How will you assess a bridge for oversized load transportation?

[TrustButVerify]

I have been quite curious about how structural engineers assess bridges when a heavy transport convoy (trailer + cargo) is being driven on top of them.

I have seen different approaches, point loads or distributed loads (with some different criteria as well).

I'm summarizing them in the picture I'm attaching, and I also adding a picture for reference.

The question will be: what is the best approach to follow? or if even several approaches have to be assessed?

Another question could be: will be the same if you are trying to assess some soil like asphalt, or ground?

 

[Smoulder]

How to apply loads in structural analysis depends on analysis model. I have used all your types except #3. And another type where you spread the contact pressure from #2 at 45 degrees to slab centroid so it will look a bit like #3 but not quite. Pavements always use actual contact pressure and area AFAIK.

 

[bridgebuster]

draw an influence line.

 

[TrustButVerify]

What is that?

 

[bridgebuster]

https://en.wikipedia.org/wiki/Influence_line

 

[BridgeSmith]

We create a simple loading configuration consisting of the axle weights and spacings, and then our line Girder analysis program (BRASS Girder) moves the series of loads across the bridge, multiplies the reactions by the Distribution Factor or Wheel Fraction calculated per the applicable specification (LRFD or Standard, respectively), and calculates the stresses on the girders.

 

[Lomarandil]

My experience is similar to BridgeSmith (assuming a relatively simple girder bridge). You'll often need to check multiple elements of the bridge structure using separate assumptions (e.g. wheel loads can be shared across the deck using an AASHTO equivalent deck strip, but are usually applied directly to the girders without that distribution). As such, while this can be done in software, there's usually a heavy element of hand calculations involved determining inputs.

(Note that the assumptions can result in loading similar to your #3 and #4, but those aren't always appropriate depending on the structure)

I see the logic of Smoulder's approach, but I wouldn't apply that together with the empirical AASHTO distribution factors.

Heavy equipment loads on grade are treated through Bousinessq or simplified earth pressure methods (assuming 1.5:1 spread through rigid pavements, 1:1 through flexible pavements, 1:2 through most soils) -- the resulting pressures at each depth need to be compared against soil capacities derived through typical geotech practices. In problem areas, timber crane pads or earth fill can be laid to increase spread and decrease pressures. If it's going to be used repeatedly, you might even use a geotextile reinforced earth fill to build a haul road.

 

[Dr_Bridge]

It depends what exactly you mean. If you are thinking about a heavy load that needs to go from point A to point B, and how the bridges on that path are evaluated, that is one thing; if you are thinking about how we design a new bridge, and if we consider large/unique trucks, that is another.

When routing a permit load, many state DOT's use software, such as AASHTOWare Bridge Rating, which maintains a database of bridges, each with a structural model of all unique elements (girders and sometimes diaphragms/cross frames) on the bridge. Where a truck is pretty unique, this software can be used to take that specific loading and essentially "run it" over the bridges along the route that a permit load will travel, and get an output of all of the ratings, where a rating is a calculated value where a number greater than 1 indicates the element is adequate and less than 1 is inadequate, and with that information, the engineer (usually in the state bridge office) will decide if the truck can travel along a specific route. The rating includes load resistance factors, and a factor for impact, so in real life, a rating < 1.0 does not mean the bridge will fall down or receive permanent damage, however, taking things to their limit, especially expensive infrastructure, is generally not wise.

One thing to consider, though, is that most permit loads do not need to be run through software. New bridges are rated as part of the design process, and although many existing bridges weren't rated when they were designed, they have since been rated based on as-built plans and/or field observations. When we perform a rating, we rate it for the current (or current at time of construction) design loading, which includes a truck load and a lane load. The truck load represents a heavy and short truck. The lane load represents essentially a line of trucks stopped in traffic. We also rate for a number of "legal loads" and "permit loads", which represent a range of axle arrangements and axles loads. Most trucks, even heavy loads, will look a lot like one of these loads, or can be modified to fall under one of those loading conditions. If a permit truck has a good analogy to an existing rating, then its a simple task to look at all of the bridges impacted, compile the already existing ratings for the truck type, and determine if there are any issues.

In terms of design, we don't necessarily think too much about unusual truck loading. The current US design load, which is called HL-93, includes a lane load of 640 lb/ft plus either a truck load (a 72,000 lb truck that ranges from 28' to 44' long) or a tandem load (two 25000 axles spaced at 4'), generally results in designs that rate very well for most permit loads. These loads are arranged in order to maximize load effects. Many states also include a permit truck load in the design process. These loads are then increased by a factor of 1.75, and typically, an impact load of 1.33 is also applied to the truck portion of the load. These design loads are intentionally conservative, and typically result in bridges that will not see damage due to overloading over their lives, even on routes where permit loads are a regular occurance.

Also keep in mind that, although permit loads can be heavy, the trucks are often modular, so axles can be added as needed in order to spread out the load. In these cases, a long length is actually helpful, because there is more room to spread out the load.