AASHTO Bridge Design_Steel Beams

[Wanna_be_SE]

Good evening,

I'm new to bridge design and have begun my first calcs for "somewhat" doing a load rating for some bridge repairs. My short career thus far has been just over 3 yrs in steel design, but primarily doing stress analysis for various structures (in a shipyard), and so I wasn't using industry standards until this past May when I decided to change career paths (within engineering) and go a civil-structural route. So now I'm doing deep dives into various code requirements trying to understand why some requirements are the way they are. AISC requirements are very similar to what I was doing at previous employer, but seems AASHTO has put their spice on many different requirements (at least the ones I've come across so far). Many similarities between AISC and AASHTO, but nonetheless definitely different.

So, I noticed the AASHTO spec only considers the steel section properties for bridge beams (girders? - I typically associate girders as a member supporting other members like joists, etc but seems various terminology is used across the board), when calculating section properties (section 6). I see there's a Kg which I guess accounts for the stiffness of deck slab (which looks like a modified version of the parallel axis theorem), but it occurred to me that given the composite section and assuming in reality the neutral axis shifts towards top flange of steel beam (probably right around the top flange), then your positive moment capacity for that beam goes DOWN, given the "c" distance to outer fiber goes up.

So, how does AASHTO account for the "true" flexural capacity by only considering the steel section properties. I did take a bridge design course in college and I know there's methods for taking an effective section of concrete deck and transforming it into an effective "block" but thought it was interesting how AASHTO (at least in the section 6 I was navigating) does these calculations and would like some greater insight into where/how the equations were derived. Do the equations account for all this in some way, or is that what the Kg factor is doing (even though it's not exactly how the parallel axis theorem works, right?)?

Any insight would be greatly appreciated.

 

[Lomarandil]

I think you're misunderstanding AASHTO.

If composite connection (typically shear studs) exist between the steel girder and concrete deck, the section properties are computed for that composite section.

If composite connection doesn't exist, the neutral axis doesn't shift (for typical configurations of girders and deck), and AASHTO is basically parallel to AISC provisions, except as modified for the more slender steel sections common in bridges.

 

[BridgeSmith]

Kudos on making the jump to bridge design. It does come with a somewhat steep learning curve, though.

We do typically call the main longitudinal members "girders", but sometimes the term "beams" is used. Transverse beams between main girders that support smaller beams are generally referred to as floor beams, and the small beams are referred to as stringers.

As far as the Kg factor, after 20 years in bridge design, I can't say I'm completely clear on what it is, but the definition ("longitudinal stiffness parameter") would indicate that it's a measure of the longitudinal stiffness of the superstructure. It's involved in determining how the weight of a truck will distribute across the superstructure (very stiff girders with more flexible connections to adjacent girders will not share as much load as more flexible girders that are more rigidly connected to the other girders). Honestly, I've never tried too hard to understand it. In the LRFD spec, it only appears in section 4.6.2, for use in defining superstructure types for analysis and the effective span length.

I'm not sure I understand your comment about the capacity decreasing for a composite section in positive bending. Compared to a non-composite girder, a composite girder, especially in positive bending, has a much larger capacity. The composite section is expected to have a lower capacity for long-term loading than short-term loading, due to the effects of concrete creep reducing the stress and compressive force in the concrete, which makes the steel carry more and more of that permanent load over time. Composite girders are typically designed for 3 stages of loading:

1) The steel girder alone - Loads applied to the girder before and during deck placement deflect the girder, inducing stress in the steel, which gets 'locked in' when the concrete cures. Loads of bracing, stay-in-place formwork (and the associated extra concrete), and the concrete deck get applied to the steel girder.

2) Long-term composite section - Permanent loads applied after the deck has cured. Curbs, railings, raised sidewalks, etc. get applied to composite section where the concrete is assumed to be only 1/3 as stiff as it is for short-term loads (concrete is transformed to an equivalent area of steel by dividing by a factor of 3*n, where n = Es / Ec)

3) Short-term composite section - Transient (AKA live) loads get applied to the composite section where the concrete is transformed into an equivalent area of steel using "n".

In LRFD, the concrete is considered effective in tension, for the calculation of stiffness, Fatigue limit state loading, and Service limit state loading, but not for Strength limit states. For the Strength limit states, only the reinforcing in the deck gets counted as part of the section, along with the steel girder.


Rod Smith, P.E., The artist formerly known as HotRod10

 

[Wanna_be_SE]

Lomarandil (Structural),

I guess I have misunderstood.

BridgeSmith (Structural),

Thanks for response. I'm getting a better understanding based on what you said.

Yea I'm not sure why I was thinking the capacity of steel member would go down once N.A. shifts, bc yea if the N.A. has shifted then it's acting as a composite section which means you have the stiffness of deck acting with steel member, and so the section capacity goes up not down. I guess I was only considering the section of steel, not composite section.

That's interesting though about what yo said for 3 stages of loading. So in a typ design, would you consider all those stages as separate load cases and analyze accordingly? Sounds like I've got plenty to learn.

- Dan, EIT (who cares)

 

[BridgeSmith]

So in a typ design, would you consider all those stages as separate load cases and analyze accordingly?

Not separate load cases, per se, but separate load effects. You would calculate the stresses in the components based on loads applied to each section, i.e. the stresses due to stage 1 component dead loads (DC) would be calculated using the non-composite section properties. For the Service, Strength, and Extreme Event (if applicable) limit state checks for the completed bridge, you would sum the stresses from the individual stages to check against the stress limits for applicable limit states.

There are limit states for constructability at various conditions during construction, from wind on the bare erected girders (loading per the AASHTO GUIDE DESIGN SPECIFICATIONS FOR BRIDGE TEMPORARY WORKS), wind on the girders with the deck forms in place, and wind during the pouring and curing of the deck (we use a reduced wind speed when combining with the weight of the wet concrete).

Rod Smith, P.E., The artist formerly known as HotRod10