Residual stress, fatigue and stress relief

Peter Cole wrote:

If there are residual stresses, there must be both tension and
compression residuals. Since the working load is tension, tension
residuals are bad for fatigue, compression residuals are harmless or
beneficial. By overloading in the working direction, tension residuals
-- wherever they occur -- are reduced.

There still seems to be some confusion about stress relief. Where
beneficial residual stresses exist (e.g. rolled threads), stress relief
will not reduce them, if anything, it will increase them. Since residual
compression stress is created by yielding other parts of the section
from tension, a uniform tension overload will reach yield in the areas
with residual tension before those with residual compression.

overload tension + load tension + residual tension yield tension,
overload tension + load tension - residual compression yield tension

This is the basis of overload stress relief. If you don't grasp that
then everything else remains obscure.

Peter Cole wrote:
Peter Cole wrote:

If there are residual stresses, there must be both tension and
compression residuals. Since the working load is tension, tension
residuals are bad for fatigue, compression residuals are harmless or
beneficial. By overloading in the working direction, tension residuals
-- wherever they occur -- are reduced.

There still seems to be some confusion about stress relief. Where
beneficial residual stresses exist (e.g. rolled threads), stress relief
will not reduce them, if anything, it will increase them. Since residual
compression stress is created by yielding other parts of the section
from tension, a uniform tension overload will reach yield in the areas
with residual tension before those with residual compression.

overload tension + load tension + residual tension yield tension,
overload tension + load tension - residual compression yield tension

This is the basis of overload stress relief. If you don't grasp that
then everything else remains obscure.


the only obscurity here is why some people are still confused about
relevance of "stress relief" where it's not observed to be the cause of
fatigue initiation!!!!

buy a magnifier and examine a broken spoke. then correlate the observed
fatigue initiation point to the residual stress profile of a spoke bend.

On Apr 24, 9:59*pm, jim beam wrote:
manufacturers actually state that spokes /shouldn't/ be bent.

Everybody I know who uses Sapim spokes bends them. Dave Ornee, what do
you do?

On Apr 24, 11:58 pm, jim beam wrote:
Peter Cole wrote:

If there are residual stresses, there must be both tension and
compression residuals. Since the working load is tension, tension
residuals are bad for fatigue, compression residuals are harmless or
beneficial. By overloading in the working direction, tension residuals
-- wherever they occur -- are reduced.


actually,

1. overload can introduce more residual stress than it relieves.
2. miner's cumulative damage rule says you're simply reducing fatigue life.

jim beam is mis-applying Miner's rule.

Miner's rule is intended to give approximate estimations of fatigue
life for repeated patterns of cyclical loads. AFAIK, it's never been
intended for evaluating the effect of a single application of a load
causing yielding.

From http://en.wikipedia.org/wiki/Metal_fatigue:

"Though Miner's rule is a useful approximation in many circumstances,
it has two major limitations:

1. It fails to recognise the probabilistic nature of fatigue and
there is no simple way to relate life predicted by the rule with the
characteristics of a probability distribution.
2. There is sometimes an effect in the order in which the reversals
occur. In some circumstances, cycles of low stress followed by high
stress cause more damage than would be predicted by the rule. It does
not consider the effect of overload or high stress which may result in
a compressive residual stress."

Note that last sentence.

- Frank Krygowski

Ron Ruff wrote:
On Apr 24, 9:59�pm, jim beam wrote:
manufacturers actually state that spokes /shouldn't/ be bent.

Everybody I know who uses Sapim spokes bends them.

that's probably because of jobst's bad advice about "correcting the
spoke line". there's no technical reason to do it though because in
reality, once the spoke becomes tensioned, and the hub holes are
deformed, it becomes unnecessary. especially on modern hubs with canted
flanges.

http://www.flickr.com/photos/38636024@N00/331112190/

that's why spokes are formed with 95 degree bends, not 90. it's not
like manufacturers are so dumb they couldn't form it right first time.


