Anodizing effect on fatigue life of aluminum alloy

wrote:
On Apr 23, 4:06 pm, Peter Cole wrote:

The sources I cited are pretty unambiguous. It's causal.

Yes, it's causal in the sources you cite. Cite a source that tests
extrusions instead of castings, and you'll have something relevant to
the discussion. CT some cracked rims to prove that they don't have
void defects which could initiate cracking along the line of
anisotropy, and you have reason to believe that anodizing breaks bike
rims.

If your contention is that extruded aluminum parts contain typically
more void defects than castings, please cite some sources.

There are several reasons (besides anisotropy) that favor
circumferential cracking at spoke holes. The 2 most obvious are that the
rim is under substantial circumferential compression and that the
extrusion is usually thinnest at the center. For some rims, cross
section hoop forces from tire pressure also add a tension component
which favors circumferential crack/fatigue. Finally, hollow section
extrusions, like those of double wall rims, will have circumferential
weld zones, formed after the metal passes the mandrel.

The fact that other factors contribute to cracking/fatigue doesn't alter
the fact that anodizing weakens rims in fatigue. Aluminum spends perhaps
90% of its fatigue life in crack initiation mode, anodizing shortens
that phase. The effects of flaws are cumulative. It may be that rim
extrusions are so crappy that anodize treatments don't affect fatigue
life, but I doubt it, real world experience shows otherwise. Again, if
you have any source that shows otherwise please cite it.

On Thu, 24 Apr 2008 05:59:30 -0700 (PDT), wrote:

On Apr 23, 7:35 pm, Michael Press wrote:
In article
,



wrote:
On Apr 23, 8:12 am, Peter Cole wrote:
Peter Cole wrote:
"Fatigue Design of Aluminum Components & Structures", Sharp, Nordmark
and Menzemer 1996

You can use the Amazon "Search inside" feature to see the graph on page
100:
http://www.amazon.com/gp/reader/0070...ref=sib_dp_pt#

The graph shows very large reductions in fatigue strength for 7075
forgings after cleaning with caustic (C22) or acid (C31) baths. It also
shows drastic reductions in fatigue strength for uncleaned, anodized
samples.

From the above graph, thick (50 micrometer) anodizing, reduced the
fatigue life by a factor of about 60 (@35ksi), while even thin (2.5
micrometer) anodizing reduced it by a factor of 6.

I should point out that these sources agree with what Jobst has
explained all along: thick anodizing has a disastrous effect on fatigue
life, and even thin cosmetic anodizing can have significant
consequences. The mechanism, as described in these sources, agrees with
his causal explanation. This is science, there can be no controversy,
except via willful ignorance.

It's only causal if you believe that there are no other factors
affecting fatigue life. You could substitute anodizing for mirror
polishing, and it's not going to improve fatigue life if your
extrusion process left internal voids. Without direct observation of
cracks appearing in the anodized layer and propagating into the metal,
it's not causality. It's correlation, and not even real correlation,
as nobody has actually bothered to pin down incidence rates.

The correlation is in the material Peter cited and quoted.
Anodized structural members are substantially more fatigue prone.

--
Michael Press

And apples are substantially redder than oranges. Again, it's only
causal if anodizing is the only factor affecting fatigue life. This
is not the case, as bicycle rims have high grain anisotropy and the
potential for extrusion induced flaws, which also make members more
fatigue prone. These factors are competing with the anodizing to
break your rim, and there's plenty of evidence that much of the time
they're winning.

No.

They don't compete - if they exist, they collude.

Anodizing is a bad idea, nomatter what "jim beam" says.

On Apr 24, 10:42 am, Peter Cole wrote:
wrote:
On Apr 23, 4:06 pm, Peter Cole wrote:
The sources I cited are pretty unambiguous. It's causal.

Yes, it's causal in the sources you cite. Cite a source that tests
extrusions instead of castings, and you'll have something relevant to
the discussion. CT some cracked rims to prove that they don't have
void defects which could initiate cracking along the line of
anisotropy, and you have reason to believe that anodizing breaks bike
rims.

