NBCC 2020 Seismic Evaluation - Table 4.1.8.18. Non-structural Components 1

[Doodler3D]

Hi,

For a vibrating screen with 3 dofs bolted on a 2-level, cross-braced w-beam structure, is it appropriate to evaluate the
component as rigidly fixed or flexible?

Thank you

 

[KootK]

Sounds like a rigid situation to me.

 

[skeletron]

Bolted into a braced structure = rigid (IMO)

CSA S832 defines "stiff OFCs" as those with a fundamental period (OFC + connection) less than or equal to 0.06sec. You could backcalculate your stiffness "k" limit based on the weight, and then determine if that stiffness is exceeded. Still some guesswork involved, though...

 

[Doodler3D]

Thank you. The screen weighs 120 tons. The NBCC does not specifically address heavy vibratory machinery unlike the ASCE 7 hence the question.

 

[TLHS]

The below is a bit of a rambling stream of thought, but this is not a small screen. This is not necessarily rigid. This is also not necessarily a non-structural component. This is likely in the Canadian hand-wavey combined system world. For all of the below, I'm assuming the screen is on its own support framing without significant other masses. The effect of the various stuff definitely depends on the flexibility of that framing, though. For a vibrating piece of equipment they could have stiffened the absolute crap out of it. 0.06s is still pretty stiff though.

Where are you assuming your datum for transition between component and the ground? Are you assuming it's rigid but the equipment is at the top of the building/frame, using the height adjustment factor? That might get you a reasonable number, but I think that's more an accident of the way the non-structural component stuff was formulated. Are you assuming it's rigid with the equipment being the bottom of the steel framing, and thus at ground level? That will be wrong unless the framing is really rigid and short.

The non-structural component section of the code is based on the assumption that the non-structural component is small enough that it doesn't affect the structural response of the structure that it's mounted to. Basically, the structure it's mounted to waves around like an inverted pendulum and the non structural component takes the motion of that pendulum as an input and moves around like it's own independent inverted pendulum. It assumes that the structure is going to move however it wants and the non-structural component is along for the ride. It half-asses a floor spectrum based on the height factor and then uses that the modify the non-structural component's zero period acceleration / peak ground motion input.

This is not that. The inertia of the screen drives the motion of the frame.

If the frame is mounted on the ground, then you could theoretically design the screen plus the frame as a combined non-structural component, but it's not really the right tool for that. The non-structural component stuff is a bit of a hack job intended to simplify design of lower criticality items with poorly defined load paths. The R values are not well supported and the amplification factors are very simplified response spectrum approximations. Realistically, if this braced frame is of any height, the design may make sense using the building seismic provisions, but with a different assumed mass distribution. Basically, but all the mass at the top and then design this as a short period structure using conventional construction. This is effectively what the US code calls a non-building structure similar to a building. This is what I would do, with maybe a two line non-structural component check at the bottom to satisfy a checker that disagrees with me.

That being said, this is an area where if you talk to five different engineers, you're going to get five different answers. I've done time history analysis on a couple of different situations for my own curiousity over the years though. If the component dominates the mass of the framed system and the equipment is reasonably rigid, the seismic motion basically ends up being a one degree of freedom inverted pendulum with the base shear being based on the overall system period. The simplifying assumption is then to just assume peak spectral unless it results in something dumb.

This is going to be a lower base shear than if you assume it's a flexible non-structural component, but on the top of a building/structure (assuming the framing moves independently and drives the screen and then the screen also responds), a higher base shear than a rigid non-structural component at ground level (this basically assumes that the component is moving at PGA), and should theoretically be similar to a rigid non-structural component on top of a building/structure (assuming the framing moves independently and drives the screen, but the screen is rigid, which is a one DOF system). The last one isn't going to be fully consistent, though, because the R value isn't based on the supporting system, and the acceleration is likely lower because the formulation of the non-structural component pulls the 2/3 reduction for base shear in that the US uses. It also approximates converting from S(0.2) to PGA and then approximates converting from PGA back to peak spectral and it gets a bit weird.

The newest S16 has seismic provisions for industrial structures. Personally, I don't think it really holds together all that well, though.

From another source, there's this:

https://www.structuremag.org/?p=11600

For rigid non-structural components on structures that where the non-structural components are 25% or greater of the total mass, it basically says to add the mass to the supporting structure and design based on the supporting structure values for R, etc. Which is basically what I was trying to say. If the frame is of any significant height, design it like a building with a mass on it. If you want, you can design the screen anchorage as a rigid non-structural component on top of a building.

Also, answering the original question, the commentary says this:

"If it is not feasible to reliably determine the fundamental period, then it would be appropriate for the designer to compute the force Vp on the assumption that the component is flexible or flexibly connected, since that case results in the larger force."

I'd generally be carefuly designating anything significant in size as rigid. You're basically aaying that it's so rigid you'd be okay designing it to PGA (and then applying an R value to that) rather than a higher spectral value. There are times where this is true, but when you start using it for bigger things like tanks, hoppers, etc, I don't think a lot of us would be doing that if we went from first principles. That being said, a big screen is probably rigid given that it's a vibrating structure.

 

[Doodler3D]

@TLHS,

Thank you for the detailed response, star for you.

The screen is installed on level 1 of the sub-structure. The sub-structure sits within a larger steel structure that supports a kiln. Additionally, the structure supports two bins for the conveyed and sifted ore. The bins are on level 0 and level 2, to feed and discharge the ore (9 m inter-storey height). The structural consultants have requested preliminary lateral seismic load calculations for each component from the equipment suppliers (Because of the 3 vibrational dofs??). I'd say it's section 4.1.8.18 of NBCC 2020, but 120+ tones of vibratory and impact loading is fairly complex as you described. Of course, this evaluation is not in my technical capability and will be sent out to qualified engineers. Nevertheless, it would be a missed learning opportunity if I did not investigate.

 

[Doodler3D]

From Metso's website, stating the obvious and a little more:

Screens, and especially larger screens, must have extremely strong and rigid supports! If the structure lacks sturdiness, the screen will dissipate its vibratory movement, causing the supporting structure to vibrate and either side of the screen to display differences in amplitude and phase.

The key to having a sturdy and reliable supporting frame is to take into account in the calculations the static and dynamic forces generated by the screen. Besides that, the frequency of resonance of the structure and its harmonics need to be very different from the vibrating frequency of the screen. If the speed of operation of the screen coincides with, or is close to, the natural frequency of one or more components of the structure, the vibration will be harmful for both parts. This can lead to serious personal injury and/or damage to the equipment and components.

We recommend that the structural frequency of the supporting structure or supporting frame is at least two-and-a-half times higher than the frequency of operation of the equipment.

The levels of vibrating structure should not exceed the max. allowed limit, measured at the lower part of the spring pads that are fastened to the support, in a tri-axial direction. The max. allowed limit can to be found in the GA drawings of each unit. These values are also inside our Screencheck software.