Irregular Box Beam Bridge 4

[JackTrades]

As you may see from the attached sketch, I have an interesting problem related to the design of a prestressed box beam bridge incorporating a large roadway radius. I have three concerns (at least until that time when I realize that I really have many more):

1. What should I do about the interface between the main and secondary bridge structures? I'm concerned about differential deflection at the junction and am uncertain that the 6" reinforced topping slab would remain uncracked. Perhaps an expansion joint?

2. Do the LRFD code methods for the distribution of live load make sense when the trucks are actually driving across the beams at an angle?

3. How would one go about checking the deck capacity at the guardrail as it follows the radius of the roadway?

Any comments, suggestions, or even dope slaps will be appreciated.

-Jack

 

[civilperson]

Tie the two bridge segments rigidly together and use NO joint. The maximum number of lanes should be used ignoring the barrier and curb geometry for lane loading. Use lane limits outside of the curbs and barriers parallel to the straight curb for maximum loading. Any other lane pattern will cause less loading.

 

[crossframe]

Some quick thoughts on your three area of concern:

1. I would think an expansion joint would be preferable. I think JackTrades' concern is the deflection of the exterior 75' box beam (at about 0.3L) compared to zero deflection at the pier supporting the 32' box beams. I don't think you'd be able to keep an enreinforced 6" slab from cracking if they were tied together.

2. My understanding is that the live load distribution factor equations in AASHTO LRFD are approximations for line girder analyses. This is why grid and 3-D analyses can produce distribution factors that will differ from those derived from the AASHTO equations. This looks like a case where a "more refined analysis" would be warranted.

3. I'm guessing this concern relates to the railing forces, and I apologize if this is an incorrect assumption. A 6" unreinforced topping slab is not much of a deck (at least in the traditional sense). Bridge rail resistance will come from the box beams through curb (for steel rails) and barrier reinforcing embedded into the beams, not from the deck. This reinforcing can follow the guardrail radius but may require you to have solid sections within the voided area to develop these bars. I don't see how this will be much worse than a typical rail connection.

And some additinal thoughts...

4. There is a lot of unused deck space in the proposed layout. I've been told by DOT personnel that this sort of unused area "will only lead to trouble" because of the possibility of people walking behing the guardrail and having access to an unprotected edge. Is there a better solution through the use of wingwalls or retaining walls? Dirt generally doesn't care which way the traffic is moving.

5. Tell the roadway guy to move his radius off the bridge.

 

[Qshake]

Sorry to be a spoilsport, but that is poor work by the Civil group.

There are many other solutions to such a problem than to dump them on the bridge group whose work will be more expensive in the long run.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[Drumchaser]

This does not appear to be a cost effective or aesthetically pleasing structure. Maybe I am not seeing the drawing correctly.

Some years ago, I was involved in a "similiar" structure with radii. (above grade intersection) We used plate girders for the radii and "intersection" portion of the bridge,and box beams for the tangent sections (transitional in width) It was a little troublesome to sequence (placing the concrete slabs) but it worked out nicely.

 

[JackTrades]

Thank you all for the informative and helpful replies.

civilperson- I'll analyze live load as affecting the entire bridge width.
Crossframe - I intend to use double railings - at the actual sides of the bridge and also placed as deliniations along the roadway. The 6" slab is reinforced and I note from a DOT detail that they can get, or assume to get, the requisite capacity by attachment of the barrier to the slab alone. But I'll check this assumption more throuoghly. I've also reconfigured the bridge so that it uses a wall abutment at the larger radius end which will elimaite the second "bridge." I've sent this revision to the hydraulic engineer for assessment.
Drumchaser- I think that you are probably seeing the drawing correcly and you're right - it's neither aesthetically pleasing and possibly not cost effective. I originally tried to fan or "splay" bulb tees or plate girders but their depth caused hydraulic conerns. I tried a ledger beam with the same results. Too deep. With the box beams, I can reduce the depth to 33", plus a 6" slab, or less. But perhaps I've misunderstood the configuration you're describing.
Qshake- I'm with you on this but there's apparently the need for the roadway to serpentine for terrorist protection purposes. This is at the entrance to a militray facility and they don't have a lot of room to work with.

Thank you all again for the help. I'm not a bridge engineer, just a structural that gets the occational bridge design, so assistance like this is golden.

-Jack

 

[Qshake]

Jack -

have you considered a concrete slab bridge? With this type of bridge you can pretty well fit the plan shape necessary. However, as you may suspect, the cost of such a bridge will increase due to the labor intensive forming. On the other hand, if it's located in the west or southeast this is par for the course.

Such briges are either box girder or voided slab to reduce weight. Structure depth can be as shallow as 24-27".

