Cross-Frame Design Questions!!!

[bridgeengineer2007]

Hello All,

I am designing my first cross-frame (K-type) for a straight, two-span continuous bridge under LRFD specs. I am looking at a typical intermediate cross-frame, and a pier cross-frame. I have several questions regarding each:

Intermediate: Per AASHTO LRFD, it says that wind loading on the top half of the superstructure is transmitted directly to the deck. Therefore, I was designing the bottom strut for wind on the lower half of the beam. I have solved the diagonals as truss members and found their equivalent loads. Is this the only load I should consider?

Pier: Do I need to design the top strut and diagonals for transfer of forces (WS and WL) back down from the deck to the bearings? Or are the girders considered stiff enough to transfer the load?

It seems like cross frames on a straight bridge are only for counteracting wind during the construction phase before the deck is hardened, but I wanted some other opinions.

Finally - what is the purpose of designing the abutment concrete diaphragms for wheel loads? Isn't the deck designed to carry loads to the beams through the bridge? I never understood why two wheel loads are placed over the end diaphragms...


THANKS!

 

[bridgebuster]

Diaphragms tie the superstructure together to distribute loads. Although some claim they serve no purpose after the deck is completed.

Regarding the design for wind, is this a DoT project? If so they should standard details to eliminate the need for designing them.

Designing diaphragms for wheel loads - we've always done it this way; and they're over designed - however, the end of the deck and joints tend to take quite a pounding. The additional stiffness helps.

 

[gwynn]

The governing load for the interior diaphragms may be the force required to brace the compression flange during construction or that required to resist the torque on the exterior girders during the deck pour. Please take into consideration the rotation of the exterior girders when spacing the diaphragms, this may make things much easier on the contractor.

I would design the pier diaphragms to transfer wind loads etc. down to the bearings.

 

[graybeach]

I also would design the pier and abutment diaphragms to transfer lateral loads to the substructure. One thing I still see in some DOT details is bearing stiffeners that are not welded to both flanges being used as connection plates for diaphragms. Maybe I am missing something, but this just seems wrong because there is no load path to carry transverse loads from the deck to the substructure. Regardless of what the DOT details show, I always show welds at the top and bottom of bearing stiffeners used as connection plates.

 

[HgTX]

I agree; there should always be positive connection at the top and bottom of any connection plates, or else you are risking failure in the web due to out-of-plane bending. I advise welding and handling whatever fatigue detail there may be (is that even a concern at the bearing?), rather than monkeying about with bolted connections to the flange.

Typically intermediate diaphragms are not designed for primary load except in curved structures or sometimes extreme skew. I don't know that I'd say they don't do anything after the deck is poured, but they are bracing, not main members. (External cross-bracing in tub girders, from what I understand, doesn't do anything after the deck is poured, and I have seen many cases where they are designed only as temporary members and removed after decking.)

Hg

 

[RHTPE]

No one has yet mentioned the fact that most bridge decks overhang the spandrel girder. That overhang concrete has to be formed (can't be SIP deck) and typical proprietary overhang brackets are used to support the remaining falsework & formwork. The loads during concreting and finishing applied to these brackets is transferred to the top flange of the spandrel girders. This applies a torsional load to the girder. What helps the girder to resist these forces?

Not to go on a rant here, but far too often this condition is thrown at the contractor to analyze and design, when it really should be a consideration in the design of the bridge itself.

I am seeing some bridges being built with very deep plate girders with thin web plates and substantial deck overhangs. Contractors do not want to install overhang brackets at 24" c/c or less, so the forces applied to the girders from these bracket can get rather high. Aside from the torsion applied to the girder, the brackets' diagonals "kick" against the web, often dimpling the web slightly.

This condition is given too little attention in the final design of the girders. I can recall a rather long bridge across a lake between 2 states where the contractor was forced to install temporary struts between the girders' bottom flanges and ties across the girders' top flanges to mitigate the torsional force applied by the overhang falsework. Had this been considered in the design of the diaphragms, they could have been lighter but more frequent, eliminating the need for temporary bracing, and the schedule and profit margin for the contractor would have been much more favorable.

