Marius Comper RO

The load test of 12 June 2023 · the 1,120 m main span

The Brăila bridge under 2,400 tonnes

Before it opened to traffic, 60 trucks of 40 tonnes each stood 32 metres apart along the whole length of the suspension bridge over the Danube at Brăila, in eastern Romania, on the carriageway on the north side of the bridge. According to the builder, the bridge "went down, overall, about 60 centimetres". For the same layout, the calculation for a cable pulled taut by the bridge's own weight gives a drop of about 35 cm at the deck centreline and 46–59 cm at the cable on the loaded side. The reported figure is closer to the loaded side's range.

Which point to track

Trucks go on the north carriageway. The calculation holds the towers fixed and ignores the deck's stiffness; the drop is drawn exaggerated.

The June 2023 test

On 12 June 2023, ahead of the July opening, the builder, the Webuild–IHI joint venture, began the load tests of the suspension bridge over the Danube at Brăila. According to Irinel Ionel Scrioșteanu, a state secretary in Romania's transport ministry, 60 trucks of 40 tonnes each were placed "32 metres apart, along the whole length of the bridge on the north carriageway", because the engineers were also testing how much the deck twisted. Four GPS stations and a system of 50 sensors fitted to the bridge tracked how it moved.

Two days later the local paper Obiectiv, citing the builder's engineer Viorel Eșanu, wrote that "the bridge went down, overall, about 60 centimetres", and Romania's public radio reported that the sensors had shown 60 centimetres. Neither report says where it was measured; IHI has published only a summary of the tests, without the measured values.

The spans, cables and deck

The main span is 1,120 m and the side spans 489.65 m on the Brăila bank and 364.65 m on the Tulcea county bank; on all three, vertical hangers connect the deck to the main cables. The two cables, 25.5 m apart, pass over the towers at +192.640 m and come down to +62.640 m at midspan: a sag of 130 m. In the main span each cable has 8,704 steel wires of 5.38 mm and is 56 cm across; in the side spans, more than 9,000.

The steel deck, in 86 segments, weighs about 21,000 tonnes (IHI; the state secretary gave 20,860), the 11 cm asphalt surfacing another 12,000 tonnes according to IHI or 14,000 according to Scrioșteanu, and the cables about 6,800 tonnes. The dead load in the calculation is about 22 tonnes per metre of bridge, cables included. At midspan each cable is pulled with about 13,200 tonnes-force, a stress of about 654 MPa, roughly a third of the wire's 1,860 MPa strength.

How far it should have dropped

The trucks added 1.25 tonnes per metre, about 6% of the bridge's own weight per metre. At 32 m spacing, 35 of them would have stood on the main span. In the cable on the loaded side the horizontal pull rises by about 1,100 tonnes-force; the cable stretches by about 33 cm and, for the longer cable to fit between the towers, sags further at midspan.

For the announced layout, the calculation gives about 35 cm at the deck centreline and 46–59 cm at the north cable, depending on how close to the edge the trucks stood in their lane; the south cable drops 12–25 cm and the deck tilts. The announced 60 centimetres or so are nearly double the centreline value and at the top of the range calculated for the cable on the loaded side. Cables 5 °C warmer would lower the middle by about 11 cm; if the reading without trucks was taken at a different temperature from the one with them, a few degrees could account for part of the gap. The temperature during the test was not published.

0.0 m0.2 m0.4 m0.6 m0.8 mdeck centrelinenorth cable, loaded side“about 60 cm”thick bar: at the reference temperature; thin bar: cables up to 5 °C warmer or colder
The calculation for the announced layout, with 35 trucks 32 m apart on the main span, all on the north carriageway: the deck centreline drops about 35 cm, and the cable on the loaded side 46–59 cm, depending on how close to the edge the trucks stood. A 5 °C difference in cable temperature moves the middle of the bridge by about 11 cm.

How its weight keeps it stiff

A suspension bridge is stiff mainly because it is heavy. A single 40-tonne truck on the north carriageway at midspan lowers the cable on its side, in the calculation without the deck's stiffness, by about 18 cm; with an estimated deck stiffness, by 14–16 cm. If the same cable carried a bridge 30 times lighter, the same truck would lower it by about 5 metres: the cable would follow the load, like a rope with a single object hanging from it.

See how far drivers would have to go if a bridge closed. The railway bridge at Cernavodă, another Romanian town on the Danube, had its own load test in 1895.

0.1 m1 m10 m5%10%25%50%100%the bridge’s own weight, as a share of the real one · drop under one 40 t truck at midspan
The same cable, with the same steel section, under a single truck, if the bridge weighed less: the lighter the structure, the more the truck changes its shape. The dot shows the calculated drop at the bridge’s real weight. The curve shows how it would change at lower weights, leaving out the deck’s stiffness, which would matter for a very light bridge; it is an illustration.

What is not known

Where the drop was measured and the cable temperature during the test are not published; they decide how well the calculation fits. The calculation holds the towers and saddles fixed and uses an estimated weight for the deck fittings; how many trucks stood exactly on the main span is inferred from the 32 m spacing. The page compares a figure announced in the press with the physics of a cable; it does not judge the bridge's safety.

Share this result

The link reopens the bridge with the same trucks, cable temperature and tracked point.

Sources and method

The calculation uses first-order deflection theory for a parabolic cable: the horizontal pull from the dead load, H = wL²/8f, and an extra pull h from the trucks, chosen so that the cable's elastic stretch is exactly what the deflected shape requires. Inputs: 1,120 m main span, 130 m sag (levels +192.640 at the tower and +62.640 at midspan, from IHI's drawings), 8,704 wires of 5.38 mm per cable in the main span (steel area 0.198 m²), steel elastic modulus 195 GPa; steel deck 21,000 t and surfacing 12,000–14,000 t over 1,974 m, cables 6,800 t, plus 1–2 t/m for other fittings (estimated). The load is shared between the two cables, 25.5 m apart, by the trucks' position on the north carriageway (its centre is 6.25 m from the bridge axis). The calculation assumes that the cable saddles on the towers stay put; the side spans were loaded too, but their effect on the saddles is not calculated. The main calculation leaves out the deck's stiffness; a separate calculation with an estimated stiffness changes the result for the 2023 test by less than 1%.

  1. I. I. Scrioșteanu, quoted by HotNews, 12 June 2023: truck layout (Romanian).
  2. Obiectiv, 14 June 2023: "Podul rezistă" (V. Eșanu); Radio România Actualități, 15 June 2023: bridge tests (Romanian).
  3. IHI Infrastructure Systems, bridge presentation at the APDP congress: general arrangement drawings, masses; IHI Infrastructure Technical Report vol. 12, §4.1 and §4.3 (surfacing, load tests).
  4. JASBC construction report, table 3-1 (cable wires, archived copy); Webuild, main cables.