Torque on a series of leveling screws 1

[TehMightyEngineer]

Got a nice job that came across my desk for a bridge deck replacement. They want to utilize precast deck panels on the existing steel bridges with ultra-high performance concrete in closure pours between each panel. The panels are fairly complex pours but otherwise should be a great project.

The item I'm looking to get a second opinion on is the panels are heavily weighted on one end from a precast barrier and need to be leveled set to the proper elevation and load distriution using a number of leveling vertical adjustment screws. The specification requires that the panel load be evenly distributed by adjusting the torque on each leveling vertical adjustment screw to a torque schedule that we're to submit with our shop drawings. The tolerance for the final installed torque is to be +/- 15%.

Here's a picture of what these leveling screws will roughly look like:

The question I have is regarding the inherent difficulty in calculating bolt tension loads based on measured torques. I can obviously calculate the torque required to raise and lower each leveling vertical adjustment screw up given a load on the bolt, but based on the RCSC recommendations these calculations have errors due to friction forces and bolt tolerances that exceed the tolerance in the specification for the contractor. This says nothing for the obvious real-world application of the bolts. Any dirt on the bolt or poorly lubricated threads will show a much higher torque which would result in the bolt being under or overloaded.

Is such a torque schedule feasible given the constraints? Is there a way to improve accuracy of calculating the torque that each bolt will need to evenly distribute the load?

I'm waiting to hear back from the engineer of record but figured I was going to make this post either way.

EDIT: Fixed terminology.

Professional and Structural Engineer (ME, NH)
American Concrete Industries
www.americanconcrete.com

 

[Ron]

The spec conflicts itself. Do you want the ends level or do you want the same torque in each bolt? You're not going to get both!

 

[TehMightyEngineer]

Yeah, I say level but it's actually a multi span panel, with the barrier side cantilevered, and sloped at a varying grade (up to 5%) in the longitudinal direction and a 2% cross-grade transverse.

So, no, the bolt torque will be different for each bolt depending on the tributary load for each bolt. I believe that's why the engineer wanted a bolt torque schedule so we can say "turn bolt A until you get X ft-lb, turn bolt B until you get Y ft-lb...".

If it were me I'd just say get the slab in the correct position then turn each bolt with a torque wrench until you saw a steady rise in torque indicating the bolt was taking load. Continue around the bolts until they all are taking load and you're done.

Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[Teguci]

Some thoughts:
- If you are "leveling" then the torque doesn't matter. Ideally you pick 3 points to level and your load at each point will be statically determinate.
- If you are distributing the load (which I think is what they mean) then you have many of these adjustments to make along the beam and you will need to consider the flexibility of the precast panel and the beam as you make your adjustments. You'll wind up torqueing one rod to value and once you do another one, the first one will no longer be at the right torque. Is that what they want you to consider on the Torque schedule? You will need initial torque and final torque then.
- I keep thinking that this is a terrible solution. Why not shim or weld? Do they really want to allow the leveling bolt to "slip" along the top of the beam? I know I don't have the full picture, but I really don't like this zoomed in part of it.

 

[TehMightyEngineer]

If you are distributing the load (which I think is what they mean)

Sorry, yes, my terminology is off. Distributing the load is the correct term for sure.

then you have many of these adjustments to make along the beam and you will need to consider the flexibility of the precast panel and the beam as you make your adjustments.

I was thinking I might have to make a FEM model of each panel given the items you mentioned. I hadn't considered the beam flexibility but you're entirely right that this is a consideration as I can't consider them rigid supports (though they will be relatively very rigid given they're only supporting the precast panels).

You'll wind up torqueing one rod to value and once you do another one, the first one will no longer be at the right torque. Is that what they want you to consider on the Torque schedule? You will need initial torque and final torque then.

I was thinking I would have to specify a bolt torquing pattern so that they avoid creating high-spots but you're entirely right that they will need to go back and verify the first bolt is at the proper torque.

I keep thinking that this is a terrible solution. Why not shim or weld? Do they really want to allow the leveling bolt to "slip" along the top of the beam? I know I don't have the full picture, but I really don't like this zoomed in part of it.

