Integral Abutment - horizontal loads question 2

[PAK_ENG]

Hello! I am working on the design of a two span bridge with integral abutments with a single row of piles and fixed pier. What are the typical horizontal loads that the integral abutment should be designed for? I have applied passive horizontal earth pressure due to backfill for the backwall design. But what about the integral abutment footing and piles? What about the typical horizontal loads of wind, earth, live load surcharge, thermal? Additionally, is there a reaction at the top of the piles due to deflection and how would I calculate this load/moment? Thank you for your help.

 

[MIKE_311]

VDOT has a pretty comprehensive design example for integral abutments:

http://www.virginiadot.org/business/resources/bridge/Manuals/Part2/Chapter17.pdf

Note- you may need to use internet explorer for this to work correctly. The example starts on 17.04-1.

If you have and know Mathcad you can recreate the calc.

 

[STrctPono]

I have applied passive horizontal earth pressure due to backfill for the backwall design.

Do you mean to say you applied at-rest earth pressures? This depends on whether you backfill before or after you pour your end beam. They may be active. If you did mean to say "passive earth pressures" then I assume you are referring to the soil springs on the backside of the abutment/end beam that will help you resist any movement should the bridge move into the backfill. Don't forget, these soil springs are compression only.

Wind: Highly unlikely that wind is going to govern your footing/pile design.
Earth: Yes, apply either your active or at-rest earth pressures appropriately for the design of your foundation elements. If the abutment walls are really tall this load can get significantly large especially if you are in a moderate to high seismic zone. Consider a stabilized cut slope and GRS backfill if you have the space for it.
LLS: Yes, apply this per the AASHTO code.... 2, 3, or 4 ft of equivalent soil height.
Thermal: This is probably your big one. Integral abutment bridges can see extremely high loads if constructed in regions with wild temperature swings. You can get away with it for short span bridges but longer span bridges really really need to consider this one carefully. I wouldn't be so worried what the temperature effects will have on your foundation elements so much as the stresses that will occur in your superstructure and abutment walls. Everything will either expand away from or contract toward your center pier (assuming it is in the middle). Spend time and run this in a FEA with a uniform temperature drop/rise and see what kind of forces/stresses you get. This may be your governing factor. I live in a tropical region so we take advantage of integral abutment bridges all the time and typically only have to design for a maximum 30 degree temperature variation. You should consider at what time of the year and at what temperature the bridge will be at when you lock it in with the end beam pour.
Shrinkage: This too can have an effect on the bridge. Pour the deck but not the end beam right away. Allow the deck to shrink substantially before you pour the end beam and lock it in to the substructure.
Seismic: This is where you want to utilize your backfill to help you out. Longitudinally, the abutment walls and end beams pour will heave into the backfill which allows you to take advantage of your compression soil springs. Transversely, you can utilize your wingwalls to engage the backfill and resist the seismic loads. This is how you can slim down your foundation and piles/shafts.

 

[Guest090822]

Different state DOTs have different requirements. New York uses full passive pressure and applied the load between the girders. In recent years, they moved to a 100% empirical method, see their Bridge Manual.

A few states, Idaho one if I recall, uses a broms-ingersol pressure distribution. They basically vary from at-rest to passive.

There is no 100% theoretically correct philosophy for IAs. I served on an integral abutment design committee and we researched many states and found many different approaches.

As stated by others, surcharge and soil will be your main loads for horizontal forces. I’ve never seen wind control or even come into play for integral abutments.

Seismic doesn’t need to be considered for abutments unless they have recently updated the code.

For a 2-span don’t use fixed
bearings at the pier. Use expansion bearings at the pier so longitudinal forces are taken by the IA - that’s the whole point of these things!

 

[BridgeSmith]

How tall is the backwall? Our fully integral abutments are typically shallow enough that we don't consider bending of the cap. The piles are assumed to form a plastic hinge, so we design the cap to diaphragm connection for the plastic moment capacity of the piles, and don't consider the soil loads, etc.

Due to the complexities of analyzing taller fully integral abutments, when we do those, we've opted for a semi-integral configuration. For those, we use at-rest earth pressure on the backwall, and passive on the end diaphragm. The pressure on the end diaphragm is of little consequence, though, since it mostly produces minor axial load in the girders.

What to use for soil loading on the wingwalls is a point of much debate, although we've used active soil pressure conditions successfully for many years. It probably works because behind most of the wingwall rear face, the backfill is fabric reinforced, with a formed air gap, so as long as the fabric holds up, there's zero pressure on the majority of the wingwall face.

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

 

[STrctPono]

Rod,

You guys allow your piles to hinge? That is a completely different philosophy from our state (and from AASHTO for that matter). I'm curious on why that is...

 

[r13]

I am not familiar to bridge design and AASHTO code. But, isn't that is conservative in designing the cap based on the maximum achievable force permitted by the piles (supports)? I guess the pile design would be based on AASHTO guidelines, and may be controlled by the service criteria. ROD can explain better then.

