To Build Steel Rafters into wall or Fix to a RC Ringbeam?

[sybie99]

Good day folks,

I have a simple building where we are using rafters to span 13m between the walls of a building. We are building a frame of reinf. concrete columns with reinf. concrete beams (ring beam) to support the rafters. The reason for this is to give a solid support into which the beams can be fixed. Is it necessary to have a slotted connection on one side for thermal expansion and contraction of the steel rafters?

If not, can these rafters not simply be be built into the brick wall without having to build a concrete frame with ringbeam? Or, at 13metres, are the rafters too long whhich may result in enough expansion to damage the wall? Temperature variations may be 30 degrees.

Basically, when should one use a ringbeam and why, for simple masonry buildings with steel rafters?

Thanks

 

[JAE]

Is there some thermal separation between the concrete support beam and the rafters?
If so, the 30 degrees of thermal variation will be the same for both.
Steel and concrete have similar thermal coefficients. That means that the whole system will expand and contract together and only very small differential movement would be expected at the rafter/beam interface.

 

[sybie99]

There is no separation, but the ring beam is a rectangle in plan, with the rafters spanning the short direction. So the main thermal movement for the concrete beam is transverse to that of the steel beam, also the concrete columns on which the ring beam rests is tied to transverse internal brick walls for stability.

The thermal expansion of the steel beam is only 5mm, that is 2.5mm each side. Would a brick wall not be able to withstand such small deflections?

I guess my question is, why is a ringbeam so commonly used to support the steel roof of a masonry structure? I know for short spans, say 3m, the rafters are usually built into the wall, but for longer spans it seems a concrete frame is constructed with brick infill panels.

Is it not easier to just cast a concrete ring beam on top of the brick walls?

Regards,

Seb

 

[marinaman]

I am not a fan of supporting long spanning steel members on masonry....especially brick masonry. Short spanning joists I'm ok with, but long spanning joists, I am not.

If the long spanning steel member is bearing on masonry, it will move, back and forth, over time due to thermal effects. Its the expansion/contraction/expansion/contraction over a long period of time on a brittle support that I do not like. I have seen many many older buildings with broken-up masonry at the steel member bearing location due to this long term action. This was especially true with regard to older buildings, as many of those are not reinforced structures but simply masonry structures with little, if any, horizontal joint reinforcement.

Today, of course, there's hardly any opportunity to use unreinforced masonry. ACI 530, in this neck of the woods, pretty much mandates at least some degree of reinforcing in masonry, regardless of the loads calculated. Today, we'd have a bond beam at the bearing elevation, likely have vertical wall reinforcing, horizontal joint reinforcing, etc......but even so, I don't like to mix long spans and load bearing masonry. Say you have 13 meter long joists bearing on a reinforced CMU wall with a bond beam at the bearing elevation. That 30 degrees of thermal expansion and contraction will happen and the forces on the bond beam will be lateral, perpindicular to the wall, and, if not properly detailed, could cause movement of the bond beam, out of plane, away from the plane of the wall (fracture the first horiztonal bed joint right below the bond beam). I've seen it detailed not to allow that to happen, but, the masonry just comes in, stops his vertical reinforcing short of the bottom of the bond beam (doesn't extend the vert r/f into the bond beam) lays the bond beam, reinforces and pours the bond beam, and sets the joists.....only to have the joists move the bond beam.....and....there's virtually no vertical uplift restraint if this error in construction is made.

I think the better detail is the concrete frame supporting the joists....with a masonry in-fill as described.

 

[sybie99]

Thanks Marinaman, you've answered some of my questions, would you provide a sliding connection at one end of the joist for long joists, so no force is transfered into the ringbeam?

Regards

 

[marinaman]

My gut feeling is no, I would not provide a sliding connection for this detail.

A couple questions come into play. For example, how is the diaphragm being connected to the "ring beams" and the rest of the structure? Do these joist provide lateral stability to the ring beam itself?

Look in your Vulcraft or other joist manufacturer's design guide and see how they recommend fastening their joists to supporting structures (LH or DLH spec 104.7). They weld the ends down without allowing any movement of the individual joist (with exception of those at a column). Joist girders though, are called by SJI to be bolted with slotted holes. The point to this is that, allowing movement depends upon what the member is and what it is doing. Personally, for repetitive joists, I would weld them down.

Think of it this way.......if you had a steel frame supporting steel joists, would you slot the top flange at every joist bearing to accomodate movement? No. We would weld the joists down to the beam flanges.

For a bigger structure, I'd think of it differently, but for one that's less than 40' wide, to me, I think its ok to weld the joists down.