I am seeing more and more SE’s not design the Steel connections and instead leave this item to be done by the fabricator. I don’t mind because we are often hired by fabricators to complete their work. But if you are hired to do a job, do it, don’t just spec “connections by others”. What do you do, design the Connections or Specify them?
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Structural Connections 4
- Thread starter BRGENG
- Start date
-
2
- #22
JennyNakamura
Structural
BRGENG,
I believe it is mostly a regional (US) thing. I'm on the west coast and it is customary for the SEOR to design all the connections. However, it is not fair to classify someone as a 90% engineer just because local custom dictates that they don't design the connections. As others have alluded to, connection design can be very time consuming, so if the SE is going to design all the connections he should be compensated for it. If the owner is refusing to do so, then let the fabricator do it.
Also, would you call an awesome bridge engineer with 30 years of bridge experience a 50% engineer because he can't design an extension for his next door neighbor's house? Or because a bridge engineer is a civil PE and not a legal SE? I think as SE's we sometimes get a little caught up in our own arrogance.
I believe it is mostly a regional (US) thing. I'm on the west coast and it is customary for the SEOR to design all the connections. However, it is not fair to classify someone as a 90% engineer just because local custom dictates that they don't design the connections. As others have alluded to, connection design can be very time consuming, so if the SE is going to design all the connections he should be compensated for it. If the owner is refusing to do so, then let the fabricator do it.
Also, would you call an awesome bridge engineer with 30 years of bridge experience a 50% engineer because he can't design an extension for his next door neighbor's house? Or because a bridge engineer is a civil PE and not a legal SE? I think as SE's we sometimes get a little caught up in our own arrogance.
JennyNakamura
Structural
An after thought I forgot to mention, connection design and detailing is either in the scope of work within the executed contract, or it is not. Simple as that. As a design engineer or project engineer, it is not your decision to make -- the decision has already been made for you before the project is handed to you. If the firm's scope of work does not include connection design and the boss walks by your cube and sees you working on connections because you want to be a 100% engineer, I guarantee you will get in big trouble. And if you get caught doing it again, you'll probably get fired.
Just curious...if structural drawings do not show the connections, what information is shown in the sections? Or do you just not draw sections? How is the design information imparted to the connection designer? List of each member with shear, moment, torsion, tension, and compression requirements, including combinations and multiaxial conditions? Seems like more work than designing the connections.
I don't believe that simple shear connections for 6-7 different depth WF shapes comprises that much of an overall building design. When you consider all of the design, not just beams, girders, and columns, but lateral system, diaphragm, meeting drift and deflection requirements, all of the design that goes along with detailing, foundations......... Not to mention the documentation. I just don't believe that 6 or 7 different connection configurations constitutes THAT much of any design.
Also, I don't know how it works out for the guys who do document connections, but there is no benefit to us to design and document them, because, as noted above, 80% of the time the fabricator will want to change them. Do you guys never get asked to change from shear tabs to double clip angles?
We never blindly publish reactions. We always check to make sure a connection will actually work without web doublers or anything crazy. I also don't think that farming that out makes you less of an engineer. Metal studs get delegated all of the time, so do stairs. Are the engineers who delegate these things less-than engineers?
As for other professionals who farm stuff out - the medical profession is noteious for this. If you have two different afflictions you can be sure you're going to see three different medical professionals. When is the last time a tax attorney tried a murder case?
Also, I don't know how it works out for the guys who do document connections, but there is no benefit to us to design and document them, because, as noted above, 80% of the time the fabricator will want to change them. Do you guys never get asked to change from shear tabs to double clip angles?
We never blindly publish reactions. We always check to make sure a connection will actually work without web doublers or anything crazy. I also don't think that farming that out makes you less of an engineer. Metal studs get delegated all of the time, so do stairs. Are the engineers who delegate these things less-than engineers?
As for other professionals who farm stuff out - the medical profession is noteious for this. If you have two different afflictions you can be sure you're going to see three different medical professionals. When is the last time a tax attorney tried a murder case?
Hokie-
I must have started typing my post before yours was up.
We don't purposefully fail to cut sections where delegated connections exist. We don't go out of our way to cut one there, either. If that's where a section wants to be cut, we cut it there. If a section is cut where a delegated connection exists, we show a generic connection (typically double angles), but don't call anything out.
