Rim fatigue failure

On 4/21/2020 3:52 AM, Sepp Ruf wrote:
jbeattie wrote:
On Monday, April 20, 2020 at 3:23:14 PM UTC-7, John B. wrote:
On Mon, 20 Apr 2020 08:18:15 -0500, AMuzi wrote:

This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

One might ask, "Is this a common failure of the bicycle wheel?"
Even if it is with lightweight, anodized, Mavic-grade, double-fried crouton
brittle alloy rims, I'm not concerned because, like suspiciously lifting
eyelets, it has always been early from outside.

Yes, very. It was not so much of an issue in the olden days with 120mm
rear hubs and 5sp freewheels and stouter, high spoke count rims.

Maybe users of disk brakes are not checking the rims regularly? And the
video does not even go l/r asymmetric spoke pulls.

I see Ralph Nader is 86. Who could ring the alarm today? AOC?


I'm still planning to do a tire burn with my Corvair in the
parking lot at his funeral.

--
Andrew Muzi
www.yellowjersey.org/
Open every day since 1 April, 1971

On 4/20/2020 8:50 PM, John B. wrote:
On Mon, 20 Apr 2020 19:59:11 -0400, Frank Krygowski
wrote:

On 4/20/2020 3:54 PM, Mark J. wrote:
On 4/20/2020 6:18 AM, AMuzi wrote:
This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

Thanks, Andy, this was the best thing I've found on R.B.T. in years!

To everybody:Â* IF you haven't watched the video yet, it's rather long
and slow, but worth the time, IMHO.Â* There is actual data obtained by
actual /measurements/ made in a /systematic/ way.

It's admittedly very wonky, but this is a TECH group, yes?


If you want to get even wonkier: It's interesting and relevant that the
higher tension wheels in that test had less variation in tension during
loading. And it reminds me that the design of a tension spoke wheel has
a similarity to the design of a bolted joint subject to fatigue loading.

As an example, think of the cylinder head bolts on an engine or piston
compressor. As cylinder pressure varies, the force in the bolts varies.
It turns out that for that sort of bolted connection, a bolt with a
thinner central section (between the head and the threads) torqued to a
higher preload stress can be much more resistant to fatigue than a
thicker bolt at a lower stress. The strongest bolt is something like the
third one in figure 12 at
https://www.fastenal.com/en/3289/fastener-fatigue

Sorry Frank, but as you state it I just don't agree.
Carried to it's logical limits you are saying that, say a 2" diameter
bolt, with the center section reduced to, oh say 1/2", is stronger
than the straight shank 2 inch bolt?

That's a very extreme example you're giving, with a factor of 16 on the
cross section areas.

I'm talking about a specific loading situation: a varying load
superimposed on a steady load. And a specific failure mode: Fatigue
failure. The other necessary feature is that the bolt (or in our case,
spoke) has to be in tension against another structural element in a way
that the relative deflections are controlled by the relative
stiffnesses. In that sort of sort of situation, a more flexible bolt is
typically more resistant to fatigue failure.

Here's an industrial example that came up once when I was teaching this.
One student working in an engineering firm talked about a bolted flanged
connection between two large pipes that was subject to vibration. They
were breaking flange bolts until they installed longer bolts, something
like 8" bolts to clamp something like two inches of total flange thickness.

How did they do that? They added thick wall tubes, something like 6"
long, between the bolt heads and the flange. The main idea was to make
the bolts longer and more flexible. With increased flexibility, the
_variation_ in stress of the bolts was much less, and they better
resisted fatigue.

For such long bolts, the long center section is often not threaded, but
has a diameter less than the minor diameter of the threads. (We'd say
"butted.") This helps two ways: by making the bolt more elastic in
tension, and by reducing stress concentration at the inner ends of the
threads.


