Bending in a Composite Beam 2

[NanoMan]

We have a 24" x 24" x 60' latticed steel "spine" to which an engineering firm is planning to attach a tower segment on the ground and then hoist the whole thing into position on the tower.

My boss has asked me to validate the approach and there's one area that I'd like to make sure I'm doing properly.

In transitioning the structure from horizontal to vertical, I'm concerned about bending in the legs of the "spine" (they are 4x4x3/8 angle).

The approach I've taken is:

1) Treat as a horizontal, simply supported beam
2) Find support reactions
3) Draw shear and moment diagrams
4) Determine I of the 4 angles using AutoCAD Region Properties
5) Determine Mc/I

Is this the proper method?
I find worst Mc/I = 25 ksi - since this exceeds .6*36 ksi I'm ready to put the kabash on this approach but I don't have the experience to know if 25 ksi is OK since this is a momentary case and my analysis didn't consider the horizontal and diagonal components of the spine, just the composite I of the 4 legs.

Any feedback is welcome.

 

[SlideRuleEra]

Be sure that your "Item 1" (simple beam) assumption accurately models how the erector will rig for the lift. Chances are he will use a more complex arrangement with more than one attachment point - if this is true the actual stresses may be lower that you have calculated.

 

[Qshake]

One thing that I will certainly say for you is good job checking the erector stresses. This is, unfortunately, an area that is not paid it's proper due.

A lot of time the engineer of design will assume that to be the problem of the erector when its is clearly inherent in the method chosen (perhaps by inadvertently eliminating other methods)by the designer.

Even if your analysis (however, simplistic) shows an overstress, I would throw the issue back to the consultant and request that it be investigated and calculations provided to show that it is not a concern. If they are the engineer of record I would assume that they would want to address this issue.

Good Luck.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[ishvaaag]

Given your intent, especially look if erection will likely produce any concurrent compression, in which case the simple beam approach wouldn't remain a safe one.

 

[desertfox]

Hi NanoMan

How did you calculate the "I" value of the four angles did you use the "Parallel Axis Theorem" to account for the spacing of the angles? or did you just sum the four "I" values together?
If you did the former your "I" value would be correct,if the latter your "I" value would be to low.
If you post details of the spacing of the angles and details of the loading I will try and check your calculation if you think it would help.

regards desertfox

 

[flamby]

NanoMan,
Reading your post, I get that you are finding the support reactions only for the dead load of the beam. But unfortunately erection is not that much of a static problem. Do it a little quicker and it is easy to see loads of several times mg, its own weight.

I would think of two things that can spoil any set of good computation work on the field.

1. What is the lifting mechanism? What acceleration is expected of the lifting device? Compute the dynamic load due to this.

2. How much "tilt" of the beam can be seen during erection? If this is uncontrolled then I would find out least I of Ixx, Iyy etc. and use it in computations.

In short, you must demand (in black & white)the complete erection procedure including machinery before you even think of solving the problem.

flame