We provide a Beam Reaction Table that lists reactions for each beam depth. Sometimes, if there is one very large reaction for a particular beam depth, we will show that one on plan and eliminate from the schedule so that all beams of that depth aren't being designed for that reaction (i.e. say all but one W16 has a 30 K max reaction, but one has a 80 K reaction; I'll show 30 K in the schedule and 80 K on plan at the one particular location).
For moment connections, our typicals call for full-pen welds, so there is no design to do there. Occasionally, we'll have a fabricator ask for design moments so that they can design them and not use full-pen welds.
We've even, on recent jobs, delegated braced frame connections. What we've had happen is we go through the design and detailing of the gussets, welds, bolts and then the fabricator comes back and says, "You know what, we prefer to make our braced frame connections with double angles field bolting to the gusset instead of the field welding you show". That's a LOT of time and effort wasted for no reason.
I'd be interested to know how the guys that design and document connections deal with those situations.
I must have started typing my post before yours was up.
We don't purposefully fail to cut sections where delegated connections exist. We don't go out of our way to cut one there, either. If that's where a section wants to be cut, we cut it there. If a section is cut where a delegated connection exists, we show a generic connection (typically double angles), but don't call anything out.
We provide a Beam Reaction Table that lists reactions for each beam depth. Sometimes, if there is one very large reaction for a particular beam depth, we will show that one on plan and eliminate from the schedule so that all beams of that depth aren't being designed for that reaction (i.e. say all but one W16 has a 30 K max reaction, but one has a 80 K reaction; I'll show 30 K in the schedule and 80 K on plan at the one particular location).
For moment connections, our typicals call for full-pen welds, so there is no design to do there. Occasionally, we'll have a fabricator ask for design moments so that they can design them and not use full-pen welds.
We've even, on recent jobs, delegated braced frame connections. What we've had happen is we go through the design and detailing of the gussets, welds, bolts and then the fabricator comes back and says, "You know what, we prefer to make our braced frame connections with double angles field bolting to the gusset instead of the field welding you show". That's a LOT of time and effort wasted for no reason.
I'd be interested to know how the guys that design and document connections deal with those situations.
Hokie... I typically spec that connections be designed for the max udl that beam can accommodate, assuming that the top flange is laterally supported with the connection including for a moment that is equal to the distance from the centreline of the supporting beam/column to the centre of the fastener.
For unusual beams... ie, those with point loads or whatever, I generally spec the factored loads.
For connections with both a moment and a shear, I always spec the factored loads and the typical detail generally shows an end plate type of connection.
For unusual beams... ie, those with point loads or whatever, I generally spec the factored loads.
For connections with both a moment and a shear, I always spec the factored loads and the typical detail generally shows an end plate type of connection.
We show standard details that show the basics of what will be expecting for any non-detailed connections within our construction documents. Similar to DIK, we have a standard note to design each end connection for 55% of the beam capacity unless noted otherwise within the design drawings. There are also some additional notes to assist with minimum thicknesses, quantity of bolts, etc. Generally we will design any connection that is not a simple span beam. After all the AISC manuals have tables for the connection designers to pull from, just as we would do. However, if we have multiple beam depths we would then have many details that all show essentially the same thing. This would be an efficient use of resources, and generally more expensive to the client. As has been mentioned this allows the fabricator to work with the contractor as to what is preferred in the field, since a majority of the time contractor is purchasing the steel. When shop drawings are submitted, we will spot review the connection shown on the drawings. If "issues" are found we will increased the quantity of connections reviewed.
I am surprised at how many of you are still using %UDL for reactions. It is ridiculously easy to just put the reactions on the drawings if you are delegating the connections and that takes all guess work out of it. Also AISC does not recommend the %UDL practice at all and highly recommends showing reactions.
connectegr
Structural
Hokie
How to show the reactions for connection design? I have seen every combination mentioned and more. Actual forces shown on the plans, 50% UDL, even 150% UDL, tables related to beam depth, and in many cases no information at all. "fully develop the beam/brace", that is a good one. I have even received an EOR's RISA analysis with all the load cases. I don't think some engineers know what to design the connections for. What is a transfer force? What is column shear related to moment frames? What is the seismic resisting system? What is the response modification? Let the RFI's begin...
How to show the reactions for connection design? I have seen every combination mentioned and more. Actual forces shown on the plans, 50% UDL, even 150% UDL, tables related to beam depth, and in many cases no information at all. "fully develop the beam/brace", that is a good one. I have even received an EOR's RISA analysis with all the load cases. I don't think some engineers know what to design the connections for. What is a transfer force? What is column shear related to moment frames? What is the seismic resisting system? What is the response modification? Let the RFI's begin...
connect-
When brought into the project in the contract stages and you are working with another firm who is handling the structural design, what do you request in terms of how connection forces are given to you?
