FTAO Shearwall Design (Wood, Residential/Light Commercial) 1

[medeek]

I am wondering how often you find yourself needing to specify force transfer around opening type of shearwall vs. the more standard segmented shearwall.

I am looking for a spreadsheet or program than anyone has found useful in the rational analysis using the Diekmann technique. Based on APA findings published in m410 (APA website, free download) this method seems to be the best method currently available for estimating detailing requirements for bracing around the openings.

If I cannot find a suitable program/spreadsheet I will probably write one and make it available to all who are interested.

 

[JoshPlumSE]

My company (RISA Technologies) put a decent amount of effort into FTAO for wood shear walls. As part of the process, we compared our FEM based results to many of the hand calc examples out there.

Honestly, none of them were very good at capturing the actual behavior of the piers. Typical assumptions made in the hand calculations just weren't valid. This includes examples done via the Diekmann method. In particular, I'm talking about the assumption that an inflection point exists at the mid-height (or mid-length) of the opening. I believe this is done in most of the methods (Diekmann method as well). But, there are other bad assumptions as well.

It understandable that these hand calcs make these assumptions. Because without them, there would be no way to do these walls by hand. However, when comparing the FEM results of a FTAO shear wall to the hand calc methods, it is not difficult to see how these methods fall short. I've always meant to put together a "white paper" on the subject. But, I never seem to find the free time.

Regardless, if you're talking about doing a single shear wall, I would recommend using an FEM based approach. That's not saying that you have to use RISA (though I would certainly wouldn't discourage it). It shouldn't be difficult to put together a well meshed plate element model with openings and such in any FEM based program. That way you can see how the shear really wants to transfer through the sheathing and what shear stresses really develop.

 

[Archie264]

Josh,

I'm not familiar with the Diekmann method but I've assumed that he used an approach similar to the portal frame method.

If so, with the portal frame method the inflection point does not need to be taken at mid-height; it can (and should) be located where it is deemed most likely to occur as per "engineering judgement." It seems to me that for a typical steel or concrete moment frame that mid-height would be reasonably accurate but I would not have faith that that would be the case with a pinned-base wood moment frame, which, I think, an opening in a wood shear wall might mimic. In that case I'd locate it much lower, even perhaps, at its base. Might that be the cause of the discrepancy?

 

[JoshPlumSE]

That's the one that jumps out the most when comparing the FEM results to the hand calc methods. But, that's just because it's easy to test. I just take a look at the FEM results to see where the real point of inflection is. If I use that location in the hand-calcs they would likely match up much more closely.

My belief is that this inflection point location is not the only problem. Though the other problems may be less significant. Like I said, I've never taken that next step towards writing a white paper for our users on the subject. If I ever get around to doing that, I'd probably have more detailed comments on the subject.

One of the main points of frustration (at least for me) with the FTAO method is that the various hand calc methods can arrive at such dramatically different results. Obviously, they can't all be correct. I imagine this frustration is worse for the average practicing engineer who has less time for such esoteric pursuits.

That's one of the reasons why I point people toward the FEM analysis. There is no question that the FEM analysis is going to fully obey statics and is going to provide internally consistent results in all cases. It just becomes a matter of properly interpreting those FEM results.

 

[dcarr82775]

No offense Josh, but it is a wood shearwall built by a framer of questionable skill, not a precision timepiece. Once our profession starts thinking FEM is the way to go on a wood shearwall I am gonna start looking out the window for four guys on horses. Wood, sheathing, nails, straps, joints in pieces..... nightmare for FEM

 

[jayrod12]

The guy's abilities on site have no bearing on the accuracy of the analysis. What Josh is trying to say (at least the way I've understood it) is the simplifying assumptions that we make when designing the wall lead to improper analysis. Using FEM to analyze your shear wall should not change the construction details.

In his specific example the inflection point of the shear wall around the opening is generally incorrectly located when hand calcs are being performed which leads to incorrect (not wholly wrong) analysis. With more accurate locations (as with FEM or another suitable method) the hand calcs might match the FEM results.

What Josh didn't mention is whether the hand calcs resulted in conservative results or whether they were allowing for undersized members. If the former then it's less of an issue, if the latter then I would be concerned.

