Steel Portal Frame: Connection Stiffness and the Assumption of Pinned Bases
A 24 m span steel portal frame warehouse has 610 x 229 mm rafters and 406 x 178 mm columns. The designer has assumed pinned bases in the global analysis, giving a sway stiffness that depends entirely on the frame action. The base plates are 500 x 400 x 25 mm with four M24 holding-down bolts in 50 mm diameter holes, grouted after erection. The frame is in a region with basic wind velocity 25 m/s and the eaves height is 7 m. During erection the contractor reports that the base plates are bearing on only part of the foundation and that some bolts are 15 mm out of position, requiring slotted holes.
1. Does the assumption of a pinned base remain valid given the observed base plate conditions, and what is the consequence for frame stability?
Read the worked answer
The pinned-base assumption is a design idealization; the real base has some rotational stiffness from the base plate bearing and the bolts. The question is whether the actual stiffness is close enough to the assumption that the analysis remains valid. If the base plate bears on only part of the foundation, the contact area is reduced and the plate may rock, which reduces stiffness and increases sway. Slotted holes allow slip, which further softens the connection.
The consequence is that the frame's sway stiffness may be lower than analysed, increasing second-order effects. Check the sway sensitivity: for a 7 m eaves height and 24 m span, the frame is likely sway-sensitive, and a 10–20% reduction in base fixity can increase the amplified sway moment noticeably. The correct response is to require full grouting under the base plate and to verify the slotted holes are designed for the bolt slip, or to re-analyse with a nominally fixed base if the foundation can take the moment. Do not accept partial bearing as a minor issue.
2. The contractor proposes welding the base plate to the foundation reinforcement to 'make it fixed'. Evaluate this proposal.
Read the worked answer
Welding the base plate to the foundation reinforcement does not make the base fixed; it makes the connection unpredictable. The reinforcement is not designed to transfer a moment into the foundation, and the weld quality and fatigue performance are unknown. A truly fixed base requires a designed moment connection: a base plate thick enough to distribute the bolt tension, holding-down bolts sized for the uplift and shear, and a foundation designed for the resulting moment and bearing pressure.
If the designer wants a fixed base, the correct route is to re-analyse the frame as fixed-base, check the foundation for the moment (which may be large at 24 m span), and detail the base plate and bolts accordingly. If the pinned assumption is to be kept, the base must be detailed to allow rotation without distress - a nominally pinned detail with a single row of bolts or a recessed plate. The proposal should be rejected as it introduces a load path that has not been designed.