Structural Stability Basic Question 2

[JD P.E.]

I am a mechanical engineer who has a layman’s interest in structural engineering. I do not engineer or advise anything structural.

My question is about bracing structures, and I know it’s very basic.

When do you know to brace a structure? Is there a rule of thumb? Or is it a combination of reducing structure deflection and reducing base moments to use a pinned type connection? Is it mostly just a design/cost choice?

I can do the basic beam/structure analysis, but nuances like bracing have me confused as to when to use and how to use.

I know it’s basic, just looking to understand more of this topic.

 

[le99]

1. When it is necessary to prevent large displacement of the joint in a frame. The bracing is similar to the diagonal member in the truss. The occasion to provide such bracing is pretty much depending on the type of the framing chosen (simply supported/connected frame vs moment frame).

2. Long beam and column tend to swing/deflect sideways, the critical stress is difficult to predict, so intermediate bracing is necessary to reduce the unsupported span length to bring the stress under control. This requirement is usually specified by the design code.

 

[Enable]

I doubt your confusion is due to the concept being basic. O'contraire, bracing can be mystifying precisely because it's hard to exactly know what we are bracing for or why!

Take a look at these sampler threads where very common bracing applications are hotly debated by some rather great structural engineers:

Is it stability bracing or unnecessary bracing
Is the 2% strength rule for LTB bracing really about strength or is it a stiffness requirement in disguise
When do clip angles provide torsional restraint at the end of a w-section (is it 50%, 60%, 75% or to infinity and beyond)
Do beams REALLY need torsional end restraint. The curious case of stacked beams that, well, stay stacked
Yet another unbraced length questions where the consensus is that...there is no consensus

If you really want to dig into it feel free to post some examples of structures (pictures, drawings, whatever) and we can all discuss the need for various bracing arrangements in that context. But as far as generic explainers go here's what I'll say:

A) Bracing is often used as "just-in-case" situations, which ultimately are "actually-the-case" conditions. In other words, the bracing is not designed to take crazy loads or even more than nominal loads really. Rather it is designed to provide restraint in the event of eccentric loading or material imperfections or anything that would cause undesirable displacements due to eccentricity. You can imagine a slender vertical member that may be able to handle the vertical loads if coming (perfectly) straight down (so totally in compression), but for which if there is even a minor eccentricity to the load it will displace far too much at the tip, due to the induced moment (load * eccentricity). This can lead to a so-called P-delta effect where deflection begets deflection! Well, if we add bracing to the top of that member there is no worries with that condition!

B) To actually take the load down to the foundation. We always have to get the load to the ground somehow, very important! Bracing whether at the roof level or in the wall or whatever can create a load path directly to the base of our structure. Basically what le99 said.

C) To control the mode of failure for given structural elements. The most obvious example here is in the case of lateral torsional buckling. By bracing the compression flange of a simply supported steel beam (see 2% requirement thread above) we can change the governing failure mode to a much higher energy one, which might allow the same beam to span the distance whereas it otherwise would not. This can be useful since bracing is cheap but heavy structural members are less so.

D) Because sometimes it just feels good to brace things. Structural engineers follow their human instincts and like to connect things together. It just feels good, and sometimes can provide an intentionally redundant load path in the case of failure of the primary load path!

 

[human909]

When do you know to brace a structure? Is there a rule of thumb?

A simple rule of thumb: ALL THE TIME AND EVERYWHERE!
More specifically: In all vertical planes between supports.
Most specifically: Through rational inspection of load paths, recognising that lateral loads are present in all structures and at least nominal bracing is normally required to about excessive deflection.

I routinely review small steel platforms and structures that are drafted by design draftspeople of a mechanical type background. Bracing is the number one item that is lacking. Connections and members are generally of sufficient if not excessive size. Bracing is often not existent or haphazard. It is all about load paths and in general for the positioning of bracing calculations are needed if you can visualise the load path to the ground.


Bracing is something I sadly even see structural engineer get wrong. I was on a site the other month and an entire row of internal columns was unbraced. Due to equipment they were supporting they were vibrating all the way down to the column base. NOT COOL.

 

[JD P.E.]

So, probably over simplifying, but would this process be fairly accurate?

1. Determine Required load
2. Size columns and beams based on ASD/LRFD
3. Brace long columns per requirements
4. Brace for lateral loads and make load paths to ground through bracing

Fair logic?

 

[human909]

That is a very rough process. But yes. Though depending on the tools used you might do most/all those steps at once. I use a structural modelling program where I'll size all my members and braces and position them accordingly then add the loads and compute the results for all load combinations.. I'll resize base on deflection and strength requirements.

Enable did a reasonably summary but I'll reiterate again it should be noted there are at least 3 distinct type of 'bracing':
-lateral bracing that resists against lateral loads on your structure (wind/seismic being the most significant) Also often required for global stability.
-compress member 'bracing' that reduces the effective length of compression members (typically columns) to increase their capacity by avoiding column buckling
-beam bracing that reduces the effective length of a beam to avoid lateral torsional buckling

I've 'observed' the performance and behaviour of too many structures with insufficient or even ZERO bracing against lateral movement. I'm talking mostly about small steel structures and platforms here. They'll normally over designed for gravity loads but if you sneeze they'll readily start rocking back and forth in a disconcerting way. With steel the pinned assumption is all too close to reality so without bracing there is really very little preventing significant wobble.

 

[271828]

It sounds like you're talking about overall building design, not member design. Sorry if I'm misinterpreting.

The bottom line is the building must have a lateral force resisting system. The basic options are moment frames, braced frames, and shear walls.

Fairly often, depending on the type of building, the architect won't allow diagonal braces or shear walls because of floor plan consideration, so moment frames are used in those cases.

When we get a chance to use braced frames or shear walls, we typically go with them because they are much stiffer and stronger for the amount of material.

This is a very coarse look at the subject.

 

[JoshPlumSE]

When do you know to brace a structure? Is there a rule of thumb? Or is it a combination of reducing structure deflection and reducing base moments to use a pinned type connection? Is it mostly just a design/cost choice?

First, let me point out that I'm talking about bracing the overall structure. I'm NOT addressing the bracing of individual members.

Next, the big question is the project design criteria. If the structure supports equipment, then does the equipment have restrictions on deflection for maintaining proper operation. If there isn't any project specific criteria to limit drift, then we would likely turn to various codes or manuals of practice. This is a necessity for occupied structures where people live or work.

There are also price considerations (will bracing be cheaper than moment frames when all the welding and bolting and detailing is done). Then there are architectural considerations.... Though this is often related more to WHERE the braces will be located rather than whether we use a moment frame or a brace frame system.

Bracing individual members, on the other hand, is all about increasing the capacity of that member. Sometimes it's about increasing the axial capacity, sometimes the bending capacity, sometimes its about increasing the ductility of the member for seismic design.