Rivet Analysis Questions 3

[DigiAlmond]

Hey everyone! We've been doing some analysis on an existing structure built circa the 1950's. We've got some rivet connections we'd like to analyze. We don't really have anyone at our firm right now with any experience with rivets. I've been exploring the forums here, looked through DG15, looked through older editions of the steel manual, etc. My main hold up with just using the older methods of analysis that I've seen is they seem a bit uber conservative? I know that the the 15th ed. Steel Manual allows one to analysis existing steel with modern method but beyond that I've got nothing. How do I analyze the dang things in a modern context?

Using the tables in DG15 (specifically table 4-3b) I end up with the rivets controlling the connection significantly, before I even apply any sort of safety factor. Heck, the carbon steel bolts also in the connection I'm looking at end up beating them out, which my gut says isn't correct. Doubly so after reading about all the love for rivets people have here.

My best shot has been to treat the rivets as bolts, and working through the connection per chapter J in the steel manual. I did use DG15 table 4-1a to find the strength for A191-49T Rivets, and then utilized C-J3-3/4 to find Fnv and Fnt from the ultimate strength. I am not entirely convinced of this method since, I mean bolts aren't rivets.

Would anyone be willing to share some insights on the "modern" analysis procedure? what sort of safety factors do you employ? any sample calculations any of you would be willing to share? Thanks!

 

[MJCronin]

Anyone who has ever done an analysis of riveted plate joints is now dead ....

MJCronin
Sr. Process Engineer

 

[SnTMan]

Heh

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[FacEngrPE]

Rivets are not bolts. They are/were hot worked in the field, and are not heat treated after working, so you can not take credit for the strength improvement that might occur randomly from quenching at installation. There is a reason why the code at the time rivets were frequently used was UBER Conservative.
If you want to try something else, as in attempting to reverse engineer the joint, to assign a "modern" strength value you might as well write a PhD paper on your effort.

 

[dik]

Some info...[URL unfurl="true"]https://res.cloudinary.com/engineering-com/raw/upload/v1657394248/tips/Rivets_uth4xk.7z[/url]

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Guest102023]

Make a plug weld where the rivet goes.
Grind flush and glue on a rivet head.
Looks like new and stronger to boot.

"If you don't have time to do the job right the first time, when are you going to find time to repair it?"

 

[Lomarandil]

Not dead yet MJC... my joints don't even ache.

I've analyzed a number of riveted joints, and a number of jobs where rivets were chipped out (individually) and replaced with high-strength bolts to improve connection capacity.

I don't recall whether we ever found a researched rationale for applying the same slip and ductility characteristics for rivets as modern bolts (e.g. for the coefficients baked into the ICR analysis method), but I seem to recall we used it, had the support of our reviewers (Caltrans), and it worked in the field. A good rivet (you can tell by the sound when an ironworker first hits on it with a chisel) is generally as or more ductile than a bolt, just weaker. So yes, I agree that "modern" analysis methods and the tables from your steel manual can be used to sharpen the analysis without issue.

A well-installed rivet will provide significant clamping force (pretension) across the joint, so they often "overperform" their specified (shear) capacity. But that's not reliable enough to count on without some other evidence.

That said... rivets are simply milder steel than HSB. So yes, the connection capacity is less. Without seeing more about your specific situation, I can't speak into the disconnect you're seeing.

----
just call me Lo.

 

[kingnero]

Have been extensively involved in a renovation project of a large, rivetted bridge some years ago.

Caan't comment on the analysis, however some observations:
- because rivets are installed in hot form, and massively deformed during installation, they completely fill the holes of the plates they join. This gives the shear capacity "from the beginning", also making all rivets in a connection work together, at the same time from the beginning, much unlike "normal" bolted joints (not slip-critical).
- Lo already mentioned it: they are better than just their theoretical shear capacity, the older and wetter they have been, the better. However how to quantify that ... ?
- If you replace them with bolts, you can fill the void inbetween the shank of the bolt and the bolt hole with some sort of expoxy, in order to "directly" transfer shear (instead of waiting for some differential movement before shear gets transferred). Or you have to change to slip-critical joints, including all the preparation work...
- What Brimstoner suggest, is indeed what has been done in the past. BUT only after veryfing the weldability of the base materials. Older steels can have large amounts of unwanted elements, and can suffer from rock candy embrittlement after welding.
- The rivets that were used in that project, were significantly better than base quality structural steel. I don't have the MTRs, but I believe UTS was around 500 - 600 MPa.

 

[dik]

I've encountered steel with high sulphur content that makes the steel hot short so it cannot be welded... just remembered that.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Guest102023]

Good point about old steel. Some is weldable but some testing would be a prerequisite.

"If you don't have time to do the job right the first time, when are you going to find time to repair it?"

 

[DigiAlmond]

Hey all! Thanks for the feedback! If I understood correctly my method may work but it seems like there was a good reason for the conservative nature of design tables at the time. Plug welding is an interesting solution if it comes down to it; I think we're planning on sending in samples of the various metals on site since we do need to weld some of the steel else where on the project. Interesting to see that this is an area of relatively low research. If I ever have any drive to get a PhD I know what to look in to ha.

 

[Lomarandil]

If you do replace with bolts, note that because of either minor joint misalignment at installation (which wasn't a huge issue for a rivet that fills the hole as kingnero mentioned) or deformation since that time, the rivet holes won't always align. (this can make it tricky to pull the rivet shank out after the head is chipped off).

Often, it's beneficial to ream the rivet hole out to the next bolt size up and clean the alignment at the same time.

----
just call me Lo.

 

[kingnero]

True. One extra suggestion I forgot to add, be prepared to add extra budget for hearing protection. Both for during removal (pneumatic tools) as installation (forging) of the rivets... And hope your desk isn't inside or adjacent to the workshop, or separated with only some thin glass or plexi!

 

[EdStainless]

We had to replace some riveted connections with bolts.
We took about half of the rivets out, reamed the holes and installed larger bolts.
Then we did the same with the other half.
And knowing the low strength of rivets we used medium strength bolts.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[PEinc]

DigiAlmond, maybe some of this will help with the rivet analysis.

http://www.PeirceEngineering.com

 

[SnTMan]

PEInc, bravo

The problem with sloppy work is that the supply FAR EXCEEDS the demand