“Structures are connections held together by members” (Hardy Cross)
Take a quick look at a truss design drawing, and you will see a reaction summary that specifies the downward reaction, uplift and a horizontal reaction (if applicable) at each bearing location. Some people are tempted to look only at the uplift reaction, go to a catalog or web app, and find the lowest-cost hurricane tie with a capacity that meets or barely exceeds the uplift reaction.
However, if uplift was the only loading that needed to be resisted by a hurricane tie, why would we publish all those F1 and F2 allowable loads in our catalog?
Of course, many of you know that those F1 and F2 allowable loads are used to resist the lateral loads acting on the end and side walls of the building, which are in addition to the uplift forces. Therefore, it is not adequate to select a hurricane tie based on uplift reactions alone.
Where does one get the lateral loads parallel and perpendicular to the plate which must be resisted by the truss-to-wall connection? Definitely not from the truss design drawing! Unless otherwise noted, the horizontal reaction on a truss design should not be confused with a lateral reaction due to the wind acting on the walls – it is simply a horizontal reaction due to the wind load (or a drag load) being applied to the truss profile. It is also important to note that any truss-to-wall connection specified on a truss design drawing was most likely selected based on the uplift reaction alone. There may even be a note that says the connection is for “uplift only” and does not consider lateral loads. In this case, unless additional consideration is made for the lateral loads, the use of that connector alone would be inadequate.
Say, for example, that the uplift and lateral/shear load requirements for a truss-to-wall connection are as follows: Uplift = 795 lb.; Shear (parallel-to-wall) = 185 lb. (F1); Lateral (perp-to-wall) = 135 lb. (F2).
Based on those demand loads, will an H10A work?
An initial look at the H10A’s allowable loads suggests it might be adequate. However, when these loads are entered into the Connector-Selector, no H10A solution is found.
Why? Because Connector-Selector is evaluating the connector for simultaneous loading in more than one direction using a traditional linear interaction equation approach as specified in our catalog.
If the shear and lateral forces were to be resisted by another means, such that the H10A only had to resist the 795 lb. of uplift, then it would be an adequate connector for the job. For example, the F1 load might be resisted with blocking and RBC clips, and the F2 loads might be resisted with toe-nails that are used to attach the truss to the wall prior to the installation of the H10A connectors. However, if all three loads need to be resisted by the same connector, then the H10A is not adequate according to the linear interaction equation.
Some might question how valid this method of evaluation is – Is it necessary? Is it adequate? How do we know? And that is where the interesting information comes in. Several years ago, Simpson Strong-Tie partnered with Clemson University on an experimental study with the following primary objectives:
- To verify the perceived notion that the capacity of the connector is reduced when loaded in more than one direction and that the linear interaction equation is conservative in acknowledging this combined load effect.
- To propose an alternative, more efficient method if possible.
Three types of metal connectors were selected for this study – the H2.5A, H10, and the META20 strap – based on their different characteristics and ability to represent general classes of connectors. The connectors were subjected to uni-axial, bi-axial and tri-axial loads and the normalized capacities of the connectors were plotted along with different interaction/design surfaces.
These interaction plots were used to visualize and parameterize the combined load effect on the capacity of the connectors. The three different interaction plots that were examined were the traditional linear relationship, a quadratic interaction surface and a cuboid design space.
The results? Not only was the use of the linear interaction equation justified by this study, but a new, more efficient cuboid design surface was also identified. It provides twice the usable design space of the surface currently used for tri-axial loading and still provides for a safe design (and for the bi-axial case, it is even more conservative than the linear equation). This alternative method is given in our catalog as follows: As an alternative, certain roof to wall connectors (embedded truss anchors, seismic and hurricane ties, and twist straps) can be evaluated using the following: The design load in each direction shall not exceed the published allowable load in that direction multiplied by 0.75.
Now we can go back to the H10A and re-evaluate it using this alternative method: Multiple each allowable load by 0.75
- Uplift = 1140 x 0.75 = 855 lbs. (>795)
- Lateral = 590 x 0.75 = 433 lbs. (>185)
- Shear = 285 x 0.75 = 214 lbs. (>135)
As it turns out, the H10A does have adequate capacity to resist the simultaneous uplift, shear and lateral loads in this example. This just goes to show that the alternative method is definitely worth utilizing, whenever possible, especially when a connector fails the linear equation.
This is an abbreviated version of a longer article by Kelly Sias that was published April 7, 2016 on the Simpson Strong-Tie Structural Engineering Blog. To read the full version, please visit the Structural Engineering Blog.