Channel Toe up or Toe Down? 1

[LowLax]

I have an existing runway beam with a cap channel. The runway needs to be reinforced to carry a new higher capacity crane, so I am adding a channel to the bottom flange to get the extra capacity.

When I did the calcs I had the bottom channel toe up so I had a doubly symmetric section (top and bottom channels are the same size). Now I'm thinking it's going to be much easier to weld with the bottom channel toed down.

Maybe I haven't had enough coffee this morning, but is there a problem with doing this? The I value would increase slightly, but there is such a small area at the extreme bottom fiber that I'm worried I am going to reach yield sooner.

 

[cntw1953]

The situation you described in the last sentence is my concern too. Can you replace the bottom channel with thicker cover plate to achieve the same result? Or a structural tee with web welded to beam bottom flange.

 

[miecz]

Yes, you'll reach yield sooner at those tips.

 

[miecz]

Apparent paradox, though. Form a stress standpoint, you are probably better with the toes up. From a strength standpoint, you are probably better with the toes down. It would be interesting to run the numbers.

 

[Compositepro]

Perhaps a valid analogy is a steel tape measure. It is far more stable as a cantilevered beam with the concave up rather than down. I believe that the reason is because the compressive buckling strength of the edges is less than the tensile yield strength.

 

[paddingtongreen]

I'm at a loss, why would toes down reach yield sooner? Sx gets bigger, Mx is the same, stress goes down. The shape of the extreme fiber makes no difference, only it's distance from the neutral axis does that.

Even so, a plate makes more sense, it's easier to clamp and just as easy to weld.

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

 

[cntw1953]

pad:

I think they mean material deterioration could be faster than other way around (legs up), because of the legs suject to the maximum stress within such small areas, provides the stress will fluctuate with position of the crane, and is repeative in nature, for which fatigue is a concern.

 

[miecz]

Paddingtongreen-

I'm guessing that c will grow faster than I, and that Sx (I/c) will go down. I could be wrong. Would have to run the numbers.

 

[BAretired]

Let's see. With toes down, f = M*y/I. M is the same. I increases a bit but not too much. The c.g. moves down slightly. The y distance to the tips of the lower channel increases by the width of the flange less the web thickness less the shift in c.g.

First yield, I think, would occur earlier but the plastic modulus would increase, so ultimate strength is slightly greater with the toes pointing down.

BA

 

[JAE]

Separate from the stress/yield issue is the constructability issue. With toes pointing up on the bottom flange channel, without knowing the relative WF flange width bf vs. the channel depth and channel bf, can you even get a stick in there at 30 degrees max. tilt to access the flange tip to weld them together? In other words, would the channel flanges impede weld access?

 

[Stillerz]

For the hell of it I ran some quick numbers (using CAD).
A W14x34 & C10x15.3 Caps.
Toes Down: Sx (bottom) = 86.9 in^3
Toes Up: Sx = 103.1 in^3

C grew faster than I in this case....so, just run the numbers and made Sx smaller with toes down.

I ran these really quick....could be wrong!!!

 

[Stillerz]

By the way, a bit of advice in welding channel caps to I-shaped girders.
I have found it good practice to plug weld the channel on the centerline of cap & girder prior to welding the flange tips to the channel web.
This helps to mitigate heat distortion/cupping of the channel web that can create a void between the beam flange and channel web.That void, when cyclically loaded fatigues the flange to web welds.

 

[LowLax]

Thanks for the replies. I ran the numbers and I get the same scenario Stillerz gets. Ix goes up but Sx goes down with toes down.

As others have suggested a plate might make more sense, so that is the way I'm probably going. Now to calculate shear flow.

 

[Stillerz]

Hmmmmm.... I remember seeing a good shear flow argument that was only a few hundred posts long somewhere....

 

[Stillerz]

Instantaneous Shear Flow = V*Q/I; use for intermittent weld design. (I don't like using stitch welds on heavy industrial or mill duty runways).

At ends use a force = M*Q/I to design the weld.
M= allowable moment of section.

 

[cntw1953]

Suggest to use continuous weld for this situation. Do not worry over-design for this is a fatique sensitive structure. Consult with person familiar with field welding practice to come up a good welding procedure, QA & QC.

 

[Stillerz]

I concurr....stay away from the intermittent welds unless a very light application where the crane is not used often.

 

[hokie66]

Just a thought...as proposed by FLCraneBuilder in another thread. Can you have the crane built differently rather than modifying the runway? Would save a lot of site welding/expense.

thread507-258757

 

[LowLax]

Thanks for the welding tips. I don't have a lot of experience with fatigue issues. Why are stitch welds bad? This is not a heavy duty crane. I'd say moderate service.

I've already suggested just putting in a new beam. Maybe once they see what sort of welding is involved they'll change their mind.

 

[apsix]

Stitch welds have a stress concentration at the ends of each weld segment, which is bad for fatigue.

Similarly Stillerz's idea of using plug welds at the beam web. I can see his argument but it would take a lot more than his recommendation here before I used it. (No disrespect intended Stillerz.)

Common wisdom is to use continuous welds at the flange tips IF fatigue could be an issue.

Depending on the industry, toes up could lead to a tray full of dust etc, with the crane ploughing thru the dust!