If your contention is that extruded aluminum parts contain typically
more void defects than castings, please cite some sources.

There are several reasons (besides anisotropy) that favor
circumferential cracking at spoke holes. The 2 most obvious are that the
rim is under substantial circumferential compression and that the
extrusion is usually thinnest at the center. For some rims, cross
section hoop forces from tire pressure also add a tension component
which favors circumferential crack/fatigue. Finally, hollow section
extrusions, like those of double wall rims, will have circumferential
weld zones, formed after the metal passes the mandrel.

The fact that other factors contribute to cracking/fatigue doesn't alter
the fact that anodizing weakens rims in fatigue. Aluminum spends perhaps
90% of its fatigue life in crack initiation mode, anodizing shortens
that phase. The effects of flaws are cumulative. It may be that rim
extrusions are so crappy that anodize treatments don't affect fatigue
life, but I doubt it, real world experience shows otherwise. Again, if
you have any source that shows otherwise please cite it.

I never said that anodize treatments don't affect fatigue life, just
that they aren't always the limiting factor. Even in the best of
extrusions, high anisotropy results in reduced fatigue life. If a
crack nucleates around an internal void or grain boundary, then it
does not matter if the part was anodized. It will fail before cracks
have a chance to nucleate in the surface layer.


Problems of fatigue crack growth in strongly anisotropic Al-alloys
Cerny, Ivo (SVUM a.s.); Ocenasek, Vladivoj; Hnilica, Frantisek Source:
Key Engineering Materials, v 251-252, 2003, p 61-72
ISSN: 1013-9826 CODEN: KEMAEY
Conference: Advances in Fracture and Damage Mechanics, Sep 2-4 2003,
Paderborn, Germany
Publisher: Trans Tech Publications Ltd

Abstract: An Al 4114 (8090 type) alloy, a material to be used in
aircraft structures, is manufactured by extrusion resulting in a
strongly oriented anisotropic microstructure. Fatigue crack growth
(FCG) data and mechanisms affecting the FCG process are an important
knowledge for an assessment of reliability and safety of structures
containing crack-like defects. As a basis, FCG characteristics in the
directions of extrusion and the perpendicular one are needed. The FCG
properties and mechanisms in the Al 4114 alloy were investigated using
CT specimens with side notches to prevent cracking in inappropriate
directions caused by the anisotropy. Unlike FCG in the direction
perpendicular to extrusion, when dependencies typical for Paris region
of stable growth were ascertained, FCG in the extrusion direction was
connected with regions of acceleration and retardation resulting in
significant scatter. Crack closure measurements and fractographical
analyses were performed to explain the irregularities and FCG
mechanisms.

Comparison of unextruded air slip direct chill cast 6061 ingot with
bar stock extruded from conventional direct chill cast 6061 ingot
Bergsma, S.C. (Northwest Aluminum Co); Kassner, M.E. Source: Journal
of Materials Engineering and Performance, v 6, n 4, Aug, 1997, p
469-472
ISSN: 1059-9495 CODEN: JMEPEG
Publisher: ASM International

Abstract: Air Slip direct chill cast 6061 small diameter ingots
(Direct Forge) were compared with 6061 extruded bar stock. The T6
mechanical properties were compared for both the Direct Forge ingot
and the extruded bar stock, as well as cold impact extruded cylinders
produced from Direct Forge small diameter ingot and extruded stock. It
was found that the tensile and fatigue properties of Direct Forge
ingot and cylinders from this ingot were significantly superior to
those of extruded stock and cylinder produced from this stock. The
improved properties are a result of higher solidification rates
leading to smaller alloy-constituent dispersed particles and, thus,
the production of smaller and more stable grain sizes. Direct Forge
has the additional advantages of (a) not requiring hot or cold work
prior to forming impacted extruded or forged parts, (b) being utilized
in the T6 temper without any prior deformation, (c) having isotropic
and consistent properties, (d) not requiring machining to remove
surface segregation or defects, and (e) having more consistent and
refined grain sizes.