--
- Frank Krygowski

On Tuesday, April 21, 2020 at 1:08:04 PM UTC-4, Frank Krygowski wrote:
On 4/20/2020 8:50 PM, John B. wrote:
On Mon, 20 Apr 2020 19:59:11 -0400, Frank Krygowski
wrote:

On 4/20/2020 3:54 PM, Mark J. wrote:
On 4/20/2020 6:18 AM, AMuzi wrote:
This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

Thanks, Andy, this was the best thing I've found on R.B.T. in years!

To everybody:Â* IF you haven't watched the video yet, it's rather long
and slow, but worth the time, IMHO.Â* There is actual data obtained by
actual /measurements/ made in a /systematic/ way.

It's admittedly very wonky, but this is a TECH group, yes?


If you want to get even wonkier: It's interesting and relevant that the
higher tension wheels in that test had less variation in tension during
loading. And it reminds me that the design of a tension spoke wheel has
a similarity to the design of a bolted joint subject to fatigue loading.

As an example, think of the cylinder head bolts on an engine or piston
compressor. As cylinder pressure varies, the force in the bolts varies..
It turns out that for that sort of bolted connection, a bolt with a
thinner central section (between the head and the threads) torqued to a
higher preload stress can be much more resistant to fatigue than a
thicker bolt at a lower stress. The strongest bolt is something like the
third one in figure 12 at
https://www.fastenal.com/en/3289/fastener-fatigue

Sorry Frank, but as you state it I just don't agree.
Carried to it's logical limits you are saying that, say a 2" diameter
bolt, with the center section reduced to, oh say 1/2", is stronger
than the straight shank 2 inch bolt?

That's a very extreme example you're giving, with a factor of 16 on the
cross section areas.

I'm talking about a specific loading situation: a varying load
superimposed on a steady load. And a specific failure mode: Fatigue
failure. The other necessary feature is that the bolt (or in our case,
spoke) has to be in tension against another structural element in a way
that the relative deflections are controlled by the relative
stiffnesses. In that sort of sort of situation, a more flexible bolt is
typically more resistant to fatigue failure.

Here's an industrial example that came up once when I was teaching this.
One student working in an engineering firm talked about a bolted flanged
connection between two large pipes that was subject to vibration. They
were breaking flange bolts until they installed longer bolts, something
like 8" bolts to clamp something like two inches of total flange thickness.

How did they do that? They added thick wall tubes, something like 6"
long, between the bolt heads and the flange. The main idea was to make
the bolts longer and more flexible. With increased flexibility, the
_variation_ in stress of the bolts was much less, and they better
resisted fatigue.

For such long bolts, the long center section is often not threaded, but
has a diameter less than the minor diameter of the threads. (We'd say
"butted.") This helps two ways: by making the bolt more elastic in
tension, and by reducing stress concentration at the inner ends of the
threads.


--
- Frank Krygowski

Also, a varying tensile load is less likely to result in failure than is a load that goes back and forth between tension and compression. Not that you'd ever get compression on a bicycle wheel spoke, but if the tension were hitting zero, I could easily see that being an issue. The whole point of butted spokes I am sure is to stress, and therefore stretch, the material more for the same tension; the tiny reduction in weight is just a happy byproduct.

On Tue, 21 Apr 2020 13:07:59 -0400, Frank Krygowski
wrote:

On 4/20/2020 8:50 PM, John B. wrote:
On Mon, 20 Apr 2020 19:59:11 -0400, Frank Krygowski
wrote:

On 4/20/2020 3:54 PM, Mark J. wrote:
On 4/20/2020 6:18 AM, AMuzi wrote:
This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

Thanks, Andy, this was the best thing I've found on R.B.T. in years!

To everybody:* IF you haven't watched the video yet, it's rather long
and slow, but worth the time, IMHO.* There is actual data obtained by
actual /measurements/ made in a /systematic/ way.

It's admittedly very wonky, but this is a TECH group, yes?


If you want to get even wonkier: It's interesting and relevant that the
higher tension wheels in that test had less variation in tension during
loading. And it reminds me that the design of a tension spoke wheel has
a similarity to the design of a bolted joint subject to fatigue loading.