Dave Ornee, what do
you do?

wrote:
On Apr 24, 11:58 pm, jim beam wrote:
Peter Cole wrote:

If there are residual stresses, there must be both tension and
compression residuals. Since the working load is tension, tension
residuals are bad for fatigue, compression residuals are harmless or
beneficial. By overloading in the working direction, tension residuals
-- wherever they occur -- are reduced.

actually,

1. overload can introduce more residual stress than it relieves.
2. miner's cumulative damage rule says you're simply reducing fatigue life.

jim beam is mis-applying Miner's rule.

Miner's rule is intended to give approximate estimations of fatigue
life for repeated patterns of cyclical loads. AFAIK, it's never been
intended for evaluating the effect of a single application of a load
causing yielding.

From http://en.wikipedia.org/wiki/Metal_fatigue:

"Though Miner's rule is a useful approximation in many circumstances,
it has two major limitations:

1. It fails to recognise the probabilistic nature of fatigue and
there is no simple way to relate life predicted by the rule with the
characteristics of a probability distribution.
2. There is sometimes an effect in the order in which the reversals
occur. In some circumstances, cycles of low stress followed by high
stress cause more damage than would be predicted by the rule. It does
not consider the effect of overload or high stress which may result in
a compressive residual stress."

Note that last sentence.


and note the krygowski snippage on #3,

"3. you can actually *activate* dislocation movement and thus initiate
fatigue. "

and i quote from an authoritative academic reference:
"while many deviations from miner's rule have been observed, and
numerous modification to this relationship have been proposed, *none
have been proven better or gained wide acceptance*." [my emphasis.]

isbn 0-07-016893-8 p414.

jim beam wrote:
Ron Ruff wrote:
On Apr 24, 9:59�pm, jim beam wrote:
manufacturers actually state that spokes /shouldn't/ be bent.

Everybody I know who uses Sapim spokes bends them.

that's probably because of jobst's bad advice about "correcting the
spoke line". there's no technical reason to do it though because in
reality, once the spoke becomes tensioned, and the hub holes are
deformed, it becomes unnecessary. especially on modern hubs with canted
flanges.

http://www.flickr.com/photos/38636024@N00/331112190/

that's why spokes are formed with 95 degree bends, not 90. it's not
like manufacturers are so dumb they couldn't form it right first time.

http://www.flickr.com/photos/38636024@N00/327722444/




Dave Ornee, what do
you do?

Ron Ruff Wrote:
On Apr 24, 9:59*pm, jim beam wrote:
manufacturers actually state that spokes /shouldn't/ be bent.

Everybody I know who uses Sapim spokes bends them. Dave Ornee, what do
you do?

I bend the heads in ones to a slightly smaller angle (depending on the
hub, crossing pattern and rim ERD) with spokes laced and nipples turned
5 full turns. Some hubs like Phil Wood have canted flanges so I don't
bend spokes on them unless it looks like it will improve the spoke line.
Yes, I saw what Sapim says on their web site about not bending, but I
find that the improved spoke line is benenficial in the long run. I
have had one returned wheel from over 1,000 built with a broken spoke...
and it broke at the end of the spoke shaft just at the elbow. I tossed
the spoke and all 31 of the others in the wheel (no pictures as I forgot
to take the time and wanted to get the wheel back to the rider) and
rebuilt it with all new spokes and new nipples.
Hub is Dura Ace 7700 rear (25th Annivesary) rim is Velocity Fusion.
I saw the spokes in Jim Beams pictures. That is interesting. Some day
I will photograph spokes before & after... inbound and outbound + spoke
holes prebuild and post build, including post stabilizing.
If any of you have such pictures already I would like to see them and
get some suggestions how to get the details photographed such that you
can really discern what is going on.


