[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [femm] Re: BH data
I'm not sure that you can "differentiate" because eddy-currents are part of
what constitutes the apparent inductance of a system. Eddy-currents always
oppose the driving current (Lenz's law) and tend to reduce flux, this
appears externally as a reduction in self-inductance. In transformers and
electrical machines operating at 50Hz I have usually found that static
analysis gives a pretty good assessment of field distribution and allows
inductances to be determined within 10% or so. This is particularly so in
saturated conditions where an ac analysis may not be possible. However,
this depends quite significantly on the construction of the problem. It
most certainly would not be the case in un-laminated situations, where core
eddy-currents usually have dramatic effects. The effects of non-uniform
current distribution in conductors and windings is also not considered. It
really is a typical design problem - deciding what can and can't be
neglected in a particular situation. For example, MikeH says "A basic model
LVDT model using mumetal core at even 100 Hz shows pronounced skin
effect.". I was surprised by this but find it perfectly believable. It
would be interesting to see how the measured inductances of such an LVDT
compare to calculation.
Sorry - I've been rambling a bit and probably have not answered your question.
Keith.
At 16:57 25/10/2002 -0700, you wrote:
I am just wondering how do you differentiate the effect of self inductance
vs. eddy current in FEMM? At 10 - 100Hz, I would think that the self
inductance is the one to consider.
Keith Gregory <k.gregory@xxxxxxxxxxx> wrote:
At 15:31 24/10/2002 +0000, you wrote:
>--- In femm@xxxx, Keith Gregory <k.gregory@xxxx> wrote:
> > Mike,
> >
> > I think you would find that static finite elements models with
>single
> > valued magnetization characteristics would give pretty good
>approximations
> > of field distribution and winding inductances in the 50-60Hz types
>of
> > applications where the materials are laminated. Although
>calculating
> > inductances in saturated conditions is complicated. Loss
>calculations even
> > at 50 or 60 Hz are not as straightforward. Operating in Khz I
>really have
> > no idea how good results would be.
>
>
>A basic model LVDT model using mumetal core at even 100 Hz shows
>pronouned skin effect. At 5kHz there is a danger of saturation near
>the material surface at normal drive levels.
The 5kHz doesn't surprise me but the 100Hz does, although mumetal is
different to the materials I would normally work with.
> >
> > >I am currently interested in trying to observe Barkhausen noise for
> > >NiFe materials. I expect to use up to about 100 Hz field for this,
>and
> > >believe that the spectrum of the noise will contain much
>information,
> > >if it can be unravelled; early days yet. Comments are very welcome
> >
> > The classic demonstration of Barkhausen jumps is done with "DC"
> > magnetization, usually with a permanent magnet. I have a vague
>memory of a
> > paper describing a different way which can produce noise displays
>on an
> > oscilloscope. I'm not sure what information it would contain
>though, what
> > had you got in mind?
>
>
>I hope to use LF ac to cycle the jumps and give larger noise output.
>Literature suggests that Barkhausen noise can indicate heat treatment
>and internal stresses, inter alia; as properties of cores are
>affected by handling stresses I believe BN can provide an insightful
>tool into known quality issues.
I could see that, since Barkhausen noise has been attributed to sudden
domain wall movement and some people say that it is impurities and voids in
the material that make the walls stick, the amount of noise could give an
indication of the material structure and quality.
It would be interesting to see if it really is "noise" or it has some
pattern during repeated cycles.
> > In addition, if my memory serves aren't there several different
>Radiomental
> > alloys differentiated by a number such as Radiometal 36?
>
>
>Yes. The 36 refers to % nickel in the alloy. Varoius types give
>different permeability / resistivity / cost materials for different
>applications.<<http://webhosting.yahoo.com/>Y! Web Hosting - Let the
expert host your web site
Yahoo! Groups Sponsor
ADVERTISEMENT
<http://rd.yahoo.com/M=212804.2460941.3878106.2273195/D=egroupweb/S=1705083663:HM/A=810327/R=0/*http://geocities.yahoo.com/ps/info?.refer=blrecs>366cf6.jpg
<http://rd.yahoo.com/M=212804.2460941.3878106.2273195/D=egroupweb/S=1705083663:HM/A=810327/R=1/*http://geocities.yahoo.com/ps/info?.refer=blrecs>366d3c.jpg<http://rd.yahoo.com/M=212804.2460941.3878106.2273195/D=egroupweb/S=1705083663:HM/A=810327/R=1/*http://geocities.yahoo.com/ps/info?.refer=blrecs>
366d79.jpg
Your use of Yahoo! Groups is subject to the
<http://docs.yahoo.com/info/terms/>Yahoo! Terms of Service.
--=====================_3573268==_.ALT
Content-Type: text/html; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<html>
I'm not sure that you can "differentiate" because eddy-currents
are part of what constitutes the apparent inductance of a system.
Eddy-currents always oppose the driving current (Lenz's law) and tend to
reduce flux, this appears externally as a reduction in self-inductance.
In transformers and electrical machines operating at 50Hz I have usually
found that static analysis gives a pretty good assessment of field
distribution and allows inductances to be determined within 10% or so.