On Apr 24, 10:51 am, _
wrote:
On Thu, 24 Apr 2008 05:59:30 -0700 (PDT), wrote:
On Apr 23, 7:35 pm, Michael Press wrote:
In article
,

wrote:
On Apr 23, 8:12 am, Peter Cole wrote:
Peter Cole wrote:
"Fatigue Design of Aluminum Components & Structures", Sharp, Nordmark
and Menzemer 1996

You can use the Amazon "Search inside" feature to see the graph on page
100:
http://www.amazon.com/gp/reader/0070...ref=sib_dp_pt#

The graph shows very large reductions in fatigue strength for 7075
forgings after cleaning with caustic (C22) or acid (C31) baths. It also
shows drastic reductions in fatigue strength for uncleaned, anodized
samples.

From the above graph, thick (50 micrometer) anodizing, reduced the
fatigue life by a factor of about 60 (@35ksi), while even thin (2.5
micrometer) anodizing reduced it by a factor of 6.

I should point out that these sources agree with what Jobst has
explained all along: thick anodizing has a disastrous effect on fatigue
life, and even thin cosmetic anodizing can have significant
consequences. The mechanism, as described in these sources, agrees with
his causal explanation. This is science, there can be no controversy,
except via willful ignorance.

It's only causal if you believe that there are no other factors
affecting fatigue life. You could substitute anodizing for mirror
polishing, and it's not going to improve fatigue life if your
extrusion process left internal voids. Without direct observation of
cracks appearing in the anodized layer and propagating into the metal,
it's not causality. It's correlation, and not even real correlation,
as nobody has actually bothered to pin down incidence rates.

The correlation is in the material Peter cited and quoted.
Anodized structural members are substantially more fatigue prone.

--
Michael Press

And apples are substantially redder than oranges. Again, it's only
causal if anodizing is the only factor affecting fatigue life. This
is not the case, as bicycle rims have high grain anisotropy and the
potential for extrusion induced flaws, which also make members more
fatigue prone. These factors are competing with the anodizing to
break your rim, and there's plenty of evidence that much of the time
they're winning.

No.

They don't compete - if they exist, they collude.

Anodizing is a bad idea, nomatter what "jim beam" says.

They exist, and collusion is unlikely. A crack may only originate
from one location. For collusion between internal and surface cracks,
they would need to form close enough together to intersect. In
reality, anodizing is not the only feature of a rim affecting its
fatigue life. It is only a bad idea under conditions in which
anodizing overrides other features as the limiting factor of fatigue
life, ie. thick, hard anodizing on a low delta extrusion.

On Apr 24, 10:42 am, Peter Cole wrote:
wrote:
On Apr 23, 4:06 pm, Peter Cole wrote:
The sources I cited are pretty unambiguous. It's causal.

Yes, it's causal in the sources you cite. Cite a source that tests
extrusions instead of castings, and you'll have something relevant to
the discussion. CT some cracked rims to prove that they don't have
void defects which could initiate cracking along the line of
anisotropy, and you have reason to believe that anodizing breaks bike
rims.

If your contention is that extruded aluminum parts contain typically
more void defects than castings, please cite some sources.

There are several reasons (besides anisotropy) that favor
circumferential cracking at spoke holes. The 2 most obvious are that the
rim is under substantial circumferential compression and that the
extrusion is usually thinnest at the center. For some rims, cross
section hoop forces from tire pressure also add a tension component
which favors circumferential crack/fatigue. Finally, hollow section
extrusions, like those of double wall rims, will have circumferential
weld zones, formed after the metal passes the mandrel.

Regarding various stresses Peter mentions: Has anyone published a
detailed FEA representation of the stresses in the rim's material?
(Or, for that matter, results of experimental stress analysis, like
brittle coating?)

It would be interesting to see what stresses looked like a) between
spoke holes, b) at spoke holes, when 1) directly at the bottom of a
loaded wheel, and 2) rotated away from the bottom. Admittedly,
modeling it correctly wouldn't be trivial, but if successful, it could
tell us a lot.


