shear flow

[pmtrevisan1]

hello folk.
I'm having brain fart.

when calculating shear show for assemblies/ stitch welding what shear value do you use as it varies along the beam length. basing it max shear is conservative but varying V doesn't seem to much sense to me.

thanks in advance.

p

 

[BridgeSmith]

It would depend on whether you are attempting to vary the length and/or spacing of the welds. If you're going to change the welds along the length, you would provide what's required for the max shear in each section you consider separately. If you're not going to change it, you would provide what's required for the max shear throughout the length.

I'm not sure what the codes or research says about the distribution of shear (using the required weld length averaged over some portion of the beam length), but for fatigue checks on shear studs, I average the required spacing over 4 ft, based on recent research and testing for stud groups on bridge girders with precast deck panels.

 

[pmtrevisan1]

makes sense

thanks a bunch

p

 

[KootK]

I vote for:

- V_max everywhere OR
- use a shear diagram that never dips below 0.25 x V_max (same logic as SJI joist shear diagrams).

It's pretty easy to have a partial load case that will actually govern shear at a particular location. In a system where the attachment is brittle, that could fail the attachment locally.

 

[EDub24]

I usually use the max shear everywhere. Unless I'm trying to shave costs it's simpler for me and simpler for the Contractor.

 

[TehMightyEngineer]

Any time I try to vary stitch welds or something funky like that the contractor asks me if I can make it all the same. Unless I'm welding with gold I'd just use V_max and have some conservative welds in your back pocket for when they get a defective weld or something.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[pmtrevisan1]

that is what jve been doing.
using max v and run with it. never has issues previously form contractors or welders
just a colleague brought it up and begun thinking about it.

 

[BAretired]

If VQ/I is not intuitive, there is another way of thinking about shear flow. EDIT: When a plate is added to the tension flange of a beam, the force 'T' in the plate at any point is proportional to the bending moment at that point. The total length of stitch weld between the support and any point must develop the force T at that point.

For a concentrated load at midspan, moment varies linearly between support and load. Thus, tension in the plate varies linearly, requiring a uniform size and spacing of stitch weld over the entire span. This is consistent with the fact that the shear V is constant over the entire span.

For a uniform load, moment is maximum at midspan (M_max = WL/8). The rate of change of moment is maximum at the support and zero at midspan. Moment is a parabolic curve, so the moment at the quarter point M_qp is 0.75M_max. If T is the maximum tension in the plate at midspan, then the stitch welds must develop T between support and midspan whereas they must develop 0.75T between support and quarter point.

Since V is the rate of change of M, the amount of weld at any point in the span is proportionate to V.

BA

 

[BridgeSmith]

BAretired, I think you took the long (and for me, confusing, sorry) way around to get to horizontal shear is proportional to vertical shear.

 

[KootK]

I think that it comes down to what does and does not speak to one's intuition as an engineer. While I get the theory behind VQ/I, the math just does not speak to my intuition. The notion that connector demand should follow the variation in required tension development in the connected part is something that I can "feel" however. BA did end up at the usual conclusion as, inevitably, he had to. But I think that it was this particular logical pit stop that he wished to share with us along the way.

 

[BAretired]

Sorry HotRod10, I didn't intend to confuse anyone, though I must admit in hindsight that my explanation did not convey my message as clearly as I intended. KootK expressed it much better.

BA

 

[pmtrevisan1]

I'm the BA camp.
as the typical vq/i gives shear flow which is fine and dandy.
but shouldnt M/h give similar results as that is the force at the interface where the assembly (bkkts or welding) take place.

 

[KootK]

but shouldnt M/h give similar results as that is the force at the interface where the assembly (bkkts or welding) take place.

No. I'm afraid that you've taken away the wrong impression from BA's comments. Connection demand is proportional to the rate of change in moment (V, dM/dx) rather than absolute moment.

 

[BridgeSmith]

With the edit BA, I can see what you're getting at now. It still makes more sense to me that the horizontal shear at the interface is proportional to the vertical shear at the same point. To each their own, I guess. As KootK wrote, what's intuitive to one is not to the next guy.

 

[BAretired]

As KootK wrote, what's intuitive to one is not to the next guy.

I can't argue with that.

BA

 

[TLHS]

It's just attacking the fact that shear is the rate of change of moment from the other direction. If you're used to thinking of moment as the defining property for beam design, working shear flow back to a moment based definition likely makes more intuitive sense.

I know that I think of that connection as restraining the slip plane between elements created by bending, so I can see why defining the calculation that way would be helpful. It's also somewhat analogous to rebar development.

 

[Agent666]

For most practical applications in welding reinforcement onto steel members, you are often going to be governed by the minimum weld sizes anyway. Calculations generally result in very small weld sizes being adequate when welding a small area of steel to a larger area member. However code you are following will generally require you to need to bump it up to the minimum fillet weld size.

I've only ever used the max shear from the analysis and kept one weld size for practicality, calculations might show a weld size of 1/2/3mm etc, then bumped it up to the minimum size based on thickness of the material. So basically, in most practical applications it usually doesn't matter what shear you take as minimum weld size covers it (still need to check though). Generally if welding say a strengthening plate to a beam flange then if you back calculate the shear from the minimum weld size you might find the shear the weld capacity can take is higher than the shear strength of the member web.

Keep in mind most codes have provisions that state you must have some length of solid non-intermittent welds at the ends of the piece of steel you are adding. The length is generally some function of the width of the piece of steel, believe its to do with the shear lag and developing the plate strength.

 

[BridgeSmith]

If a significant portion of the loading is cyclic, bear in mind that components connected using a non-continuous weld have a considerably lower fatigue resistance than those that are continuously welded.