The Post to Beam and Post to Pier Connections
This what it looks like under your house. The bottom of the posts are sitting on concrete piers and the tops of the posts hold up beams also known as girders. Floor joists rest on top of the girders and the floor you walk on is nailed to the joists. The building code makes no distinction between posts that are 2 feet tall or 10 feet tall.
As shown below, sometimes the posts support beams that the floor is nailed to and there are no joists.
Below are photos of some metal post connectors. Our first alert something might be wrong here is that this is not recommended by any seismic retrofit codes and guidelines.
What Does the Building Code Say?
Here are three examples of drawings given to contractors to show what the engineer wants done. In the drawing to the left, the red box contains the steel “T” strap post connector we showed you earlier. The blue box shows the post to block connection. The arrows point at the hardware and bolts connecting the post to the concrete lock. The image to the right from a different engineer shows the exact same thing.
The drawing at the bottom requires a new steel reinforced concrete foundation. It is buried in the ground and is 18″ wide and 18″ deep. It also has two types of steel connectors at the top of the post and one steel connector at the bottom. This contraption is so complex it would probably cost $1200 or more to build.
Mudsill Blocking and Retrofit Guidelines
Engineers often recommend wood blocks which connect the house to the foundation, as well as to plywood bracing. Retrofit guidelines limit the number of nails in these blocks to 4 to reduce the risk of splitting them, as shown below. A seismic retrofit engineer should be aware of the potential for splitting.
Many misinformed engineers and contractors recommend using this Nailed Blocking Method. In this example of a construction detail drawn by an engineer, TWELVE large nails are called for in each block, which will certainly split the blocks. An engineer who recommends nailed blocking is NOT the engineer you want to hire.
The Sub-Floor to Joist Connection
Here is another example of an expensive error often seen in engineer’s plans. The detail below tells the contractor to install a piece of steel that has been bent into a right angle called an L90. This is supposed to strengthen the connection between the sub-floor (which is another layer of flooring under the floor you walk on), and the floor joists (the supporting lumber the the sub-floor is nailed to).
The red arrows show where a connection of the sub-floor to a joist is to be made with the L90. The top of the L90 is nailed up into the sub-floor and the other leg is nailed into the side of the joist. Once the L90 is installed in this way the sub-floor to joist connection is complete.
Failures in the Sub-Floor to Joist Connection Never Occur
While houses can fall off of their foundations, it is never the sub-floor to joist connection that is the cause of this. There is not a single case in all the earthquakes in the United States of this connection ever failing. This is based on the author’s personal experience evaluating damage to homes after the 1989 Loma Prieta and subsequent earthquakes, as well as on my interviews with building inspectors and members of the Structural Engineers Assosication. The numerous photographs in the Steinbrugge Collection also confirm this.
Moreover, this connection is never reinforced for code compliant homes built directly on top of a known earthquake fault. Nor is this procedure found in any of the national or regional seismic retrofit codes and guidelines.
In contrast, look at another construction detail below. This detail is found in all six seismic retrofit guidelines. The blue arrow points to the fact that no steel sub-floor to joist connectionis shown at the joist to sub-floor location. This is because the members of all 6 committees, consisting of some of the finest structural engineers in the country, believed it is not important.
Compare this image to the one with the red arrows and you will see where the images are dissimilar in this single connection. The black arrows point to the bottom and necessary L90 Shear Transfer Tie recommended by all the retrofit guidelines.
Angle Iron Braces
This video discusses another ineffective and untested retrofit method that your contractor or engineer may recommend. It demonstrates why only tested retrofit hardware and methods should be used. Otherwise, what will happen is anyone’s guess.
The information in this video was created after consultations with numerous structural engineers, especially Bay Area-based Thor Matteson, and Josh Kardon Ph.D. Both of these structural engineers told me they would be happy to tell people about them if you want to call them.
Kelly Cobeen was kind enough to do actual calculations regarding their effectiveness. In conclusion she discovered that an Angle Iron Brace has the strength of a 1/4″ bolt. Numerous structural engineers were kind enough to answer questions I had about their efficacy. In addition, Buddy Showalter with the American Wood Council, the largest wood products research center in the world, was willing to give his expert opinion. All of these engineers are highly distinguished in field of wood-frame seismic retrofitting.
The Redundancy Factor
Let’s look at which connections fail in actual earthquakes. In earthquake engineering there is a factor known as the “redundancy factor.” Redundancy factors are those factors in a home’s original construction that protect it during an earthquake. These factors are unique to each home, so they can be are difficult to measure. Therefore engineers, contractors and building codes simply ignore them, and assume there is NO earthquake resistance in the original construction. The house in the photo below is an excellent example of redundancy factors at work.
According to the what the building code tells us about the strength of existing construction materials, the house should have collapsed. The sub-floor to joist connections should have also failed. Because of redundancy factors, neither of these things happened. Although the house fell off its foundation, the walls, roof, and even the windows held together and remained intact. Some engineers would probably recommend expensive reinforcement of the roof and walls, even after they did NOT fail in an actual earthquake.
Of course we need to look at the connections that did fail. First, the floor joist to cripple wall connection failed. This connection is strengthened using Shear Transfer Ties. Second, there was a weak cripple wall. This is strengthened with plywood. Thirdly, the mudsill to foundation connection failed – this connection is strengthened with foundation bolts.
Let’s Do the Numbers
Floors are only 3/4 inch thick. When doing engineering calculations with this connection the floor is called the main member. When we nail an L90 into the main member the L90 is referred to as the steel side member.
When anything is nailed into the 3/4″ floor, the penetration length of the nail can be no greater than the floor main member thickness. Once it exceeds 3/4 inch, the nail drives through the other side and does not penetrate anything.
The penetration length into solid wood determines the strength of the connection. In other words, when we drive nails through the top of the L90 into the 3/4″ inch thick floor main member the nail penetration can only be 3/4 inches because this is the thickness of the floor.
The American Wood Council Online Line Calculator
When we plug variables such as thickness of side member, type of wood, type and thickness of side member into the Online Calculator, we discover that we need a main member (floor) that is thicker than 3/4″ for the L90 to have any value at all. This is one more reason not to consider strengthening this connection.
The information in the red box which says “Try selecting a longer nail, or a thicker main member, or a thinner side member” tells us the problem. In this example, the 3 inch nail size in the chart is already far greater than the 3/4″ floorthrough the sub-floor so no need to go longer, there is no side member thinner than a steel L90, which leads to the conclusion that we need a thicker sub-floor main member which is fixed at 3/4″.