Splices to long span portal frame rafters

[nd0586]

Hi all,

I have designed a 20m bay portal structure which is currently about to go into fabrication. 20m rafters can be transported to site, but the steel fabricators have requested a splice in the rafters so that they can store them more easily in their yard.

The rafters have been designed very close to full utilisation, and I am nervous about allowing them to provide a splice as I feel it would invalidate the plastic hinge history calculations (to British Standards) by not providing a uniform section.

I am aware that rafter splices are used on portal frames - has anyone given any consideration to their effect on the rafter design calculations? Do you provide a bit of fat in your rafter size to make up for it? Or would you be happy with a rafter designed to 100% utilisation to be spliced?

I have already queried the additional weight of the connection, which is negligible, and would only allow the splice near the points of contraflexure (which however will move around with different load cases)

Thanks for any help!

 

[hokie66]

It is normal practice to splice rafters near the points of contraflexure. One way is to fabricate the low part of the rafter as a weldment with the column, then make the rest of the gabled rafter as one piece. So the portal frame is transported in 3 pieces.

 

[nd0586]

Thanks for the response! I understand that it is normal practice; however do you make any design allowances for it, or would you be happy for a continuous rafter in a portal frame designed to 100% utilisation (more or less...) to have a splice put in it without any further checks?

 

[KootK]

I'd be okay with the splice I think. To convincingly preserve your hierarchy of plastic hinge formation, however, it might be prudent to design your splice for a moment somewhat greater than the expected demand. Maybe 115%?

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[civeng80]

If you splice the rafter the moment capacity in the splice should still be equal to the section moment capacity so I really so no problem in splicing even if the section hasn't got a lot of extra capacity.

 

[nd0586]

Thanks All!

CivEng80 - I'd be more worried about the stiffness throwing the plastic hinge hierarchy, not the capacity. Say you got a couple of degrees slip out of the splice - wouldn't that throw your calcs? Or if the splice was significantly stiffer that the rest of the section, again that might throw your hinge formation away from an assumed position?

 

[KootK]

Remember that the deformation of a member is the result of the integration of M/EI along the entire length of the member. A localized stiffness discontinuity won't have much of an impact on the overall behaviour. I do worry about the connection slip. However, if your splice involves a fair number of pretensioned bolts, I think it should be okay.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[hokie66]

You need not develop the full member capacity at the splice, just a reasonably conservative amount. And remember to brace the bottom flange near the splice location. I wouldn't be concerned about the deflections being affected by joint slippage. Just make sure the bolts are tightened.