Shear Wall Hold Down Distance Apart

[CurlyQue]

When designing a wood shear wall with gypsum board or wood diaphragms, is there a maximum distance the hold-downs can be from each other? If a shear wall is 20' long versus 100' long, the uplift is more for the 20' than the 100' so longer shear walls give you better leverage to prevent overturning. However, 100' long walls seem more prone to buckling than 20'.

Thans

 

[msquared48]

Yes, but there is probably a lot more dead load to resist the overturning too.

Depending on the number of stories, I would be surprised if a Holddown was needed for a 100 foot wall, considering the dead load resisting moment.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[CurlyQue]

Thank you msquared48,

For the sake of argument, let's say that the lateral force is greater than or double the dead load resisting moment. Is there a maximum distance between hold-downs? I understand that a shear wall has a height:width ratio. I don't find a length:height ratio. At what point does a shear wall quit acting as a unit and buckle? I understand the mechanical advantage of having a longer shear wall but how long can a shear wall be?

If the wall is viewed as a beam; at some point would there not be buckling in the wall or crippling in the web? I know some perpendicular walls would help prevent this. What if there are no perpendicular walls? Mathematically, the wall could be as long as required but at some point the wall has to buckle if loaded, correct? If so, what controls this? I'm looking at the wall as being like a web in a beam.

If I also look at the wall in plan and consider it like a column with the lateral load being a vertical load, then l/r controls. How does the theory of a shear wall control this buckling?

The inverse is understandable, the closer and closer I put the hold-downs, the less chance there is for buckling. The 'height' of the column becomes less and less and the l/r becomes less and less. This configuration will have higher overturning moments, yes, but the forces in the hold down and the end posts are purely tensile (compressive on the other end). A 2x6 shear wall is going to have more stoutness than a 2x4 wall. What prevents this buckling is lateral support at the floors and the resistance to bending by the studs but, for a high load condition, the studs would have to resist both bending and compression from the dead load at least for part of the wall.

My engineer does not believe in long, 'flimsy', shear walls which supposedly buckle like paper. Consequently, he designed OSB wall modules 4' long with two hold-down, all-thread rods per panel, virtually doubled every 4', going the full height of a 3-story building. This is causing much grief with my client. He also does not believe in gypsum board shear walls which, in mid-Indiana, are used quite often. I understand his reasoning and I agree that OSB is a better material but gypsum board is part of the cost of a building. OSB is extra, $160,000 extra! If I have enough gypsum board walls to handle the shear; then why use OSB if the client doesn't want to pay 'extra' because they 'always use gypsum'?

Design-build, you gotta love it ... or starve.

My question still stands; what controls the distance to hold-downs if anything?

Thank you

 

[woodman88]

I would have to take a close look at any shearwall with a stacked length:height ratio of 2.5/1 that had a uplift over 300#.
As for a single story shearwall 100' long, it is being braced by the diaphragm/foundation top and bottom so buckling due to shear is not a problem IMHO. Hopefully the diaphragm(s) is rigid enough or there are perpendicular shearwalls to maintain a vertical wall for the lateral loads perpendicular to the shearwall.
As for gypsum shearwalls the biggest problem with them IMHO is that in a code event they may crack. But the owner/contractor repairing the cracks will not know which walls are shear/braced walls (unless they review the plans). So they would not know which walls they can putty and tape vs the ones they need to replace the gypsum boards on. This is especially true of residential homes.

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.

 

[CurlyQue]

Thank you woodman88,

Your response focuses more on the problem as I see it.

I'm thinking now that buckling due to uplift is not an issue because, theoretically, the uplift puts the whole wall in tension, correct?

Buckling due to shear or lateral load seems to be more of the issue. I agree that the diaphragm(s) provide lateral support at the top and bottom. Can you imagine any situation in which buckling of the shear wall would occur in shear; perhaps a 3.5:1 OSB wall? Perhaps this is why shear wall diaphragms pull the nails out or the diaphragm pulls out around the nails in test failures. If the horizontal diaphragm(s) buckle because of shear then they weren't designed properly.

