Accommodating Truss Movement (Besides Vertical Deflection)

Vertical deflection resulting from live and dead loads – of both roof and floor framing components – is an important serviceability consideration in the overall design of the building. And while this could be a blog topic in and of itself, this post is instead going to focus on two other types of truss movements that often prompt questions: seasonal up-and-down movement (of the trusses relative to the walls) and horizontal movement (of scissor trusses).

On the one hand, these are completely different topics. But on the other hand, they both deal with movement; which needs to be properly addressed when incorporating trusses into the overall building.  So it’s sensible to discuss them together in one blog post.

Seasonal Up-and-Down Movement

This type of movement goes by many different names that might sound familiar – truss arching, truss uplift, partition separation, or – to use the most formal name – ceiling-floor partition separation. All of these names describe the separation that develops between interior partition walls and ceiling finishes, which can cause gaps in the drywall to open in the winter and close in the summer. This movement is often considered to be a truss issue; however, it is not always the trusses that do the moving, but rather the walls or floors, or both, beneath the trusses.

This issue is also not limited to truss construction, but can also occur with other types of wood construction. The truss industry has information on this topic to help educate the market about the causes of ceiling-floor partition separation, best practices and construction techniques for minimizing the movement, and how to accommodate this movement in the structure to prevent drywall cracking.

trussmd1

For those who are interested in a very thorough and technical discussion of this issue and all of the factors that can contribute to it, there is a Technical Note available from the Truss Plate Institute (TPI) called Ceiling-Floor Partition Separation: What Is It and Why Is It Occurring? Although it was written several years ago (by the Small Homes Council-Building Research Council), the information remains relevant because the problem and its causes are the same now as they were then. The Technical Note discusses the potential causes of ceiling-floor partition separation, which may include one or more of the following: attic moisture (and the differential shrinkage and swelling of truss chords due to seasonal changes in moisture content), foundation settlement, expansive soils, excessive cumulative shrinkage of wood framing members and errors made during the construction process such as pulling the camber out of a truss to attach it to a partition. There is even an Appendix with a brief discussion of longitudinal shrinkage and an example calculation showing how much upward deflection results when a truss arches because of differential shrinkage.

For a condensed version, there is also a document available from the Structural Building Components Association (SBCA) called “Partition Separation Prevention and Solutions (How to Minimize Callbacks Due to Gypsum Cracking at the Wall/Ceiling Interface)”. This single-page document is particularly useful for educating the industry to take the appropriate preventive measures during construction, which help minimize problems later.

For example, the use of slotted roof truss clips – such as our STC (see below) – is one preventive measure, since these clips allow for vertical movement, but still provide lateral support at the top of the wall. DS drywall clips can be used in conjunction with the STC clips to secure the drywall to the wall. Then, to allow the drywall ceiling to “float,” the drywall is not fastened to the bottom chord within 16” from the wall. Taking these steps allows movement between the truss and the wall, without causing cracking in the drywall at the wall/ceiling interface.

trussmd2

It is important to note that, while foundation settlement may indicate a structural problem and can be prevented by proper design, truss arching resulting from the natural shrinking/swelling of wood does not indicate any structural problem and cannot be avoided in the design process.

Horizontal Movement of Scissor Trusses

In the typical design of a scissor truss, a pin-type bearing is used at one end, and a roller-type bearing is used at the other end, which results in some amount of horizontal deflection at the roller bearing.

trussmd3

trussmd4

The bearing assumptions used in the design of a scissor truss are important not only to the truss, but they also have design implications for the building as well. Using a pin-type bearing at both ends of the truss has undoubtedly been a temptation to every truss technician at one time or another, when the same scissor truss that is failing the analysis suddenly works as soon as the bearings are switched from pin-roller to pin-pin. Unfortunately, that isn’t a valid option unless the walls are infinitely stiff (which they typically aren’t), or unless special measures are taken to resist the horizontal thrust that develops at the pinned reactions. In most cases, such measures won’t be taken which means with the exception of some rare cases, scissor trusses must be designed with pin-roller bearings.

The horizontal deflection that results when a scissor truss is designed with a roller bearing on one end prompts further questions and discussion. What happens when a scissor truss is rigidly secured to the walls of the building – how does that horizontal movement happen? How much horizontal movement is too much? Should the scissor truss be attached to the wall with a sliding (roller-like) connection?

First, a scissor truss that is rigidly secured to both walls will still experience horizontal movement due to the flexibility of the building’s construction in most residential and light commercial construction. How much horizontal movement is too much for the building? This is definitely a question that the Building Designer needs to answer based on his/her evaluation of the overall structure. However, there are a couple of resources that can provide some insight.