The fact that other factors contribute to cracking/fatigue doesn't alter
the fact that anodizing weakens rims in fatigue. Aluminum spends perhaps
90% of its fatigue life in crack initiation mode, anodizing shortens
that phase. The effects of flaws are cumulative. It may be that rim
extrusions are so crappy that anodize treatments don't affect fatigue
life, but I doubt it, real world experience shows otherwise. Again, if
you have any source that shows otherwise please cite it.

I'm sure that this must have been mentioned before, but: All the
influences Peter describes have always existed in bike rims, to one
degree or other. If rim cracks became much more frequent once
anodized rims became popular, that's strong indication (but not proof)
that the anodizing was a big factor. Not the only factor, but perhaps
the one that pushed things over the limit.

- Frank Krygowski

wrote:
On Apr 24, 10:42 am, Peter Cole wrote:
wrote:
On Apr 23, 4:06 pm, Peter Cole wrote:
The sources I cited are pretty unambiguous. It's causal.
Yes, it's causal in the sources you cite. Cite a source that tests
extrusions instead of castings, and you'll have something relevant to
the discussion. CT some cracked rims to prove that they don't have
void defects which could initiate cracking along the line of
anisotropy, and you have reason to believe that anodizing breaks bike
rims.
If your contention is that extruded aluminum parts contain typically
more void defects than castings, please cite some sources.

There are several reasons (besides anisotropy) that favor
circumferential cracking at spoke holes. The 2 most obvious are that the
rim is under substantial circumferential compression and that the
extrusion is usually thinnest at the center. For some rims, cross
section hoop forces from tire pressure also add a tension component
which favors circumferential crack/fatigue. Finally, hollow section
extrusions, like those of double wall rims, will have circumferential
weld zones, formed after the metal passes the mandrel.

The fact that other factors contribute to cracking/fatigue doesn't alter
the fact that anodizing weakens rims in fatigue. Aluminum spends perhaps
90% of its fatigue life in crack initiation mode, anodizing shortens
that phase. The effects of flaws are cumulative. It may be that rim
extrusions are so crappy that anodize treatments don't affect fatigue
life, but I doubt it, real world experience shows otherwise. Again, if
you have any source that shows otherwise please cite it.

I never said that anodize treatments don't affect fatigue life, just
that they aren't always the limiting factor. Even in the best of
extrusions, high anisotropy results in reduced fatigue life. If a
crack nucleates around an internal void or grain boundary, then it
does not matter if the part was anodized. It will fail before cracks
have a chance to nucleate in the surface layer.


Problems of fatigue crack growth in strongly anisotropic Al-alloys
Cerny, Ivo (SVUM a.s.); Ocenasek, Vladivoj; Hnilica, Frantisek Source:
Key Engineering Materials, v 251-252, 2003, p 61-72
ISSN: 1013-9826 CODEN: KEMAEY
Conference: Advances in Fracture and Damage Mechanics, Sep 2-4 2003,
Paderborn, Germany
Publisher: Trans Tech Publications Ltd

Abstract: An Al 4114 (8090 type) alloy, a material to be used in
aircraft structures, is manufactured by extrusion resulting in a
strongly oriented anisotropic microstructure. Fatigue crack growth
(FCG) data and mechanisms affecting the FCG process are an important
knowledge for an assessment of reliability and safety of structures
containing crack-like defects. As a basis, FCG characteristics in the
directions of extrusion and the perpendicular one are needed. The FCG
properties and mechanisms in the Al 4114 alloy were investigated using
CT specimens with side notches to prevent cracking in inappropriate
directions caused by the anisotropy. Unlike FCG in the direction
perpendicular to extrusion, when dependencies typical for Paris region
of stable growth were ascertained, FCG in the extrusion direction was
connected with regions of acceleration and retardation resulting in
significant scatter. Crack closure measurements and fractographical
analyses were performed to explain the irregularities and FCG
mechanisms.

What's the point? All this says is that crack growth rate is uniform
perpendicular to the extrusion axis and erratic parallel.


