Footings on Blasted Ledge and a Steel Beam Redesign in Barkhamsted, CT

What the owner wanted

A new house on a lot in Barkhamsted that had just been blasted out of granite, and a straight answer on one question before the footings were formed: what goes under them. The rock along most of the footing lines was a few inches below where the footing bottoms would be, the site contractor had started spreading clean stone, and in one place, at the rear wall, the rock dropped away into a pocket about four feet deep. The town wanted a letter.

Then the scope grew. Another engineer had sized the steel beams, a column and two footings for the owner, and he wanted changes: shallower beams for headroom, a different post layout, and the kitchen beam run the full length of the bump-out. He asked me to take over the steel, redesign it the way he wanted it, and carry the whole package under my seal. All of it from my desk.

What we found

I did not set foot on the site, so the letter says what it stands on. The owner walked the footing lines and photographed along each one from the corners, put a tape in the deep pocket, lettered the corners on a copy of the foundation plan and wrote the depth from footing bottom to undisturbed ground along each wall. Then he wrote that the pocket had been cleaned out and bottomed on bedrock a few inches below the 44 in he had measured.

What they showed was a blasted granite site with the rock fairly flat along the lines: bedrock 3 to 8 in below the footing bottom along the right wall, ground he described as soil or rock 3 to 8 in below it along the others, and one pocket at the rear wall about 11 ft by 6 ft and about 4 ft deep. The pocket sat directly under the pier between the two garage doors. That is where the center beam’s rear post comes down onto the frost wall. Water stood three or four inches deep in it after heavy rain, and a curtain drain ran to daylight from the corner above the bottom of the pocket.

What goes under the footings?

The footings bear on a bed of compacted crushed stone placed on the cleaned rock, which the Connecticut code allows as engineered fill. Before any stone goes in, the loose and shattered rock left by the blasting comes off, and topsoil, clay, silt and standing water come out. The letter spells out how to build it so a site crew can follow it: clean angular stone, 3/4 to 1-1/2 in with no fines; lifts no thicker than 8 in loose; a vibratory plate or roller with the vibration on, at least four passes, until the stone stops moving; at least 4 in thick after compaction and never more than 24; at least 12 in past each footing face, held on a 1 to 1 shoulder wherever the rock falls away so the stone never ends at a free face. A static drum roller alone does not compact clean stone; the vibration does the work. Where sound rock is at the footing bottom or within 4 in of it, the footing is poured down onto the rock instead. The bed is checked at 3,000 psf, half the code’s value for gravel, and the highest pressure anywhere in the package is about 2,450 psf, at the bump-out corners with the girder truss bound on them, about 2,500 at one corner once the stair was added on SK-1, so the stone governs and the ground under it never does.

The pocket gets 3,000 psi concrete, the footing mix, placed on the cleaned rock after the water is pumped out and brought up to the footing bottom, about 10 cubic yards, no forms and no bars. The continuous footing is then built on top of it as a separate pour.

The engineering

The steel came out as three W16x31 beams. Two on the center line, each about 26 ft, meeting end to end on one HSS 4x4 post at the middle of the house, the front end in a pocket in the gable wall and the rear end on a second post standing on the frost wall inside the framed wall between the garage doors, so no beam lands on wood at the rear. And one on the kitchen line, running the full length of the bump-out from pocket to pocket. The owner had asked for steel the full length with a post at the end of the stairs. The same beam that works on the center line spans the whole bump-out without it, so the post, its footing and 4 ft of leftover wood girder all went away. The beams are 16 in deep against the 24 in of the scheme they replace, with the top of steel 1-1/2 in below the joists and a 2x6 nailer on the flange. The center beams deliver about 17,300 lb to each end and the kitchen beam about 15,000, service loads, and the center post carries about 34,600 lb.

The corners were the hard part. The roof over the kitchen bump-out is carried by a girder truss along the main wall line, and the two ends of that girder bear on the bump-out’s end walls at the corners. The trusses are not designed yet, so the letter takes that reaction at 22,100 lb a side, a bound built from the drawings with the roof, the snow, the attic floor and the bump-out roof all on it at once. The girder’s post comes down the main wall onto the sill at each corner, exactly where the architect’s pocket for the kitchen beam would have cut the wall. So the kitchen beam moved 12 in outboard and both pockets with it, and the post stands on 15 in of solid wall beside the pocket. Even so, the kitchen beam end and the girder post together put about 37,100 lb within a foot of wall at each corner. The 20 in footing carries that only if the 10 ft wall is credited with spreading it at 45 degrees, and I was not willing to lean on that at both corners. Footing F2 is 36 in wide and 14 in thick under the full length of each end wall and 2 ft along the main wall past it, with three #5 bars the long way lapped into the footing on either side, bearing at about 2,450 psf with the whole corner on it.

One thing came off the drawings, too. Earlier in the job I had described a widened footing under the pier between the garage doors, sized for the 34,500 lb the deeper beam scheme brought to that pier. With the W16 pair the load there is about 22,200 lb, the pier bears at about 2,230 psf on the 20 in footing that runs around the house, and the widening was withdrawn in writing.

The letter closes with what it does not change, which is almost everything else on the architect’s foundation sheet, and with one condition that matters: the truss supplier’s layout comes to this office before the trusses are ordered, because a truss design that bears on either beam line, or puts more than 22,100 lb on the girder’s ends, is a change to this design.

How it got delivered

The job came in on a Monday by email and phone. The owner sent his photographs and his lettered plan two days later, and he had the stone answer by email within the week. When the scope grew to the steel, I gave him a date for the whole package and the letter, sheet S-1 and sheet S-2 were signed and sealed on that date.

That same morning he wrote back: his stairs land on the kitchen beam, and would I look. Cutting the I-joists in the stair band short onto a header is an ordinary stair opening, but where the stringers meet the steel it wanted a detail. The stamped one-sheet stair detail at the beam went out that afternoon, inside the fee. It was ours to catch, so it was ours to fix.

If you are building on ledge in the Litchfield hills and the question is what goes under the footings, or you have a set of beams you want a second look at before the steel is ordered, call me. A remote job like this one takes your photographs and your tape, and the stone question here was answered within the week.