Integrated Buildings & Structures

Our bridge comprises of 5 Nos. steel plate girders, each of 61.4m length and 60.8m c/c of bearings + 225mm thk concrete deck slab. Overall Bridge width is 11.15m. Each girder weighs 140 tons including all the steel work.

Deck cross section comprises of 5 plate girders at 2125mm c/c with two cantilevered concrete footpaths of 1325mm. Deck cross section has cross gradient of 2.5% on each side from ridge located along center line of bridge.

60.8 span is made of 7 parts of 2 end parts of 5400mm each and 5 parts of  10000mm each. Each part is welded in factory in China with camber of 170mm at center. After erection in position parts are bolted by HSFG bolts. Camber provided at center is 170mm. 

After bolting, intermediate temporary supports used during erection were removed. It was noted that all 5 girders deflected differently as follows:

Deflections at center span after several rounds of surveys by different parties are:

Two edge girders: 54mm & 177mm
Two pen-ultimate girders: 68mm & 138mm 
Center girder: 77mm

We propose to rectify increased deflections as follows:

1. Prior to concreting the deck we plan to raise the girders by jacking up one end girder and one pen-ultimate girder to match deflections with corresponding girders.
2. Provide jacks under all girders during concreting.to maintain levels. 
3. Commence concrete from both ends of bridge simultaneously.
4. Maintain the jacks in place during full curing period.

Comments/suggestions welcome.



------------------------------
Shashank Pendharkar Aff.M.ASCE
Pendharkar & Associates Ltd
Dar Es Salaam
255754461650
------------------------------

Dear Shashank,
 
Before looking into the next construction procedure, we need to understand why the deflection differential is so large? Is that due to some deficiency in the girders? Last time I had something like that (it was a different kind; prestressed precast girders) it turned out that some strands were not stressed. The girders were discarded.

In this case, what possibly caused it? Slippage in the joints coupling the girder pieces? Or inferior material? Or wrong plate thicknesses? Or …?

By the way, using so many pieces per girder, relies heavily in the connections. The connections at 5m from the centre will be carrying 97% of the maximum bending moment; a high risk situation. 

On the other hand, could the camber be faulty at manufacturing?

Do the girders have the required strength? If that was my bridge, I would have demanded performing load/deflection test.

If the above was checked and found safe; say,  the problem was only in faulty camber at production, then construction procedure can be discussed. 

The 5 steps you listed are reasonable, except that step 1 needs to supported by analysis: on a 3D model, check the effect of the girders that were jacked up on site, being loaded DOWN by the force that was used in jacking. That's because those girders will try to go back to their original shape before site-jacking. On that 3D model, check the stresses and deflections of the concrete slab as the released girders will try to rebound back; adding to the stresses in the slab.

If that passed the 3D analysis check, then the remaking 4 steps you listed are good. In my steel bridge construction, I made sure that the whole concrete deck was poured before any part of it was set; for a long bridge, I used concrete cure retardant so that setting happened after the end of the whole pour. 

Good luck.

------------------------------
Neil Kazen, M.Eng., M.Sc., P.Eng.
FASCE, FCPCI, FEC,
Retired Structural Engineering Manager, Transportation Division, SNC-Lavalin
Toronto, Ontario, Canada
------------------------------