 

[dcarr82775]

Maybe I think different things when I hear FEM. All sorts of slip between materials and plys that would really need to be accounted for if going down the FEM road. Correctly modeling where your sheathing joints are, straps, nail slip, etc. is crucial if you are going to say it is a better answer. If FEM really just means a 2D frame analysis that is a different story, but not necessarily correct either.

I believe that in this profession it is not only prudent but necessary to keep in mind that no matter how fancy and detailed the analysis, it still had to be built be the common worker. The accuracy of the analysis must account for that. Just too many things going on at the corner of a window opening.

My apologies for wandering off topic.

 

[RFreund]

It has been a while since I looked into this but if I remember correctly the simplified methods for FTAO become difficult when you start introducing more than one opening. I as well, wanted to create a spreadsheet for this application, but again it became difficult to handle multiple openings and so I scraped the idea.

Whether practical or not, a white paper on wood shear walls and or diaphragms with openings and how to properly model and interpret the results would be awesome. Keep me/us posted. Do you know of any literature that discus this already? I'm curious on how you set up the model to consider the parameters mentioned by dcarr.

EIT

http://www.HowToEngineer.com

 

[Archie264]

Dcarr,

That’s an interesting topic unto itself, to be sure. In going through various sample problem sets that sometimes accompany new code releases (or are available for purchase separately) I’m often struck by how often the problems’ solutions rely on the results of a computer output. Mind you, that’s not computer output that is given as part of the problem but, rather, one they expect you to generate as part of solving the problem. In other words the problem can’t be solved with a pencil and calculator.

If that’s the state of our profession then I think it’s a shame and a hazard, even. If a computer must be relied upon to solve part of the problem then why not simply use it to solve all of the problem? Why involve an engineer at all? Simply write software that takes user input and spits out a code-compliant output, with little understanding of the structure’s response.

Engineering software is a wonderful tool, one which I use constantly. I’m just concerned that in some cases it’s used as more than just a tool.

Or maybe I’m a tool. Apologies for taking it further off topic.

 

[medeek]

My first exposure to FTAO was with Design of Wood Structures by D. Breyer and Associates. The best white paper I have found out there that compared some of the methods of rational analysis of this type of shear wall compared with actual experimental data is here:

APA / USDA Joint Research Report on FTAO

Their conclusion was that all of the hand calc methods were basically flawed but the Deikman method seemed to give the best answer out of the bunch.

I would like to setup a nice little spreadsheet or program that can handle and FTAO shearwall with one or two openings (windows).

I will post this to the board once I have something presentable and let everyone way in on its usability and accuracy.

 

[JoshPlumSE]

No offense taken.... In general, wood + FEM is a challenge. So, I understand the general reluctance to use it with wood. However, in this case, I'm talking specifically about the FTAO procedures and really only thinking of a single wall.

The comparisons that I have done have been mostly for a single opening wall. The various hand calc methods can give wildly different results from each other. See the WoodWorks presentation on the subject for example.

What hand-calc method is best, I can't say conclusively. I'd think the two most widely used examples seemed to be the one from Breyer's 6th edition (which was removed from future editions of the book) and the one from the SEAoC Seismic Design Manual - volume 2. And, those methods seem reasonably accurate and mostly rational. Though I remember that both methods have some internal inconsistencies in their assumptions. I don't have my detailed notes available to me right now, but the chicken scratches I've written down suggest something like a 10 to 30% difference from the FEM results. Some panels show higher shear and some lower shear.

My real issue is that those two procedures are not easy or straight forward. They actually get very cumbersome even for simple walls with a single opening. When the geometry gets more complicated (i.e. multiple openings, asymmetry in the opening placement, et cetera), I can't imagine how painful they would become. You have to make so many (possibly questionable) assumptions just to get results that I tend to lose faith in the final product. Sure, you can run the calc multiple times with multiple different assumptions to try to "envelope" the results. Maybe that's what others are doing.

I tend to think that a simple FEM plate analysis of the sheathing will shed some light on the expected behavior of the wall more easily than fumbling through multiple hand calc assumptions. Of course, that's coming from someone who is an expert in interpreting FEM results. At the very least, I would think the FEM analysis would shed light on which hand calc assumptions are reasonable and which are not.

