by Phillip Knodel, Lyndsay Cross, and Chris Rodes
Aggressive timelines are creating a scourge of incomplete construction drawings, leaving manufacturers and fabricators of joists and steel decking guessing at the designer’s intent and unclear about critical details. With compressed project timelines, the specification of critical details can get pushed downstream to the structural engineer, who pushes them to the fabricator. In reality, this method of saving time upfront by leaving out information in the drawings just transfers that work to a different stage of the project and can lead to mistakes, requests for information (RFIs), and cost overruns. The answer: upfront collaboration with all members of the team and a shifting of responsibility back to the appropriate team member.
Specifically with regard to the manufacture of steel joists, incomplete construction drawings and a lack of the specifics required by subcontractors can slow down a project. This article addresses some common areas of concern for steel decks and joists.
Avoiding the RFI
While often a necessary part of construction projects involving multiple teams, RFIs complicate and slow down project timelines. Pushing critical decisions to later in the project cycle to compress schedules can lead to more RFIs, which, ironically, slows the project down even more.
Poorly labeled sections, incorrectly labeled sections, or conflicting information on drawings is all fodder for RFIs. It is a good practice for the manufacturer or subcontractor to try and identify all problem areas at once, prior to sending construction drawings out for approval, so that all of the team’s concerns can be addressed simultaneously. If the drawings come back from approval and not all of the concerns were addressed, or the answers are unclear, then an RFI is sent. Frequent areas of concern for subcontractors with respect to steel roofs and decks are: net uplift value, rollover forces, fatigue design, diaphragm capacity, cantilevers, and welds.
On ballasted roofs, uplift is not generally a concern once the construction is complete. But with other roof types, when uplift is a concern, in the authors’ experience, it is often provided as a gross value right out of the American Society of Civil Engineers (ASCE) 7-10, Minimum Design Loads and Associated Criteria for Buildings and Other Structures code, instead of the net uplift required for the joist design. It is then up to the subcontractor to calculate net uplift using the actual roof dead loading, which may not be specified in the design loading.
If no uplift is indicated, it is up to subcontractors to determine whether uplift needs to be considered or not, and to do the calculation. It then needs to be confirmed with the engineer of record (EOR). This ends up being more work on the back end of the project.
Vague references in the design documents to rollover forces create similar problems for subcontractors. For roofs and decks, it needs to be clearly stated on the construction drawings if the joist seats need to be designed to handle the rollover forces. If it is not clear how the rollover forces are going to be resolved, it adds time to the process as the subcontractor has to determine the best method to address these forces. For example, if the design does not indicate channels, tube steel, or something similar in between the joists to pick up the diaphragm, then the joist seats must be designed to handle the rollover. This can become challenging, depending on the size of the forces.
Fatigue design and cranes
Another example of introducing uncertainty into a project because of a reluctance to make a decision early in the process is related to the use of cranes. This function is controlled by the American Institute of Steel Construction (AISC) Design Guide 7. When crane loading on the joists is specified, the joist design must account for the cyclic loading, otherwise known as fatigue. Joist specifiers need to know the crane’s classification, the type of crane, and how it is being controlled. These factors determine the impact and fatigue factors that are used in the joist design. The fatigue experienced by joists supporting a crane is similar to bending and unbending a paper clip repeatedly. Eventually the paper clip will snap. If the EOR does not have details about the crane at the time of design, it is advisable to provide the information as soon as it is available.
Another area in which drawings often lack critical specificity is transferring diaphragm capacity through deck attachments. The Steel Deck Institute (SDI) has standard practices regarding diaphragm transfers. SDI’s Code of Standard Practice 1.4.2 and its commentary explicitly state that deck attachment is to be specified by the specifying professional (often the EOR). That lack of information could easily lead to RFIs on the issue. Instead of trying to delegate the design of diaphragm transfers, it would be more efficient if the EOR or other professional would design it at the outset.
The joist manufacturer is the expert on the capabilities of joists and joist girders. Joists with cantilevers are situational and there can be a knowledge gap on the part of the architect and engineer. Joists can have a significant cantilever, but in most cases, if there is a long cantilever, it is not advisable to camber the joist. If a long cantilever joist is cambered, before it is even loaded, the curvature of the joist will cause the end of the cantilever to dip significantly, which is likely not what the EOR expects. In this case, the subcontractor would likely question the design and send out RFIs or approval notes. Earlier collaboration may avoid these types of errors.
Rooftop units and snow
Snow drifts are often specified around rooftop units (RTUs). When there is a square mechanical air conditioning unit on a roof, it is advisable for the EOR to look carefully at the specifications for snow drift requirements all around those units. In accordance with ASCE 7-10, Section 7.8, “If the side of a roof projection is less than 15 feet long, a drift load is not required to be applied to that side.” The authors have experience with one project in which the manufacturer was able to save the client approximately $70,000 while still remaining within the code, by carefully considering the details of the rooftop units and the need (or lack thereof) for drift load. It is possible that some design software may not account for exceptions regarding drift load.
Specifying welds is another area that often leads to unnecessary costs when early project team collaboration is skipped. Specifically, unnecessarily large weld sizes are often specified on chord toes or on chords attaching to columns or tie plates. Weld size can drive material thickness. For example, if a 9.5-mm (3⁄8-in.) fillet weld at 102 mm (4 in.) long is specified on the bottom chord of a joist with 8-mm (5⁄16-in.) thick chords, the chords would have to be bumped up to 11 mm (7⁄16 in.) thick to receive the weld. To save money, an EOR could specify a smaller weld that is longer: instead of 9.5 mm, a 6.3-mm (1⁄4-in.) weld could be specified. A 6.3-mm weld only has to be 50 percent longer to be the same strength and would use one-third less welding material. Having a conversation with the joist manufacturer about chord thickness and weld size early in the project can save time and money.
Revisions on drawings
The quality of construction drawings goes a long way in determining the overall outcome of a construction project both in terms of quality and project length. This includes the proper designation of revisions using clouding. When revision marks on drawings are not called out with clouding, finding revisions becomes a tedious process. As a result, a trade may miss something. Additionally, Section 3.5, “Revisions to the Design Documents and Specifications,” of the AISC Code of Standard Practice for Steel Buildings and Bridges 15th edition, states, “… all revisions, including revisions that are communicated through RFI responses to RFIs or the annotation of the approval documents … shall be clearly and individually indicated in the contract documents.” Hunting through 100 pages of detailed drawings to find one changed area requires a lot of unnecessary time and effort.
In the authors’ view, attempts to save time and money by compressing project timelines will often have the reverse effect. In the case of deck and roof design, leaving critical decisions until later often leads to additional questions and avoidable RFIs from manufacturers and subcontractors. Early collaboration can streamline the process.
One comment on “Common problems in steel deck and joist design”
We are investigating a noise complaint with a client that has a fully adhered fleeceback TPO roof installed in hot asphalt over 1/2″ gypsum cover board and 3 1/2 inch isocyanurate insulation mechanically attached to a type B metal deck. The noise sounds like someone in beating on the metal deck and varies in intensity as it travels throughout the building starting on the West side. The noise gets so bad sometimes they cannot hold meetings in their conference roof. The roof is only 2 1/2 years old. Weather conditions sometimes play a factor but the noise is reported even in the morning before the sun comes up. I’m looking for anyone who has experienced a similar problem and the resolution.