Ralph
Structures Consulting
Northeast USA

 

[OSUCivlEng]

Cross frames are need to brace the bottom flange in negative moment regions even after the deck is hardened.

RHTPE

You are right about the overhang brackets. I always include a beam bracing plan that requires 4x4 hardwood struts at 4'-0" centers at the bottom flange and 1/2" tension ties at the top flange. I also require the kicker leg to bear within 6" of the bottom flange. Also a good rule of thumb is for the overhang to not be any larger than the beam depth, this keeps a maximum 1:1 slope on the diagonal of the brace. The DOT I do work for has started requiring this after they ended up with a pretty wavy deck due to excessive overhang formwork deflection.

The Kansas DOT has come up with a pretty good program for torsional analysis of exterior girders, TAEG. I think TAEG 2.1 is the latest version.

 

[RHTPE]

OSUCivlEng - I have used the Kansas software. It is limited in its application though.

There is a deck overhang limit for 'off the shelf' brackets though. There are vendors that can do custom applications, but the longest 'std' bracket is around 7', and a maximum vertical leg height of 5'-6. Space has to remain along the edge of the deck for guardrails, a bit of walkway & edge form braces.

The typical applications I've dealt with having thin webs were often a case of not being able to achieve the vertical leg height to reach near to the bottom flange because of the equipment that was chosen by the contractor.

With respect to torsion, I prefer provide my calculated falsework reactions to the Bridge Designer and request that he evaluate. When dealing with horizontally curve girders or bulb tees or even tub girders, I feel the Bridge Designer is in a better position to evaluate the girder that he designed.


Ralph
Structures Consulting
Northeast USA

 

[gwynn]

Although I did not address the issue of the diagonal or bottom foot from the overhang jack landing mid point on the web of the exterior girder, I did mention that the torsional loads and deflections need to be taken into account, though not as explicitly as you have.

I agree that both are an overlooked issue in bridge design - a typical example would be that almost every steel girder bridge I see has a load diagram for the proposed overhang formwork with the diagonal coming in near or at the bottom flange. On deep girders this is not realistic. The standard overhang brackets (Dayton C49s and NCA HFOBs) have a maximum depth of 1270mm. With the exended version you can get that to 1780mm and the exended version may be for purchase only. There are a few other available overhang brackets typically available in North America, but the only one that would result in a deeper bracket would be from EFCO (in this case more expensive and where I practice for purchase only).

With deep girders and substantial overhangs it is fairly common that the contractor will need to install timber struts between the web of the exterior girder and the junction of the web and flange on the first interior girder just to limit the defelections at the edge of the soffit.

Not to go off on my own rant, but the big issues for the contractor in terms of deck overhangs are:

1. Wide overhangs on shallow girders. This is where the torsional loads mentioned by RHTPE come into play the most. Ideally the contractor would like to see overhangs limited to ~2/3 the girder depth or 5', whichever is less (for large bridges, it is expected that you may go over 5', but for a typical structure it would be nice if you could avoid it). When I see 1.9m overhangs on 0.6m deep girders I cringe.

2. Flange and web thickness. If you have a small overhang the contractor is going to be placing up to an 18.9kN point load on the top flange at every overhang bracket. Aside from NU girders, this generally is not an issue. If you have larger overhangs, the contractor will want to place a 36kN point load on the edge of the top flange at every overhang bracket. Check that the flange and web can handle that.

3. If you are using NU girders, customise them and make the top flange 2" thicker. Some jurisdictions do not allow anything related to deck formwork to be cast into the girders. Placing the aforementioned 36kN point load, or even the 18.9kN one, on the edge of the flange is a no go with NU girders. Unless you know the contractor and know that they will get the inserts in, do it.

 

[OSUCivlEng]

Sometimes your hand is forced. For example I have a "friend" who is designing a bridge that has a 200' simple span plate girder that is 8' deep with a 3'-4" overhang. He had no input into the span configuration. The foot of the bracket is probably going to hit more than 2 feet above the bottom flange. The only thing I...err he knows to do is add more wood struts behind every overhang bracket.

 

[bridgebuster]

I have a rehab project under construction on which we are replacing the fascia stringers - about 8000 LF. There was no overhang on the original fascia. The new fascias will have overhangs from 2 to 4 feet; the stringers are W36x150. We don't tell the contractor what to do - there are standard notes warning him about LTB in the plans - but we suggested struts, based upon our analysis.

