grating which should help reduce sliding friction.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[C. Lodge]

@SlideRuleEra, thank you for your response.

Due to the height of the building we have some 2g+ horizontal earthquake forces so the lateral force from a 500lb panel itself is 1000lb. This is ignoring any live load on the panel.

I am concerned about the ability of the saddle clips bolts to take it. They are only 1/4" diameter and the way the saddle clips are connected, the lateral force would induce bending on the bolt. For a 1 1/2" grating the "bending" capacity of a single bolt is on the order of 40 lb. If you think it's ok to load a bolt in bending. I don't.

 

[BridgeSmith]

We allow for bolts to bend under seismic loading, so we check the tensile capacity under the "prying" action of the movement and check the shear capacity. If there is concern about bending, you could add a plate washer that fits the openings in the grate. That will severely limit your tolerance on the bolt placement, though. Or you could use bigger bolts. 1/4" bolts seem small to me for a grate of that weight, although we don't see grating all that often and ours are secured by 1/2" high strength bolts through the grate from the railing post base plates to the support structure.

 

[JoshPlumSE]

Am I the only one who thinks this would be a great question to ask a Hilti Tech guy? Just to see if they've got documentation or testing concerning lateral force resistance....

You could look into other products as well. My thinking is to start with the Hilti guys just because they seem to have a bit more technical data and testing available.

 

[SlideRuleEra]

C. Lodge - Thanks for the additional information. Our electric generation stations may be somewhat different than a typical building (I don't do buildings).

1) There are acres of (banded) bar grating and some of it is surrounded by concrete floors. For those areas, as long as the grating is held down (saddle clips,) it is not going anywhere.

2) All grating has to be (potentially) easily removable for operational reasons. Don't want to weld it down.

3) For all grating areas, including those surrounded by concrete floors, the grating is not needed for any structural reasons. Structural steel framing takes care of all bracing.

Given those constraints, I am comfortable allowing friction between the (saddle clipped) grating and the supporting structural steel to provide horizontal seismic restraint.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[JLNJ]

If you are not buying SRE's argument, could you shop weld some tabs, "stops" in effect, which would keep the grating from translating horizontally?

 

[cvg]

use more saddle clips, 12" OC instead of 24"

 

[MotorCity]

Not sure how you calculated the seismic load of the panel, but it can never exceed the weight of the panel (2g does not mean 2x the weight). If you are in the US, take a look at ASCE 7 for the seismic load of components. The seismic load from a 500 lb piece of grating will be very small.

 

[TLHS]

Sure it can exceed the weight of the item (at least in the basically identical Canadian formulation of the non-structural component equation). At full building height in areas where the mapped short period acceleration is in the 1g+ range you'll see that sort of reaction from an item that responds flexibly (specifically, when it falls into certain types of resonance with the main structure, but the formula in the US/Canada doesn't check that).

It seems unlikely for this type of item, though, which is fairly rigidly restrained to the structure.

A floor should have an Ar value of 1, an Rp somewhere around 2 or 3. If you're at full building height I would generally expect a design seismic coefficient of no more than around 0.5g for something like grating... Unless the grating is part of a secondary structure that could dynamically amplify.

Even then, I wouldn't be expecting 2g.

I also think ASCE-7 has a cap on the maximum amplification that needs to be accounted for, which probably makes 2g impossible unless you're doing a more advanced analysis that would indicate it. With the Canadian cap, you end up limited to about 1.3g in the bad seismic regions.

 

[C. Lodge]

Thank you all for your responses.

@JoshPlum, we have the Hilti data and the Hilti discs can take about 65 lb laterally without permanent deformation. This is slightly better than saddle clips but we would need a lot.

@JLNJ, we did look at this but we have some complicated layouts in some areas so the practical ability is somewhat limited, it looked like we could end up needing 6 or more of these per panel due to geometry. However, if you have an example of this you would be willing to share I would certainly look and see if we can make it work.

@TLHS- So, the seismic load that was calculated took into consideration a multi-modal factor of 1.5. I need to look into this and see if I agree with that factor. Without the multi-modal factor, the acceleration is 1.4g. This is based on an advanced analysis (finite element model) that was performed by others within my company. We are governed by this analysis, we do not use ASCE-7 for loading.

The grating is supported by steel beams that are relatively long and in the weak axis direction they have a low frequency around the order of 7 hz. These connect into the main concrete shear wall structure.

@TLHS, if you were dealing with grating in an area with a 1.3g acceleration, what would you use for connecting the grating to the structural steel? We probably only have 0.2 for a friction factor so we would not be able to overcome all of the seismic lateral load with that friction unless we took a clamping force into account from the clips to increase friction.

Thanks

Colleen

 

[TLHS]

So you've developed a custom floor spectrum?

In that case, you should still be able to use the Rp value. Your custom acceleration would just replace the 0.4 x Short Period Accel x height factor in the non-structural component formula, since that represents the local peak acceleration at the connection point of the component. Are you currently designing to elastic loads without an R value? Or have you incorporated the building's overall R value through the building analysis? Neither of those are necessarily valid.

I'd only be expecting them to use a modal factor for a static analysis. A static analysis wouldn't give you a better floor spectrum than using the non-structural component forumulas, generally. The non-structural component formula is already based on those same assumptions. Basically, it sounds like people might be overcomplicating the problem without getting a benefit for it.

I wouldn't design to that load, because it isn't a realistic requirement except in extreme situations. I'd only expect to see numbers that high for grating in areas where they might be asking you to design to full elastic loads. Critical infrastructure and similar.

If I were a third party engineer being asked to do the design, I'd send a letter telling them that the loads look excessive, and then if they verified them I'd probably just use a weld detail to secure them. At the end of the day, in that situation, they're my client asking me to overdesign. If the loads came from somewhere inside my company, I'd chase them and try to figure out how we got to that point, because it doesn't mass the smell test.

 

[C. Lodge]

@TLHS- There is proprietary information that I am not able to share with you but the area I am working with is on a high floor for a nuclear power plant in a high seismic zone. The data I quoted are based on a dynamic analysis and I know that the accelerations appear stunningly high but they _are_ correct.

If anyone has any suggestions for how to anchor this grating to withstand the significant lateral forces it would be appreciated.

 

[bones206]

At one of the nuclear plants I've worked on, I remember seeing threaded studs welded to beam flanges, which fasten to the grating through square steel plates shop-welded to the grating bars. The plates fit between the bearing bars were set about halfway down the bearing bars. You can make this detail as heavy duty as you want.

 

[SlideRuleEra]

C.Lodge - I can help you. You mentioned that the grating is 1 1/2" with 500 lb weight for a typical piece.
How many square feet is that?
If the grating happens to be 19-W-4(1 1/2 x 3/16) Steel, that would be about 45 ft² - fairly large.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net