Always Riveted Girders For Railway Bridges? 2

[IJR]

My dear friends

A detailer I know is detailing a very old design of railway bridge. The design makes use of riveted plate girders, formed from double angle as flanges and web plates, all well riveted. When I asked him why not simply a well designed plate girder he reported faithfully that welding and bolts are prone to fatigue which I agree with him.

But arent there railway bridges with welded plate girders and high strength bolts?

I request a discussion on this subject.

 

[JedClampett]

Of course.
Maybe he's trying to match an existing design or there's some architectural requirement to make it look antique. Good luck finding anyone with experience, equipment or material for riveting.

 

[HgTX]

Tell your acquaintance to take a look at the AREMA manual. (Unfortunately it's a huge 4-volume monstrosity and all he really needs is the chapter on steel design, but if he's gonna be doing railroad design he needs to have it. I can't believe the RR doesn't require design in accordance with that manual.) There are provisions for riveted (but I'm with JedClampett--good luck getting it done), bolted, and welded construction.

Typical practice these days is welded in non-fracture-critical areas and bolted in fracture-critical.

Concerns about fatigue and welding just aren't what they used to be. There have been significant advances in welding technology and inspection since the 19th century when your friend was apparently trained.

Are bolts really more susceptible to fatigue than rivets? Why?

Hg

 

[IJR]

HgTx

Most engineers like me tend to visualize nuts as loosening up crazy nuts. Even effects of preloading simply doesnt register well in our minds. And there are times we wish we were in labs to see it by ourselves that yeah those nuts wont start dancing faster than the beams.

I guess this vision is what makes us still think of rivets as superior.

Question: In a simply supported highway steel girder which zones would be fatigue critical? If you mention tension flange to web connection then we are going to have a problem with not welding that one.

Thank you for the great post

IJR

 

[HgTX]

You may want to get yourself to a lab, then. Slip-critical connections aren't known for working loose. The force holding the plates together is the same as the force holding nut thread to bolt thread. By definition, no room to wiggle.

Fracture-critical is defined by AASHTO. Tension components of nonredundant members, basically. So if you have a RR bridge made of at least 4 parallel plate girders, it's non-FC. On the other hand, yer classic thru-girder brige, with only two, is FC.

What I've seen for FC RR girders is welding of top flanges to web, welding of stiffeners to web (although sometimes they avoid welding in the tension zone of the web, 1/6 or so of the depth), and then bolted connections to the bottom flange.

My understanding is that they're more worried about maintenance than actual collapse. FC members require relatively frequent and intensive inspection, and more immediate action if repairs are required, and RRs can't just route a local detour or shift traffic to another lane if they need to fix a bridge.

Hg

 

[IJR]

Thanks HgTX

I have a copy of AASHTO 4th edition 1989 which I will use to follow up your great notes. And yes, I personally have a lot of faith on preloaded bolts, perhaps because every number in tensile strength simply fascinates me.

And if you have more to share, please do. For example you can give me a website showing a case of a well designed railway bridge with steel plate girders or the like.

Respects
IJR

 

[HgTX]

I don't know of a website but our RR guru has a powerpoint presentation on CD. All the personal profiles are down at the moment, so I don't know if Eng-Tips has some built-in means of one-on-one communication. My email address is posted on this document:
ftp://ftp.dot.state.tx.us/pub/txdot-info/cmd/mpl/electrod.pdf

Hg

 

[HgTX]

Now that my attention's been drawn back to this thread, what came of this bridge? Did they manage to find a skilled riveting crew, perhaps next door to the local blacksmith?

Hg

 

[IJR]

Thanks HgTx

Yes we got into a lot of trouble doing rivets. We got to a point where we had to stop and try to persuade the owner to give up rivets.

But the owner simply did not yield and we had to do rivets. After a lot of searching we did find someone who understood the craft, and we had to spend more weeks looking for the guy who could custom make rivets(large diameter rivets are nowhere to be found in this region- Europe). So we custom made them using ASTM specs. And a number of simple tests because no one has the right tools to do intensive tests.

We have installed the bridge. But the whole thing was made with a phantom factor of safety and the accompanying costs, and an equally phantom time span.

best regards HgTx

 

[dbuzz]

FYI, the Australian Steel Institute published a paper on the limit state assessment of rivets:
Barker, G.F
"Assessment of existing structures: A strength limit state appraisal of hot driven rivets"
Steel Construction, Vol 34, No 1, 2000
(

http://www.steel.org.au/)

Barker suggested a capacity reduction factor of 0.8 for new rivets and between 0.6 and 0.8 for existing rivets (and if in doubt use 0.6).