As an example, think of the cylinder head bolts on an engine or piston
compressor. As cylinder pressure varies, the force in the bolts varies.
It turns out that for that sort of bolted connection, a bolt with a
thinner central section (between the head and the threads) torqued to a
higher preload stress can be much more resistant to fatigue than a
thicker bolt at a lower stress. The strongest bolt is something like the
third one in figure 12 at
https://www.fastenal.com/en/3289/fastener-fatigue

Sorry Frank, but as you state it I just don't agree.
Carried to it's logical limits you are saying that, say a 2" diameter
bolt, with the center section reduced to, oh say 1/2", is stronger
than the straight shank 2 inch bolt?

That's a very extreme example you're giving, with a factor of 16 on the
cross section areas.

I'm talking about a specific loading situation: a varying load
superimposed on a steady load. And a specific failure mode: Fatigue
failure. The other necessary feature is that the bolt (or in our case,
spoke) has to be in tension against another structural element in a way
that the relative deflections are controlled by the relative
stiffnesses. In that sort of sort of situation, a more flexible bolt is
typically more resistant to fatigue failure.

As I wrote, "but as you state it I just don't agree".

Here's an industrial example that came up once when I was teaching this.
One student working in an engineering firm talked about a bolted flanged
connection between two large pipes that was subject to vibration. They
were breaking flange bolts until they installed longer bolts, something
like 8" bolts to clamp something like two inches of total flange thickness.

How did they do that? They added thick wall tubes, something like 6"
long, between the bolt heads and the flange. The main idea was to make
the bolts longer and more flexible. With increased flexibility, the
_variation_ in stress of the bolts was much less, and they better
resisted fatigue.

For such long bolts, the long center section is often not threaded, but
has a diameter less than the minor diameter of the threads. (We'd say
"butted.") This helps two ways: by making the bolt more elastic in
tension, and by reducing stress concentration at the inner ends of the
threads.

Which is a totally different explanation that you originally posted so
I'll stand by my original statement.

--
cheers,

John B.

On Tuesday, April 21, 2020 at 6:37:17 PM UTC-4, John B. wrote:
On Tue, 21 Apr 2020 13:07:59 -0400, Frank Krygowski wrote:

On 4/20/2020 8:50 PM, John B. wrote:
On Mon, 20 Apr 2020 19:59:11 -0400, Frank Krygowski wrote:

On 4/20/2020 3:54 PM, Mark J. wrote:
On 4/20/2020 6:18 AM, AMuzi wrote:
This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

Thanks, Andy, this was the best thing I've found on R.B.T. in years!

To everybody:Â* IF you haven't watched the video yet, it's rather long
and slow, but worth the time, IMHO.Â* There is actual data obtained by
actual /measurements/ made in a /systematic/ way.

It's admittedly very wonky, but this is a TECH group, yes?


If you want to get even wonkier: It's interesting and relevant that the
higher tension wheels in that test had less variation in tension during
loading. And it reminds me that the design of a tension spoke wheel has
a similarity to the design of a bolted joint subject to fatigue loading.

As an example, think of the cylinder head bolts on an engine or piston
compressor. As cylinder pressure varies, the force in the bolts varies.
It turns out that for that sort of bolted connection, a bolt with a
thinner central section (between the head and the threads) torqued to a
higher preload stress can be much more resistant to fatigue than a
thicker bolt at a lower stress. The strongest bolt is something like the
third one in figure 12 at
https://www.fastenal.com/en/3289/fastener-fatigue

Sorry Frank, but as you state it I just don't agree.
Carried to it's logical limits you are saying that, say a 2" diameter
bolt, with the center section reduced to, oh say 1/2", is stronger
than the straight shank 2 inch bolt?

That's a very extreme example you're giving, with a factor of 16 on the
cross section areas.

