I have 24 glulam beams that are 3″x12″ that have been installed in a post-and beam shed roof. The connection to the column below is via a 9/16″ hole plumb down the centreline of the beam that continues through into the column where a 2″ sleeved through-hole allows installation of a nut and flat washer onto a length of 1/2″ threaded rod. A plate washer and nut on top of the beam complete the connection. The high and low ends of the roof both have the same detail.
Prior to installation, I braced and plumbed all the low roof columns. After two days of crane time, the assembly is complete. The problem is that the glulam supplier has incorrectly calculated the line length between the two vertical holes in the beam and as a result all of the high-end columns have been forced outward. They are out of plumb by 7/8″. After leaving it to the architect and engineer to sort out, I’ve been asked to ovalise the rod holes in the low roof connection, allowing the beams to be pulled back and thus bringing the columns back to plumb without losing roof overhang at the high end.
My question is of course how to achieve this objective. Accurate drilling over a 12″ depth is not easy. My notion so far is to use a 3/4″ HSS twist bit and try to follow the threaded rod down its length. I suspect that if I take the rod out that the bit will break through into the existing hole and it will be very hard to drill another hole on the correct line. The beams are 25 ft long and weight about 250 lb each. They could be removed one at a time, but that shouldn’t be necessary as there is a flat roof conveniently located for easy work access at the low end. There are a further 6 beams that have not yet been installed (due to other reasons) but will require the same remedial action.
The attached pdf shows the detail described above. It’s on page 4, detail 4. This is a government job so all dimensions are in metric; that 13 x 650 rod is really 1/2″ x 26″.
Although this work is outside my contract and so is an extra for me, as usual I want to minimise cost and expense to client, as well as get moving forward on putting decking on this 5000+ SF roof (winter is getting close). What’s the best way to field-drill accurate 3/4″ holes?
Replies
I would not use a twist bit in wood. It is a poor choice due to a tendency to wander and clog with shavings. Forstner bits are perfect for this sort of job, but Forstner bits are short so an extension would be needed, which could introduce errors in drilling. A long ship auger bit would work, but it will probably need a jig to drill accurately. The jig can be made of plywood quite easily.
Won't a ship auger will be
Won't a ship auger be damaged from touching the steel threaded rod in the exiisting hole? Their design makes them less stiff because there's less material. My thinking was a twist bit with a full 12" length will be stiffer and less prone to wandering off line. Happy to be corrected here though.
Jig design ideas?
I didn't realize the threaded rod would be in place while drilling. Yes, ship augers probably would be damaged with contact with the rod. My idea of a jig is two plywood squares with 3/4" holes aligned to produce a nearly perfect hole at right angles to the plate. Basically you would have a 5 sided box (one open side) that would be clamped or screwed to the plate while starting the hole. The two holes would be about 6 inches apart and after starting the hole, the jig could be removed to finish the hole.
The ideal thing would be to get a jig made up consisting of a piece of 1/2" or 17/32" rod welded to a bushing to hold a Forstner-style bit just out of touch of the rod. Make the rod maybe 18" long, with the bushing 2" from one end. Put the bit in the bushing with the shank facing the short end, install extensions as needed on the bit, and drop the rod in the hole. Attach Vice-Grips to the bottom end of the rod to give you something to "steer" with. Then drill.
Of course you'd have to stop often to clear the hole, but you could probably minimize it if someone worked a vacuum while drilling.
That is a very interesting problem, and I am trying to think of what I would do if confronted with it. If I understand, you would like to leave the gluelams in position with the threaded rods in place, and then oval the holes in the gluelams, and then shift the threaded rods over to the newly cut part of the ovaled hole; so that would bring the columns back to plumb.
If that is what you are planning to do, that seems like an incredibly difficult proposition. If the hole is ¾” and the rod is ½”, there is 1/8” clearance all around. If you try to run a twist drill down alongside the rod, it is going to crowd the rod to the opposite side of the hole it is in. That would mean that when you drill a new ¾” hole, the center-to-center shift will only be ½”, however, you want to achieve a shift of 7/8”. Even if you crowd the threaded rod all the way to the far side of the new hole, the center-to-center shift will only be 5/8”.
Even if these numbers would possibly work, I think it will be impossible to control the drill bit going down 12”. The threaded rod would probably resist the bit to keep it from running into the existing hole. However, I suspect the bit will try to go around the threaded rod, by walking sideways. The bit will develop friction as it makes shank contact with the threaded rod, and this will cause a reaction force that will encourage the bit to walk sideways and follow around the threaded rod.
I assume the original holes were drilled in a shop with some type of holding fixture. Even if you could take the pieces back to the shop, I suspect that ovaling the holes would be a major challenge. The holes might have to be plugged in order to get the new holes to track straight.
I would be thinking of a design modification of that joint that would get around the problem. You could make a bent offset in the rod where it passes from the column into the beam, and chisel out a pocket for the offset to be housed in the top of the column. The problem is that the rod would have to have each end fed into the each of the adjoining holes during assembly, and that might not be possible, unless you disassembled the other end of the gluelams. It might require a boom truck to lift the gluelam up, so you could put the rod between it and the column. Actually, that would probably be the best option, but you would have to get the rods reworked or remade for the offset.
