Rafter connection 1

[SteelPE]

We are currently in the process of removing some columns from an existing building. The client wants to remove the columns in order to increase the amount of open space. The existing columns are spaced at 15’ o.c and the client would like to have a 45’ open space (so we are removing 2 columns). This happens at multiple locations. The columns are supporting a wood roof.

The idea we have come up with is to place new steel beams under the existing joists on each side of the existing columns. Place the steel and then remove the existing columns (to eliminate shoring). We would have a wood nailer placed on top of the steel beam.

I have a little bit of a detail problem. The roof slopes and average of 1.75/12 (it varies but this is the average). How would you attach the sloping joists to the top of the steel beam? Is this something you would just fill with blocking? I was looking at using Simpson VPA connectors but they require a minimum slope of 3/12.

 

[ToadJones]

I suppose you could also cut wedges and slip them under the rafters. This would also allow you to drive the wedge in as far as necessary until tight. Then use some kind of Simpson sheet metal clip angle to connect the rafter to the sill.

 

[msquared48]

I know you guys will hate me, but what's wrong with notching the existing rafters to sit flat on the plate on top of the new steel beams? I think a 1" max deep tapered notch would do just fine with toenailing and blocking.

Yea Toad, I know the contractors will bitch, but, frankly, contractors are paid to bitch. It's part of their game.

I do not like the sloped bearing as it is imparting a lateral force to the system. Flat is better, at least in this case.

Mike McCann
MMC Engineering

 

[hokie66]

I would agree with the contractor about that solution. Too hard. If you can connect the rafters to the beams for uplift, surely you can connect them for a bit of downhill force. I don't see why this force would be any greater globally than with the sloped bearing.

 

[csd72]

I have a suggestion,

Why not put a channel on each side of the column and support the column directly thus avoiding the need for any new roof connections.

Not sure if you can get channels big enough but it is worth a thought.

 

[ToadJones]

csd-
Good idea.
Basically a transfer girder.
Seen this done in mills many times to remove columns and open a bay for new equipment.

 

[SteelPE]

Here is a scheme of what is planning to be done. Hope this clears some things up.

 

[ron9876]

Still not clear if existing beams are steel or wood. Either way I would be looking at connecting to the existing beams/columns and not the existing rafters.

If the beams are steel it would be easy to weld supports off the new beams to connect to the existing beams. If they are wood then csd had the best idea of attaching to the existing columns (maybe even if they are steel). You can add cross pieces each side of the existing columns that span between the new beams. Connect the cross pieces to the columns.

 

[SteelPE]

4thorns

You basically have the section correct. We were just thinking about spacing the beams closer to the original support beam.

csd,

Good idea but I believe clear height is going to be a problem, At the lowest point we only have 10' or so to the underside of rafter.

When I originally toured the building I said (tear down)..... but again nobody listened.

Great suggestions from everyone. I was originally hoping for someone to suggest a Simpson XXXX connector. I have three major concerns.

1)Connection of the beam to the rafters to resist LTB of the beam.
2)Connection of the beam to the rafters for uplift.
3)Getting proper bearing for the rafters on the steel beam.

I guess ripping a new "sill" for the rafters would be the best. I could then use a Simpson hurricane clip (H3) to attach the rafter to the beam

 

[RFreund]

I would think they will most likely rip a 2x plate on top of the beam or cut a "birds mouth" in the joist.

EIT

 

[hokie66]

Another option would be to site measure all the rafter spacings at each new beam, then fabricate cleats on the beams for bolting to the rafters. Similar to purlin cleats.

 

[4thorns]

How about something like this with a field weld at each rafter. Not sure of the connection but might be able to handle all your concerns at once.Again it seems like a lot of work but so will ripping and bolting down a 2x. Just a thought.

 

[dhengr]

I’m still perplexed by the idea that you would mess around at every joist bearing in the bldg., when they are nicely seated on an existing beams, which could be picked up (supported) from the new beams, on either side of the columns you want to remove. Then remove the column and leave the old beam in place. The idea of cutting a new birds mouth in the joists at each new bearing, would likely get someone shot if they ever showed up on the job site.

In the OP you said you were going to nest two new steel beams alongside the existing beam, aren’t these new beams as deep or deeper than the existing beam for the same load and triple the span and some reasonable deflection? Now, all of a sudden you have a headroom problem with structure below the 2x jsts.? Some sort of saddle off these two new beams would pick up the existing beam, I called that a 1' or 2' cross beam bolted to the bottom of the new beam webs. Maybe this has to be a stiffened plate off the bottom flg. of the new beams to pick up the existing beam in bearing. Then, at this same location, btwn. the 2x jsts. and over the existing beam, provide another cross beam, or two, to the new top flgs. to resolve the torsional loading problem, and again at mid span for top flg. stability. My new beams are both at the same elevation and do not touch the 2x jsts., they relate solely to that existing beam.

Now go ahead and apply hold downs at the existing jst. bearings, if needs be, before you put in the new beams, so you don’t have to reach over them. And, then your top cross beams can be made to hold down the existing beam too.

 

[4thorns]

dhengr, you just got me wondering. Is the original beam capable of carrying it's own dead weight in a span 3 times the original design? If not then it's load will have to be carried by the connections at the existing bearing of the rafters. Sounds as if the new beams might have to be attached to the old beam regardless of which design is used. At the very least, to prevent excessive deflection of the old beam.

 

[4thorns]

hokie66 (Structural)

"Another option would be to site measure all the rafter spacings at each new beam, then fabricate cleats on the beams for bolting to the rafters. Similar to purlin cleats."

Missed this reply hokie. Didn't mean to sound repetitive.

 

[hokie66]

No problem. I think it would be possible to fabricate the beams with the cleats in place. Needs careful measurement, but saves lots of site work.

 

[RFreund]

dhengr - Brilliant!

I think I would go that route.

I guess I was think more along the lines of new construction. Cutting birds mouth though is not a huge deal - make 1 pattern, start tracing. But if the joists are already seated well then I agree, make use of what is there.

EIT

 

[jaypartain]

After reading all of these posts at one time, dhengr's stands out to me as requiring the least material, least labor, and least modification to the existing structure. One add on to this is could you somehow use the existing columns (cut off near the top) in the attachment to the cross beams.

Jason A. Partain, P.E.

http://www.myconstructionfinds.blogspot.com

 

[MattJM]

I agree with dhengr - don't support the rafters; support the thing already supporting the rafters.

Slick solution. Would love to see the final detail.

 

[msquared48]

OK.

Then use one steel tube section on each side of the existing beam and support the existing wood beam at or near the existing beam joints or at or near where the columns used to be. Temporary shoring may be required here though and the OP's original idea was to not use shoring. To get around the shoring, the beam supports would have to be installed to either side of the existing columns, not at the columns. The use of steel tubes will take care of the torsional loading problem.



Mike McCann
MMC Engineering