Comparison of unextruded air slip direct chill cast 6061 ingot with
bar stock extruded from conventional direct chill cast 6061 ingot
Bergsma, S.C. (Northwest Aluminum Co); Kassner, M.E. Source: Journal
of Materials Engineering and Performance, v 6, n 4, Aug, 1997, p
469-472
ISSN: 1059-9495 CODEN: JMEPEG
Publisher: ASM International

Abstract: Air Slip direct chill cast 6061 small diameter ingots
(Direct Forge) were compared with 6061 extruded bar stock. The T6
mechanical properties were compared for both the Direct Forge ingot
and the extruded bar stock, as well as cold impact extruded cylinders
produced from Direct Forge small diameter ingot and extruded stock. It
was found that the tensile and fatigue properties of Direct Forge
ingot and cylinders from this ingot were significantly superior to
those of extruded stock and cylinder produced from this stock. The
improved properties are a result of higher solidification rates
leading to smaller alloy-constituent dispersed particles and, thus,
the production of smaller and more stable grain sizes. Direct Forge
has the additional advantages of (a) not requiring hot or cold work
prior to forming impacted extruded or forged parts, (b) being utilized
in the T6 temper without any prior deformation, (c) having isotropic
and consistent properties, (d) not requiring machining to remove
surface segregation or defects, and (e) having more consistent and
refined grain sizes.

This is about a special billet casting process. It has nothing to do
with the subject under discussion. Please stop wasting our time with
irrelevant references.

wrote:
On Apr 24, 10:51 am, _

They don't compete - if they exist, they collude.

Anodizing is a bad idea, nomatter what "jim beam" says.

They exist, and collusion is unlikely. A crack may only originate
from one location. For collusion between internal and surface cracks,
they would need to form close enough together to intersect. In
reality, anodizing is not the only feature of a rim affecting its
fatigue life. It is only a bad idea under conditions in which
anodizing overrides other features as the limiting factor of fatigue
life, ie. thick, hard anodizing on a low delta extrusion.

Whatsa "low delta" extrusion?

On Apr 24, 12:43 pm, Peter Cole wrote:
wrote:
On Apr 24, 10:51 am, _
They don't compete - if they exist, they collude.

Anodizing is a bad idea, nomatter what "jim beam" says.

They exist, and collusion is unlikely. A crack may only originate
from one location. For collusion between internal and surface cracks,
they would need to form close enough together to intersect. In
reality, anodizing is not the only feature of a rim affecting its
fatigue life. It is only a bad idea under conditions in which
anodizing overrides other features as the limiting factor of fatigue
life, ie. thick, hard anodizing on a low delta extrusion.

Whatsa "low delta" extrusion?

Sorry, I was under the impression that I was talking to someone who
knows something about material processing. Delta is the ratio of
thickness to die contact area. It's the kind of thing that's going to
increase if you're trying to make a very light rim. As it increases,
so does the probability of internal void formation.

wrote:
On Apr 24, 12:43 pm, Peter Cole wrote:

Whatsa "low delta" extrusion?

Sorry, I was under the impression that I was talking to someone who
knows something about material processing.

I'm flattered. Also embarrassed I didn't make the same assumption.


Delta is the ratio of
thickness to die contact area. It's the kind of thing that's going to
increase if you're trying to make a very light rim. As it increases,
so does the probability of internal void formation.

Don't you mean decrease? Funny, I've never heard the term and can't seem
to find any on-line references. Know any?

On Apr 24, 2:25 pm, Peter Cole wrote:
wrote:
On Apr 24, 12:43 pm, Peter Cole wrote:
Whatsa "low delta" extrusion?

Sorry, I was under the impression that I was talking to someone who
knows something about material processing.

I'm flattered. Also embarrassed I didn't make the same assumption.

Delta is the ratio of
thickness to die contact area. It's the kind of thing that's going to
increase if you're trying to make a very light rim. As it increases,
so does the probability of internal void formation.

Don't you mean decrease? Funny, I've never heard the term and can't seem
to find any on-line references. Know any?

The term is pretty standard, and particularly well described in
Hosford and Caddell. Unfortunately, you're probably going to have to
track that down in good old analog format.