I'm talking about a specific loading situation: a varying load
superimposed on a steady load. And a specific failure mode: Fatigue
failure. The other necessary feature is that the bolt (or in our case,
spoke) has to be in tension against another structural element in a way
that the relative deflections are controlled by the relative
stiffnesses. In that sort of sort of situation, a more flexible bolt is
typically more resistant to fatigue failure.

As I wrote, "but as you state it I just don't agree".

Here's an industrial example that came up once when I was teaching this.
One student working in an engineering firm talked about a bolted flanged
connection between two large pipes that was subject to vibration. They
were breaking flange bolts until they installed longer bolts, something
like 8" bolts to clamp something like two inches of total flange thickness.

How did they do that? They added thick wall tubes, something like 6"
long, between the bolt heads and the flange. The main idea was to make
the bolts longer and more flexible. With increased flexibility, the
_variation_ in stress of the bolts was much less, and they better
resisted fatigue.

For such long bolts, the long center section is often not threaded, but
has a diameter less than the minor diameter of the threads. (We'd say
"butted.") This helps two ways: by making the bolt more elastic in
tension, and by reducing stress concentration at the inner ends of the
threads.

Which is a totally different explanation that you originally posted so
I'll stand by my original statement.

Sorry, John. If you think the two explanations were totally different,
you didn't understand. Perhaps it needed more explanation.

I can go into this as deeply as you like. IIRC, it took a week or
more in one of my courses, and was applied frequently after that.

Interestingly, when I took the test for the Professional Engineer's
license, one of the problems involved this stuff.

- Frank Krygowski

On Tuesday, April 21, 2020 at 6:33:47 AM UTC-7, AMuzi wrote:
On 4/21/2020 3:52 AM, Sepp Ruf wrote:
jbeattie wrote:
On Monday, April 20, 2020 at 3:23:14 PM UTC-7, John B. wrote:
On Mon, 20 Apr 2020 08:18:15 -0500, AMuzi wrote:

This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

One might ask, "Is this a common failure of the bicycle wheel?"
Even if it is with lightweight, anodized, Mavic-grade, double-fried crouton
brittle alloy rims, I'm not concerned because, like suspiciously lifting
eyelets, it has always been early from outside.

Yes, very. It was not so much of an issue in the olden days with 120mm
rear hubs and 5sp freewheels and stouter, high spoke count rims.

Maybe users of disk brakes are not checking the rims regularly? And the
video does not even go l/r asymmetric spoke pulls.

I see Ralph Nader is 86. Who could ring the alarm today? AOC?


I'm still planning to do a tire burn with my Corvair in the
parking lot at his funeral.

--
Andrew Muzi
www.yellowjersey.org/
Open every day since 1 April, 1971

Don't do that....
That would be Unsafe at Any Speed.

pH

On Tue, 21 Apr 2020 16:57:06 -0700 (PDT), Frank Krygowski
wrote:

On Tuesday, April 21, 2020 at 6:37:17 PM UTC-4, John B. wrote:
On Tue, 21 Apr 2020 13:07:59 -0400, Frank Krygowski wrote:

On 4/20/2020 8:50 PM, John B. wrote:
On Mon, 20 Apr 2020 19:59:11 -0400, Frank Krygowski wrote:

On 4/20/2020 3:54 PM, Mark J. wrote:
On 4/20/2020 6:18 AM, AMuzi wrote:
This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

Thanks, Andy, this was the best thing I've found on R.B.T. in years!

To everybody:* IF you haven't watched the video yet, it's rather long
and slow, but worth the time, IMHO.* There is actual data obtained by
actual /measurements/ made in a /systematic/ way.

It's admittedly very wonky, but this is a TECH group, yes?


If you want to get even wonkier: It's interesting and relevant that the
higher tension wheels in that test had less variation in tension during
loading. And it reminds me that the design of a tension spoke wheel has
a similarity to the design of a bolted joint subject to fatigue loading.