Have you heard or seen or been involved in a design where perpendicular walls are used to laterally brace a shear wall? What controls this; when are lateral walls required? To me, that is the equivalent of a web of a beam needing web stiffeners because of high load or point loads. Am I incorrect in viewing the wall as like a beam?

I view a building as nothing more than a cantilever beam divided into sections resisted by shear walls. In some sense, the mechanism is either a box beam or series of 'I' beams in plan with flanges loaded and the web being a shear wall. Doing the math that way is roughly equivalent to the floor+floor method.

Also, my engineer says dead load cannot be considered in resisting wind load. Yet, I found a calculator that considers the outside perimeter wall weight and foundation for helping to resist or completely negate overturning. That method seems reasonable but it also seems reasonable that the depth of the building and the associated mechanical leverage would go a long way too. I've also used other engineers who include the dead weight to offset the overturning and resolve the remaining uplift into a proprietary T-Wall connection that transfers that load back into the foundation using 2x and lots of 16d nails and avoiding 'costly' Simpson hold-downs. Yet others prefer the hold-down because it's much easier to inspect than counting hundreds of nails. Which or who is correct? Perhaps all since we're dealing with forces, code, and the fact that all engineers are different in how they resolve loads. One shoe does not fit all.

Thank you

 

[JAE]

I can't even imagine how a 100 ft. long shear wall would even need hold downs.

 

[woodman88]

CurlyQue
First, I think you are focusing on one element of the building (the shearwalls) without considering the overall structure. I do not see the shearwalls as beams, rather I see the structure as a cantilevered column.
Second, so far most of my long shearwalls have had perpendicular shearwalls attached to them. No I did not analysis where and when they were needed. Since they were there I new I could justify the long shearwall due to them if asked.
Third, whether you can use dead load to resist wind depends on which building code and method you are looking. As you have not given the code/method I can not answer this.

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.

 

[CurlyQue]

I can imagine a vehicle going to Mars. The question is how to get it there.

If a shear wall is 100' long, 100' tall, and has 100 mph load on a 100' wide wall contributing to it, will it overturn? I imagine so; I don't have to do the math. You might want to resolve it into smaller components but the point is still the same.

My point is at what length and or load combination will cause a shear wall to buckle because it's too long. Maybe this situation will never happen. Maybe shear walls just don't buckle because the overturning moment causes them to all be in tension. Could the shear of such a wall cause it to buckle? At some point it seems you have to consider the thickness and ability not to buckle.

My engineer doesn't want to use long shear walls because he says they will buckle like a piece of paper. Gypsum board is not paper. OSB is better than gypsum board. Multiple layers on two faces or configured for 3 diaphragms of OSB is better than 1. Masonry is better than OSB. Concrete is better than masonry. The thickness and the shear capacity of the material and configuration have some contributing factor.

The Fv of a steel beam is better than wood. The Sx and I values are better for a beam with a thicker web than for a thinner one. Yes, most of the work is done at the flanges but that is where the analogy breaks down somewhat. The question I have regards the web, the thickness or stoutness of the material in the web or its lateral stiffness. It seems that the depth to thickness ratio is important. Or, is a piece of super-tough, carbon-fiber paper simply good enough. It will never buckle when glued to toothpicks at 0.1" oc? Thickness has to be a consideration, does it not? Maybe I'm missing something (or a lot of things).

Diagonal bracing is steel frame. Although I can't understand why diagonalized metal straps can't be used to improve shear or get rid of it altogether. That's what metal stud manufacturers and steel frames do. Why not wood? Shear gets resolved at the base and is resisted by anchors or fasteners. Diagonalizing it all puts all the forces in compression or tension; there is no shear except in the diaphragm except in the deck which has a certain value of shear for members spaced a certain distance. Wood panels have this requirement too, 16" or 24" center spacing minimum. If this is true for plywood, then isn't it also true for a 2x4 wall versus a 2x6 wall or a 2x8 wall.

What is the advantage of having a 4', 'panelized', modular system with all-thread tie rods at each corner going completely up through the building. Sure, it's much more 'stout' and eliminates dead-men at the hold downs but it's tremendously more expensive so I fail to see the economic advantage when enough gypsum board walls will do the same thing at virtually no cost especially for a 3-story building ... for god's sake.

