Anchor Bolt Spacing 8"OC seems overkill

[Rich Zimmerman]

I'm trying to check the work of my structural engineer. He's got 5/8" anchor bolts spaced at 8"OC attaching a sill plate to the top of a basement 6" ICF wall. Joists hang off the sill plate. 1-1/8" floor sheathing attaches to the top of the sill plate with 2 rows of #8 screws 4"OC. Say's it's necessary to restrain the top edge of the wall. I've NEVER seen spacing at 8" for an anchor bolt. This structure isn't particularly unusual (other than I think it's a bit weird to hang the joists from the sill plate rather than on a ledger board.)... 8' ceiling basement. 7.5' unbalanced backfill behind the wall.

So here's the big question I don't understand. Given the spacing there are 4 #8 screws attaching the plywood to the sill plate for every 5/8" anchor bolt. If we look at the cross section of the associated fasteners, the 5/8" bolt has a shear area of 0.218in^2. The four #8 screws have a shear area of 0.039in^2 COMBINED. And that assumes the screws shear before they pullout. It would seem we have nearly an order of magnitude discrepancy here.

 

[jittles]

Sounds just fine to me.

You are way off base and your comparison of shear areas isn't anything to do with the failure modes you could expect in either the screws or the anchor bolts.

 

[Rich Zimmerman]

OK. That makes me feel better. That said, the engineer told me the spacing was to restrain the top of the wall through the plywood due to the backfill behind the wall. It would seem you need to transmit that force from the wall to the sill plate through the AB, and then to the plywood via the #8 screws. It would seem to me the weak link here BY FAR is the #8 screws, but as you mention I must be missing something huge. I'd love to learn what that is...

 

[atrizzy]

Is there a stud wall above the sill plate? If so, the sill plate would also need to potentially transfer wind and seismic forces from above to the wall, whereas the floor sheathing would only need to transfer the force of the backfill, and perhaps its own seismic load component.

Just a thought. 8" o.c. does seem a little tight, but it really depends on a lot of factors.

 

[jittles]

Since you're honestly curious and my first answer was a bit short, one thing you're missing is that the strength of the anchor bolt embedded into the 6" concrete wall will be governed by the failure of the concrete breaking out well before the steel anchor gets anywhere close to its own strength in shear.

 

[Rich Zimmerman]

Yes. There is a stud wall above that. It's a walkout basement so it's only retaining in one section. The portion that's retaining has the 8"OC. The other side only has 32"OC spacing. So I figure the 32" is enough for the the forces from the wall. The movement from 32"->8" is purely for the retaining portion. (BTW: We're in southern AZ, so not a lot of seismic activity down here). Again, I'm concerned that if the forces are so great to go to 8"OC, then 6 #8s per foot isn't going to move that force into the plywood. Perhaps the joists should have been attached directly to the wall with a ledger rather than hanging off the sill plate?

 

[atrizzy]

jittles, I agree with your breakout comment, but then why not use 1/2" bolts?

Rich, is it possible that the retaining wall top is partially restrained by the bearing against the joists rather than the plywood alone?

 

[Rich Zimmerman]

I don't think so. The joists are hanging off the sill plate with top mount hangers. The joists are at a 45 degree angle WRT the wall so there's actually a not insignificant gap there. If the wall started pushing against the ends of the I joists it would likely just skew them. Plus they're ITS hangars so the joists don't even have fasteners to keep them from sliding back and forth in the hanger.

 

[BridgeSmith]

I'm not familiar with the configuration you're describing (not that I've seen much in the way of building construction), but a few things that may be contributing to restraining the wall are:

1) The attachment of the joists to the sill plate.
2) The bottom plate of a wall attached to the floor (and maybe the sill plate as well).
3) Interface friction between the sill plate and the flooring
4) Construction adhesive applied to the joint between the sill plate and the flooring.

This is just my speculation of things that, if present could provide restraint to the wall (#4 may seem inconsequential, but can actually provide considerable shear capacity).

Rod Smith, P.E., The artist formerly known as HotRod10

 

[BridgeSmith]

jittles, I agree with your breakout comment, but then why not use 1/2" bolts?

I agree as well with regard to breakout. The use of 5/8" bolts may just be a matter of availability (especially if they're J-bolts).

Rod Smith, P.E., The artist formerly known as HotRod10

 

[oldrunner]

I gather that the bolts are in a ICF wall, where just the cells are filled. Does this mean a bolt in each cell? The wall is 6" thick - then the cell size is pretty small. Are the bolt placed off center? I imagine that the backfill is being resisted by this wall and there can be either seismic force from the house and maybe some seismic force from the backfill. I don't really have enough info and haven't done an ICF structure for quite a while.

 

[atrizzy]

oldrunner brings up a great point. Perhaps 16" is not close enough, and due to the cells, 8' is the only other sensible spacing in order to avoid issues with the cell dividers.

