Basic truss chord splice question

[jbuening]

A bit of office discussion around here and surprisingly the opinions are split down the middle, so curious what you all think.

If you were to analyze a Pratt truss in a program like RISA and determine the max tensile force (lets say 100k) in a lower chord midspan at a node, for a connection design is that 100k on each side of the node or the combined force at that node? What I'm getting at is lets says you have a splice plate on the lower chord with 2 bolts total (one each side of node), is the shear force on each bolt 100k or 50k? The 100k force is shown in the detail for the axial force at the end of the member on each side of the node.

 

[sekwahrovert]

Using RISA, I don't know how you determine the force at a node, but you can determine the force in a member. Look at the force near the spliced end of the chord. That is the force that needs to be transferred into the bolts and plates that you will design to form the splice.

 

[jbuening]

Yes, that 100k force is the force at the end of the member (which essentially is at that node) from the member detail/output. Member to the left of midspan has 100k tension, and member to the right of midspan has 100k tension. For a bolt on each side of the splice (2 bolt for entire splice), is that 100k shear per bolt or is the 100k force distributed between the two total bolts?

 

[MDStruct]

I second that. I don't know how you check forces in a node in RISA. I would use the adjacent members to determine the force in each. If this is the case then you have 100k in each member equal and opposite, so 100k in each bolt, one bolt holding each member. Your node itself should give a 0 result due to equilibrium unless you are looking at a support.

 

[bootlegend]

Maybe the node is confusing the issue. You've already acknowledged that you have 100 kips tension at each member end. So you need to develop 100 kips on each side of the splice. 100 kips per bolt in your example.

 

[sekwahrovert]

You need to transfer 100 k from the left chord to the left bolt. Then 100 k from the left bolt to the splice plate(s). Then 100 k from the splice plate(s) to the right bolt. Then 100 k from the right bolt to the right chord. I think a free body diagram will easily settle the question.

Is it possible that the bolt shear question is not related to the issue you have brought up, but instead is about single vs. double shear of the bolts?

 

[jbuening]

Agree bootlegend, my description including the node is likely confusing others. It is 100k on each side of the node (each member). I thought it was 100k per bolt as well, but others thought it was additive at the joint and therefore divided among the bolts. They argued how you can have 200k total on a connection when only 100k is physically present. That then had me questioning my own logic!

Lets kick the discussion up a notch (see attached Solidworks model picture), and see where the entire discussion is being had. We have prefabricated truss sections that are being joined end-to-end. First we have a lower chord splice, with the aformentioned two bolts in double shear. Then the verticals are angles that bolt together, three near the bottom that would be in tension and three near the top that merely keep the compression chord aligned. Analyzing this connection has been a headache, since the vertical angles in my opinion will act somewhat like a base plate in which the angles will bear on each other above the neutral axis. The lower three bolts as well as chord splice bolts will take up the tensile forces. Load path is through one splice bolt, up through the three lower angle bolts, and back down through the second splice bolt.

How would you approach this analysis? Neglecting one side of the joint, I'd have a compressive force from the upper chord, tensile force from lower chord, the full vertical angle with compression only springs throughout the length, and then have the three thru-bolts and single splice bolt all as pinned supports. Reactions from pinned supports would be used to determine bolt tension/shear. That was my thought. But following the load path, the splice bolt would have to first accommodate 100k shear, then the 3 bolts would resist the uplift/tension from a bearing analysis. So including that single splice bolt in the model would be underconservative/incorrect, no?

 

[bootlegend]

I think the axially loaded chords will be much more stiff than the vertical angles. This will mean the chord splice will carry the majority of the load. If you strengthen behind the angle to create more of a "box" you can probably do what you describe, but at that point I'd probably just try to design the chord splice for the full load.

 

[SAIL3]

I would also design the bottom chord splice for the full load and make the top & bott chord splice the same detail....

 

[XR250]

It is amazing how computers make us forget (or never learn) simple statics. Do a method of sections on the truss and in 60 seconds you should have an answer.