What's the problem? Code allows gyp but one structural engineer won't and so the building cost goes up $160,000? For what reason? So he can sleep better at night. Why not fire the engineer and spend another $16 or $16,000 for a one who'll use gypsum if the owner only cares about the cost. The point, I thought, was to optimize; sometimes gilded buildings are better; most times they're not.

I understand the value of having OSB. If cracks appear in the gypsum board, then someone will stop to think about this wall being a shear wall, an important wall. Red paint effectively says the same thing or red nails or a big sign: "Shear Wall! Do not touch without consulting an engineer!".

The value-engineers and their supporting civil engineers are having a hay day with this item; like sharks in chum-filled water.

Thank you,

Thanks

 

[CurlyQue]

Woodman88,

I too see a building as a cantilever structure. But, in plan, when the shear walls and their contributing load walls are isolated, they do act like separate 'beams' in section; you do have tensile and compressive 'flanges' and a web which is what I see as the shear wall.

I'm interested that you did not analyse the lateral support but 'could justify' long walls because of the same lateral walls. So, you see that they must or could have some contribution; you just don't feel it's necessary to include them in your calculations. That's the 'feel' of a structural engineer for the largeness or smallness of a problem; the ability when to say, "nah! simplify, simplify."

The code is IBC. The Method is ASD or LRFD, I don't know for sure.

I would be interested in your thoughts on the panelized system in my other post.

Thanks

 

[JAE]

100 ft. long WOOD shearwall - 100 ft tall? I have no imagination I guess.

 

[CurlyQue]

JAE

I appreciate your input.

I really do.

I imagine you have an imagination, I guess, even if you don't believe you do. Of course you were being facetious.

I am pursuing a point; not an absolute specific condition. Obviously a 100' long shear wall 100' tall is a most probable improbability. Try a 10'x10'x.001' thick shear wall instead with a 26,000 lb lateral load. Say the dead load counter force is 0 which it is in some municipalities.

Let's do the math: 100 mph = 26psf. 100' high x 100' long load wall x 26 psf = 260,000 lb. 260K x 100'/2 = 13x10^6 lbft overturning moment. / 100' = 130,000 uplift. / 150pcf = approx 10'x10'x10' chunk-o'- regular weight concrete, a big boat anchor. Shear = V = 260,000lb / 100' = 2,600 plf. There, that was fun. Hedgehogs and Foxes are now both happy or foxes and hedgehogs are now both happy; however you want to put it. ;-)

Back to paragraph 2: 26,000 lb / 10' = 2,600 plf. 26,000 lb x 10'/2 = 130,000 lbft overturning moment. 130,000.00001 lbft / 10' = 13,000 lb uplift. / 150 pcf = approx 4.425381248 chunk-o'-regular weight concrete.

So now we've proven that a 100'x100'shear wall with 100 mph load acting on a 100' wide wall induces the same amount of shear in a 100' wall as a 10'x10' shear wall with 100 mph load acting on a 100' wide wall. It's all in the width of the load wall; reduce that and you reduce shear; plain and simple.

The overturning moment is obviously 10x less mainly because the load is applied 10x closer.

The uplift is 10x less for the longer wall because it's 10x longer. That's the mechanical advantage of a longer shear wall.

If the 100' long load wall is unimaginable, reduce it to 10' and chop off another zero (i think) from shear, moment, and overturning, it doesn't matter for my point, my question.

The point is that the shear remains the same for the tall condition versus the 1/10 condition.

The difference is that the load acts on a 100' long section versus a 10' long section. That is, shear is shear and plf is plf. Right?

Now, take that same 2600 lb load and apply it to a column 100' tall. You'd choke on the l/r, right? But, if you applied it to a 10' tall column you'd sleep at night ... except you'd wake up in a fright when you realize ... :

Now, Woodman says the diaphragm is braced at the floors, which it is. That's the same as saying the column example immediately above is braced continuously and that column is 10' wide or whatever the floor to floor is.

The problem is we're talking about a wall that is .001' thick or a column that is .001 thick. You wouldn't put a 2600 lb load on a column that's 3 1/2" thick in one direction and .001' in the other because your radius would be infinitesimally gyrational; rather irrational, right? I know you can imagine that JAE.