ANSI/TPI 1 has the following provision:

trussmd5

Per ANSI/TPI 1, a scissor truss can have up to 1.25″ of total horizontal deflection in the absence of stricter limits from the Building Designer. Scissor trusses may even be designed with more than this amount of horizontal deflection, along with a warning that special provisions for lateral movement may be required. It is important for the Building Designer to be aware of the calculated horizontal movement of the scissor truss, as reported on the truss design drawing, to ensure that it is an acceptable amount of horizontal movement for the supporting structure and/or to determine whether special provisions for the lateral movement need to be made.

trussmd6

While 1.25″ of total horizontal deflection may seem like a lot of horizontal movement, these calculated horizontal deflections are considered to be conservative; many Designers agree that the predicted movement from the pin-roller bearing combination is greater than will actually occur in the constructed building. This is based on the fact that the design loads may be overstated and the contribution of the sheathing (and drywall if applicable) to resist the horizontal movement is not taken into account during the analysis of the truss.

The National Building Code of Canada (NBC) references Section 5.4.4 of the 2009 Engineering Guide for Wood Frame Construction, which limits lateral movement at the top of each wall to h/500. This correlates to a total allowable horizontal movement of 3/8″ for 8ˈ walls. However, the Canadian truss design standard (TPIC-2014) permits trusses to have a horizontal deflection (at the roller support) of up to 1″. In this case, since the horizontal deflection of the truss exceeds the allowable horizontal deflection of the wall, a sliding connection needs to be used between the truss and the wall.

trussmd7

There are different opinions on the use of sliding connections, such as the slotted TC24 or TC26 connectors (see below), which allow for horizontal movement of the trusses without pushing out the wall, and also provide uplift resistance. The use of these clips also varies greatly by region. There are many places where these clips are used regularly and successfully. However, some Designers prefer to restrict the truss horizontal deflection and require the use of a positive connection between the scissor truss and the wall plate due to concerns regarding the transfer of lateral loads from the top of wall to the roof diaphragm. When TC connectors are used, they are often used on alternating ends of the trusses so that there is a positive connection along each wall at every other truss. Some Designers feel this approach minimizes the horizontal movement between the truss and the wall after the building is constructed and fully sheathed and braced.

trussmd8

There is not a single correct answer to address horizontal truss movement for every building. The amount of horizontal movement that is acceptable for the structure and whether or not a sliding connection should be used will depend on the building, the loading conditions, the designer’s experience and/or judgment, and, in some cases, the local building jurisdiction. What is more important than the decision to either restrict horizontal deflection or utilize sliding connectors like the TC24/TC26 (both have been successful) is that the bearing assumptions used in the design of the scissor truss are accounted for in the design of the building. The worst-case scenario is when a scissor truss is designed with a pin-pin bearing and installed in a building where absolutely no measures have been taken to supply the needed resistance to the calculated horizontal thrust.

What are your thoughts or experiences with either seasonal up-and-down movement or horizontal movement?  Let us know in the comments below!

The New Truss Design Standard: Enter to Win A Copy of ANSI/TPI 1-2014 National Design Standard for Metal Plate Connected Wood Truss Construction

If you are like me, then you enjoy this time of the year. Instead of looking back and reviewing the events of the past year, this is the month for looking ahead at the year to come and what’s in store. So what is in store for 2015?

For the truss industry, there is a new truss design standard that was just released the last week of December. Still hot off the press, the ANSI/TPI 1-2014 standard is a revision to the 2007 edition and is referenced in the 2015 International Building Codes.

While the 2015 I-Codes might take some time for some municipalities to adopt, others are gearing up for adoption of the 2015 I-Codes as early as mid-2015. Either way, it is always good to know what is in the latest and greatest code-referenced design standards. So here’s a look at the new ANSI/TPI 1-2014 truss design standard:

The New ANSI/TPI 1-2014 Standard
The New ANSI/TPI 1-2014 Standard

First, here is a brief primer on the TPI 1 standard. The Truss Plate Institute (TPI) published the first truss design criteria in 1960. Many updates to these design criteria followed after that, and in 1995, TPI published its first ANSI-accredited truss design standard, ANSI/TPI 1-1995. Subsequent editions of this American National Standard have included ANSI/TPI 1-2002, ANSI/TPI 1-2007, and now ANSI/TPI 1-2014. All of the TPI standards, including archived copies going all the way back to TPI-60, are available from TPI (www.tpinst.org). Here is a link to the overview of non-editorial changes from ANSI/TPI 1-2007 to ANSI-TPI 1-2014.