 

[HgTX]

Did the owner learn the lesson that they were being complete idiots and that they should have a different plan next time?? Odds are they have a much LESS safe bridge now.

Crap like that makes me so mad. Engineering decisions dictated by undereducated, uninformed, closed-minded IDIOTS! I don't know whether I'd have been able to bring myself to seal the design in your (or your friend's) place.

RRRRrrrrrrrrrrrrrr................

Hg

p.s. *Could* there be a next time RR bridge with this owner?? Is the owner a municipality or a RR?

 

[IJR]

Yes my dear friend HgTx

You prompted a short discussion from me.

What happens when a highly technical and sole authority in a given field of engineering practice suddenly drops dead along with others in the same field for reason of market or trend and then for reason of current political changes has to make a comeback in so little time? They tend to go back to where they left the job at, and it does take time to adapt to new practices. Seems like doesnt make sense but there are cases when this happens. The authorities in this project had piles of old "prequalified" good designs and wanted those to be applied immediately.

For me it was a special experience helping this detailer friend of mine, because I got to understand how things can be influenced so much by non-engineer parties/situations.

The other side of life is interesting and important.

My respects.
IJR

 

[HgTX]

"They tend to go back to where they left the job at, and it does take time to adapt to new practices. "

A little awareness that perhaps the state of the art has changed in the last several decades would be well within the bounds of sound engineering judgement--and I would go further and say that refusal to acknowledge this change is irresponsible.

Hg

Eng-Tips guidelines: faq731-376

 

[TFL]

rivets!!!! - dear god are we really having this conversation. i need a beer!

 

[trac]

Just some observations/comments-


I was told that even up to the 60’s, a Canadian fabricator (Montréal based I think) still used rivets for shop fabrications, as in some cases riveting was more economical than either welding or bolting.

I am not aware of any major problems with riveted structures subject to long term dynamic loading e.g. the substantial number of riveted railway bridges that are still being used to this day around the world as opposed to welded structures (fatigue) or bolted structures (loosening). With this in mind, a Client may consider the higher CAPEX costs for a riveted structure compensated by reduced OPEX (maintenance) costs.

I know that there are innumerable posts concerning bolts, torquing, loosening etc., etc., i.e. slip critical connections, but as yet little emphasis on ply contact. If you have 8 mm gussets plates that sandwich a diagonal strut I assume that initial snug tightening of the bolt group is likely to pull the gusset plate to flange to achieve a contact friction interface. But if you have 30 mm thick gusset plates that have a twist or separation between the flange interfaces – can you guarantee that snug tightening the bolt group will deform the bent gusset plate and/or bowed/twisted flange enough so that the required contact between the slip critical faces is enough so that the required friction load can be developed.

regards
Trac

 

[HgTX]

That's where quality control and inspection come in. Legally, the contractor may be the one responsible to do everything by the code and make sure everything works as it should, but if the owner wants that to happen, they need to ensure that inspectors know how to check for snug-tightness (if there's a gap in there, it ain't snug-tight), and that fabrication standards ensure that the connection is indeed boltable (meaning don't let them ship something with a severe twist in it to the jobsite, and make sure they are well aware of proper straightening techniques when they fix it).

The 1960s were a long time ago, even though it might not always seem that way.. I would have SERIOUS worries about the quality of a new riveted structure because of the lack of truly skilled labor in this area. You so sure you're getting better quality through older technology? There's no ASTM standard for rivets any more; what quality standard are you going to enforce for material? What toughness or hardness? What current standard establishes testing for as-riveted material?

There are well-established ways to design for fatigue, and there are well-established ways to bolt, not to mention stringent material controls for both fasteners and plate. Anyone who thinks bolting, welding, bridge design, or even riveting is still in the same state it was half a century ago is irresponsibly and unprofessionally ignorant.

Hg


Eng-Tips guidelines: faq731-376

 

[twofingerstoshort]

What about ship building yards? I know they still use rivets when not welding.

 

[frankiee]

I remember from my class on "How to break a ship in half 101"
Cut a notch with no radius and watch it progress and dont tell anybody.


From

http://www.uscg.mil/hq/g-m/nvic/3_68/n3-68.htm

"In this age of welded ships hot driven rivets are no longer the principal method of connecting ship structural components. Yet there are locations where riveted connections (or their equivalent) are necessary or desirable. For. example, riveted joints are used as "crack arrestors". There is also a significant amount of repair work on existing riveted structure."