 

[Rich Zimmerman]

It's a flat wall ICF system (FoxBlocks), so there's not really "cells" like in a waffle grid ICF. You get a nice flat 6" concrete wall inside, so it's easy to use all the normal concrete wall math. Bolts are dead center.

The bolts only attach the 2x sill plate. The plywood floor gets 1-1/4" holes bored over the bolts to clear the nut/washer. Then the wall goes on over that. I only mention that so you can see there's no interference issue with wall joists or anything that might require a specific interval of J-Bolt spacing.

BTW: Is there a rule of thumb (or a IRC, AISC or ACI reference) for calculating what capacity an anchor bolt has in shear at the top of a 6" wall?

 

[atrizzy]

ACI annex D has procedures for calculating bolt shear strength for this situation. It's somewhat involved.

 

[XR250]

Seems tight to me. How many feet of backfill is against the wall? I have never spec'ed them closer than 16" O.C.
Prescriptive IRC code is 6 ft O.C. IIRC. The numbers never work out for 6 ft. O.C. but historically, it has proven to work reasonably well (at least in my area).

 

[BridgeSmith]

Do the screws from the wall bottom plate into the floor extend through into the sill plate (It's only 2-5/8" to the sill)? If so, those screws would count too, as well as the frictional resistance due to the permanent load carried through the wall to the interface of the floor and sill plate. Again, that is neglecting the resistance gained from a bead or 2 of construction adhesive, which is considerable if applied.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[kissymoose]

XR250 - I agree the numbers never work out for 6ft o.c. This is something that has bugged me for a while, because this being the minimum spec in the code = the standard expected by contractors and any more is met with pushback. The IRC spec for attaching the joists to the sill plate is also just (3) 8d toenails, and I've seen a good number of floor joists that break that connection, the wall leans inward, and the joists end up cantilevering over the wall by a couple inches. It's hard to get in there and see what's going on, so it could be that the joists/rim board do not satisfy this connection requirement to begin with. As for designs historically proving to work but not by the numbers is a concept that is frustrating, but I do so the validity in.

BridgeSmith - I agree that friction and glue would contribute to the lateral strength, but I haven't heard of anyone using friction in their calculations and wouldn't feel confident in doing so, and wood glue creeps under load and cannot be used in calculations.

Rich_Zimmerman - Per the NDS, (4) #8 screws have a 370lb± capacity and a 5/8"Ø anchor bolt has a 530lb± capacity. The failure mode for the anchor bolts is not in the concrete or the bolt, it's in the sill plate with load being applied perpendicular to the grain. The capacity of a 5/8"Ø anchor bolt is at least 1000lbs for breakout accounting for edge distance. So if you were to account for strength in the connection from the wall plate to the sill plate below, the screws/nails capacity probably comes close to the anchor bolt capacity, or at least isn't orders of magnitude off.

 

[Rich Zimmerman]

Thanks kissymoose. I went thorough this with my engineer and got the same numbers (530lb & 370lb). He actually needs close to the 500lb on each bolt (60pcf/f, 8' unbalanced backfill, 10' wall height, math...) and was a little taken aback when I stepped him through the screw capacity. (I think his exact words were "s**t, I missed that". So he's going to think on the matter and come up with a solution... Thanks everyone with your help.

In retrospect I really wish I had asked him to design that back wall as a cantilevered retaining wall. Couple extra yards of concrete and some extra rebar and I wouldn't be dealing with all this shear in the floor and (a MASSIVE advantage) I could backfill before the floor is built! Right now I'm having a hard time figuring out exactly how I'm going to backfill this wall. Once the structure is up I can't get a skidsteer back there anymore, and I can't just dump over the wall if the floor is up....ugh! I'm sure 20-30 yards of dirt will be easy to push uphill in wheelbarrows....

 

[BridgeSmith]

The failure mode for the anchor bolts is not in the concrete or the bolt, it's in the sill plate with load being applied perpendicular to the grain.

That's an excellent point. That's probably what we've been missing is this, that made it look like the anchor bolt connection was so much stronger than the screws.

BridgeSmith - I agree that friction and glue would contribute to the lateral strength, but I haven't heard of anyone using friction in their calculations and wouldn't feel confident in doing so, and wood glue creeps under load and cannot be used in calculations.

Agreed on both counts, although my personal experience with the stuff leads me to believe the adhesive will still provide significant resistance, even if it does creep, albeit with a 1/2" or so of movement.



Rod Smith, P.E., The artist formerly known as HotRod10

 

[XR250]

He actually needs close to the 500lb on each bolt (60pcf/f, 8' unbalanced backfill, 10' wall height, mat

That is quite a load. Surprised he could make 6" thickness work. I hope he has a plausible escape route for all of the shear being dumped into the floor diaphragm.