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[JackTrades]

Oshake - adding to the complexity of the design, there is also a very short period of time that the DEQ (State Environmental folks) will let us work within the creek banks. The pre-cast boxes will probably be seen as the way to go.
Also, with reducing the span to around 50 feet, I believe that I can reduce the box depths to 27" with the 6" slab, even with the HET overload traffic.
Thanks.

-Jack

 

[Drumchaser]

Jack


What about slab beams? Not sure what the design max. length available is. Could require another bent(s).
Although if you are reducing spans to 50 ft, they are tailor made. Slab beams are pretty thin, and you can overlay with ACP (option to conc. slab)

I still say curved plate girders (and tangent) utilizing PMD would be a good option. There would be no beams protruding out from under the bridge (which I have not encountered) If I saw your drawing correctly.

Not sure what your local regulatory agency hypocrisy is there, usually the corps. (USACE) allows plenty of time to complete a bridge. It is usually the enviro. doc, EIS, and clearance from the corps. that takes forever.

 

[JackTrades]

Drumchaser - I'll look into slab beams. Thanks for the suggestion. The curved plate girder may very well be the way to go but, without much of a background in LRFD steel design and an RTA of mid-May (along with several concurrent smaller projects) I'd better stick with a more familar material type. I've informed the architect (this bridge is part of a much larger project) about the excess bridge deck. Maybe he can put a planter on it or something.
I'm not sure of the reason for the DEQ construction contraints but will check into it. Thanks again.

-Jack

 

[Qshake]

Jack - One such way to address this and use conventional bridge construction is to extend the endbents (abutments) and use I-girders that are placed normal to the endbents where possible and that are ultimately skewed as you move toward that large radius curve. This is done frequently with Single Point Urban Interchanges (SPUI). The benifit of the I-girders is that you can use 3" precast prestressed concrete panels which serve as the bottom of the deck and the formwork for the 5" cast-in-place concrete deck. The panels can be formed to match the skew for the variable spaced girders.

This is quick and inexpensive. It also eliminates the two separate structure issues.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[JackTrades]

Qshake - I took Crossbeam's advice, re-looked at the abutments, and came up with what's shown on the attached sketches. I'll look at using the I-girders as you've suggested as see if I can get an acceptable structure depth. Thanks.

 

[Drumchaser]

Jack

Looking at your drawing....at the opposite side of the creek from the radius...the roadway, widen the roadway to fit your bridge beam width (72'-0" shown). Transition your roadway from existing width to bridge width.

Eliminate the 32' box beams and associated bents. {This should allow for the cost of addition box beams, see below}

Shift the entire bridge towards the radius, by adding a few more beams (72 footers) you should be left with only a small piece of the radius, which should be negligible to support with conventional methods. You could change the radius to assist.

I am familiar with box beams...I do not recall any having overhang: all thread inserts cast inside (outside edge of outer beams only) to permit edgeform placement.


Also, the outside beams are cast with the bridge rail steel inserted. (due to the thin slab thickness, "our normal" bridge rail reinforcing steel will not fit).

[Always include a North arrow on all drawings, preliminary or otherwise.]

Sorry to keep nagging on this but...I do not think it would be wise to design (or construct) a bridge that has sections of beams sticking out.

Hope this helps with your "design"
----------------------------------------------------------

Mr. Qshake
Good point on the I-Girders, they are much more stable. I believe they can be cast as short as 28". Though I have only seen the larger variety.
Our PCP (panels) are cast a min. of 4" thick, with a min. deck thickness of 8". What I prefer most about using PCP's is the indication (tell tale)of final deflection after the panels are set into place.

 

[Qshake]

The last schematic is good but I would not place the end spans on fill. I would use a one way slab spanning the longitudinal direction. A sleeper beam (grade beam) at the furtherest end for support and near the abutments support the slab on a corbel. This will eliminate any work the contractor has to do to shape the fill underneath the bridge.

If you're span is really down to 32' you might try half I-girders (popular in the southeast US for short spans). This will reduce the total number of beams required. You'll still use PCPs.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[JackTrades]

Thanks Qshake.

I didn't really intend to support the right side of the slabs on fill but I didn't add the support beam on the drawing. I may actually have to support that right side on piles as there is a weak clay layer about 25' below the creek bed and very prone to inducing settlement. Anticipated settlement with only 1800 psf applied was 2.5". But I may good with a sleeper beam if I can disregard LL. What are your thought's on doing so?

I'll look a the half I-girders. I ran across them recently on a local precaster's web site but didn't think to consider them for this.

 

[Qshake]

Opinions vary, of course, on the LL or rather the impact of LL on sleepers and corbels. Live load is a given really, but impact is optional.

Some argue that the typical resultant dip in the road at the bridge approach warrants the use of impact. Others, of course, dismiss it due to dissapation through the element right into the soil.

You may want to check the practice of the DOT in the area before concluding for yourself.

Good luck.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.