 

[BridgeSmith]

We do allow the piles to hinge, or at least assume that they do; they may or may not, depending on the stiffness of the top 10-20 ft of embedment soil. Based on research I've seen (from TXDOT, I believe) hinging does not reduce the axial load capacity for piles that are embedded in stiff soil and driven to refusal. We haven't seen anything in AASHTO that we believe we're violating with this approach. We also have decades of successful experience with this approach, as do many other states.

We don't consider any service load limit states for piles driven to refusal, which is most of the ones we do. For friction piles, we consider the service load combinations, but if the cap or footing on top of the piles is expected to move, we would ignore the effect of the embedment soils around the piles where they would be expected to move.

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

 

[STrctPono]

Theoretically, I don't see any issues with allowing the piles to hinge, but from a code perspective and a practical and serviceability standpoint there is no way to inspect the level of damage and then go in and make the necessary repairs since they are beneath the ground. The inelastic straining that takes place in the pile will usually necessitate a repair.

I was thinking specifically of AASHTO 3.10.9.3 and 3.10.9.4.3 which essentially require all hinging to occur in the substructure elements above the pile/shaft. For Seismic Zone 3 it specifically states that foundations shall be designed for R=1

 

[BridgeSmith]

For the way we use them, the hinging of steel piles doesn't compromise their functionality, under service conditions or seismic. The first 3 highlighted portions only address the calculation of seismic design forces, and the last one addresses inelastic hinging of columns. If you have a column that loses capacity due to inelastic hinging (such as a concrete column), then of course, it would need to be in an inspectable location, so that the reduction in capacity can be assessed.

Anyway, the 2nd highlighted area would indicate that the foundation elements are expected to reach their plastic moment, thus the reason for dividing by an "R" factor reduced by half, which effectively doubles the design forces.

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

 

[r13]

Rod,

So simply to say, there is a safety factor of 2, is it?

 

[BridgeSmith]

It's not a safety factor, per se, r13. It's more of what we'd call an overstrength factor. The actual ultimate moment capacity of a column or pile is considerably more than the moment at first yield (elastic capacity), so when the goal is to ensure that under a loading too large for the structure to survive elastically, such as a large inertial loading due to a seismic event, that the failure is where we want it to be. That way it is a ductile failure that keeps the structure from total collapse, and in a location where the damage is visible and the remaining capacity can be assessed. We accomplish controlling the point of failure by ensuring that the desired failure point is much weaker than other parts of the structure.

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

 

[Guest090822]

I don't know what provision of AASHTO you think you are violating? Take a look at 3.10.9.1 (7th Edition is all I have in front of me). For a single span bridge you only need to meet the minimum support length requirements, and design forces for bearings follow the equation for As. However, for an integral abutment you don't have bearings and don't need the minimum support length since the girders are cast into the abutment. You also don't know exactly how an IA behaves so assuming a hinge in the pile versus the abutment rotating about the top of the piles is a moot point - nobody knows what happens. AASHTO is virtually silent on Integral Abutments so don't read too much into the code as it isn't very helpful.

Again - check out the New York State Department of Transportation's Bridge Manual. They have a whole section dedicated to IAs and have moved to an empirical design.

 

[BridgeSmith]

You also don't know exactly how an IA behaves so assuming a hinge in the pile versus the abutment rotating about the top of the piles is a moot point - nobody knows what happens.

Actually, some full-scale testing was done (in Texas, I believe) that showed 12" of embedment of an HP 12 x 53 pile into a reinforced cap was sufficient to make it hinge without breaking the concrete. We recently increased the standard embedment we use to 2' (since a fair portion of ours are HP 14 x 73), so we're fairly confident the piles will hinge if the rotation is large enough and the soil is stiff enough (otherwise they just flex elastically). Some of our existing integral abutments with #4 bars at 12" on the faces have separated at the cap to diaphragm interface, so we've bumped up to #5 bars at 9" or 10" spacing for the bars crossing the interface.

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

 

[Guest090822]

BridgeSmith - are you referring to a hinge developing where the piles meets the abutment?

All the models our state developed showed a few different ways that IAs could behave. There are a lot variables to condsider so while that’s great that a full scale model was tested, it doesn’t really predict how all IAs will perform. NYSDOT used to limit the reveal height at the abutment to 4 feet. Other states went higher. There will be a different behavior between a low reveal and a high reveal. Now that more states are building these I’m hopeful that in the next 20 years there is some good data out there to implement something into the code so there is consistency.

 

[BridgeSmith]

All the models our state developed showed a few different ways that IAs could behave. There are a lot variables to consider so while that’s great that a full scale model was tested, it doesn’t really predict how all IAs will perform.

All I was saying is that if the embedment is sufficient, the end of the pile doesn't break out of the abutment, which is what we want to avoid. There is a continuum possibilities , ranging from no bending, to fully plastic, all of which, according to my understanding, are acceptable. The limiting case is a fully plastic, so if we assume that is the moment the piles transfer to the cap, we can design the abutment to ensure that the abutment remains intact. The only unknown of interest is the actual rotational resistance, which will affect the deflection calculations for the girders in the end spans. We've found few instances where the affect of the restraint on the deflection calculations was sufficient to warrant an assumption other than the pinned connection we use for design of the girders.

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