As an example, think of the cylinder head bolts on an engine or piston
compressor. As cylinder pressure varies, the force in the bolts varies.
It turns out that for that sort of bolted connection, a bolt with a
thinner central section (between the head and the threads) torqued to a
higher preload stress can be much more resistant to fatigue than a
thicker bolt at a lower stress. The strongest bolt is something like the
third one in figure 12 at
https://www.fastenal.com/en/3289/fastener-fatigue

Sorry Frank, but as you state it I just don't agree.
Carried to it's logical limits you are saying that, say a 2" diameter
bolt, with the center section reduced to, oh say 1/2", is stronger
than the straight shank 2 inch bolt?

That's a very extreme example you're giving, with a factor of 16 on the
cross section areas.

I'm talking about a specific loading situation: a varying load
superimposed on a steady load. And a specific failure mode: Fatigue
failure. The other necessary feature is that the bolt (or in our case,
spoke) has to be in tension against another structural element in a way
that the relative deflections are controlled by the relative
stiffnesses. In that sort of sort of situation, a more flexible bolt is
typically more resistant to fatigue failure.

As I wrote, "but as you state it I just don't agree".

Here's an industrial example that came up once when I was teaching this.
One student working in an engineering firm talked about a bolted flanged
connection between two large pipes that was subject to vibration. They
were breaking flange bolts until they installed longer bolts, something
like 8" bolts to clamp something like two inches of total flange thickness.

How did they do that? They added thick wall tubes, something like 6"
long, between the bolt heads and the flange. The main idea was to make
the bolts longer and more flexible. With increased flexibility, the
_variation_ in stress of the bolts was much less, and they better
resisted fatigue.

For such long bolts, the long center section is often not threaded, but
has a diameter less than the minor diameter of the threads. (We'd say
"butted.") This helps two ways: by making the bolt more elastic in
tension, and by reducing stress concentration at the inner ends of the
threads.

Which is a totally different explanation that you originally posted so
I'll stand by my original statement.

Sorry, John. If you think the two explanations were totally different,
you didn't understand. Perhaps it needed more explanation.

I can go into this as deeply as you like. IIRC, it took a week or
more in one of my courses, and was applied frequently after that.

Interestingly, when I took the test for the Professional Engineer's
license, one of the problems involved this stuff.

- Frank Krygowski

I think one of the things that was a bit off putting was the mention
of cylinder head bolts. Not that I've seen every head bolt that ever
existed but I have worked on engines up tp 1,500 KW and can't remember
ever seeing a "stepped" head bolt, or stud. However the through bolts
that hold the 8 sections of an R-4360, 3,750 hp engine "crank case"
together are :-)
--
cheers,

John B.

On Tue, 21 Apr 2020 16:57:06 -0700 (PDT), Frank Krygowski
wrote:

On Tuesday, April 21, 2020 at 6:37:17 PM UTC-4, John B. wrote:
On Tue, 21 Apr 2020 13:07:59 -0400, Frank Krygowski wrote:

On 4/20/2020 8:50 PM, John B. wrote:
On Mon, 20 Apr 2020 19:59:11 -0400, Frank Krygowski wrote:

On 4/20/2020 3:54 PM, Mark J. wrote:
On 4/20/2020 6:18 AM, AMuzi wrote:
This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

Thanks, Andy, this was the best thing I've found on R.B.T. in years!

To everybody:* IF you haven't watched the video yet, it's rather long
and slow, but worth the time, IMHO.* There is actual data obtained by
actual /measurements/ made in a /systematic/ way.

It's admittedly very wonky, but this is a TECH group, yes?


If you want to get even wonkier: It's interesting and relevant that the
higher tension wheels in that test had less variation in tension during
loading. And it reminds me that the design of a tension spoke wheel has
a similarity to the design of a bolted joint subject to fatigue loading.

As an example, think of the cylinder head bolts on an engine or piston
compressor. As cylinder pressure varies, the force in the bolts varies.
It turns out that for that sort of bolted connection, a bolt with a
thinner central section (between the head and the threads) torqued to a
higher preload stress can be much more resistant to fatigue than a
thicker bolt at a lower stress. The strongest bolt is something like the
third one in figure 12 at
https://www.fastenal.com/en/3289/fastener-fatigue

Sorry Frank, but as you state it I just don't agree.
Carried to it's logical limits you are saying that, say a 2" diameter
bolt, with the center section reduced to, oh say 1/2", is stronger
than the straight shank 2 inch bolt?