No I agree with you. The end goal is this is a new bridge deck on existing steel beams. They're also replacing the bearing under the girders as well. My assumption is they can't anticipate the camber the steel girders will have and they want to be able to use the adjustment in the precast panels to set the final deck elevation. I'm assuming they don't want to use shims as they figure a leveling screw will be a more rapid method to set the panels at the appropriate elevation. In addition they're placing the ultra-high performance concrete under the panels to form a haunch that will be the final support for the panels onto the steel beams, so these leveling screws are not the final support for the bridge deck.

Overall this is just guesses though and really I don't have a much bigger view than you do. I'm not the EOR on the project so I have to get any changes approved. I'm definitely thinking that there must be a better way to get an even load distribution than using bolt torque, though.

Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[BUGGAR]

How accurate do they want to be? Are they considering relaxation creep that will tend to redistribute the dead loads?

 

[Brad805]

That detail seems overly complicated and will require a lot of manual effort assuming the PC panels are in the realm of 30,000lb.

 

[TehMightyEngineer]

BUGGAR: There's no tolerance listed for the elevation of the bridge deck panels; though they do give a detailed table of target final bridge deck elevations to 0.01' accuracy. Here's the actual spec. requirement:

The panels shall be set to the elevations detailed on the plans. Final panel elevations
shall be attained by adjusting the torque on leveling screws to promote an equal distribution of
panel dead load to all girders. The torque schedule shall be submitted with the shop drawings for
the panels. The torque tolerance shall be +/- 15% using calibrated torque wrenches. Panels shall
not be adjusted to final elevations until all panels for a span are in place.

They do not appear to consider creep, though the construction schedule is quite tight and I imagine that typically the closure pour which will support the panels will be done shortly after the final leveling.

Brad: The typical panel weight will be around 20,000 lb. I agree it will be more work but it does seem to be equal or less work than shimming it.

Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[Teguci]

This is what I was thinking for a welded connection (I don't think there is a reasonable way to avoid composite action). Another thing I noticed, the threaded stud extends outside of the precast form. They will probably want to provide embedded threads which can receive a bolt.

 

[dhengr]

TehMightyEngineer:
I agree with the others, you can have level (elevations) or you can have constant (identical) leveling rod loads (“evenly distribute the load”), but not both. And, from the way you’ve described the precast panels you won’t get the latter in any case; not with a continuous panel (“multi span panel”) over several supports and one edge/end cantilevered with a heavy barrier on it.

I think you have to help the EOR rethink and rephrase his real needs. And, what they’ve shown so far is a very expensive detail (how many per panel?) and a very expensive and complex process. And then, bolt torque is just not the way to do it, and then remove the bolt and patch that hole afterwards? You want to bring the panels to the proper elevation, prior to the casting of the haunches and filler strips, which then become the final mating and bearing btwn. the stl. girders and the precast panels. And you want to do this without hurting the tip of the girder flgs., at discrete temp. bearing points, or hurting your panel supported at only 6 or 8 or 10 points. How far can your panel span btwn. these temp. bearing points without too much deflection? What is the max. load your panel can take from one of the bearing points? I would think the slab elevation and its tolerance is the important thing not the load on the leveling rods, although there may be some max. for this too.

Show us a plan of a few panels and their support girders and these discrete leveling rod locations on the panels, all dimensioned and in proportion. Show the barrier weights and locations, and some sections through those pour strips and the haunches. What’s the range of that vert. gap btwn. the top of the stl. and your panel?

 

[TehMightyEngineer]

This is what I was thinking for a welded connection (I don't think there is a reasonable way to avoid composite action). Another thing I noticed, the threaded stud extends outside of the precast form. They will probably want to provide embedded threads which can receive a bolt.

Hmmm interesting. There's actually bolt pockets cast into the precast slab that will also be used as access holes to fill the haunch between the steel beam flange and the precast panel but will also have steel studs for true composite action.

As it's a public project you can actually download all the bid documents of the MaineDOT website so I'll just provide those here.
Here is the plans:

http://www.maine.gov/mdot/contractors/projects/2015/020483.00-fairfield/fp020483.00.pdf

And here is the specs:

http://www.maine.gov/mdot/contractors/projects/2015/020483.00-fairfield/bp020483.00.pdf

There's two bridges on the same project so ignore the second set of plans.

