Measuring bridge member stresses in-situ

[AlanLord]

I have to measure the existing stresses (tension or compression) of bridge cross-bracing members on an existing bridge under dead load, preferably in a non-destructive manner. There is a bewildering range of methods and equipment out there to accomplish this. What would be the easiest, cheapest way to accomplish this?

 

[Ron]

Alan...check out residual stress gauges. Check with MicroMeasurements division of Vishay Instruments.

 

[AlanLord]

Thanks Ron!

 

[MotorCity]

Since you are interested in the dead load stresses only, does that mean that the bridge will be closed to traffic while these measurements are taken (i.e. no live load)?

If that is the case, the method selected may depend on your allowable time frame to take the measurements (set up equipment, take readings, breakdown equipment, etc)

In any case, my first thought was to install strain gages

 

[jayrod12]

The issue with strain gauges is they don't account for any of the current strain that has happened to that point. Your beam may be loaded to 75% capacity already and the strain gauge wouldn't tell you that until it reached a "premature" yielding.

 

[rb1957]

i don't think you can stick strain gauges on an existing structure and measure dead loads (they're already acting when the s/gauge is installed). the problem is finding the zero.

maybe install the strain gauge, then add some extra load, get a difference reading, now go to your FEM and see what happens there when you add the same load.

or add some load that represents some proportion of dead load, i'd do several tests.

another day in paradise, or is paradise one day closer ?

 

[JStephen]

I wasn't aware that such a thing could be done.
With the "residual" stress gauges, I would think there's potentially a good bit of residual stress in an "unstressed" member, so I'm not sure how you'd separate the residual stresses from dead load stresses.

 

[AlanLord]

Here's a bit of background: the steel for 2 skewed slab on girder bridges has been erected, and the steel is presently bare while we investigate why the girders have all been installed crooked. The fitting was forced together, and one of the main girders was jacked longitudinally to correct placement; I'm trying to determine the locked-in stresses included with the dead load forces that are present in the x-bracing members, for code checking purposes later at ultimate. Is a residual stress gauge going to be able to give me the existing load in a x-bracing member?

 

[jayrod12]

was it jacked into position after the bracing was installed? I would expect not. So there may be residual stresses in the beam itself but not the bracing members.

JStephen makes a good point about residual stresses, there are likely a bunch present on an "unloaded" girder just from fabrication. If it's bare steel it shouldn't be too difficult to get an accurate stress due to current dead load.

 

[rb1957]

how would you get the stress due to the dead load (already applied) ?

i've never heard of a residual stress strain gauge. i looked at the Vishay site, didn't see anything there ? didn't see anything on the micromeasurements site ... a little s/gauge info buried in their "KIS beam", but nothing yelled "residual stresses".

another day in paradise, or is paradise one day closer ?

 

[AlanLord]

Yes, the girder was jacked AFTER the x-bracing was in place.

 

[jayrod12]

Would it not be cheaper to remove and replace the cross bracing? Just a thought, considering what the costs of testing and analysis may be. Fabrication cost should be minimal, and I would also assume they could likely reuse the majority of the material.

Still doesn't solve the current stresses on the girder, but I would bet that if you removed all of the bracing, that the out of plane bending stresses would then be minimal (provided it never reached any plastic deformation).

 

[dhengr]

AlanLord:
There are strain gauge methods for measuring some residual stresses. Strain gauges normally measure strains/stresses that occur after the gages are applied, zeroed out, temperature compensated, etc. They don’t know or help very much as regards what happened stress wise prior to their installation. I think you would be better off (do better by) studying and trying to understand what the girder jacking/moving process did to the various components. How did that inch or two of girder movement change the geometry, and what does that movement do the various members. You can take a good, sound, stab at analyzing that. To a great extent these will be localized stresses, strains, deformations in the immediate areas of connections, they peter out fairly quickly as you move away from the connections. There will likely be some localized yielding (plastic deformation). They could be ongoing fatigue problems on down the line. This analysis is required to even know where you might want to put strain gauges. It is also required for you to know how to study and monitor this situation. This study might also help you determine/see how you can modify things to minimize the problem. What details can be changed easiest to mitigate the problem.

 

[AlanLord]

It's becoming clear to me that no method exists to give the in-situ member stresses; I'll have to ask the erector for his precise sequence of installation, and add up the stresses for each member manually in a spreadsheet, and that this at best will be an approximation of the actual stresses.

Thanks all.

 

[BridgeEI]

Has anyone checked to see what the fit condition of the girders was supposed to be? Depending on the design or fabricators specified fit, girders can be laying over before the deck is poured.

 

[BUGGAR]

Let's see if I can make this clear. Pick the highest dead load stressed part and location. With only the dead load on the structure, start cutting the part at that location with a torch. When the depth of cut is deep enough that the part goes into yield, as measured by advanced deflection, measure the section modulus of the remaining section. Now you know the dead load yield vs. cut section modulus. From here you can back into the dead load stress in the uncut structure. Jack it up, weld it all back together, and send the Contractor a bill.