That's a very extreme example you're giving, with a factor of 16 on the
cross section areas.

I'm talking about a specific loading situation: a varying load
superimposed on a steady load. And a specific failure mode: Fatigue
failure. The other necessary feature is that the bolt (or in our case,
spoke) has to be in tension against another structural element in a way
that the relative deflections are controlled by the relative
stiffnesses. In that sort of sort of situation, a more flexible bolt is
typically more resistant to fatigue failure.

As I wrote, "but as you state it I just don't agree".

Here's an industrial example that came up once when I was teaching this.
One student working in an engineering firm talked about a bolted flanged
connection between two large pipes that was subject to vibration. They
were breaking flange bolts until they installed longer bolts, something
like 8" bolts to clamp something like two inches of total flange thickness.

How did they do that? They added thick wall tubes, something like 6"
long, between the bolt heads and the flange. The main idea was to make
the bolts longer and more flexible. With increased flexibility, the
_variation_ in stress of the bolts was much less, and they better
resisted fatigue.

For such long bolts, the long center section is often not threaded, but
has a diameter less than the minor diameter of the threads. (We'd say
"butted.") This helps two ways: by making the bolt more elastic in
tension, and by reducing stress concentration at the inner ends of the
threads.

Which is a totally different explanation that you originally posted so
I'll stand by my original statement.

Sorry, John. If you think the two explanations were totally different,
you didn't understand. Perhaps it needed more explanation.

I can go into this as deeply as you like. IIRC, it took a week or
more in one of my courses, and was applied frequently after that.

Interestingly, when I took the test for the Professional Engineer's
license, one of the problems involved this stuff.

- Frank Krygowski

I think one of the things that was a bit off putting was the mention
of cylinder head bolts. Not that I've seen every head bolt that ever
existed but I have worked on engines up tp 1,500 KW and can't remember
ever seeing a "stepped" head bolt, or stud. However the through bolts
that hold the 8 sections of an R-4360, 3,750 hp engine "crank case"
together are :-)
--
cheers,

John B.

On 4/21/2020 7:33 PM, pH wrote:
On Tuesday, April 21, 2020 at 6:33:47 AM UTC-7, AMuzi wrote:
On 4/21/2020 3:52 AM, Sepp Ruf wrote:
jbeattie wrote:
On Monday, April 20, 2020 at 3:23:14 PM UTC-7, John B. wrote:
On Mon, 20 Apr 2020 08:18:15 -0500, AMuzi wrote:

This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

One might ask, "Is this a common failure of the bicycle wheel?"
Even if it is with lightweight, anodized, Mavic-grade, double-fried crouton
brittle alloy rims, I'm not concerned because, like suspiciously lifting
eyelets, it has always been early from outside.

Yes, very. It was not so much of an issue in the olden days with 120mm
rear hubs and 5sp freewheels and stouter, high spoke count rims.

Maybe users of disk brakes are not checking the rims regularly? And the
video does not even go l/r asymmetric spoke pulls.

I see Ralph Nader is 86. Who could ring the alarm today? AOC?


I'm still planning to do a tire burn with my Corvair in the
parking lot at his funeral.

--
Andrew Muzi
www.yellowjersey.org/
Open every day since 1 April, 1971

Don't do that....
That would be Unsafe at Any Speed.

pH


If true, one may as well drive fast!

