[TheForge] riveted joints aren't flexible

[Thomas A. Troszak]

On Mar 2, 2005, Dave Smucker wrote:

> Use of the riveted construction that allowed ships to "work" was an
> extension of the timber construction that did the same.


Dear Dave,

I have done a lot of riveting since age thirteen, mostly on locomotive 
and traction engine boilers, including constructing riveted and welded 
pressure vessels where I designed the joints myself.

In a riveted pressure vessel, the riveted joints are actually NOT as 
flexible as the rest of the plate(s) due to the added thickness at the 
joints, but the material in the joints and the rivets themselves are no 
more or less likely to crack from flexure than the plates themselves, 
only from internal wasting (from various causes) or caustic 
embrittlement.

The only reason that the early riveted hull and boiler joints were 
assumed to be more "flexible" than welded structures is because the 
early welds and welded steels were excessively brittle, and couldn't 
survive thermal cycling, especially cold temps. Nowadays, a GOOD welded 
steel joint will bend and flex and expand and contract right along with 
every other part of the structure, whatever it may be, and survive 
environments that would could damage a riveted joint.

Hand (hammer) riveted joints are not watertight or pressure tight 
unless properly caulked, but they may still be mechanically rigid, and 
are definitely stiffer than the rest of the structure. Pneumatically 
riveted joints still need to be caulked at the seams, and the only 
occasional rivet head. The hydraulically clinched rivets found on late 
model locomotive boilers generally do not leak at all, but the seams 
can.

Very early riveted ship hulls probably did work (move) some visible 
amount at the joints, but not intentionally, the scale of the structure 
simply exceeded the technology of the joinery. You may find that the 
accounts of leakage are often precursors to immediate or catastrophic 
failure of the particular joint, which is more likely evidence of 
inadequate structure or execution of the joints, not the amazing 
flexibility of riveted structures. The poor bastards were simple 
floating around in some very poorly assembled hulls.

Very early riveting sucked. Boilers exploded, bridges fell. After 
decades of practice, riveting technology fully matured. By the time 
electric welding processes were first used in large scale practice, 
such as the Liberty ships of WWII, structural riveting had been a 
science for more than a century, and the Liberty ships were stick 
welded together outside in the open air by housewives with three days 
training, then tossed into the icy Atlantic. They actually performed 
fantastically well considering the circumstances.

A riveted joint is in fact structurally very similar to a bolted joint, 
but rivets are much less expensive than bolts, and they can be made 
water tight (sort of). No engineer would intentionally design or 
tolerate a bolted structure with every bolt "working" visibly in it's 
hole as the structure flexed, the bolts would soon fail from a variety 
of causes.

Please save all flames. I am very aware of the fact that all structures 
move and flex to a great great degree constantly. Some of the boilers 
that I worked on physically grow 1-1/2" or more in length from cold to 
working temp. By "working" is specifically refer to joints or fasteners 
that move visibly, or cause visible wear of the components. Fasteners 
that move in their holes are always a bad thing, IME (In My Experience)

The wing of a 747 flexes up very visibly during flight, but not because 
all of the thousands of little rivets are squirming in their holes, but 
because the entire structure is flexible, just like a bridge. A 
composite wing can flex even more than an aluminum one, and it has no 
rivets at all, just intentionally flexible material and structure.

All of the early riveted ship hulls were hand riveted, and manually 
bucked, and the science of designing riveted structures was in fact 
still very primitive. The simple lap seams of very early hulls were 
simply scarfed at an angle and bashed together. This means that every 
plate and joint is subject to constant bending with every load and 
change in temperature. In short: early riveted joints really sucked, 
and often failed.

The double butt strap of the 1920's is a 'ting o' beauty, and is in 
fact stronger than the plate it joins.

In summary (from my experience):

Properly executed riveted joints are always LESS flexible than the 
surrounding material.
All riveted joints can and will leak water no matter how stiff they may 
be, but they sure aren't flexible.

Just my two cents worth,

Tom Troszak