I-35W Bridge Collapse

Riveted connections are not stronger than the parent material but perhaps the connection itself is stronger than the element to which it is connected.

Rivets can be analysed simply as shear connectors or, if they are hot placed rivets, with account taken of the thermal shrinkage which gives rise to friction forces between the assembled elements. Cold placed rivets do not have this advantage (at least to a very minor extent depending on the fixing technique).

The hot rivet fills the hole under the blows from the riveting process. It is initially hot and so shrinks on cooling to compress the connected elements together. This friction can mobilise the thickness of the two connected elements thereby creating a stronger element.

The weakest point is the last line of rivets where nearly all of the forces have been transmitted from one element to the other but we have a line of holes for the rivets which subtract from the capacity of the elements. Can the friction process be sufficient to counter this capacity reduction? This is also the place where rust is most likely to occur.

Having done rivet design in the past, and having been made aware of the above, we were only asked to calculate the rivet pattern based on simple shear.

Today, we use High Strength Friction Grip bolts (HSFG bolts) to transmit forces from one element to another but still do not consider that the two elements work in tandem.

This is a picture of the structural engineering side of things. Is mechanical engineering any different?

Thank you OMW7 for you comments. It never occurred to me that the shrinkage of the hot rivets added strength to the member. My thoughts were that the purpose of heating was to make them soft and relieve stress from forming.

In the bridge photos, it appears obvious that all the riveted sections are designed to be in shear mode so In the initial design, shear mode stress would be expected to be the prime concern.

In looking at the photos, the splice failures appear to having been subjected to stress modes other than shear and since you have called my attention to it, this make sense that, under these conditions, failures were to be expected.

Comment: In aircraft work we frequently use "ice Box Rivets" The aluminum rivets are heated in boiling water and then chilled quickly, (to anneal) then they are stored on ice until driven. The driving action fills the hole and work hardens the rivet in position.

Once again, thanks for taking the time for your comments,

Snakers

What aluminum alloy are these rivets made of? And, what aluminum alloy(s) are you riveting with them? Just curious.

I am presently 80 years old and retired, I haven't driven an Ice Box rivet in 20 years but they are still used extensively in aircraft manufacture.

Any of the 2024 T-6 rivets must be treated (annealed) prior to driving or they will crack or you cannot head them up.

In the war song "Rosie the Riveter" Rosie drove Ice box rivets.

Best you go to your browser and search "ICE BOX RIVETS" There is a world of information there. far better than I can explain. Boeing aircraft uses them extensively.

The advanced E-rivet MS20426E work hardens on the first blow. If you don't get it on the first hit, you drill it out and replace it.

Some inexperienced riveters used the set and buck system. The driver calls set, the bucker sets the bar along side the river. and a light blow is given to seat the rivet in the hole. He then hollers Buck and the bucker holds the bar over the rivet and he drives it home. With Ice Box rivets the first light blow hardens the rivet and all future blows destroy the rivet. I have inspected a number of these failures on home built aircraft.

Remember that aircraft rivets are driven from the Head (shop end) and the bar is on the back side. Not like copper rivets that are beat down on the back side.

I don't know how hot steel rivets are driven, it's so high up I have never been close enough to watch them.

Hope this helps.

Snakers

Snakers,

I appreciate your taking the time to address my post. The information you provided is extremely interesting to me as, in spite of being involved in the riveting of aluminum structures for over 25 years, I have never heard the term "ice box rivets". I gather they were common in the aero industry and I have not been involved in aluminum airframe construction. Again, thank you!

FKIA

addendum.

I need to correct and clarify my comment>

Aircraft rivets have a factory end and a shop end. The factory end is the round headed end on of the rivet as it comes out of the box, The shop end in the straight end that the bucking bar goes against. The rivet is driven from the factory side with the bar on the shop side. This swells the rivet out where is protrudes from the base metal. It the bar is held properly, the rivet expands into a nice round circle. The ideal rivet protrusion length prior to bucking is 1 1/2 times the rivet diameter. I may not have said this correctly in my previous comment.

Snakers

This is probably a naive question, but I'll ask it anyway;

If the rivet is not driven properly, the shop end of the rivet is not going to extrude into a nice circle. I would imagine the end result is more tear-drop shaped. The resultant contact area would still be the same though (ie the same volume has been flattened out). Does this affect the rivets ability to resist shear stress ? Apart from distortional work hardening of the material, I can't readily see how it could if contact area is the same. The resistance to bending stress would be compromised though.

My description of the shop end of the river was not very good. You are correct. it bulges the rivet. If the bucking bar is not held squarely, the rivet lays sideways and in not a concentric circle. Sorry about that.

Interesting. Just this past week, there has been an on-going arguement re one of Boston's newest bridges, where the cable plates have "apparently" warped under the stress. Others argue that the steel plates were deformed during welding. Stay tuned to see if we have a real problem! Also, note that much of the assembly of aircraft is now done using adhesives, primarily (as I understand it) to limit the creep and stress-crack formation that riveting may eventually cause.

Hi guys, I'm a blacksmith type level III UT, ICC welding and bolting special inspector, CWI, certified welder, etc., etc. I like rivets better than bolts or welding. Take a look at ballardforge.com to look at some modern hydraulic methods of hot riveting. I think engineer's get too hung up on pretensioning, slip critical, etc. and automatically think bolts are the way to go. The bottom line is riveting worked- quite well.