Truss connection

[Aye Min Khaing]

How to provide pin connection in practical as per design concept. Direct welding or gusset plate connection. Which one right concept??

[Hafsa Azmat]

either you do welding or make a bolted connection, your scheme of welds and bolts will gonna decide if your connection can takeup the moment or not

[Aye Min Khaing]

Can you give me some reference about scheme of welds and bolts to get pin connection. Thanks.

Edited April 9, 2020 by Aye Min Khaing

[ANStructs]

Practically, the behaviour of a bolted connection resembles to a "pinned" connection more than a welded connection for a connection designed as "pinned". Because welded connection are more rigid than a bolted connection. However in reality,  a bolted or even a welded connection designed as a pinned connection, provides some degree of resistance to the moment. Very few connections in reality are truly pinned or truly fixed. And, while a hinge connection could be considered a “worst case scenario” for the beam itself, it does not hold true for the connection, and the supporting member. Because, a "wannabe" hinge connection will in fact transfer "some" moment and it's possible that the connection, or other elements that the connection is attached to may be too weak to resist that moment.

Now coming to your question, a gusset plate is a mean to connect a member to a supporting member. A member can be fastened to Gusset plate using either bolts or weld. 
Direct welding can also be done, and is definitely a more cost effective option but, it could results in field fit-up problems. As to the degree of resistance to moment, If i were you, I would ask following questions to myself.
1. Do I need to be concerned with the possibility of end moments?
Depends on how large an end moment I'll get.
To check this, I'll perform the analysis for full joint stiffness (i.e., no end releases) and see how "big" the end moment I'm getting. If very small, I might neglect these end moments after cursory making sure it'll not cause any unwanted yeilding in my connection/supporting member. For large magnitude, I'll either provide a fixed connection or will properly model the provided joint stiffnesses. 

2. How rigid is the provided connection?
Depends on the thickness/size of connection plate and spacing of bolts or weld length/distances.
For exact estimation, I believe there's a method specified in euro code with detailed guide lines. (And before you ask, I don't remember the reference off my head, you'll have to surf the net for this one)

 

[Aye Min Khaing]

Thank you sir 🙂

[Simple Structures]

Let’s consider a steel beam to steel column connection – to respond to your query:

1. The connection first needs to be designed for the Shear Force on beam end - this vertical force will dictate - number of bolts or length of weld.

2. I have designed a pin connection, with two members, being connected by a 75mm (3inch) diameter pin/bolt – a large single bolt and a gusset connection gave me the pin behaviour, and full rotation at the connection. Full rotation at the connections means you will get close to a "pure pin" as you can get. Such connections are often used for "special" architectural features structures.

3. For a beam connected to a column say, a 2-bolt connection would be considered as 'pin' connection; Even a 4-bolt connection would be considered a pin; The higher the shear, the higher the number of bolts. Still pin connection!

4. For a moment or fixed connection, the end plate on the beam will be thicker, will have welded stiffeners, on both end plate/beam, and on column web too! The forces are much higher, and rotation of join is restricted. If you have a building where no shear walls or X-bracing was allowed anywhere then you would design all beam/column connections as fixed "moment connections" - sized to suit the moment you obtained from your structural analysis. The moment connections provide the lateral sway resistance also in this case.

5. Welding in a quality-controlled workshop is good - welding on site, often overhead welding situation, requires special procedures and specially trained welders! In most developed countries there are special procedures for site welding followed by onsite weld testing - site welding is therefore avoided where possible. 

[Simple Structures]

6. Ooops… the question related to a truss connection... just to add, same applies to a truss end connection, where top chord and bottom chords are fixed to a column, the depth of truss (lever arm) the moment connection, and resistance to sway. 

7 ...the internal connections, nodes in a truss are workshop welded, are pinned. …unless the truss is too big for transporting on a lorry, and then a bolted "splice connection" is provided to suit transport length to get the full truss length. These splice connections are bolted on the ground at the site location, and then the truss is lifted by crane into place (members also designed/checked for this temporary crane lifting forces!).

