Bolted column splice type 3

[GD_P]

Hello forum once again,

How to select type of bolted column splice between
- Covet plate splice (require preloaded bolts for rigid connection)
- End plate splice ( do not require preloaded bolts for rigid connection)
Ref. SCI P398

FYI:
-The columns are considered continuous and support lateral load through moment connection with beam in elastic global analysis.
- The floor above the splice support rotary machine such that dynamic load ( though small in magnitude) acts along the plane of floor.
- The splices are located below the floor within 500 mm

Please share your opinion

GD_P

 

[BridgeSmith]

Using splice plates (cover plates) to the flanges is more rigid and you can get whatever capacity you need by adding more bolts and thickening the plates as required. The end plate connection will be at least a little less rigid for bending, and the moment capacity will be limited by the placement and number of bolts that will fit in the end plate. So, the feasibility of the end plate likely depends on required moment capacity and the magnitude of the compression. For a column with high axial compression and small moment, the end plate connection may be as good or better, but for a column with light axial load and required moment capacity near the bending capacity of column, designing an adequate end plate connection could be difficult.

I'm not familiar with fabrication and construction methods for 'vertical' construction, but I would guess that the using splice plates to the flange would be the more economical choice, since it requires no welding and likely fewer bolts (at least in high moment applications). I assume the end plate connections are typically used to avoid having protrusions outside the flanges.

 

[Agent666]

For an end plate splice I'd question if you are not using tensioned bolts if you are transferring moment. Tensioning the bolts helps draw the plys together and creates the rigidity required.

I'd agree with HotRod10 though, usually splice plates are a better option that offers more design flexibility, in fact any connection that uses bolts and plates in shear is preferable to placing bolts in tension and plates in bending in my mind.

Both connections will require 'end bearing preparation' to ensure even bearing (usually a code requirement). The splice plate is usually cheaper in terms of fabrication as there is no welding, and only cutting and drilling plate and sections (quite easy to process in a beam line process). For erection use of end plates is generally easier and cheaper due to being able to rest the column on the last section and easily locate a few bolts to secure it, vs needing to balance the column on top of the previous column lift and place and bolt the flange and web plates to secure it in place, and of course many more bolts to tension.

I'd also question the location of the splice being below the floor, generally for erection its easier to locate them above the floor as its potentially easier for the erectors to work on it from the floor level. Not sure on Eurocode approach, but locally we are required to put them near the middle of the column length away from the maximum design actions for flexure (exact requirement is locating them within the middle third of the column length).

 

[GD_P]

Thank you for your valuable posts and sorry for the delayed response.

As depicted by HotRod10, yes I am using the end plate spice to avoid the protrusion outside the column and moment to be transferred is less than the 25% of the column moment capacity.
Regarding the location of the splice,
Current structure application is shop fabricated chemical structures (skids), which require structure splitting (hence column) just below the floor, as floor to floor spacing not too high (2.4 mts)and which ensures erection from the floor.
Agent666 (Structural)6 Dec 18 19:03
For an end plate splice I'd question if you are not using tensioned bolts if you are transferring moment
Actually I referred the SCI book P398, which explicitly specify that end plate splice dont require preloaded bolts whereas coverplate splice do require.
I am a bit confused now.
May I know what what will happen if we dont use the preloaded bolts.
One more question, how to evaluate the minor axis moment capacity of the splice, any reference will be helpful.


GD_P

 

[Agent666]

If you don't tension the bolts, the connection won't be as stiff, as the plys might not be drawn together. Perhaps for your application (not a building) it won't matter too much. Our code in NZ specifically required tensioned bolts in these applications to maintain stiffness (assuming of course you are designing the connection with continuity of the member in mind.

Another aspect would be that tensioned bolts are better for fatigue, for smaller forces there is no change in the bolt force/stress state.

For minor axis capacity, from the major axis calculation you'll have the capacity of each row of bolts as a group (all bolts one side of the web with a collective yield line pattern). I'd halve this with the simplifying assumption that the bolts on one side of the connection are yielding the plate in a similar pattern to the major axis direction. You could add in the other row using the elastic linear force assumption based on location and yield force of the opposite row. Then work out moment capacity (forces x lever arms) similar to the major axis direction.

 

[canwesteng]

Snug tight bolts will be drawing the plies together by definition, pre tension isn't required unless it is designed for. I'd have no reservations with respect to fatigue using snug tight bolts in flush end plates unless you are supporting vibrating machinery or cranes, which it appears you are. I would generally want to see pre tensioned bolts in splice plates if there is any possibility of load reversal, similar to gussets for wind bracing.

The best resource by far is design guide 16 from AISC.

During construction, end plates are generally easier to install and require fewer bolts, the trade off being that it is not feasible to achieve 100% moment capacity.

 

[GD_P]

@Canwesteng
In this case, I find that even if I assume friction coefficient 0.1, the friction force is more than the lateral force due to machinery (few centrifugal pumps).
My doubt is, can I rely on this assumption, for not using preloaded bolts?
Agent666
Thanks for the idea, I will work on it.

GD_P

 

[Agent666]

You'll be checking your bolts for combined shear and tension either way with an end plate. Tensioning doesn't affect this.

 

[BridgeSmith]

It's not only the friction value, which determines whether there is slip between the plates, it's also the tension fatigue of the bolts to be concerned about with non-tensioned bolts. The stress on pretensioned bolts doesn't change much at all when the connection is subjected to cyclic loading. Not so for non-tensioned bolts. For bridges, AASHTO requires that any bolted connection subject to cyclic loading meet the requirements for a slip-critical connection (all fully tensioned high-strength bolts).

 

[canwesteng]

The concern with vibrating machinery is that the bolts will become loose over time. Pumps maybe you could ignore the vibration as I doubt the actual connection sees it depending on your layout, I assumed it might be screens or fans, in which case I would be pre tensionning the bolts.

Re: Agent - the AISC design philosphy lumps all the shear into the bolts in the compression zone in order to use the full tension capacity of the bolt for moment resistance.

 

[Agent666]

I guess that will only work until it slips. For an endplate with minimum number of bolts the slip force might be quite low.

The code I work to doesn't allow this mechanism, always requires the bolts to be checked for the shear/tension interaction. Though you can always assign bolts on the compression side to carry the shear, and tension side to carry the tension only.

 

[GD_P]

Eurocode also uses the full shear capacity of the bolts in compression and 28% shear capacity of other bolts in tension.
This 28% value is derived from the shear and tension interaction formula such that the bolts can be used for their full tension capacity.
Thank you for your valuable comments.

GD_P

 

[GD_P]

Hello forum,

I am evaluating the minor axis moment capacity of the end cap splice, I have already evaluated the effective length of the equivalent T stub. But I am stuck on the following points:
What should be the lever arm for calculation of the moment capacity?
a) compression flange edge to bolt centre?
b) Centre of compression area (at a dist 1/4th from compression edge) to bolt centre?
c) Any other
I personally think, it should be option b)

For minor axis, we reduce the tension capacity of bolts rows, if it is greater than resistance of compression flange, which is calculated considering the moment capacity of beam is resisted by flanges only: M / (h-tf), where M = moment capacity of beam ; h = depth of beam; t= thk of flange.
For minor axis BM capacity less as compared to major axis, so it is very likely that we will have to reduce the tension capacity of bolt rows, but how to calculate it for minor axis?

I went through guide design guide 16 from AISC, but it only deal with major axis connections.

Any guidance will be very helpful.


GD_P