Haunched connection 1

[LeonhardEuler]

I'm analyzing a structure that holds an over-head crane and am having trouble finding the correct method to analyze the "haunched" connections that connects at the column line and supports the runway beam. My first thought is to simply analyze the 3/4" plate for buckling as it should be the limiting piece, but I know the I-beam will surely add strength in composite action. Attached is a drawing of the connection.

 

[LeonhardEuler]

Clearer picture

 

[jimstructures]

The PEMB industry would just use an I section as a haunch/crane bracket and then analyze it with a shear load to the column and a concentrated moment at the face of the column, with full width stiffeners across the supporting column at the flanges of the haunch/bracket. They would stop using a haunch/bracket with any cranes having a capacity of 25 Tons or greater and go to separate columns or stepped columns to support the cranes.

Jim

 

[LeonhardEuler]

This is an existing structure that is upgrading the crane from 60 tons to 65 tons. I have the structure modeled in Risa, but I was only able to put in that I-beam as the "support" without the haunched plate under it. In Risa the I-beam is failing code checks due to bending about it's major axis. I'm just wondering how to calculate how much bending strength the angled plate adds to the short I section. By the way my drawing was a little misleading. The I beam in question has another I- beam on top of it that runs under the rail.

 

[hokie66]

jimstructures is right. A heavy crane should never be supported on brackets.

 

[Tmoose]

What is the column, etc that the butt of the W18X130, 3/4" and 1" plates are (I assume) welded to?

 

[LeonhardEuler]

Unfortunately I can't change the current design, but I am tasked with analyzing it. Interesting that the brackets can't be used for heavier cranes though. What is the reason? Does anybody have a resource that shows how to calculate the strength of these brackets?

 

[LeonhardEuler]

By memory I think it is a 14x80 column and it does have 1/2" stiffeners running from flange to flange. The W18x30 is a "WUF-b" FR Moment connection and the plates are welded to both the column and W18[

 

[JAE]

How is the crane's longitudinal force resisted? With the bracket off the column there would be torsion induced into the column with longitudinal force applied out at the bracket.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[LeonhardEuler]

JAE this is definitely one of my concerns. This connection is responsible for taking longitudinal, transverse, and gravity loads. There is no other support for the crane forces near this point. The current 60 ton crane has been in use for over 20 years. I'm not sure how to analyze this shape "by hand" and I'm beginning to wonder if this is an FEA situation.

 

[jimstructures]

Take the vertical load at the top of the bracket and apply that load to the column then add the concentrated moment (vertical load times the eccentricity to column center) and apply that as well to the column at the bracket elevation. Obviously check all the stiffeners for their respective local point loads (i.e. moments resolved into a couple). PEMB companies might use a W 18 x 50 as a bracket for up to a 25 Ton load so your situation is way over that. Check all welds and fatigue limits.

Jim

 

[LeonhardEuler]

Thanks Jim. I think I have the method down for the weld strength, but I am more confused about the strength of the steel itself. I want to know how others would cheek the strength of the haunch and W18 combination

 

[jayrod12]

To me that would depend on how close the W18 is to working alone. I'd likely check the vertical plate for the entire compression load. The horizontal 1" plate might not be doing much at the geometry you show. The haunch is typically detailed such that it doesn't really see true cantilever bending, but rather inclined compression transfer back to the face of the main column and a tension tie at the top, like a truss.

 

[KootK]

I think that your first pass ought to be simply ignoring the presence of the lower two flanges and seeing what you can milk this thing for as a straight up stiffened seat connection using the lesser thickness of the two available web plates. Here's why:

1) Handy evaluation methods for this abound. AISC manual, textbooks,...
2) Original designer probably didn't do much more than that if she was working to a budget.
3) This provides a reasonable lower bound benchmark for whatever you do next.
4) If this works out, you can stop there and save a bunch of fee and effort.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[LeonhardEuler]

Another piece of information I should have provided. I checked the W18 by itself and it failed in bending at 1.48 times the code check using Risa.

 

[KootK]

Frankly, I don't feel that a flexural check makes sense for this other than perhaps as an index/benchmark. At these proportions, I believe that the analysis should morph into something analogous to how we use strut and tie in concrete. You're in the Blodgett-o-sphere now, not Bernoulli-town.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[LeonhardEuler]

KootK I think you really nailed it! I followed your suggestion to analyze as a stiffened seat connection and in my research I came across a design guide for triangular bracket plates, which recommends analyzing the plate as a strut of a certain width. I will analyze the plate only for now and hope it passes. After reading this guide I think if the plate buckles under the full load it should be a failed connection, since the beam should not add appreciable strength to this connection other than some of the force being dissipated into the beams welded connection as the top will be in tension as a "tie".

 

[KootK]

Thanks for coming back to post the paper that you found. Classy. I'd love to hear what you come up with for the stiffened seat connection value too.

KootK I think you really nailed it!

Even a blind squirrel finds a nut once in a while.

I think if the plate buckles under the full load it should be a failed connection, since the beam should not add appreciable strength to this connection

Depends. If the vertical plate fails by buckling, I think that the bottom flange of the W18 would help to restrain that buckling. The trick would be coming up with a way to evaluate that. Obviously, it's not going to be a standard AISC clause addressing that.





I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

Vinnakota was actually my thesis adviser. Small world.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[LeonhardEuler]

I will use the bottom flange and the plate as a T shape when considering torsion and buckling. I will post the results. Thanks KootK!