I was mildly surprised to learn that the (automotive) SAE grade 2, 5,
8, etc bolt specifications do not include brittle fracture ratings.
This straight from the relevant spec committee chair people about 5 or
so years ago. I've never seen (or if I did I forgot :-() the actual
SAE bolt grade specs. I wonder if the metric bolt specs have brittle
fracture ratings. Aerospace fasteners better though or I'm sticking
to the bus. :-)
On Sat, 19 Jan 2008 00:21:47 +0000, Christopher Tidy
Automotive seal belt anchors are GENERALLY designed in such a way that
the floor pan will deform before the bolt or the belt breaks. The thin
floor is doubled with a heavier plate that retains the extra strength
If the two bars are properly joined with a properly engineered joint,
the bars themselves will almost stretch before either the bolt or hole
Posted via a free Usenet account from http://www.teranews.com
In a way, yes.
We design to a give load limit.
Anything beyond that is excess weight.
So we don't ever expect an unbreakable structure.
It will survive up to the yield point - beyond which the structure
is damaged - but not broken.
At teh ultimate load point the structure breaks.
And yes, we know the elongation of both the joint material and the bolts
(AN bolts are typically grade 5).
But the real deal is simply this...
WHEN the structure is over stressed - i.e.: it has been loaded beyond
the yield point - how do you know? And - what gets damaged?
Do you want the bolt to bend visibly?
Or a few hundred rivets (or welds?) to be invisibly damaged?
I kind of see your argument. The grade 8.8 bolt has a stretch of 12% before
he breaks (no numbers for the 5.8, should be something around 17%).
But lets get back to the "I'll never use a grade 8 bolt":
If the joint is *designed* the bolts are calculated to take the load. If
that structure is overloaded, the two parts that are held together will
lose contact and now we do have to distinguish two cases:
* The bolt was stretched, and the joint will rattle and no longer work
(you'll see that later).
* The bolt was elastically stressed and the joint is back in contact after
the overload and is still working (albeit maybe in a bit different
Now we have a closer look at the second case:
All joints with bolts relay on the elasticity modulus of the bolt. The
modulus is the same for any grade. The bolt keeps two parts together with
the preload given by the bolt (by the torque it was tightened with) and
they will lift/move/shift as soon as the outer force is bigger than the
Now there are two cases:
A joint that is stressed with shearing forces:
Design flaw: The *bolt* is designed to take the shearing force. You are
fired! That is wrong by design! It always are the two parts and the
friction between the two and the preload given by the bolt. There is no
difference between different grades of bolts (-> modulus). OK, there are
*rare* exceptions where a shearing pin (which needs to have tolerances in
the diameter together with the hole it goes into) and a bolt are merged
into one piece. But that is *not* a normal bolt.
A joint that is stressed by pull:
Lets take two hypothetical bolts. One with 500 N/mm^2 and one with 1000
N/mm^2. I call them grade 5 and grade 10 for now.
For a given designed load, the grade 10 bolt can have half the cross-section
of the grade 5. Now if we overload the joint, the grade 10 acts *softer*
than the grade 5 (half the cross-section, same modulus. acts like a spring,
half the spring rate) that looks to be an advantage, because parts can move
easier with the grade 10, before we do have a plastic deformation.
Now if we come into the pastic region of stress, the grade 5 bolt was
elongated *half* the distance of the grade 10. The fact that it has about
30% (17% vs. 12%) more plastic elongation than the grade 10 bolt doesn't
help, because it has double the spring rate of the grade 10.
So by design -assuming propper design- a grade 10 is more forgiving than a
And now finally to the case a grade 5 is replaced by a grade 10 without
changing the diameter:
Well, something will bend/stretch. It makes a difference what will bend. But
it is already a failure. We will find enough examples where the slight
bending of the structure is better (and keeps the whole structure in a
still perfect working condition) contrary to where a rattling joint will
have its advantages. Uh? Read above, a joint that is loaded by shearing
forces no longer works *at**all*. A joint that is stressed by pulling
forces is just rattling and not keeping things together.
So you are suggesting that the bolts aren't tightened? Maybe you have to
re-read what I wrote. Friction & clamping forces are the keywords to look
If it is shear only, you are at the *second* failure of the joint.
I guess I expressed it a little TOO simply (?)...
However, there are a lot of cases where the clamping force is minimal
and the bolt is simply a shear pin.
Ultralight wing tube style spars - for one.
Can't clamp much without crushing the tube.
On Fri, 18 Jan 2008 13:49:32 -0600, cavelamb himself
And VERY poor design when built that way. The proper way is to sleeve
the hole so the bolt is supported for it's full length instead of just
by the skin of the tube, and the bolt can be torqued to it's proper
torque to provide maximum strength.
Then they could likely even get away with lighter bolts.
Posted via a free Usenet account from http://www.teranews.com
On Fri, 18 Jan 2008 17:37:45 -0600, cavelamb himself
Not off the top of my head, but I think the Kolb 500 trainer (2
seater) is sleeved.from the factory. A friend a few years ago had one
that was sleeved
Also I think the Beaver 2 seater does. I know a friend's Beaver does,
might not have been from the factory.
I know a lot of ultralights have SCARY engineering (or lack there-of)
and many owners modify them to make them significantly safer.
By "sleeved" I don't meed doubled tubes. I mean the hole is drilled
overside and a piece of tubing is welded or brazed into the hole, the
right size to fit the bolt snuggly.
Many ultralightes also have holes drilled through the tubes the wrong
way, seriously weakening the tube.
Posted via a free Usenet account from http://www.teranews.com
He doesn't like to see holes drilled vertically in a spar tube.
Actually, he's right in that removing material from the top and bottom
do reduce the amount of bending load the tube can ultimately take before
The reason is that the top and bottom are the highest stressed areas.