There are certainly some pitfalls with the FEM results (localized stress risers for example) that could be problematic for many engineers. At RISA, we've really tried to address that with the way we post process of our FTAO wall results. But, I didn't really want to get into the details here because then I come off as a sales / marketing guy who mostly wants to sell software. I prefer to come across as fellow engineer who has spent lots of time with FTAO walls and has something useful to contribute to the forum....

 

[Michel60]

I've regularly designed shear walls with openings using a modification of the Deikman approach as presented by Malone and Rice in "The Analysis of Irregular Shaped Structures". I like their method in that it allows including dead loads in the analysis and the procedure is simple to do in a spreadsheet format just involving the solving of a series of static free-body diagram equations. Of those using an FTAO hand calc approaches, how have you addressed the following:
1) how to handle shallow spandrels depths above openings (aspect ratio exceeds 3 1/2:1)? - I've defaulted to the cantilever beam approach and just accept the conservative results for the strap forces
2) Do you include dead loads in the panel shear forces? - As mentioned I like to include them but it does drive up the panel shears
3) Consideration of multiple openings? - It is really a beast and I still haven't completely debugged a spreadsheet yet that can consider them.

I've also played with FEM a bit but haven't gone so far as presenting it as the design basis for plan review. For the FEM contingent (Josh):
1) How have you resolved forces for the tie straps? Use the nodal force at the corner of the opening?
2) Do you include any boundary elements in the FE models? At ends/opening?
3) What sort of integration procedures do you use for the panel shear determination?

 

[JoshPlumSE]

Michel-

1) We use plate corner force to determine the total moment and axial force in the block above and below the openings. It is then assumed that this force must be resisted by a tension compression force couple. That couple force is taken as the required tie or strap force.
2) Not sure what you mean by boundary elements in this context. Do you mean modeling of studs, or chords or such? If so, then there is no direct modeling of the studs or chords or such. That was deemed to sophisticated for what we were trying to do. We do use an orthotropic plate element so that the vertical stiffness of the wall is independently manipulated to reflect the stiffness of the studs. But, we don't explicitly model in vertical king studs or posts / chords.
3) The program uses an area weighted average of the plate Fxy forces to come up with an "average" shear force for each block / panel.

 

[rn14]

meedeek, I was wondering if you ever made progress with this? Also, any chance you could point me in the direction of some good examples? Thanks.

 

[medeek]

I haven't given up on this I just have not found the time to put it all together yet. I'm hoping to actually make a nice little calculator for FTAO shear walls similar to the one I'm currently working on for perforated shear walls found here:

http://design.medeek.com/resources/shearwall/perfshearwall_calculator.pl

I was initially just going to put it into an Excel spreadsheet however the power of procedural coding is very attractive and offers far more capability IMHO.



A confused student is a good student.

 

[woodman88]

Yes, but have any full scale tests been done to confirm if any of these shearwalls work as design by FTAO or FEM methods?

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[medeek]

This paper I posted above involved full scale tests that showed the Diekman method was reasonably accurate.

http://www.apawood.org/SearchResults.aspx?tid=1&q=M410

A confused student is a good student.

 

[woodman88]

The M410 is a general and limited testing for research data. On page 142 it states "Table 3... ... The Drag Strut method can both under predict and over predict the maximum FTAO. The cantilevered Beam, Coupled Beam, and Diekmann's methods on the other hand seem to be very conservative..."

I would be very careful about analyzing shearwalls with other than the standard methods typically being used now.

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[rn14]

Medeek, that is a pretty clever way of doing your calculators. I guess the only down side is that all inputs have to be re-entered when changes are needed. Most of my spreadsheets rely heavily on VBA which I never liked much. Recently I have started playing with having excel use functions I have written in Python, which I much prefer to VBA. Anyway, what references were you planing on using for the design method?

 

[manstrom]

Note the the M410 paper relates FTAO to actual values, but some predicted values are 200% of actual.

Still, Diekmann is the best method thus far. I have been in contact with the APA regarding their research and have hopefully given them some ideas on how these are used and what research would be most helpful.

In calculating FTAO forces, I try to limit my analysis to a single opening (multiple is not realistic). However, on my drawings, I will often use a series of openings. The actual calcs are approximate at best.

When I am working on a problem, I never think about beauty but when I have finished, if the solution is not beautiful, I know it is wrong.

-R. Buckminster Fuller