Aren’t We Done Testing Yet?

If you’ve been following the Structural Engineering Blog for any length of time, you’ve probably noticed that we like to run a lot of tests around here. Building test setups and breaking them is one of the things we enjoy the most in R&D at Simpson Strong-Tie, but often load rating a product is not as simple as just running a test. At times the process requires significant time and effort, so much so that we start asking ourselves, Aren’t we done testing yet?

We’ve posted on how we test connectors, holdowns, and screws. Most products have an ASTM standard or Acceptance Criteria that sort of tells you what to test. Yet figuring out how to test a product can be a challenge – in other words, how do we make sure our test simulates installed conditions? Or maybe a product can be used in so many different ways that it is unreasonable to test every possible installation, so what to do?

We have been challenged with these situations over the years, but probably never as much as with our connectors for curtain-wall construction that we introduced about two years ago. In particular, testing our bypass framing connectors to resist in-plane loads (designated as F1 loads) presented testing challenges.

For curtain-wall designs, in-plane loads from the curtain-wall system are often transferred to the supporting structure through slide-clip and fixed-clip connectors. But the capacity of these connectors to resist in-plane loads varies widely based on the details of installation, forcing designers to make assumptions about the location of the applied load and the amount of rotational restraint at each end of the clip.

Bypass connector used to resist in-plane loads
Bypass connector used to resist in-plane loads.

For seismic loads, the location of the center of mass of the system, and therefore the location of the applied force, will vary with different weights of cladding materials – resulting in different allowable loads for the same connector. For a heavy cladding material, the in-plane load will be located near the exterior flange of the stud, while for lightweight cladding it will be closer to the supporting structure.

Location of seismic force for heavy cladding system.
Location of seismic force for heavy cladding system.
Location of seismic force for lightweight cladding system.
Location of seismic force for lightweight cladding system.

Also, the presence of sheathing materials, blocking, and other installation details will affect the rotational restraint between the stud and connector. In the case of a flexible system with little or no restraint, the connector will bend as a cantilever. However, if the curtain wall system provides rotational restraint, the connector may bend in double curvature, resulting in a higher in-plane allowable load.

Cantilever bending for connector with no rotational restraint.
Cantilever bending for connector with no rotational restraint.
Double curvature bending for connector with rotational restraint.
Double curvature bending for connector with rotational restraint.

After some preliminary testing and extensive planning, we completed a little bit different test program to establish allowable loads for the multitude of possible design conditions. Instead of focusing on testing specific installations, we tested conditions to determine the capacity of critical failure modes – flexural yielding of the clip in bending, capacity of anchorage to the supporting structure, capacity of fasteners to the stud including prying, etc. Testing also simulated both cantilever bending and double curvature bending. Using this data, we were able to tabulate allowable loads for a wide range of installations (http://www.strongtie.com/ftp/letters/generic/L-CWC_F1LOADS.pdf)

Test setup for cantilever bending.
Test setup for cantilever bending.
Test setup for double curvature bending.
Test setup for double curvature bending.
Connector bent in double curvature.
Connector bent in double curvature.

Aren’t we done testing yet? Nope. Next we ran cyclic load tests to confirm ductile behavior (which the tests did confirm). These tests required a new setup that allowed us to both push down and pull up. And if that wasn’t enough, we have even dabbled with tests where we are attempting to define system behavior including restraint provided by sheathing.

Test setup for cyclic loading.
Test setup for cyclic loading.
Test setup for system test.
Test setup for system test.

As you can see, we are continually performing tests to better understand the behavior of our products. This allows us to offer unparalleled technical data for designers and building officials. In some cases, testing is never really “done yet.”

How important to you is product testing that represents a range of field conditions and design assumptions? Let us know by posting a comment.

– Paul

Print Friendly, PDF & Email
Paul McEntee

Author: Paul McEntee

A couple of years back we hosted a “Take your daughter or son to work day,” which was a great opportunity for our children to find out what their parents did. We had different activities for the kids to learn about careers and the importance of education in opening up career opportunities. People often ask me what I do for Simpson Strong-Tie and I sometimes laugh about how my son Ryan responded to a questionnaire he filled out that day:

Q.   What is your mom/dad's job?
A.   Goes and gets coffee and sits at his desk

Q.   What does your mom/dad actually do at work?
A.   Walks in the test lab and checks things

When I am not checking things in the lab or sitting at my desk drinking coffee, I manage Engineering Research and Development for Simpson Strong-Tie, focusing on new product development for connectors and lateral systems.

I graduated from the University of California at Berkeley and I am a licensed Civil and Structural Engineer in California. Prior to joining Simpson Strong-Tie, I worked for 10 years as a consulting structural engineer designing commercial, industrial, multi-family, mixed-use and retail projects. I was fortunate in those years to work at a great engineering firm that did a lot of everything. This allowed me to gain experience designing with wood, structural steel, concrete, concrete block and cold-formed steel as well as working on many seismic retrofits of historic unreinforced masonry buildings.

5 thoughts on “Aren’t We Done Testing Yet?”

  1. From my perspective as a construction engineer, there’s no such thing as too much testing. Knowing how things fail helps to improve them. Knowing how they behave under loading helps to improve the testing methodology. Improving the testing methodology leads to better and more realistic test results. The cycle continues. Perhaps at some point for some components we simply run out of new data obtained from the testing. Until someone discovers a slightly different application.

  2. Looks like the loads you are testing for are not what is important. The tie back load with lateral and vertical movement capacity is what is impotant. We do not want the panel to fall off. Keep on testing.

    1. We did the out-of-plane and vertical load testing and published those loads first, since those are the loads designers are most commonly addressing. The in-plane, F1, loads were the next priority and were challenging tests as far as determining proper boundary conditions in the lab.

      1. David K poses an interesting question. What is the
        tie-back (out-of-plane) capacity during (or subsequent to) the in-plane movement? That is to say, not the in the initial state, but for example after cycling.

        1. For combined in-plane plus out-of-plane loading, our experience with connectors has shown that a simple unity equation works very well although can be too conservative. For hurricane ties, we did an extensive test program at Clemson University to get a better (less conservative) solution for multiple simultaneous loads on a connector. A similar program with curtainwall clips may yield similar results.

Leave a Reply

Your email address will not be published. Required fields are marked *