While the 2007 edition included many significant revisions to the previous edition, the 2014 standard has relatively few substantive changes to the 2007 edition, which is good news for those who are still trying to catch up. Chapter 2 covers the design responsibilities involved in metal plate connected wood truss construction and looks different at first glance because it has been reorganized. However, the actual “Design Responsibilities” as they were defined in TPI 1-2007 have not changed.

In short, two separate sections in TPI 1-2007, which address design responsibilities in projects that require registered design professionals and projects that do not, have now been combined into one section. The “Truss Design Engineer” is simply referred to as the “Truss Designer” and the “Registered Design Professional for the Building” is simply the “Building Designer.” If the project requires registered design professionals, then the Truss Designer and Building Designer will be registered design professionals. Regardless of whether or not those two parties are registered design professionals, their responsibilities relating to the design and application of metal plate connected wood trusses are the same, so defining those responsibilities once within the TPI standard simplifies things and makes more sense.

Not new to the wood industry, but new to TPI 1-2014, are provisions for Load and Resistance Factor Design (LRFD). AF&PA incorporated LRFD provisions into the 2005 National Design Specification (NDS) for Wood Construction, and the TPI standard has followed suit, using the same basic approach as the NDS.

The section in TPI 1-2014 with the most changes is the section on deflection criteria. The deflection criteria have been revised in the last three editions of the TPI standard. Starting in TPI 1-2002, a requirement was added to consider creep in total deflection calculations. However, specific creep factors were not specified in the standard and were only presented in the Commentary. In the 2007 edition, creep factors were moved into the standard, and the total deflection calculation explicitly specified a component due to creep of no less than 50 or 100 percent of the initial deflection for long-term loads for dry and green (wet service) use, respectively. This was consistent with the 1.5 and 2.0 creep factors specified in the NDS for total deflection calculations for seasoned and unseasoned conditions.

Between the 2007 and 2014 editions, an inconsistency was discovered between the TPI 1 deflection criteria and the deflection limits in the U.S. model building codes. While the intent of the TPI standard was to present the same basic L/xxx deflection limits for Live Load and Total Load as the model building codes, it was discovered that the IBC deflection limits for “DL + LL” were actually intended to address only the creep portion of the dead load deflection plus the immediate live load deflection. So although long-term deflection including proper creep considerations can be an important consideration in the overall design of the building, it is not intended to be used to limit the design of a truss with respect to building-code established limits on vertical deflection.

Excerpt from the ANSI/TPI 1-2014 Commentary
Excerpt from the ANSI/TPI 1-2014 Commentary

To resolve the issue of inconsistent methods used in the building industry to specify deflection limits, the 2014 edition now distinguishes between the following:

• “Deflection due to Live Load Plus Creep Component of Deflection due to Dead Load” for purposes of meeting the IBC deflection limits for DD + LL, which is defined as

ΔCR = Δ LL + (Kcr ‐1) x Δ DL

• “Long-Term Deflection”, which includes the full effect of creep but for which there are no explicit deflection limits specified in TPI

• “Deflection due to Total Load”, which is based on the full load (including both dead load and live load), but includes no explicit creep factors. The deflection due to total load has the same deflection limits as the IBC deflection limits for DD + LL, but this is not a mandatory check in TPI; it only applies to trusses if the Building Designer specifies that such a check due to total load be performed. Further, any consideration for creep in that calculation would also have to be specified by the Building Designer.

In recognition of the increased creep in trusses compared to solid sawn beams, the creep factors have been increased to 2.0 and 3.0 for dry and green (wet service) use, respectively. For purposes of deflection checks in accordance with the IBC, these factors reduce to 1.0 and 2.0, respectively, since the equation for “Deflection due to Live Load Plus Creep Component of Deflection due to Dead Load” uses KCR-1 rather than KCR as the factor on the immediate deflection due to dead load.

What does this all mean? For the majority of truss applications (e.g., dry-service), the effect of switching from TPI 1-2007 to TPI 1-2014 will be a change in creep factor from 1.5 to 1.0, unless additional requirements are specified by the Building Designer. Those additional requirements may include a limit on long-term deflection or a check for total load deflection (subject to the TPI deflection limits), including any considerations for creep.

A complete listing of the changes in TPI 1-2014 and more discussion about these changes are available in the TPI 1-2014 Commentary.

Now is your chance to win a copy of the ANSI/TPI 1-2014 standard for your own design library! Simply post a truss-related question, comment or idea for a future truss-related blog topic, and we will enter you into a drawing during the week of Jan 15-22. One winner will be picked at random. We look forward to hearing from you!