LRFD Bridge Design 1

[minorchord2000]

I am trying to gauge how widespread LRFD bridge design has spread throughout the United States.

I just attended a very intense LRFD Bridge Substructue Class by the FHWA (3 days). The learning curve seems steep to this pre-LRFD bridge engineer, but after talking to the other bridge engineers in the class, they all stated that the engineering hours per sheet will go up because of this new code. FHWA has mandated the use of this new code for all projects scoped starting in October 2007. I have not even attended the Superstructue Class yet.

Any thoughts or comments about this?

 

[civilperson]

If it was easy, anyone could do it.

 

[BradPL]

As far as how widespread LRFD is, I would venture to guess 100% for Federal Aid projects. Although I think FHWA is allowing some wiggle room for states in that if 'design' has begun it can be carried through using the original code. That can leave a wide interpretation as to when 'design' begins.

Several states have been using LRFD since its 1993 introduction while others have held off hoping the Feds would back off the requirement (Metrification).

We haven’t noticed anything too drastic in LRFD other than the Extreme Event Limit States. The truck collision with the substructure can be quite difficult to reconcile. Serviceability is also more strict with the crack width model in the 2006 Interims.

 

[minorchord2000]

Thanks, Brad. This is what I was looking for. Here in AZ, the state DOT is dragging their feet on this because the bridge engineers are not familiar with the new code. It appears that the new code is continually be revised with updates and amendments. We know it is coming, so we have to get up to speed.

Have you found that is it costing your firm more to do these LRFD designs that using LFD or ASD?

 

[BradPL]

A lot of the updates and amendments that have been incorporated into the LRFD spec were the contentious or controversial issues that were more or less put off for a later date while states clung to the Standard Specification. I think there will be a continuous stream of interim revisions now that LRFD has states’ full attention particularly where there was a radical departure from past experiences.

There are some design costs associated with adopting LRFD. However, much of the commercially available software has been dual specification for quite a while to appeal to designers in both the LRFD and Standard Spec states. That by no means insinuates that it’s a matter of flipping a toggle switch as many templates, design aids, and spreadsheets have to be updated to the new spec.

The funny thing is I encounter the most difficulty with the code-independent issues like clearances, geometry, foundation conditions, constructibility, etc. If you can work through those problems, LRFD is cake.

 

[Umassengr]

From what I know LRFD design will be required for all new projects recieving federal money starting in October of this year. It sounds like it depends on the interpretation that your DOT is using as to when LRFD should be used. If the structure design has not begun by October or if the project has recieved federal funding for engineering by October.
My state DOT's thinking is that if the project has been awarded and preliminary engineering funds have been spent on the project, then it does not have to be designed with the new code. Only projects awarded after October 2007 will be designed using the LRFD code. This allows the DOT to buy time and the first LRFD projects will not be designed until mid-2008 at the earliest and more likley closer to 2009. Some projects with a fast burn delivery method such as design build may fall through the cracks though and have to be designed with the new code.

 

[Dinosaur]

In my state, the decision was anything that does not have an approved TS&L before October 1st will be LRFD. Anyhting going to Ad after 2013 will be LRFD regardless of completion status. You would think the second requirement would be irrelevant because of the first, but sometimes politics moves projects off into the future. Anything else will be considered on a case by case basis due to complexity and whether it is a consultant design or in-house.

I do agree that foundation conditions or the Corps of Engineers or something can be a bigger pain in the end than the new code. But I believe the new code has some real problems with consistency. I was looking at something the other day and asked the design group about the new Code. They responded, there are no changes to that design method in LRFD. Translation ... this component is still designed by WSD.

The argument that much of the work required by LRFD goes on inside the computer is extremely dangerous. If folks can't check an answer by hand, then we are running full steam ahead with our eyes shut. I have caught serious errors before we had LRFD because I knew about what the answer should look like and I could get an answer with only a calculator and a pad of paper. LRFD will only widen the understanding gap between reality and theory. I fear for the future.

Good Luck

 

[Andy10]

I agree with Dinosaur and his wisdom has obviously come from observations over his career!

We (our province - yeah north of the 49th //) have been using the AASHTO LRFD for our designs since the LRFD code was first introduced. We have found that the code has become somewhat more complicated and is leaning more toward the heavy use of computer software for the design as Dinousaur was mentioning.

The software that we use (BRASS LRFD - girder/culvert as well as OPIS, among others) does provide enough back up to be able to trace and check the work quite well, although this can be very labour intensive depending on how thourough you want to go. Going into a design blindly trusting the software to do what we as Engineers are trusted and responcible is down right unspeakable!

Obviously one must have a good handle and understanding of what you expect to obtain in a design prior to using the software is a must. One must have confidence in one's tools prior to using them. As such we thouroughly test a new release with a previous design and evaluate the differences (if any) and check the calculations. Once we have confidence in the software we perform less thourough checks. Bugs are always present in any given program not to mention the most common problem which is the wrong input.

In our experiance we have found that the LRFD method of foundation design have produced excessive foundation designs (very conservative => expensive). As a result we do not use the foundation design of the LRFD code for design. As a result we have gone back to a more conventional design method for the foundations.

 

[bridgebuster]

Dinosaur - Amen. It's difficult to check a lot of the LRFD designs by hand.

Andy10 - I ran into the same problem with foundation design about seven years ago. At the time, I called a professor from Rutgers who worked on the code. He said LRFD still wasn't ready for substructures. The client agreed to use LFD using loads obtained from LRFD.

 

[yakpol]

LRFD Foundation design sometimes produces too conservative designs. The geotechnical engineer shall have a very good understanding of the code to provide you with appropriate soil parameters and resistance factors. There're a few inconsistances and vague explanations to add confusion.

After year of iterations and changes steel design by LRFD became very clear and consistant. Concrete design covers many more subjects than 17th edition did.

Open EuroCode and you will find out that AASHTO LRFD is a breeze. Building codes change every two years (offten drastically). So, I would not complain.

Yakpol

 

[VoyageofDiscovery]

I am surprised that the LRFD method results in a more conservative design, my experience with Limit States Design in Canada is opposite. What is the point in a new method that appears to reflect a lower level of confidence?

That being said I still have personal reservations about Limit States and LRFD methods.

 

[BRGENG]

In Missouri the DOT is now requiring LRFD on all bridges on state routes but not county bridges, even if federal money is used. Our office has decided to switch LRFD for all projects. You would not design a building with the 1993 BOCA code if the local jurisdiction did not require a code, a good engineer would use the latest IBC code. Although many engineers still use ASD for steel design (guilty) instead of LFRD.

Did they explain why the new truck load includes the lane load? I have yet to get an answer from anyone. Also watch out for the software while using LRFD. Some of the code checks are incorrect per the new code. Do the first by hand until you understand the new code.

 

[cwrugrad]

LRFD is more time-consuming than the 16th or 17th Edition - which should be no surprise to anyone who has hefted the two documents!

In concrete design, the biggest difference is shear-capacity calculations. All that work to get a crack angle and beta factor!