I don’t understand how that nut and flat washer work inside of that round steel tubing.
What did all that have to do with my suggestion? ;)
Get the beams down
and set them on sawhorses. Fill the holes with hardwood dowels and gorilla glue. Once dry, drill new holes in correct spots.
Reaming the holes with the all-thread still in place... that sounds tough.
Obviously I've been thinking a lot about this. Taking the beams down would require at least a full day with a crane, and some elaborate bracing of the high roof columns. The columns on the low side are only 3 ft or so tall and they bear on a flat roof, so access to perform the work in situ is good (for a change). It would take a lot for me to consider anything but an in-place solution.
The rods are removable at this time, and their holes could be indeed be plugged. A new hole 7/8" up the slope from the existing would in theory be all that was needed, but the likelihood of being accurate enough is the matter at issue. I'll chew over the jig that was proposed - using the existing hole as a guide.
On a diverging tack, what about using a small chainsaw with a plunge-cutting bar? That would produce a larger than needed slot but it would be a solution, albeit an inelegant one. The structural engineer might have something to say about slotting rather than a hole that locks the beam in one location. Visuals are not an issue as this connection is covered by finished ceiling.
Seems like
the issue is at the bottom of the beam, where the misalignment is causing the columns to be forced out of plumb. If you can get the beam in position to drill from the bottom, drill another hole as close to the existing as possible, but far enough away so that break-through into the existing hole doesn't occur for several inches. Use a chisel to break out the waste between the holes. Re-set the beam and use a long clamp(s) to pull the column into alignment, flexing the rod to do so. Those are only 1/2" rods, so should bend easily.
best suggestion so far
This is doable. Remove the rods from both ends of the beam and roll the beam upside down for easier and more accurate drilling/slotting. Reset the beam, install the high end rod and nuts, then use a turnbuckle to pull the column back to plumb, and finally, tighten the low end nuts.
Let's see, 24 beams already in place, that will be a solid day's work for two guys, plus drilling the other 6 beams before they get flown in. Client wants an estimate, should be somewhere between $2,000 and $2 million...
QUESTION:
Can't you just move all of the short posts 7/8" toward the downslope direction while keeping them plumb? Doing that would pull the tall posts back into plumb, and then you could just leave the bolted connection as it is. You would have a slightly larger building, but so what?
You might find
that with a little practice the hole drilling gets quite accurate and little if any turnbuckling is needed.
Or it might turn into a total cluster that costs twice the estimate. I would hesitate to price it, myself. I would tell the owner you will fix it and let him know the cost when done. You probably already hard-bid 99.7% of the job, so he should have no trouble going T&M on the fix. It's not like you have unit costs on this type of remedy on previous jobs...
You're right, this is a fixed price contract for installation of the glulams as well as all the framing. The problem is not the glulam supplier's either, they used the elevations and roof pitch given in the architecturals, but those elevations don't work out to the slope that was specified. That made the line length incorrect and so here we are. It's already been agreed to as a T&M change order.
QUESTION AGAIN:
Slykarma,
Again I ask why you can't just move the short posts 7/8" and be done with the problem. That seems like the obvious and incredibly simple solution compared to ovaling the holes. What am I missing?
wish I could
Most of those short columns are exposed, as there are clerestory windows in the spaces between. I'd have a problem with finishes not lining up. There are also a pair of Simpson L50 angles at the base of each column that are now obscured by infill framing. Would be many hours of work just to get the columns to a point where they're able to be moved.
Thanks for all the comments and suggestions. I have a jig being welded up that will let me drill new 3/4" holes. Will try from above first and see if that gives the results I'm looking for. If that's not accurate then we'll remove a beam, roll it over and drill from the underside. I have some dowel available for plugging the existing hole if the need arises.
The first thought would be to rent a mag drill, clamp a metal plate to your wood and have at it. I would still be afraid of the bit drifting and the weight of the mag drill might be a problem during initial set up.
My second thought was to fabricate a jig with bushings and use a pistol drill. Sinilar to what others mentioned. I still think the bit will want to walk into the adjacent hole.
My third thought would be to fabricate oversized plugs with an offcenter hole. Then open up the existing holes and glue in the oversized plugs. Your original hole would be the pilot.
All done
The work is now complete. I had a jig welded that utilised a 3" length of 3/4" ID pipe welded at a 7 degree angle to a saddle sized for the beam. A couple of test holes showed that the jig was aligned very slightly off square and this was corrected with a thin shim. All holes were plugged with 5/8" dowels and wood glue. After that, work proceeded well and the results were satisfactory with drilling from the top only. The oversized new holes provided sufficient play to get the final locations bang on. All rafter tails were strung in straight and that's it. As I estimated, this took two carpenters a full day. Now we need to drill the remaining uninstalled beams and get them into place. Crane booked for tomorrow.
Thanks again to all for your ideas.
Sly