The top skin is normally in compression and the bottom in tension.
Drilling here reduces the amount of material to carry the load.
But, like the sleeved bolt holes he mentioned, it is unnecessary -
if the stresses at that point are below what the structure will stand.
Now, the two seaters that Clare mentioned (there is no such thing as a
two seat ultralight in the US) are considerably heavier airplanes.
At that point they probably do need the sleeves (bushings?) in the tube
to take the compression load that the wing imposes on the root connection.
At the other end of the spectrum, the mast on my sailboat (and on the
bigger boats too) have only a simile bolt pinned through an unsleeved
aluminum tube to secure the base of the mast. AND - the compression
loads on it are right near the same as the wing root compression loads
on the airplane's wing root! Interesting.
To bring it back to the thread topic...
I used grade 8's on my airplane - for the landing gear axles.
5/8" diameter by 5 to 6 inches - to fit the wheels you want ti use.
The heads are cross-drilled for a 3/16 (AN-3!) bolt that attaches to the
shock absorber tubes (telescoping tubes with bungees).
After drilling a 3/16 hole through the heads, there is not a lot of
metal left - but we've never had a failure there.
So we can conclude that the stresses imposed here (gear loads are the
highest on the whole airplane) are lower than what the bolt head can
And - that - is all that matters.
Most joints work as you have described. Most structures are designed for
ridigity, not for strength, and the stronger the fasteners, the more rigid
is the structure. But ductility/elongation come into play in structures that
require high ratios of strength to weight, as is the case with some bridges
and many aircraft, as well as some other things.
Where that's the case, designing for ductility can be an important issue for
the overall strength of the structure in at least a couple of different ways
(not counting the visible near-failure that Richard is talking about, which
also is of practical importance). The primary way relates to distribution of
loads by avoiding point loads that will cause progressive failure. Some
riveted sheet structures and shear panels are examples. It can also be an
issue where there are multiple bolts in a heavier plate structure, with
holes that have some reasonable production tolerance and where the bolts are
going to be loaded in shear as the limits of strength are approached. This
is what I had in mind with the trailer-hitch example. As with the sheet
structures, the value of some ductility here can be distribution of the load
among multiple fasteners, where harder/stronger ones would localize the load
until the first hole tore out, and then then second, and so on.
As I said, I was just looking for an argument. <g> Most of the time, what
you're talking about is the important design parameter. But not always. And
aircraft designers, as well as designers of many other types of highly
loaded structures, must design around ductility to avoid excessive point
loads. One well-known example is in the welding of tubular space frames. A
pure space frame does not load joints in bending but a *real* space frame
always does. If the weld is too strong it will not allow plastic deformation
of the joint. That will cause a failure where the tube joins the weld,
because of the high point loads. The situation with bolts and rivets is
Sometimes, weaker is stronger. That's why we have structural engineers.
OK. Now I got you. The image were several rows of bolts (classical sheet
overlapping joint) that are calculated more like rows of rivets. Not yet
fully convinced that a stronger bolt has an disadvantage (except the
But anyhow, the point is more the hole in the sheet and its projected
surface area that has to accept a certain pressure (when shearing) and you
can't make the bolt smaller because of the pressure. The clamping force is
(almost) ignored in that setup. But maybe that was an old approach in
design. I know, that joints like these are now glued (to take the shear)
and screwed/riveted) to take the peeling forces.
But I also got your point about distributing load and what happens if one
joint fails and you get an avalanche failure of the neighboring joints.
Yes! And you've identified an issue here that most people miss: Those
rivet-bonded wing skins and so on are *not* designed to share the load
between the glue and the rivets. The rivets are just there to prevent peel
and cleavage, failure modes in which high-strength epoxy is very weak. But
the glue is much stronger at resisting shear than an all-riveted joint would
be, even one that's optimized for resolving the shear loads in the rivets.
Yes, but that's only one of the issues with ductility in fastening. As I
mentioned, I'm too rusty to get into the whole schtick, but there are other
reasons you need ductile fasteners, as well. Sometimes.
I'm inclined to agree with Nick here. The few times I've made this
mistake myself in the past, the joint has come loose. Sure, there are a
few instances in which it can be okay to load a bolt in shear (a shackle
but they're cases in which loosening isn't an issue. Mostly it's a bad idea.
Sure I'm right! ;-)
No, I got the point now. We had different pictures in our minds.
The grade-5 fraction had this in mind:
Overlapping sheet that is held together with rows of bolts. You have that
picture of several rows of rivets?
OK, here it really pays to have soft bolts. Because what happens when you do
have overload is, that the bolt most stressed and being at his second
failure (plastic deformation) still takes some load but also is partially
giving in and thus handing the overload to his neighboring bolts/rivets. If
that bolt would be too strong, the sheet would start to tear and this is
certainly more catastrophic than a bent bolt with an oval hole in the
As soon as you are going away from sheet metal and do have more solid
constructions with longer bolts things are getting different and the grade
5 fraction is getting wrong.
Is that acceptable? :-)
Well, better anyway,
We can't judge this part right or wrong without putting numbers on it.
Like the example one fellow mentioned - hard bolts on an oil pan that
started breaking when they replaced the old hard gaskets with silicone.
But this outta be a good place to start for general sizing to loads...
We made the most common error in discussion. We didn't specify what we were
exactly talking about. So everyone had his own picture in his head and
defended it. :-)
Furthermore, each field (be it planes or machines or whatever) has his
tradition in how to look at things and what they want to avoid or achieve
under all circumstances. The classical math to solve those problems always
needs some simplifications and so we should have agreed upon that in the
Anyhow, interesting discussion with something to learn from!
Better than ... oh you know! ;-)
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.