Column Splice Connection Calculator

Design a column splice - the bolted cover-plate joint that gives continuity between two columns - to Eurocode 3 (EN 1993-1-8) with SCI P358 §6. Choose a bearing splice (the prepared ends transfer compression in direct contact, through a division plate where the serial sizes differ) or a non-bearing splice (the cover plates and bolts carry the full axial force and moment). The calculator runs the net-tension test (M_Ed vs N_Ed,G·h/2), then verifies the flange cover plates and their bolt group (with the packing and long-joint reductions), the web cover plates, the 25% minimum-resistance rule and the structural-integrity tie - each with its utilisation and a clear PASS/FAIL on an interactive 3D model.

Free online calculator with step-by-step working and one-click Excel (.xlsx) and PDF export.

Using this column splice calculator

What it checks

A column splice joins two column lengths, typically just above a floor level. This calculator follows EN 1993-1-8 and SCI P358 for both bearing and non-bearing splices: it determines whether net tension develops across the joint, then checks the flange cover plates and their bolts, the compression path, the web cover plates and bolts, the minimum-resistance requirement and the tying resistance for robustness.

The critical question it answers first is whether the splice ever goes into net tension. In a bearing splice the two column ends bear directly on each other, so compression passes through the contact rather than through the cover plates - and if the axial load is large enough relative to the moment, no part of the section ever goes into tension at all.

When no tension develops, the 25 per cent rule governs

Running the tool on its defaults - a 305 x 305 S355 column, external bearing splice with 12 mm S275 cover plates and M20 bolts - gives:

Net tension develops?
No
Moment at which tension would start
235.9 kNm
Net tension force
0 kN
Minimum-resistance bolt group
800 / 1129.0 kN, utilisation 0.709 - GOVERNS
Minimum-resistance cover plates
800 / 1980.0 kN, utilisation 0.404
Flange cover tension
0 kN - not engaged
Lever arm between flanges
314.5 mm

Because no net tension develops, every tension-related check returns zero and would pass trivially. The splice is instead governed by the minimum-resistance requirement - the code rule that a splice must be able to carry a defined fraction of the member capacity regardless of the calculated forces, which the tool evaluates at 800 kN against a bolt group resistance of 1129 kN.

This is the single most misunderstood aspect of column splice design. A designer who checks only the applied forces will size the splice far too lightly, because in a well-loaded column those forces are often zero across the cover plates. The splice still has to exist and still has to be substantial - the rule exists so a column is not left with a joint that has no reserve against the unforeseen.

Bearing versus non-bearing

In a bearing splice the column ends are prepared so they make contact and compression passes directly across the joint. The cover plates then carry only tension, shear and the minimum-resistance requirement. In a non-bearing splice - where the ends are not prepared to bear - the cover plates must carry the full compression as well, which makes them substantially heavier.

The tool asks which type you have because it changes the whole load path, not just a factor. Bearing requires the ends to be sawn or machined flat within a tolerance and erected in contact, which is a fabrication and site requirement, not something the calculation can assume for you.

Formula reference

Moment at first tension
Mnotension=NEdh2M_{no\,tension} = N_{Ed}\,\dfrac{h}{2}
Net-section tension
Nu,Rd=0.9AnetfuγM2,netN_{u,Rd} = \dfrac{0.9\,A_{net}\,f_u}{\gamma_{M2,net}}
Gross-section tension
Npl,Rd=AfyγM0N_{pl,Rd} = \dfrac{A\,f_y}{\gamma_{M0}}
Minimum-resistance demand
Nmin=0.25NEdN_{\min} = 0.25\,N_{Ed}
Bolt shear
Fv,Rd=αvfubAsγM2F_{v,Rd} = \dfrac{\alpha_v f_{ub} A_s}{\gamma_{M2}}
Bearing
Fb,Rd=k1αbfudtγM2F_{b,Rd} = \dfrac{k_1 \alpha_b f_u d t}{\gamma_{M2}}

Note gamma_M2,net = 1.1 rather than 1.25 for the net-section tension check - a distinct partial factor that is easy to get wrong and was corrected during verification against SCI P358.

Assumptions and limits

A bearing splice assumes the column ends genuinely bear. That requires the ends to be prepared flat within tolerance and erected in contact - if a gap remains, or if packing is used carelessly, the compression path does not exist and the splice behaves as a non-bearing one it was never designed to be.

The tool checks the splice, not the column. Column buckling either side of the joint, the effective length implied by the splice location, and the frame stability that depends on it are separate design questions. Splice location is usually a few hundred millimetres above a floor for access, which is a practical choice that also keeps the moment low.

Tying resistance for structural robustness is checked as a separate requirement with its own load case, and it can govern the detail in a building where accidental-action design applies. Fatigue, fire and seismic detailing are outside the scope.

FAQ

Yes. The minimum-resistance rule requires the splice to carry a defined fraction of the member capacity regardless of the calculated forces - 800 kN in the worked example, where the calculated tension is zero. Sizing a splice from the applied forces alone is the most common error in column splice design.

In a bearing splice the prepared column ends make contact and carry the compression directly, so the cover plates handle only tension, shear and the minimum requirement. In a non-bearing splice the cover plates carry the full compression too, and are much heavier as a result.

When the moment is large enough relative to the axial load that part of the section goes into tension - beyond N_Ed h/2, which is 235.9 kNm in the example. Below that the whole section stays in compression and the cover plates carry no tension at all.

Because EN 1993-1-8 uses a distinct gamma_M2,net for that check rather than the 1.25 used for bolt resistances. It is easy to apply the wrong one, and doing so was a real bug caught when the engine was verified against SCI P358.

Typically a few hundred millimetres above floor level, for erection access. That also places it where the column moment is usually small, which is convenient - but the location affects the effective length assumed for buckling, so it is a framing decision rather than a purely practical one.

Yes - against SCI P358 Section 6 worked examples, with 21 of 21 assertions passing. The verification caught the net-section partial-factor error noted above.

Rate this
No ratings yet
Sign in to join the discussion.
Loading…