 

[RHTPE]

My issue with overhang conditions is really simple.

I'll provide a falsework design & details that work for the construction loads, within the constraints of the components involved. I will summarize the reactions applied to the girders from the falsework for the various load cases considered.

But since I am NOT involved in the design of the girders, I prefer to have the bridge design engineer evaluate torsional effects on the spandrel girder. After all, he/she is more intimately in tune with the design parameters.

I often wonder if fewer (but heavier) diaphragms (cross frames) are actually more cost effective than more frequent, lighter diaphragms. Closer diaphragm spacing better mitigates the torsional effect from the falsework.


Ralph
Structures Consulting
Northeast USA

 

[OSUCivlEng]

I think that's fair. When you are designing a girder you can only make assumptions about the construction loads. Checking against the actual loads would be nice. However what do you do when your falsework design over stresses the girder and the girder has already been fabricated?

 

[RHTPE]

OSUCivlEng - The only time I have heard of it being a problem was with the Missisquoi Bay bridge in VT. To mitigate the torsion induced from the brackets, the contractor had to install tension ties across the bridge above the top flanges, along with timber struts on the bottom flanges. This bridge had a substantial overhang and the diaphragm/cross frame spacing was substantial. Don't remember the dimensions as it was some time ago and I was not involved in the falsework design. This was the first hint to me that the material savings of fewer diaphragms might just be false economy, given the material & labor costs to install the ties & struts.

Bridge decks using SIP metal deck have a bit of an advantage in that the metal deck provides some resistance to the tension needed to offset the brackets' horizontal load component. When the deck is done with removable forms the tension force gets more difficult to resolve. Dropping in timber struts on top of the bottom flanges is generally no big deal.

The peak construction load on bridge girders occurs during placing & finishing the concrete deck (obviously). Generally the highest load is confined to an area near the finishing machine and is transient, as it follows the machine.

Which give rise to another overhang issue: supporting the screed/finishing machine on the overhang brackets. I have actually had projects where the load from the machine applied to the brackets was impractical to accommodate and I had to insist to all involved that the rails the machine rode on HAD to be located over the spandrel girder. The bracket spacing would have been absurd (like 12-16" c/c). Many DOTs in the Northeast insist that the decks be machine screeded and finished. I understand why, but the means & methods to accommodate can be a challenge.


Ralph
Structures Consulting
Northeast USA

 

[gwynn]

@ OSU
That situation is one I can't see the contractor having an issue with. Maximum 4' overhang on an 8' girder is on the low side for overhang widths. Depending on web thickness, diaphragm spacing, finisher weight, falsework arrangement, etc., those struts may or may not be needed, but I will leave determining that up to you.

Bridgebuster
It may be different where you are, but many of the contractors I work with would look at that situation and bid based on hanging strongbacks under the stringers, putting a deck at the strongback level and building pony walls up to support the overhang soffit, eliminating the torsion on the exterior girder. Unless there are clearance issues. Bridge rehabs are a also different ballgame from original design in terms of what variables you can change.

RHTPE

Was the case where you had to place the finisher on the girders with concrete or steel girders, and was this using standard off the shelf overhang brackets? Just curious. I have seen some fairly significant deck finisher loads, but I have usually had to go to tight spacings and/or custom overhang brackets to accomodate them. The only exception being with NU girders, where the top flange could not accomodate the overhang bracket loads unless the finisher was on the girder.

General comments:

In general, aside from the above instance where underhung beams are used to support the overhang deck, I don't see many situations where the torsional loading on the exterior girder cannot be taken into account by the design engineer. The torsional load from the deck overhang concrete is known, the live load on the walkway can be taken into account and the weight of the deck finisher is based on the overall width of the deck. The weight differences between Bidwell and Gomaco machines are small, and both companies will supply this information if you ask. The spacing and geometry of the overhang brackets does not significantly effect the torsional loading on the girder.