8. Finally remember, in a truss the loads in members in purely axial (compression or tension).

9. In a Vierendeel girder, where there are no diagonal members, then the vertical member and top and bottom chord connections are moment connection. But the site "splice" connection on the Vierendeel girder is also bolted.

[abbaskhan2294]

@Simple Structures can you provide any reference for the design procedure of fully and partially moment resistant connections in frames? 

[Simple Structures]

Full and Partial moment resistant connections in steel frame:

First thing to note is that steel frame design is all about understanding and mastering joint behaviour and joint design!..... whilst selection of beam or column size is easy.... it’s the joints that determine how the frame will behave under vertical and later loading. 

Steel column to beam joints are characterised by "stiffness" (rigid joint, semi rigid or pinned) and "strength" (full, partial or pinned).

For moment resisting frames the joints are MOSTLY "continuous or full moment resisting". As "semi-continuous or partial moment resisting" joints are discouraged as it relies on rotational stiffness of the joint. This is not easy to do, and thus most codes do not recommend it.

I have previously in another post said that one connection that I may design for "partial fixity" is the column to foundation connection - only for lateral sway purposes. Please read that post. In such cases the column is provided with minimum of four bolts spread apart, to assist during erection of the column as first item of frame. With four bolts it will stand up under temporary case (as a cantilever till beams are fixed to it), and often the most onerous loading case.

Here are pictorial examples of "full moment resistance connections in steel frames".... detail design of connection is an art. The connections require thick end plates, bolts with large lever arm, and cap and backing plates etc. 

Hope this helps the understanding the "moment resisting steel frame joint".

By the way, I started my career as a structural steel welder, reinforcement steel fixer, then a draughtsman, then Engineer .... thus I have welded and erected many steel frames in my early years!  The reason for mentioning this is that Structural Engineers, who have become engineers through gaining degrees and have missed out on practical skills, should make a habit of regularly visiting construction sites regularly, look at the details, talk to the trades men (or women) on site, and gain valuable experience from their see and hear….

 

[Simple Structures]

@abbaskhan2294   One book to reference on connection design is: Handbook of Structural Steelwork Connection Design and Details by Akbar R. Tamboli.

[reoman45]

can I say that you can assign any type of connection (welded or bolted) to any joints? then it would be up to the engineer to design this connection if it could take shear, tension, and moment? for example, you can implement a welded connection to act as a pin if it can take the forces but not the moment.

[Simple Structures]

 

The answer to your question is broadly YES.  Normally the consulting engineer does frame analysis and provides joint / node forces to the steelwork fabricator (a specialist) who will have his own design office / connection specializing design engineer just designing connections. He will design the connection to suit (i) their workshop way of working, (ii) & how they will assemble (erect) the steel frame on site.

In the western countries bolted connections are preferred on site, unless there is a special (architectural or other specific) reason for welding. 

Each country has its preference - where you are it may be different to the way its done where I am based.

Here are few points to remember:

1.You assign the forces for the connection - shear, tension and moment as you get them from your frame analysis - and put them in a table on on the drawing for each frame (grid line) or each joint / node. On the drawing is best, give your +/ve -/ve notation. You state if the forces are factored (ultimate) or un-factored (working) loads.

2. Endplates to beams are always welded (most cases) in the workshop, and the beam and column connections are generally bolted on site location.

3. Welding on site in most countries is avoided as it requires good quality control system to be in place on site and also requires an experienced specially trained "coded welder" to undertake the work on site. 

4. Avoid site welding at all cost - unless its the norm in your country. Bolted connections can also be designed as moment connection, and are used 95% of the time on site!

5. Give loads for each worst case, i.e Max& Min Dead, Live (Superimposed), Wind and Earthquake loading clearly stating whether loads are factored and un-factored and +ve -ve notation for avoidance of doubt. Connection designer will then combine them into the most onerous combination.

6. Lastly welded connection can be pinned also- diagonal wind brace welded connection has tension and shear forces - no moment. In a steel truss connections can be welded (or bolted) and there is pure tension or compression! Yes "connection design engineer" will know what to do .... you just give him the connection/node forces in tabulated form.

Hope this helps.

 

[reoman45]

Understood. Thanks for your reply!