Buckled steel roof trusses 1

[Teguci]

What repair would you recommend for this situation?

Existing 100 ft span steel roof trusses (fink with monitor) have been exposed to a high temperature and are buckled. The trusses are laterally braced and have only sagged maybe 12". Clearance is not an issue. Buckles occur in the top and bottom chords. The chords (double angles) are twisted and bent out of plane but still hold load. Replacement is not an option so be creative in ruling out future failure modes.

 

[ztengguy]

Hmm, gonna be tricky one. Anyway to weld plate on the sides, or sister up other trusses? Pictures available?

 

[Teguci]

Sistering will be tricky due to interferences with lateral bracing, etc.

This is hypothetical, no pictures.

 

[dik]

You can model the trusses with the 'deformations' and see how this would affect the compression chord and web members.

Dik

 

[MiketheEngineer]

I have seen trusses with this kind of damage. Theoretically, there should be a list of fixes that might work. But when tried - it just led to more problems. Eventually, replacement was the ultimate and best answer.

Sorry.

 

[ToadJones]

Well, if they have yielded then they have failed, no?

 

[SLTA]

For me, the key part was "only sagged maybe 12 inches". Only? At some point, they're just no longer good.

Why is replacement not an option?

 

[Teguci]

dik - so the tension chord is not a concern?

MiketheEngineer - thanks for your experience.

ToadJones - Lets call it plastic hinging.

slta - I don't want the really smart thinkers on this website debating a simple decision like replacement. There is a process we all follow before replacement becomes an option.

 

[BAretired]

The problem is hypothetical?

BA

 

[paddingtongreen]

I assume the buckled region was the same one that was heated. If so, the steel yielded at a much lower stress than at normal temperature.
I also assume that there are straight sections with the bends being short. I would be tempted to try the effect of adding diagonals from the bend points to the adjacent trusses. Make it look like part of the original design.
Other than that, investigate heat straightening. I say investigate because I haven't heard of it on double angles but an expert might know how to do it.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[dik]

Tension chord is less of an issue... biggest problems are the compression elements with an 'initial' eccentricity.

Dik

 

[Ron]

Hypothetical or not.....

First. I would inspect EVERY weld in the truss for cracking. When redistribution of stress took place during the plastic deformation, overstressing/cracking of the welds was likely.

Next. Metallography needs to be done in the area of heating. At the least, there is likely a carburized zone at the surface that can mask a significant change in grain structure/strength, so that needs to be checked before going on to the next step. Some metallography can be done in the field (depth of carburized layer, polish for grain and look at grain with portable microscope). Coupons should be removed for full lab metallography.

With the information from the metallography, check the in-place hardness with a portable hardness tester. Correlate with strength testing.

A combination of re-heating/jacking and welding reinforcement can then be employed to correct the issues.

It will not likely ever achieve original straightness. After the repairs, I would load test the truss.

 

[saplanti]

Ron is right about his concerns;

1. If steel exposed to high temperature you need to make sure it still contains the same mechanical and chemical properties. If does not you need to find out the current properties. If the exposure is local you may need to replace the exposed members if you are not sure about the properties in case the calculation looks for the stresses/forces to the capacity of the exposed members.

2. If the large displacement on the roof may be resulting with "SNAP THROUGH" type buckling after the exposure you need to replace it entirely.

 

[msquared48]

At only 12" deflection, L/100, being a fink truss configuration, snap-thru buckling of the entire truss is not likely until the centerline of the truss top and bottom chords falls below a horizontal line. Likely this does not play into the problem here.

It has to be metalurgical in nature as previously stated. Steel begins losing strength significatntly upon seeing 800 to 1000 degrees farenheight. Before that temperature is reached, it will tend to expand along its axis length, and after that it will start bending/buckling as the ability to take load is reduced.

A further key here are the buckles of the tension chord.

No mention is made of either welded or bolted connections in the original problem, so these factors do not play in either.

I assume that the truss end conditions are hypothetically pin-roller.

I assume that the buckling occurs at the panel points.

Mike McCann
MMC Engineering

 

[ishvaaag]

Once material concerns and damage to connections are dealt with, you may try to convert the system in a wholly 3D spatial structure with the addition of diagonal and parallel members, but you still will have to deal with any initial imperfections remaining; furthermore the stiffness of the original system should be of such stance that from a material employment might make scarce sense.

Yet it can help with dealing with initial imperfections standing since you can add fanning diagonals from hardpoints to forestall the worst impending deformations that, starting from the current shape, would be exacerbated by further loading to non tolerable values for the structural members.

 

[csd72]

Agree with most of the above this is definately a cause for concern and I imagine that proving the adequacy of the existing would be more costly than replacement.

There is more to the job than just structural mechanics, things like this are the reason why we are also taught materials and chemistry at university.

You need to sort this out in your own mind before you try and explain it to your client in plain simple english.

 

[dik]

In addition to the chemical changes that Ron noted, with high temperatures there is a significant reduction in the yield strength and Young's modulus. This gives rise to failure from lack of strength or buckling.

Because of the 'slender' nature of trusses, a lack of a complete collapse may indicate that temperatures may not have been severe. Can you check if there are any fire reports that may indicate the duration and temperatures?

The original design may have been very 'tight'. It is not likely that you can return your original trusses into somthing that may be similar to the original condition and your roof supporting system may be toast.

Dik

 

[steellion]

Sometimes the smartest solution is to punt. Too many variables come into play when trying to create such a fix, and the fix would likely be more expensive than replacement anyway.

 

[Teguci]

In summary, possible failure mechanisms that need to be controlled:
- Problem: Loss of elasticity (potential brittle failure) due to carburizing and change in grain structure due to high heat. Solution: Check affected members in the field, further test a coupon or rule this out based on historical information (length and type of high temp.)
- Problem: Failure of connections (welds were mentioned) due to overstress. Solution: Field verify every connection for cracks or breaks.
- Problem: Further buckling of the top chord and compression webs under load. Solution: Provide additional bracing, heat straighten the offensive members, or analyze members with deflections included.

Anything else?

 

[Ron]

Load test after repair