--
daveornee

daveornee wrote:
Ron Ruff Wrote:
On Apr 24, 9:59*pm, jim beam wrote:
manufacturers actually state that spokes /shouldn't/ be bent.
Everybody I know who uses Sapim spokes bends them. Dave Ornee, what do
you do?

I bend the heads in ones to a slightly smaller angle (depending on the
hub, crossing pattern and rim ERD) with spokes laced and nipples turned
5 full turns. Some hubs like Phil Wood have canted flanges so I don't
bend spokes on them unless it looks like it will improve the spoke line.
Yes, I saw what Sapim says on their web site about not bending, but I
find that the improved spoke line is benenficial in the long run. I
have had one returned wheel from over 1,000 built with a broken spoke...
and it broke at the end of the spoke shaft just at the elbow. I tossed
the spoke and all 31 of the others in the wheel (no pictures as I forgot
to take the time and wanted to get the wheel back to the rider) and
rebuilt it with all new spokes and new nipples.
Hub is Dura Ace 7700 rear (25th Annivesary) rim is Velocity Fusion.
I saw the spokes in Jim Beams pictures. That is interesting. Some day
I will photograph spokes before & after... inbound and outbound + spoke
holes prebuild and post build, including post stabilizing.

this is the hub those spokes came from:
http://www.flickr.com/photos/38636024@N00/104463818/

that hub hole deformation is *definitely* going to affect exit angle.
and thus "correcting" the spoke line /before/ tensioning [and
stabilizing], as most people do, is going to be premature.


If any of you have such pictures already I would like to see them and
get some suggestions how to get the details photographed such that you
can really discern what is going on.

jim beam Wrote:
daveornee wrote:
Ron Ruff Wrote:
On Apr 24, 9:59*pm, jim beam wrote:
manufacturers actually state that spokes /shouldn't/ be bent.
Everybody I know who uses Sapim spokes bends them. Dave Ornee, what
do
you do?

I bend the heads in ones to a slightly smaller angle (depending on
the
hub, crossing pattern and rim ERD) with spokes laced and nipples
turned
5 full turns. Some hubs like Phil Wood have canted flanges so I
don't
bend spokes on them unless it looks like it will improve the spoke
line.
Yes, I saw what Sapim says on their web site about not bending, but
I
find that the improved spoke line is benenficial in the long run. I
have had one returned wheel from over 1,000 built with a broken
spoke...
and it broke at the end of the spoke shaft just at the elbow. I
tossed
the spoke and all 31 of the others in the wheel (no pictures as I
forgot
to take the time and wanted to get the wheel back to the rider) and
rebuilt it with all new spokes and new nipples.
Hub is Dura Ace 7700 rear (25th Annivesary) rim is Velocity Fusion.
I saw the spokes in Jim Beams pictures. That is interesting. Some
day
I will photograph spokes before & after... inbound and outbound +
spoke
holes prebuild and post build, including post stabilizing.

this is the hub those spokes came from:
http://www.flickr.com/photos/38636024@N00/104463818/

that hub hole deformation is *definitely* going to affect exit angle.
and thus "correcting" the spoke line /before/ tensioning [and
stabilizing], as most people do, is going to be premature.


If any of you have such pictures already I would like to see them
and
get some suggestions how to get the details photographed such that
you
can really discern what is going on.


Here are links to 6 pictures I took today of a Dura Ace HB-7700 32H
wheel I built today with Sapim Race 14/15 DB spokes. One side I used my
previous pattern of "correcting the spoke line" and the other side I
made no such correction.
Judge for yourself.


http://www.flickr.com/photos/17085834@N08/2443679334/


http://www.flickr.com/photos/17085834@N08/2443679236/


http://www.flickr.com/photos/17085834@N08/2442851273/


http://www.flickr.com/photos/17085834@N08/2443679416/


http://www.flickr.com/photos/17085834@N08/2442851411/


http://www.flickr.com/photos/17085834@N08/2442851349/


After this experiment I think the next build will be done without
correcting the spoke line.


--
daveornee