There can be many combinations for which a connection need be designed.
When brought into the project in the contract stages and you are working with another firm who is handling the structural design, what do you request in terms of how connection forces are given to you?
There can be many combinations for which a connection need be designed.
Willis... I agree that it's more correct to show reactions in a beam schedule... and the AISC want this to minimise connection costs... it's more convenient (and cheaper from a design point) to spec the UDL... It distresses me a tad, however, to see a W24x68 being attached with 2 bolts...
Dik
Dik
connectegr
Structural
When working directly with the EOR we typically work with envelope forces. If the reaction is an interaction of forces, moments, axial, torsion, and/or shear then we work with envelope of each. This is conservative since these many not occur simultaneously or in the same load case. If the resulting connection is unreasonable we may compare different combinations of actual interactions and determine the best connection. This is a lot of back and forth, but is the extreme case, and a few hours in engineering can save days of fabrication and material.
Most of our time is spent on relatively normal moment connections, vertical bracing, and seismic connections. Designing for real forces and discussion of member shapes and sizes can significantly decrease stiffeners and reinforcement. When working directly with the EOR, all the these options are on the table. We can compare results and find quick solutions. This interaction is rarely available when our contract is with the fabricator.
Most of our time is spent on relatively normal moment connections, vertical bracing, and seismic connections. Designing for real forces and discussion of member shapes and sizes can significantly decrease stiffeners and reinforcement. When working directly with the EOR, all the these options are on the table. We can compare results and find quick solutions. This interaction is rarely available when our contract is with the fabricator.
connectegr
Structural
yes...
Seismic is a special condition. Connections are design to exceed the plastic capacity of the member. But, there can be a large diffence in design for the actual "maximum force" or RyAgFy. I have only seen one project that provided the "maximum forces that can be developed by the system" to the fabricator for connection design. Usually the seismic inform shown on contract drawings is conflicting or confusing.
Seismic is a special condition. Connections are design to exceed the plastic capacity of the member. But, there can be a large diffence in design for the actual "maximum force" or RyAgFy. I have only seen one project that provided the "maximum forces that can be developed by the system" to the fabricator for connection design. Usually the seismic inform shown on contract drawings is conflicting or confusing.
WillisV- Where is it that AISC does not recommend the %UDL? Per 13th Ed (9th is similar) section J1.1 "...or shall be a proportion of the required strength of the connected members...". I take this to mean %USL is an acceptable method. Several years back I looked into revising our standards and determined it was not needed.
Recently we have bee doing some work for a connection only engineer (he hires us when he is overly busy) and the %UDL is used by him, unless noted otherwise on the design drawings.
Recently we have bee doing some work for a connection only engineer (he hires us when he is overly busy) and the %UDL is used by him, unless noted otherwise on the design drawings.
Toad - its for simple shear connections only.
SrVaro - AISC as an organization (not necessarily in the Manual) does not recommend it as it is poor practice that can result in ridiculously oversized connections (for instance in the case where say you are using a very short deep member to support another deep member, the UDL on the short deep member will most likely be way higher than necessary). On the other hand, if you had a point load from a beam framing in near the very end, using a % of UDL could be unconservative. Both issues resolved with putting the actual shear reactions on the drawings, which again is about 20 mins worth of effort in many cases.
The case for showing reactions:
From AISC Solutions Center:
See Myth 3:
Read this entire paper which gives real world examples of what can happen with UDL:
And just about every article ever published in Modern Steel etc. on steel economy mentions showing reactions on the drawings as a method of economy - just one recent example here:
SrVaro - AISC as an organization (not necessarily in the Manual) does not recommend it as it is poor practice that can result in ridiculously oversized connections (for instance in the case where say you are using a very short deep member to support another deep member, the UDL on the short deep member will most likely be way higher than necessary). On the other hand, if you had a point load from a beam framing in near the very end, using a % of UDL could be unconservative. Both issues resolved with putting the actual shear reactions on the drawings, which again is about 20 mins worth of effort in many cases.
The case for showing reactions:
From AISC Solutions Center:
See Myth 3:
Read this entire paper which gives real world examples of what can happen with UDL:
And just about every article ever published in Modern Steel etc. on steel economy mentions showing reactions on the drawings as a method of economy - just one recent example here:
- Status
Similar threads