I am seeing this done most of the time lately in some areas, with the location of the finisher, weight of the finisher and maximum walkway width shown schematically on the drawings. Though it is still always noted that the contractor is responsible for stability of the girders during construction. The only other thing I have to say on that, is that it the designer should know the standard practices of the construction engineers where they are. Here the construction engineer will check the girders regardless and they may be using higher load factors. Control of deflections at the edge of the soffit is left to the contractor, as the amount of deflection due to deformation of the girders will depend on the bracket spacing and geometry.

Closer diaphragm may result in a more cost effecient structure at times. But in some cases, generally with wider bridge decks, the load on the diaphragms during the pour is largely governed by the weight of the deck finisher. If that governs diaphragm design, unless the diaphragms are very closely spaced you may not see a large reduction in diaphragm member size.

I have seen a 5-10 cases where supplementary bracing of steel girders was required due to torsional loading on the girders. This was usually due to one of two things. The first would be what I mentioned above, where there is a large overhang on a shallow girder. Usually this has been due to an overpass having the turnout lane for an off ramp start midspan - much of the bridge is fine but as you near the abutment the overhang widths go up dramatically. The second is varying flange width in multispan bridges without varying the diaphragm spacing. There were two cases I can think of where we had to brace the first 20m or so at each end of a long span bridge, until the section changed and the flanges widened.

Concrete girders are a different story, but this thread seems to be mostly about steel girders so I will leave that for now.

 

[RHTPE]

Gwynn - I would love to share project experiences with you some time - sounds like we have some common interests.

The problems I've encountered with the finisher usually involves one or more of the following: shallow spandrel girders; excessive overhangs; contractors who insist on using brackets and/or deck material they have in the yard; lack of clarity wrt finisher wheel locations. It has occurred with both steel and concrete girders. Finding out after you've completed (and submitted) the details that the finisher WILL be riding on the edgeform is always annoying.

I think we've hijacked this thread enough from the construction engineer's perspective - we've kinda strayed from the OP's original question. My intent was to spotlight some of the problems that have to be dealt with by the contractor's (or vendor's) engineer with the hope that the bridge design engineers (& maybe a few contractors) might recognize how their design impacts the construction solution.


Ralph
Structures Consulting
Northeast USA

 

[pikku14]

Gwynn/RHTPE,

Interesting comments

Are most contracts design only in the US? Down under there's a lot more Design and Construct or Alliance projects so the designers have to work closely with the contractor and account for their construction methods etc. (well that's how it's supposed to work, doesn't always work 100% in practice)

 

[HgTX]

To continue the tangent, I may be misremembering, but I thought there was recent talk within AASHTO circles of setting out some limits of "normal" overhang bracket configurations that wouldn't require special analysis, and requiring it for others. (But still not putting it on the designer. But I may be misremembering.)

Hg

 

[RHTPE]

pikku14 - I can't speak for all projects in the U.S., other than those I have been involved with. Those projects all have been designed by outside consultants to DOT specs. The degree of involvement of the consultant with the construction means & methods varies from state to state and project to project.

HgTX - I am not aware of such a movement within AASHTO. I think it would be quite difficult to establish "normal" configurations, as girder depth, finisher rail location, bump-outs for poles and the contractors' choice of deck materials all influence the design of the brackets.


Ralph
Structures Consulting
Northeast USA

 

[gwynn]

RHTPE

Feel free to be in touch. There aren't that many around who do this sort of stuff, so it would be fun to swap experiences.

Pikku14

It varies by scope, location and owner in the parts of Canada and the USA I have done work. I would seperate contract arrangements for infrastructure projects into three groups from what I see. Local public roads could be either design build or design only, depending on where they are. Some cities/municipalities favour design build and some design only. State or Provincial tend to be design only up until a certain dollar figure. The largest projects (say $500 million plus) all seem to go design build, and some of these will also include the maintenance contract. Private roads tend to go the design build route as well. This could be entirely different elsewhere in Canada or the USA.

I agree, both that theoretically the design build jobs should go smoother and that this isn't always the case.

HgTX
I am not aware of this, either in Canada or the US. If I remember correctly, there are a couple of DoTs that have guidelines for overhang widths on the AASHTO prescast girders and I know there are MoTs that are working towards constructability guidelines for provincial projects, but I am unawware of anything ongoing that relates to the overhang formwork.