Rivet Design

[ToadJones]

Anyone ever seen rivets designed as see in the attachment? Seems overly simplistic.
(I know that no one uses rivets anymore).

 

[ishvaaag]

Certainly a beautiful way to describe the shear flow to be taken by the rivets. It doesn't say anything about what a concentrated load would require (probably a separate structural device fitted to the web); otherwise would add to shear stress at the rivets.

However for a modern design better one would check for one of the latest specifications dealing with them, that would have the latest knowledge incorporated. Yet awareness of the old pertaining facts would be paramount if for rivet works predating the era of the specification.

 

[rb1957]

yeah, it looks ok ... we rivet all the time (and wouldn't think to weld !)

 

[ToadJones]

rb1957-
Numbers I'm getting using a classical shear flow formula and including the web are considerably lower. But I guess that makes sense if I am using VQ/I and 'I' includes the web.

 

[rb1957]

yes, i'd expect VQ/I to be less than V/h (if that's what you mean !?)

 

[dhengr]

No, not exactly that way, but it seems reasonable.

As an aside, but not for Toad’s eyes: The interesting thing about the ‘old fashioned’ rivet (which some people do still use) was, that when driven, the rivets actually tended to fill the holes, so they did take there proportional share of the load. Unlike bolts, and bolt holes and fit-up, where some percentage of the bolts are either yielding in shear, or causing plate/hole bearing yielding, to bring the other half of the bolts into real loading and bearing, thus joint slip.

 

[ToadJones]

We have actually discussed that very topic in my office a time or two. I have heard some people incorrectly refer to that phenomena as "shear lag".

 

[BAretired]

Toad,

For Case I, "Web Member Disregarded", V/h is identical to VQ/I. For Case II, "Web Member Considered", presumably on the next page, they probably use VQ/I.

BA

 

[ToadJones]

Yep- Thanks BA

I was being a dummy... I checked VQ/I using 'I' for the pair of angles only, matches exactly.

And yes, you are correct, they show two cases "Web MOMENT Disregarded" and "Web MOMENT Considered".

For the case where they take into account the web, they don't use VQ/I but rather an approximation taking into account 1/8 of the web depth....

H.I. = V/h x {Af/(Af + 1/8Aw)}
Af= area of flange
Aw= area of web

 

[rb1957]

yes, the simplification is the web is effective for shear only (so the angles are doing all the bending), and the shear in the web is V/h.

the web Might be fully effective in bending, depending on the buckling of the thin sheet under the bendng compression stresses. At a minimum the web stabilised by the angles won't buckle, so a estimate of h/8 doesn't sound unreasonable. but then the tension side would be fully effective.

the problems of using old textbooks ...

 

[msquared48]

This is a very riveting thread...

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[ToadJones]

must still be feeling sick huh Mike?

 

[ishvaaag]

I LOVE the simplicity of the old books.

 

[BAretired]

Taking into account the full web depth,

Q = Af.h/2

I = Af.h^2/2 + t.d^3/12
= Af.h^2/2 + Aw.d^2/12

I/Q = h + (Aw.d^2/12)/(Af.h/2) = h + Aw.d^2/6Af.h

If we assume that d = 1.22h, then d^2 = 1.5h^2

so I/Q = h + Aw.h/4Af = h(Af + Aw/4)/Af

and Q/I = Af/h(Af + Aw/4)

VQ/I = V.Af/h(Af + Aw/4) seems reasonable for Case II. I wonder why they didn't come up with that. Anyone know?

BA

 

[ishvaaag]

Maybe they did, BAretired. Make Aw=0 and you have a marginally bigger "shear flow" estimate, their formula V/h

 

[BAretired]

ishvaaag,

I was referring to the Case II formula at the end of Toad's 15 April 11 16:07 post. They had Aw/8 instead of Aw/4.

BA

 

[ToadJones]

I'll look at the text when I get back in my office....I think they said it was a "reasonable estimate"

 

[ToadJones]

any of you ever check the actual bearing stress on a rivet itself in a connection?

 

[BAretired]

Not me!

BA

 

[ToadJones]

BA-
Is that because you are not old enough to have designed a rivet, or for some other reason