Sheet 23 of 32 shows the typical plan dimensions of a panel. Sheet 25 of 32 shows a typical transverse section and sheet 26 shows a typical longitudinal section.

I agree with the others, you can have level (elevations) or you can have constant (identical) leveling rod loads (“evenly distribute the load”), but not both. And, from the way you’ve described the precast panels you won’t get the latter in any case; not with a continuous panel (“multi span panel”) over several supports and one edge/end cantilevered with a heavy barrier on it.

You're entirely correct, I was using the wrong terms but definitely meant load distribution and not leveling.

And, what they’ve shown so far is a very expensive detail (how many per panel?) and a very expensive and complex process.

Actually this wont be too expensive for us and we knew it would be more than a typical deck panel project (this is not our first rodeo) so we're covered easily. Though we're trying to avoid items sticking out of the forms and will probably use cast-in threaded coil inserts instead of the headed studs shown on the contract drawing (assuming approval from the EOR). You're probably right on this being a big headache for the contractor though and I could easily see them not even bothering checking torques and just getting it set at the proper grade and moving on. Can't say I would blame them either.

How far can your panel span btwn. these temp. bearing points without too much deflection? What is the max. load your panel can take from one of the bearing points? I would think the slab elevation and its tolerance is the important thing not the load on the leveling rods, although there may be some max. for this too.

This is actually where my scope begins. Stripping loads, transporting, lifting and placing, and these vertical adjustment devices design is all under my scope and needs to have stamped design calculations submitted to the EOR. The torque schedule is supposed to be included with our stamped shop drawings and this is where my sticking point began. Definitely going to address panel deflections, stresses, and other items in the near future.

I agree that my thought was that setting the deck elevation and grade was more important as the ultra-high performance concrete will act as grout and support the panel. In addition we're removing these vertical adjustment screws after the UHPC hits 6.5 ksi so there's really no need for these to be evenly distributed except for steel girder deflections, panel strengths/deflection, and capacity of these vertical adjustment screws.

Show us a plan of a few panels and their support girders and these discrete leveling rod locations on the panels, all dimensioned and in proportion.

Done, see above. The gap is nominally 1.5" if I recall correctly, our drafter is working on the preliminary shop drawings right now and we'll have a better idea of the actual anticipated gap between the panels then.

Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[BUGGAR]

Thanks for sharing the Construction documents. After a quick scan, I didn't see the weld specified for the threaded coupling for the adjuster. Chose this weld and the coupler threads so welding distortion doesn't bind it up. We had to chase a bunch of threads in a similar situation. That was expensive.

I also saw hardwood shims holding the panels. With hardwood wedges and a large hammer and the right man on the adjuster, you could probably wedge those panels to make the "screw jacks" read any load you wanted.

 

[TehMightyEngineer]

Great call BUGGAR I'm going to give a star for that! I hadn't even considered the effects of heat distortion on the threaded coupling. You're entirely right that it could derail everything like crazy if we ordered 200 of those vertical adjusters from a fabricator (or made them in-house) and then they all bound up when put under load.

Yeah, I'm not even sure why we have hardwood shims combined with the vertical adjusters. You're entirely right that they could adjust this with the hardwood shims and get all sorts of crazy effects.

Hopefully the EOR will call me back today.

Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[TehMightyEngineer]

Talked with the lead engineer on the project and got some good information.

The intent the engineer had was to not invalidate their assumption of even load distribution to each girder. I expressed concern that this isn't at all realistic for these panels as they're not tied together at the center line of the bridge, but that's not really my concern.

The engineer clarified that the spec was more or less copied from other similar jobs from other DOTs and the details of actually establishing it was left to the contractor/fabricator. In the end it was clear that the intent of it was to ensure that there's not a bolt left floating or similar unbalanced load.

So, in lieu of a torque specification, I'm going to submit a bolt torquing pattern and a specification for the contractor that they set the panel elevation then torque each bolt to verify it not floating and, if it is, torque it until they see a rise in indicated torque which should show that the bolt was now supporting it's tributary weight of the panel. This should meet the intent of the specification and be much simpler and easier to perform for us and the contractor.



Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[bridgebuster]

The danger of blindly copying and pasting. let's face it, we're all guilty of it from time to time.

Teh - You should ask the DOT to issue a written response to the question. This way you have something to fall back on if the issue comes up during construction.