Anchor Testing for Light-Frame Construction

I started off doing a four-part series on how connectors, fasteners, concrete anchors and cold-formed steel products are tested and load rated. I realized that holdown testing and evaluation is quite a bit different than wood connector testing, so there was an additional post on holdowns. We have done several posts on concrete anchor testing (here and here), but I realize I never did a proper post about how we test and load rate concrete products per ICC-ES AC398 and AC399.

AC398 – Cast-in-place Cold-formed Steel Connectors in Concrete for Light-frame Construction and AC399 – Cast-in-place Proprietary Bolts in concrete for Light-frame Construction are two acceptance criteria related to cast-in-place concrete products.

Cold-formed steel connectors embedded in concrete are not considered in ACI 318 Appendix D, so it was necessary to create criteria for evaluating those types of connectors. Some examples of products covered by AC398 are the MASA mudsill anchor, CBSQ post base, and the STHD holdown.

MASA mudsill anchor

MASA mudsill anchor

CBSQ post base

CBSQ post base

STHD holdown

STHD holdown

ACI 318 Appendix D addresses the design of cast-in-place anchors. However, the design methodology is limited to several standard bolt types.

ACI 318 Appendix D Cast-in anchors

ACI 318 Appendix D Cast-in anchors

There are a number of anchor bolt products that have proprietary features that fall outside the scope of ACI 318, so AC399 fills in that gap by establishing test procedures to evaluate cast-in-place specialty anchors. Simpson Strong-Tie SB and SSTB anchor bolts are two families of anchors we have tested in accordance with AC399.

SB Anchor Bolt

SB Anchor Bolt

SSTB Anchor Bolt

SSTB Anchor Bolt

SB and SSTB anchors have a sweep geometry which increases the concrete cover at the anchored end of the bolt, allowing them to achieve higher loads with a 1¾” edge distance. The SSTB is anchored with a double bend, whereas the SB utilizes a plate washer and double nut.

AC398 (concrete connectors) and AC399 (proprietary bolts) are similar in their test and evaluation methodology. AC398 addressing both tension and shear loads, whereas AC399 is limited to tension loads. Testing requires a minimum of 5 test specimens. These are the allowable load equations for AC398 and AC399:

AC398 - Allowable Load Equation

AC398 – Allowable Load Equation

AC399 - Allowable Load Equation

AC399 – Allowable Load Equation

For comparison, here is the standard AC13 allowable load equation for joist hangers:

Allowable Load = Lowest Ultimate / 3

Without getting into Greek letter overload, what are these terms doing?

Nu (or Vu) is the average maximum tested load. Calculating averages is something I actually remember from statistics class. Everything else I have to look up when we do these calculations.

(1 – K x COV) uses K as a statistical one-sided tolerance factor used to establish the 5 percent fractile value with 90% confidence. This term is to ensure that 95% of the actual tested strengths will exceed the 5% fractile value with 90% confidence. COV is the coefficient of variation, which is a measure of how variable your test results are. For the same average ultimate load, a higher COV will result in a lower allowable load.

The K value is 3.4 for the minimum required 5 tests, and it reduces as you run more tests. As K decreases, the allowable load increases. In practice, we usually run 7 to 10 tests for each installation we are evaluating.

Table A2.1 – K values for evaluating the characteristic capacity at 90% confidence

Table A2.1 – K values for evaluating the characteristic capacity at 90% confidence

Rd is seismic reduction factor, 1.0 for seismic design category A or B, and 0.75 for all others. This is similar to what you would do in an Appendix D anchor calculation, where anchor capacities in higher seismic regions are reduced by 0.75.

Rs and Rc are reduction factors to account for the tested steel or concrete strength being higher than specified. There are some differences in how the two acceptance criteria apply these factors, which aren’t critical to this discussion. Φ is a strength reduction factor, which varies by failure mode and construction details. Brittle steel failure, ductile steel failure, concrete failure and the presence of supplemental reinforcement.

The α factor is used to convert LRFD values to ASD values. So α = 1.0 for LRFD and α = 1.4 for seismic and 1.6 for wind. Both criteria also allow you to calculate alpha based on a weighted average of your controlling load combinations. This has never made a lot of sense to me in practice. If you are going to work through the LRFD equations to get a different alpha value, you might as well do LRFD design.

MASA - Test Setup MASA - Test Failure

Rse is a reduction factor for cyclic loading, which is applied to proprietary anchor bolts covered under AC399, such as the SSTB or SB anchors. A comparison of static load and cyclic load is required for qualification in Seismic Design Category C through F. Unlike the cracked reduction factor, manufacturers cannot take a default reduction if they want recognition for high seismic.

Due to the differences in AC398 and AC399 products, the load tables are a little different. AC398 products end up with 4 different loads – wind cracked, wind uncracked, seismic cracked and seismic uncracked.

CBSQ Load Table

CBSQ Load Table

AC399 products are a little simpler, having just wind and seismic values to deal with.

SSTB Load Table

SSTB Load Table

What are your thoughts? Let us know in the comments below.

New Simpson Strong-Tie Anchor Designer software

Remember back to the days when you used allowable stress design for designing anchorage to concrete? Once you had your design loads, selecting an anchor was quick and easy. The 1997 UBC covered the anchorage to concrete in less than two pages, so the calculation was painless. Post-installed anchors were even easier, since allowable loads were tabulated and you just needed to apply a couple of edge distance and spacing reductions.

Since the introduction of strength design provisions and the adoption of ACI 318 Appendix D, first in the 2000 IBC, designing code-compliant anchorage to concrete has become much more complex. At least once (and probably not more) armed with a pencil,  calculator, and an eraser, most of us have set out to design a ‘simple’ anchorage to concrete connection using the  Appendix D provisions. Several pages of calculations later (and hopefully with a solution to the problem), most of us, I imagine, came to realize that designing anchorages to concrete by hand required much more time and effort than we anticipated or could allocate time for. As a result, many of us probably created an Excel template to speed up the design process using built-in functions and some Visual Basic programming.

I'm never going for looks on my spreadsheets.

I’m never going for looks on my spreadsheets.

The question is: are you still using the template?

For me, the answer is an emphatic “NO”, mainly because the spreadsheet I created has limited capability given the complexity in adapting the design methodology to complex anchor layouts, changes to the design provisions with each new code edition, and the need to add/modify data each time a new post-installed anchor product is introduced.

Continue reading