So how come it's okay for a shear wall to be loaded laterally with a large load but not a column with the same proportions? How come one engineer sees the wall is a piece of paper and another says that's okay? Who's missing what ... because who's on first and what's on third?

I'm interested in what keeps the .001' shear membrane from buckling. Is it the toothpicks at .01' o.c. or the thickness of the shear wall or what, the nails or the glue? The tests and tables and all of that say plf, plf.

I haven't heard a reasonable answer to my question: "Shear wall hold down distance apart?" One engineer says 4' max; another says 20' or 50' or JAE says never 100'. What's the difference and how do you calculate it? It's got to do with thickness. Why else would shear panels require 2x versus frames or spacing at 48" o.c. I think it has to do with the 10' wide column example also being braced 16" oc vertically such that a 2x6 or a 2x8 should give more shear value up to the shear capacity of the panel itself of course. It also has to do with how much you nail the shear diaphragm to keep that puppy from poppin'. What if I put studs at 12" o.c. or 6" o.c.; I should get more value, right? What keeps a 100' wall from buckling (or a 4' vs 40' for JAE ;-))?

I also think it has to do with l/r, the length of the hundreds of mini-columns 1" wide and .001' thick spanning from stud to stud. If that's the case, then a 100' shear wall will not buckle any more than a 10' wall.

The problem is, I don't believe that for the same reason I won't balance a book on a straw but I will on a straw/2.

That's why I've asked the question.

A lot of the problem is always in the question, right? Put another way, a lot of the answer is in the problem.

CQ

 

[JAE]

Well, for a wood stud shearwall with some kind of sheathing there are a couple of checks in the design that would mitigate "buckling" of the shearwall.

Say you have a 20 ft. long (not 100 ) shear wall that is 10 ft. tall. Lateral wind/seismic load along the top in plane. You also perhaps have perpendicular wind occurring with the lateral wind.

Most stud wall shear walls have built-up end posts - say 3 2x studs - and then typical studs between the end posts. The end posts have hold downs (if needed - as the wall gets longer relative to height the hold down requirements does go away).

With the vertical gravity loads (0.6 x Dead) and the lateral wind load, the wall assembly, in plan, is taking forces similar to a beam/column. There is axial load (P/A) and axial force due to bending (My/I) where M is the overturning moment equal to the lateral force times the wall height.

y is the distance from the center of the wall (in plan) to the end posts. I is a moment of inertia of the wall assembly. This I value can be calculated based on the areas of the studs and end posts relative to the center of the wall.

With P/A +/- My/I you can calculate the maximum tension and compression forces in the end posts and in the individual studs. Each stud will have a compressive force that begins to diminish as you approach the center of the wall. The end posts, of course, have the highest.

You can individually check each stud against this compressive force and see if it is capable of resisting that axial load (and perhaps in combination with any perpendicular wind load that would apply). If the studs all work, I don't see how the wall would ever buckle.

Now most engineers don't do the My/I for the hold downs, or even the end post axial checks, but rather take the overturning moment and force all the load into the end posts alone - ignoring the in-between studs.

Hope this helps.

 

[CurlyQue]

JAE,

Yes, your approach helps by considering the assembly, the I value, and the Moment about the centroid of the wall. These views at least address the bending in the wall due to uplift.

I want to discuss why an engineer who sees long sheer walls as 'flimsy' which would crumple 'like a piece of paper' would design a 4' panelized system with hold-downs at the corners of each panel. How is this method stronger?

That's what generated my initial question. If hold-downs at 4' is stronger, then, conversely, aren't hold-downs at 10' or 20' or (god forbid ;-)) 100' weaker?

I'll get back to you when I get a little more time.

Thanks again.

 

[AELLC]

A 3-story building with shear panels only 4' long will have very large uplift or compressive forces due to overturn at the end of the panels, and the tension rods will go slack over the years as the studs shrink slightly.

I have never heard of any limitation of maximum distance between hold downs, or the concept of long shear walls being "flimsy" and prone to buckling.

