Multispan bridge kinked girder erection (apologies for the repost)

[gwynn]

I posted this a couple of weeks ago in the Brige forum, but as that forum seems nearly dead and I have had no responses since then I was hoping it wouldn't be too frowned upon to repost it in here in the hopes (yeah, hope, bad thing to do as an engineer) that it would garner some response in the moretravelled areas of the board.

I have been doing a fair number of steel girder erection plans lately and I have one for a multi-span bridge with kinked girders coming up a couple of months from now. A total of five spans and eight kinks with each kink at a field splice (one kink in each of the end spans and two in each of the interior). Eight different girder sections in total, with the section changes and splices at approximately the same location on the interior and exterior girders (four girder lines and the section changes and splices look to line up radially on the curve of the bridge deck). The section changes include an increase in girder depth from ~5' to ~8'. The piers run square to the girders and are along the radius of control line curvature, but the abutments are skewed.

The bridge has diaphragms located either side of each kink. All of the diaphragms are detailed to fit under total dead load and the camber differential between girders is significant. I know this will cause problems for the erection.

At first thought I see a few erection options:
1. Erect the girder segments on shoring and twist/layover the girders enough to install the diaphragms. This will result in the flanges and webs of adjacent girder sections being out of plane when installing the splices and require jacking on the shoring or other means to twist the connected set of girders back to a near plumb condition before splicing.

2. Erect the girder segments on shoring with the minimal amount of bracing/diaphragms required for stability. Assemble the splices and then install the diaphragms. This will increase the forces required to orient the girders in the geometry required to install the diaphragms, as torsion on one girder segment would now be partially resisted by strong axis bending of the girders.

3. Similar to option one, but leave out the diaphragms nearest the splice locations. This would require a combination of the means that would be needed in options one and two, but may lower the demand required for each.

I would appreciate any thoughts on how to deal with this situation.

 

[SlideRuleEra]

gwynn - Its been a long time (35 years) since I worked in the bridge construction industry, but I do have some general thoughts on how to approach this problem:

There is not much information on erection of (horizontally) kinked girders, that I could find, and I don't have first hand experience with them. Since kinked girder have the same type of stability issues as (horizontally) curved girders, use available info on curved girder erection as a source for your investigation of how to proceed. Fortunately, erection of curved girders has been well studied. Much of the research is academic, but you probably can get insight and ideas from it. For example, the Texas DOT / U. S. Federal Highway Administration technical report "Guidance for Erection and Construction of Curved I-Girder Bridges" may be a good place to start:

http://www.utexas.edu/research/ctr/pdf_reports/0_5574_1.pdf

Also, take a look at erecting the girders in pairs. The pairing of girders is a good way to reduce stability issues.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[paddingtongreen]

I'm not a bridge man but something here bothers me.

If the two beams connected at the kink are horizontal, the cut ends are single angle. If they are not horizontal when forced into their deflected shape when connected, the cuts must be compound angle cuts. If they are compound angle connections they would seem to negate the geometry of the original calculations (they would not spring back to the original design geometry if the loads were removed).

If they are using some kind of equivalent the "cold spring" used in piping, they should say so.

Now to find out that I have attacked something endemic to bridge building.

Michael.
"Science adjusts its views based on what's observed. Faith is the denial of observation so that belief can be preserved." ~ Tim Minchin

 

[gwynn]

Sliderule:

I have been going through the Tenesee DOT and NCHRP reports on curved and skewed girder bridges. NCHRP report 725 deals extensively with the issues associated with detailing the diaphragms for total dead load fit. You are correct that I should be able to transfer much of this over to the horizontally kinked situation. I don't think either specifically deal with splice fit up though, it seems this is more an issue for kinked girders only. I will do some more digging and see if I can find anything that specifically addresses this.

The stability issues will likely be resolved with shoring towers. I think erecting the girders in pairs may make things more difficult in this case, both due to girder weight and the difficulty of adjusting the cross slope of a pair of girders versus a single girder if the diaphragms do not line up exactly right.


Paddington

I don't think you could be attacking anything endemic. Horizontally kinked girders are not very common in my experience.

I would have to check the design and shop drawings to be sure, but I believe the cut ends are a single angle (detailed to fit once all of the camber has been removed).