This is particularly so in saturated conditions where an ac analysis may
not be possible. However, this depends quite significantly on the
construction of the problem. It most certainly would not be the case in
un-laminated situations, where core eddy-currents usually have dramatic
effects. The effects of non-uniform current distribution in conductors
and windings is also not considered. It really is a typical design
problem - deciding what can and can't be neglected in a particular
situation. For example, MikeH says "A basic model LVDT model using
mumetal core at even 100 Hz shows pronounced skin effect.". I was
surprised by this but find it perfectly believable. It would be
interesting to see how the measured inductances of such an LVDT compare
to calculation.<br><br>
Sorry - I've been rambling a bit and probably have not answered your
question.<br><br>
Keith. <br><br>
<br>
At 16:57 25/10/2002 -0700, you wrote:<br><br>
<blockquote type=3Dcite class=3Dcite cite>I am just wondering how do you
differentiate the effect of self inductance vs. eddy current in
FEMM? At 10 - 100Hz, I would think that the self inductance is the
one to consider. <br><br>
<br><br>
<b><i>Keith Gregory <k.gregory@xxxxxxxxxxx></i></b> wrote:=20
<dl><tt>
<dd>At 15:31 24/10/2002 +0000, you wrote:
<dd>>--- In femm@xxxx, Keith Gregory <k.gregory@xxxx> wrote:
<dd>> > Mike,
<dd>> >
<dd>> > I think you would find that static finite elements models
with
<dd>>single
<dd>> > valued magnetization characteristics would give pretty
good
<dd>>approximations
<dd>> > of field distribution and winding inductances in the
50-60Hz types
<dd>>of
<dd>> > applications where the materials are laminated. Although
<dd>>calculating
<dd>> > inductances in saturated conditions is complicated. Loss
<dd>>calculations even
<dd>> > at 50 or 60 Hz are not as straightforward. Operating in Khz
I
<dd>>really have
<dd>> > no idea how good results would be.
<dd>>
<dd>>
<dd>>A basic model LVDT model using mumetal core at even 100 Hz=20
shows
<dd>>pronouned skin effect. At 5kHz there is a danger of saturation
near
<dd>>the material surface at normal drive levels.<br><br>
<dd>The 5kHz doesn't surprise me but the 100Hz does, although mumetal is=20
<dd>different to the materials I would normally work with.<br><br>
<dd>> >
<dd>> > >I am currently interested in trying to observe
Barkhausen noise for
<dd>> > >NiFe materials. I expect to use up to about 100 Hz
field for this,
<dd>>and
<dd>> > >believe that the spectrum of the noise will contain
much
<dd>>information,
<dd>> > >if it can be unravelled; early days yet. Comments are
very welcome
<dd>> >
<dd>> > The classic demonstration of Barkhausen jumps is done with
"DC"
<dd>> > magnetization, usually with a permanent magnet. I have a
vague
<dd>>memory of a
<dd>> > paper describing a different way which can produce noise
displays
<dd>>on an
<dd>> > oscilloscope. I'm not sure what information it would
contain
<dd>>though, what
<dd>> > had you got in mind?
<dd>>
<dd>>
<dd>>I hope to use LF ac to cycle the jumps and give larger noise
output.
<dd>>Literature suggests that Barkhausen noise can indicate heat
treatment
<dd>>and internal stresses, inter alia; as properties of cores are
<dd>>affected by handling stresses I believe BN can provide an
insightful
<dd>>tool into known quality issues.<br><br>
<dd>I could see that, since Barkhausen noise has been attributed to
sudden=20
<dd>domain wall movement and some people say that it is impurities and
voids in=20
<dd>the material that make the walls stick, the amount of noise could
give an=20
<dd>indication of the material structure and quality.<br><br>
<dd>It would be interesting to see if it really is "noise" or
it has some=20
<dd>pattern during repeated cycles.<br><br>
<dd>> > In addition, if my memory serves aren't there several
different
<dd>>Radiomental
<dd>> > alloys differentiated by a number such as Radiometal 36?
<dd>>
<dd>>
<dd>>Yes. The 36 refers to % nickel in the alloy. Varoius types give
<dd>>different permeability / resistivity / cost materials for
different
<dd>>applications.<<a href=3D"http://webhosting.yahoo.com/";>Y! Web
Hosting</a> - Let the expert host your web site
<font face=3D"Courier New, Courier" size=3D2 color=3D"#003399">
<dd>Yahoo! Groups Sponsor</b></font> <font face=3D"arial" size=3D1>
<dd>ADVERTISEMENT</font>
<dd><a href=3D"http://rd.yahoo.com/M=3D212804.2460941.3878106.2273195/D=3De=
groupweb/S=3D1705083663:HM/A=3D810327/R=3D0/*http://geocities.yahoo.com/ps/=
info?.refer=3Dblrecs"><img src=3D"cid:.0"; width=3D185 height=3D250 alt=3D"3=
66cf6.jpg"></a>
<dd><a href=3D"http://rd.yahoo.com/M=3D212804.2460941.3878106.2273195/D=3De=
groupweb/S=3D1705083663:HM/A=3D810327/R=3D1/*http://geocities.yahoo.com/ps/=
info?.refer=3Dblrecs"><img src=3D"cid:.1"; width=3D115 height=3D250 alt=3D"3=
66d3c.jpg"></a><a href=3D"http://rd.yahoo.com/M=3D212804.2460941.3878106.22=
73195/D=3Degroupweb/S=3D1705083663:HM/A=3D810327/R=3D1/*http://geocities.ya=
hoo.com/ps/info?.refer=3Dblrecs"> </a>
<dd><img src=3D"cid:.2"; width=3D1 height=3D1 alt=3D"366d79.jpg"><br><br>
<dd>Your use of Yahoo! Groups is subject to the
<a href=3D"http://docs.yahoo.com/info/terms/";>Yahoo! Terms of Service</a>.
</tt><br><br>
</dl></blockquote></html>
--=====================_3573268==_.ALT--
