When to Split a Machine Base Into Bolted Sections

Why an oversized machine base gets built as bolted sections instead of one piece, what the joint costs in stiffness, and how alignment is held across the split on site.

Most machine bases ship as one welded or cast piece. Above a certain size, that stops being an option, and the design decision becomes how to split the structure rather than whether to.

What forces the split

Three limits compete, and the tightest one governs.

Shipping envelope. A standard 40-foot flat rack handles cargo up to roughly 12 metres long; beyond that, or beyond typical out-of-gauge width and height allowances, the load moves to break-bulk or project cargo terms with different routing, cost and lead time. Road transport from port to site adds its own ceiling: width, height and axle-weight limits vary by country and by permit, and they are frequently tighter than the ship-side limit.

Shop capacity. A fabricator’s crane hook height and safe working load, planer-mill or boring-mill table length, and stress-relief furnace internal dimensions all cap what can be built and finished as one piece, independent of how the finished part would ship. A base that ships fine can still be too long for the machine that has to finish it.

Handling at destination. Even a base that clears both of the above can exceed what the buyer’s own crane, forklift or rigging crew can safely place in one lift.

Any one of these can force a monolithic design into sections; check all three before committing to a build strategy, not just the one that seems most obviously limiting at first glance.

The stiffness cost of a joint

A continuous weld transmits bending and shear as if the material were one piece, because at a structural level it is. A bolted flange joint cannot do that as cleanly: contact faces have compliance, and a joint loaded predominantly through bolt friction can creep under reversing loads. The assembled structure will always be somewhat less stiff at the joint than an equivalent continuous section; the practical question is how much that matters where the split falls.

Two decisions control the loss:

• Where the split falls. Locating it near a point of low bending moment, rather than at the point of highest load, costs far less stiffness for the same joint design. This is a placement decision that belongs in the same load-case analysis used for rib design and stiffness targets, not an afterthought once the split location is already fixed by convenience.
• How the joint is made. A generous flange with fitted (reamed) bolts on a tight pitch, plus dowels sized to carry shear in bearing rather than relying on bolt clamping friction, recovers most of the practical stiffness. A thin flange with a handful of clearance-hole bolts does not, regardless of how many bolts are used.

Match-machining: the step that makes it work

The mating faces of both sections are machined together, in the same setup, at the fabricator’s shop, not machined separately to a shared drawing tolerance and assumed to fit. Dowel holes are reamed at the same time, through both halves, so the pin locations are guaranteed rather than merely toleranced. The two sections are then match-marked, disassembled for shipping, and reassembled in the same relative position on site.

This is the difference between a joint that is proven before it ships and one that is hoped to work once it arrives. A fitted joint made from two independently machined faces, each within its own tolerance, can still stack those tolerances into a gap or misalignment that a crew has to correct in the field with shims and patience. A match-machined joint removes that variable before the crate is built.

Site assembly sequence

1. Set and level the first section on the foundation, checked against the same flatness targets used for a one-piece base (see levelling and grouting).
2. Crane the second section into position using the match-marks as the primary alignment reference.
3. Insert dowels, then draw the fitted bolts to specification in a staged, cross-pattern sequence rather than fully tightening one side first.
4. Re-check flatness and alignment across the joint against the drawing tolerance before proceeding to final grouting.
5. Grout only once the assembled structure passes the same checks a monolithic base would have to pass at the same stage.

Deciding early, not late

The decision to segment a base is cheapest when it is made at the design stage, where the split location can be chosen for load, and most expensive when it is discovered after a monolithic drawing is already committed and the shipping quote comes back as project cargo. If a base is approaching either the shop or shipping limits above, raise it during RFQ rather than after quoting, along with the drawing referenced in the RFQ checklist: segmenting is a design change, and it is far cheaper to make on paper than on a crated part already at the port.

Common questions

← All guides