For a multistory, mutifamily wood structure in a low seismic, non-tornado nor hurricane-prone area, usually the endwalls are fully sheathed with OSB, the front and rear elevation walls have OSB walls as long as possible using perforated shear wall design around the windows, and the center long wall(s) (party or common wall - there may be 2 depending on fire rating requirements) are gypsum shear walls the full length of the structure, with few or no hold downs specified.

 

[AELLC]

I should amend my above post by adding that there would also be interior shear walls at certain partition walls aligned front to back, usually gypsum both sides of the wall, and perhaps OSB in some rare cases on the bottom story.

 

[CurlyQue]

AELLC

Thank you for your post.

Your description (in two posts) is the design I think most were anticipating; especially the design-builder and the sub-contractor. The owner is kind of hidden behind the design-builder.

I'm not sure if I made clear that the design has 4' panels at 4' oc. There are several walls but the one I have calculations for is 48' with (12) 4' panels. The configuration of the building is somewhat convoluted (architects ;-)) but these shear walls basically occur every 40' so one wall that jogs would be about 24' long with 6x4=24 all-thread locations, 2 per 4'.

The original design had the same configuration with 1 layer OSB on each side of double stud walls (4 total layers) which is mostly problematic for construction but can be done. That got changed to a few more walls (in the long direction which makes no sense) with 1 layer each side of each double stud (2 layers) but with all-threads completely up through the building, 2 per panel at each corner or basically (2) every 4'. There is no accommodation for shrinkage. Now the design has anchor bolts at the corners of the 1st floor panels with 3"x4" square washers and straps at the 1-2 and 2-3 floor boundaries with A23 hold downs at 2' oc for the trusses all due to contractor budget who freaked when he saw the first design which cost $160K more (for 10 buildings)! All panels are sheathed in OSB at 2" o.c. perimeter and 4" in field; basically maxed out all the way up.

This design all seems a bit cobbled but my main focus is on why one would design a 4' panelized system if it's not any stronger or cost effective since you've never heard of such. I have pictures of test panels with all-thread but never in a multiple-panel configuration like I'm saying.

Thanks

 

[AELLC]

In cases like this, usually the Owner or Builder submits the finished structural plans to a second, independent SE for value engineering.

 

[AELLC]

I did some research and found that the Simpson Strong-Tie people have something called the ATS system, and it does compensate for wood shrinkage. Is your SE spec-ing this? Maybe he had such a difficulty in this case he felt it was necessary.

Ar eyou saying the buildiong is very long, but the exterior walls are jogged every few feet? Maybe your SE is saying he doesn't believe a long, jogged shear wall is effective without buckling. I am not sure how the Code addresses jogged shear walls.

 

[CurlyQue]

Yes,

In this case the project was on a tight timeline, foundations are in, and framing has begun. The design-builder is getting letter from another engineer which they're going to wave in front of the inspector. That won't do much good because the inspector won't say anything if it's not in the code or on the drawings. I don't think the design-builder is going to get very far in trying to change anything because the engineer has to sign-off and, since the foundations are in, it's kind of late.

 

[Cadair]

I see a few problems with how your engineer is seeing this.

First, he is saying that the wood shearwall will buckle because it is "too long." That assumes that wood shearwalls act as units. They don't. Each individual panel acts as an individual resisting element with its own buckling capacity included into its listed capacity.

Second, the "too long" theory is viewing the shearwall like it is a column with a point load at the top. It isn't. It is being loaded along its length, and it is being unloaded along its length.

Third, by basically installing holdowns every 4 ft., the engineer is trying to resist the uplift each panel has to individually resist. This is understandable, except he is missing the other half of the equation. The panel "upstream" of the shear force provides the compression force needed to do this.

The three points above basically suggest that he is being extremely conservative by saying wood shearwalls act as a unit when it might be to their disadvantage but they also act as individual panels, without the influence of neighboring panels, when it might be to their disadvantage.

Lastly, how does he explain how having more holddowns will prevent buckling? If viewed in plan, holddowns resit forces in and out of the page, not in plane with the paper, which I assume is the direction you are talking about buckling.

If what you have presented and I understood it correctly, the engineer does not understand wood design and should not be doing this.