Lateral restraint in angles. 2

[WARose]

I have a situation where I am code checking a [equal leg] angle under a bending load (a concentrated load on the tip of a cantilever). The thing about it is: section F10 in the 13th edition only allows checking the member about the major geometric axis if continuous lateral-lateral torsional restrain is provided. I have substantial torsional restrain but it is not continuous (it is at the tip and at the mid-point; it’s a 4’6’ cantilever). So I was thinking: if I could prove that the equivalent torsional restraint I was providing exceeded that [cumulative] required in Sect. 6.3.2b [p. 16.1-195]....that might do it.

Only problem with that is 2 things: one [the obvious] is that really the equivalent of what the code wants? And secondly when you think about an angle bending about its geometric axis there are 2 components of deflection (vertical and horizontal)….so you have to wonder: for angles if you don’t restrain it laterally, is that really acceptable for bracing (i.e. does torsional restraint really do the job for angles)?

Thanks in advance for any insight.

 

[ishvaaag]

The last part of your question self answers: it is effective if it REALLY provides torsional restraint. The question is that sometimes the overall torsion might make the overall torsion going in the total structure, say, at panel points in a tower, go in the same sense of rotation than the torsional-flexural buckling of some members, being the equivalent of the castle of cards for such kind of structure, a kind of failure seen in overstressed truss towers.

In any case, it is a matter of how the total and local rotation allowed by the outfit at the pertaining level of load per the code relate one to the other that will indicate if you, even in such circumnstance, may be getting effective torsional bracing of the members at the connected points; and certainly there may be cases where improperly conceived braces may not show either the required strength or stiffness, quite likely due to one badly considered path to hardpoints at foundations.

It is of help that the direct strength method when including the bracing shows you (details apart) what is expected of your whole structure under the design (typically, factored, except you do it ASD, yet direct design) IF you have modeled BOTH main members and braces. This way you can get some start on reliable behaviour and then can concentrate on detailing the restraints. But first get a structure that stays stable in direct design with everything modeled, braces included, at factored level.

 

[msquared48]

The design torsion will be minimized if you load the angle through it's shear center.

Mike McCann
MMC Engineering

 

[WARose]

Thanks for the answers (so far). Mike, it’s basically a bracing problem as much as anything. The code allows you to provide bracing for beams 2 ways: with lateral restrain or torsional restraint. But the section of the 13 edition I was working in was requiring continuous bracing. So that presented 2 problems: I had substantial bracing but it really wasn't continuous (although it was at all the critical points); and I wasn't sure if that type of bracing [rotational] was really ok for angles (given their displacement under flexural loads).

 

[ishvaaag]

Since you don't provide continuous bracing will concentrate on how to proceed for such case, single angles,

AISC 360-05, p. 16.1-58

"Single angles with continuous lateral-torsional restraint along the length shall be permitted to be designed on the basis of geometric axis (x, y) bending.

Single angles without continuous lateral-torsional restraint along the length shall be designed using the provisions for principal axis bending except where the provision for bending about a geometric axis is permitted."

So the code directs you (except "where the provision for bending about a geometric axis is permitted", to find if one applies to your case) to design your member "using the provisions for principal axis bending".

By this spec, you can forfeit any effort on considerations of some discrete bracing subtituting the effect of continuos bracing (no guide is given on how to, even for the cases "permitted", to be found), and directly pass to treat your problem in a bending about principal axes setup.

At p. 16.1-280

"When bending is applied about one leg of a laterally unrestrained single angle, the angle will deflect laterally as well as in the bending direction. Its behavior can be evaluated by resolving the load and/or moments into principal axis components and determining the sum of these principal axis flexural effects. Section F10.2(i) is provided to simplify and expedite the calculations for this common situation with equal-leg angles."

That is, the existence of the specs for LTB on geometrical axes is only a device for simplification; the general procedure must go along principal axes.

Reading the rest of the commentary where the most recently quoted paragraph is placed makes one infer that the intent of the code AISC 360-05 is that you resolve your solicitations in x,y geometrical axes unto principal axes, then use eqs. (F10-2) and (F10-3) with Me (for equal legs' angles) from F10.2(iii)

 

[WARose]

Thanks for the reply ishvaag. I guess that kind of clears up what I have to do. [I wasn't 100% comfortable with my original approach anyway.]