--
Andrew Muzi
www.yellowjersey.org/
Open every day since 1 April, 1971

On 4/21/2020 7:42 PM, John B. wrote:
On Tue, 21 Apr 2020 16:57:06 -0700 (PDT), Frank Krygowski
wrote:

On Tuesday, April 21, 2020 at 6:37:17 PM UTC-4, John B. wrote:
On Tue, 21 Apr 2020 13:07:59 -0400, Frank Krygowski wrote:

On 4/20/2020 8:50 PM, John B. wrote:
On Mon, 20 Apr 2020 19:59:11 -0400, Frank Krygowski wrote:

On 4/20/2020 3:54 PM, Mark J. wrote:
On 4/20/2020 6:18 AM, AMuzi wrote:
This was in my overnight email:

https://www.youtube.com/watch?v=qKeeHDuoFq8

Thanks, Andy, this was the best thing I've found on R.B.T. in years!

To everybody: IF you haven't watched the video yet, it's rather long
and slow, but worth the time, IMHO. There is actual data obtained by
actual /measurements/ made in a /systematic/ way.

It's admittedly very wonky, but this is a TECH group, yes?


If you want to get even wonkier: It's interesting and relevant that the
higher tension wheels in that test had less variation in tension during
loading. And it reminds me that the design of a tension spoke wheel has
a similarity to the design of a bolted joint subject to fatigue loading.

As an example, think of the cylinder head bolts on an engine or piston
compressor. As cylinder pressure varies, the force in the bolts varies.
It turns out that for that sort of bolted connection, a bolt with a
thinner central section (between the head and the threads) torqued to a
higher preload stress can be much more resistant to fatigue than a
thicker bolt at a lower stress. The strongest bolt is something like the
third one in figure 12 at
https://www.fastenal.com/en/3289/fastener-fatigue

Sorry Frank, but as you state it I just don't agree.
Carried to it's logical limits you are saying that, say a 2" diameter
bolt, with the center section reduced to, oh say 1/2", is stronger
than the straight shank 2 inch bolt?

That's a very extreme example you're giving, with a factor of 16 on the
cross section areas.

I'm talking about a specific loading situation: a varying load
superimposed on a steady load. And a specific failure mode: Fatigue
failure. The other necessary feature is that the bolt (or in our case,
spoke) has to be in tension against another structural element in a way
that the relative deflections are controlled by the relative
stiffnesses. In that sort of sort of situation, a more flexible bolt is
typically more resistant to fatigue failure.

As I wrote, "but as you state it I just don't agree".

Here's an industrial example that came up once when I was teaching this.
One student working in an engineering firm talked about a bolted flanged
connection between two large pipes that was subject to vibration. They
were breaking flange bolts until they installed longer bolts, something
like 8" bolts to clamp something like two inches of total flange thickness.

How did they do that? They added thick wall tubes, something like 6"
long, between the bolt heads and the flange. The main idea was to make
the bolts longer and more flexible. With increased flexibility, the
_variation_ in stress of the bolts was much less, and they better
resisted fatigue.

For such long bolts, the long center section is often not threaded, but
has a diameter less than the minor diameter of the threads. (We'd say
"butted.") This helps two ways: by making the bolt more elastic in
tension, and by reducing stress concentration at the inner ends of the
threads.

Which is a totally different explanation that you originally posted so
I'll stand by my original statement.

Sorry, John. If you think the two explanations were totally different,
you didn't understand. Perhaps it needed more explanation.

I can go into this as deeply as you like. IIRC, it took a week or
more in one of my courses, and was applied frequently after that.

Interestingly, when I took the test for the Professional Engineer's
license, one of the problems involved this stuff.

- Frank Krygowski

I think one of the things that was a bit off putting was the mention
of cylinder head bolts. Not that I've seen every head bolt that ever
existed but I have worked on engines up tp 1,500 KW and can't remember
ever seeing a "stepped" head bolt, or stud. However the through bolts
that hold the 8 sections of an R-4360, 3,750 hp engine "crank case"
together are :-)
--
cheers,

John B.


That's quite a piece. Just snooping around I found a guy who
actually made one from scratch!

https://www.nyemachine.com/pratt_whitney_r4360.php


--
Andrew Muzi
www.yellowjersey.org/
Open every day since 1 April, 1971