Sinotruk Intelligent Connected Heavy Truck Assembly Workshop

Sinotruk — China National Heavy Duty Truck Group — is the country’s oldest and one of its largest heavy-truck manufacturers, producing trucks whose frames, cabs and drivelines are assembled in long linear process lines. This project delivers a 77,000 m² intelligent connected heavy truck assembly workshop: the shell within which chassis frames travel through frame stations, cab marriage, driveline installation and final engineering checks without once being forced to deviate around a column.

The roof system engineered, fabricated and erected by Permanent Steel is a double-layer planar grid — two parallel chord planes connected by diagonals — configured on an orthogonal module and supported at the bottom chord level. The selection of bottom-chord support was not an aesthetic choice or an afterthought; it was dictated by the process inside. This case study examines the physics and the economics of that decision, the integration of heavy crane loads into a three-dimensional grid, and the joint strategy (bolted spheres for the distribution network, welded spheres for the load skeleton) that kept programme and integrity in balance.

Truck assembly is distinguished from passenger-vehicle assembly by scale and weight. A heavy-duty chassis line handles frames up to 12 metres and gross weight classes in the tens of tonnes; the tooling is massive; overhead cranes lift complete cab and powertrain assemblies; and the same line must accommodate multiple truck platforms with minimal reconfiguration. None of that is possible behind a forest of interior columns. The building had to be, in effect, one enormous unobstructed volume — and the steel had to deliver it economically enough to justify a grid structure over a conventional frame on a construction budget measured with engineering discipline.

A double-layer grid can be delivered to its supports in two fundamentally different ways. In top-chord support, the columns rise through the depth of the grid to connect at the upper chord plane; in bottom-chord support, the column caps terminate at the lower chord plane and the grid bears directly on them. The difference sounds like a detail. It changes the entire structure.

Column economy

Supporting at the bottom chord removes the grid depth from the column length. With a grid depth of roughly 3–4 m at this span class, every column is 3–4 m shorter — and in a 77,000 m² hall with perimeter and internal column lines at close spacing, that is thousands of tonnes-metres of column material removed, reduced buckling length, smaller foundations and reduced wind moment. The saving is real and it is distributed across every support in the building.

Clear height and service integration

The clear height that matters to a production engineer is measured from the finished floor to the lowest obstruction — which in a grid building is the bottom chord plane. Bottom-chord support keeps that usable datum at its designed level without forcing longer columns; the grid depth effectively sits above the support plane rather than consuming the usable volume. Suspended services, crane runway brackets and mezzanine structures all attach to the column lines below the chord plane, where the geometry is simple and the steel is straight.

Member and node coordination

In top-chord support, columns pass through the grid band and must be detailed to avoid conflicting with diagonal members on the way up — a coordination problem at every column, with structural and aesthetic consequences. Bottom-chord support eliminates the conflict entirely: the column cap bolts onto the bottom-chord node, the diagonals above it are uninterrupted, and the load path into the column is direct and concentric. The connection detail at the cap is simple, inspectable and standardised — exactly the attributes that keep an erection programme honest.

When is top-chord support still right?

Intellectual honesty requires the caveat: top-chord support is not always wrong. It is typically chosen where the grid must be held high above its base for structural reasons — for example when the roof plane is deep enough to justify a Vierendeel-like interaction, or when architectural expression wants the grid hovering above glazed walls. For a heavy industrial assembly hall, none of those motives apply, and the engineering case for bottom-chord support is unambiguous. The building is 3–4 m of columns lighter for it, and the process space is exactly as deep as the process needs.

The hybrid joint philosophy applied on this project is the same one Permanent Steel applies across its large industrial grids, and it deserves an explicit defence because it is frequently misunderstood as a compromise between two competing technologies. It is not a compromise; it is an optimisation with two variables — connection capacity and connection cost — and the two technologies occupy different parts of both scales.

The bolted population

Bolted sphere joints dominate the interior of the grid: machine-turned spheres with tapped faces, connected to member-end conical fittings by high-strength bolts torque-controlled on site. They are the fastest, most weather-tolerant and most measurable connection in industrial steelwork — a bolted connection is verified by controlled torque, not by welder craft. In a grid with tens of thousands of joints, the programme effect of bolting the ordinary population is the difference between a twelve-month roof and a twenty-month roof.

The welded population

Welded sphere joints take over where member forces exceed the efficient envelope of any practical bolted node — the high-compression chord zones near supports, the crane-reaction nodes, corner and edge regions with unbalanced forces, and every junction between the grid and its support columns. A welded node transfers force through continuous material with no bolt-group limit; its capacities track the member capacities instead of capping them. Every welded node on this project was produced in our workshop or by certified site welders, with 100% NDT coverage as specified, because a welded joint’s integrity belongs to the weld — and the weld is never left to chance.

The optimisation rule

The practical engineering rule we applied: classify every node by its worst-case force envelope; bolt where fasteners carry it with reserve, weld where they cannot. The result is a structure whose welded skeleton guarantees global rigidity exactly where rigidity is needed, and whose bolted majority buys speed, substitutability and ductility everywhere else. This is the engineering answer to the false question of “bolted or welded?” — and the cost model of this 77,000 m² project is the evidence that it is the right answer at industrial scale.

The governing cases of a heavy-truck plant are not exotic, but they are unforgiving: deflection must not disturb conveyor alignment or crane runway level; seismic response must exploit — not fight — the grid’s redundancy; and thermal movement over continuous bays of several hundred metres must be absorbed without locking stresses into the structure.

• Deflection and camber: the grid was analysed for the full envelope of dead, live and crane-induced loads; pre-camber was built into the fabrication geometry so the soffit reads level under dead load and service deflections remain inside the code envelopes that protect cladding and suspended services.
• Seismic: the two-way grid’s high redundancy and ductile bolted population give it a favourable seismic energy-dissipation profile. The design followed GB 50011 with the structure modelled as a three-dimensional coupled system — grid, columns and craneway — so that no member was checked in isolation from the system it works within.
• Wind and thermal: large roof planes attract pronounced uplift and suction; uplift-critical connections were verified for tension, and expansion provisions at the support system allow the structure to breathe over temperature cycles without accumulating stresses in the chord planes.
• Crane-related durability: details subjected to repeated crane cycles were proportioned beyond the static checks, with weld geometry and connection stiffness selected to control stress ranges at the fatigue-sensitive locations.

Completion verification followed the same discipline: independent surveying of node coordinates against the design model, camber confirmed after de-propping, and a documentation package that ties every member, weld, bolt lot and coating batch to its inspection record. That package is not bureaucracy; it is the evidence on which a production-critical building is trusted for the next four decades.