WIMPS?

Op donderdag 27 juni 2013 22:10:10 UTC+2 schreef Steve Willner:
In article ,

Nicolaas Vroom writes:
5% is ordinary matter, 27% is dark matter and 68% is dark energy.

That should be "baryonic matter" and "non-baryonic matter" to be
accurate in terminology.
I fully agree.
Is it not time to modify the text in Wikipedia?

The numbers 5 and 27 come from the CMB radiation.

Not exactly; the numbers include a variety of data, all of which are
consistent. In particular, the supernova distances and baryon
acoustic oscillation data are important, as is the local measurement
of the Hubble parameter.
My impression is that an accurate measurement of the Hubble parameter
based on local data is the most difficult.

To explain a galaxy rotation curve based on a concept of dark matter
seems to me not very plausible.
Whatever is causing the flat rotation curves isn't visible in any
form we can detect.
I agree.

The most obvious candidate, I have
expressed before, is dust i.e. baryonic matter.

Dust won't get you anywhere; it would have a visible effect on star
colors. You might try cannon balls, though. In fact, any baryonic
particles from submillimeter to megameter size wouldn't be easily
detectable.
That is really what I meant.

Last time I checked -- and data have no doubt improved
since then -- if you took the full 5% allowance of baryons and packed
them into spiral disks, you could just about explain the rotation
curves.
My simulations more or less give identical results. You do not need
hugh amounts to simulate flat galaxy rotation curves. At least much
less mass than is directly visible.
If you consider certain galaxy examples the bulge does not require
anything.
There are, however, several problems: if 70% or so of the
baryon mass is hydrogen, why don't we detect it? etc
What about planet sized objects ?

Even if you can explain those problems, you still have to account for
galaxy cluster velocity dispersions and gravitational lensing.
What do you mean with velocity dispersions?

Nicolaas Vroom

In article , Nicolaas Vroom
writes:

Even if you can explain those problems, you still have to account for
galaxy cluster velocity dispersions and gravitational lensing.
What do you mean with velocity dispersions?

Look it up. Historically, this was the first hint of dark matter.
Fritz Zwicky noticed that the velocities of galaxies within clusters are
too large for them to be gravitationally bound to the cluster if only
the mass in stars is taken into account. Since there are other reasons
to believe they are bound, the obvious answer is that there is more
matter than we can see.

[Mod. note: http://en.wikipedia.org/wiki/Velocity_dispersion -- mjh]

In article ,
Nicolaas Vroom writes:
Is it not time to modify the text in Wikipedia?

I haven't looked at it, but by all means modify if you think it needs
it. Isn't that the Wikipedia idea? There's probably more than one
topic that needs modifying.

My impression is that an accurate measurement of the Hubble parameter
based on local data is the most difficult.

It's gotten a lot better in the past several years. A 2012 paper
(Freedman et al. ApJ 758, 24) gives 74.3 with a systematic
uncertainty of 2.1. The main improvement is using infrared
magnitudes for Cepheids. I haven't looked at exactly what they've
done, for example whether they include the trigonometric parallax
distance to NGC 4258 or not. If not, the local value can be improved
further even with existing data. Even as it is, the uncertainty is
comparable to that from the CMB.

My simulations more or less give identical results. You do not need
hugh amounts to simulate flat galaxy rotation curves. At least much
less mass than is directly visible.

I'm surprised by that result. Where are these simulations published?

SW There are, however, several problems: if 70% or so of the
SW baryon mass is hydrogen, why don't we detect it?

What about planet sized objects ?

If you can make them out of (mostly) hydrogen, no problem, but the
individual planet masses have to be small enough to make the planets
unobserved in the lensing observations. It may be hard to make
hydrogen planets that small. Hydrogen objects as small as cannon
balls seem impossible.

On another topic in this thread, someone asked about non-baryonic
matter having a non-zero interaction cross section. By coincidence,
I just received the following invitation:

We are writing to invite you to a workshop on self-interacting
dark matter and challenges to LambdaCDM cosmology. Our goal is to
bring together a mix of astrophysicists and particle theorists to
discuss the possibility that dark matter has interesting
self-interactions, as well as possible observational problems
(like the core/cusp problem or the recent "Too Big To Fail"
problem) with the standard picture of cold, collisionless dark
matter.

We intend this to be a workshop, not a conference, with ample
opportunity for free discussion and collaboration.

I don't plan to attend; this isn't my field. Nevertheless, we can
see that the theorists are having fun with this question.

--
Help keep our newsgroup healthy; please don't feed the trolls.
Steve Willner Phone 617-495-7123
Cambridge, MA 02138 USA

On 6/28/13 12:19 AM, Phillip Helbig---undress to reply wrote:
Faster-than-light neutrinos and the Pioneer anomaly have both been
explained by rather simple mechanisms. While we should be open-minded,
we shouldn't be too quick to conclude that "the textbooks need to be
rewritten", especially if there is a discrepancy between only a couple
of measurements.

The Pioneer acceleration data used to test the thermal recoil hypothesis
retains an acceleration residual with time
that is not explained within the thermal recoil hypothesis.

[Mod. note: reference, please -- mjh]

On 7/4/13 11:22 AM, Richard D. Saam wrote:
On 6/28/13 12:19 AM, Phillip Helbig---undress to reply wrote:
Faster-than-light neutrinos and the Pioneer anomaly have both been
explained by rather simple mechanisms. While we should be open-minded,
we shouldn't be too quick to conclude that "the textbooks need to be
rewritten", especially if there is a discrepancy between only a couple
of measurements.

The Pioneer acceleration data used to test the thermal recoil hypothesis
retains an acceleration residual with time
that is not explained within the thermal recoil hypothesis.

[Mod. note: reference, please -- mjh]

the Pioneer deceleration anomaly
as thermally hypothetically induced
is based on classical exponential decay to zero (dx/dt = -kx)
as an expression of changing on board thermal energy source(RTG).

Ref: 1 http://arxiv.org/abs/1107.2886v1
2. http://arxiv.org/abs/1204.2507v1

Visually and mathematically in Ref2 Figure 3,
the thermal force has less contribution
to Pioneer 10 deceleration with time
with the Pioneer 10 approaching a constant deceleration (not zero)
This constant deceleration was originally suggested by John Anderson.
The Pioneer thermal reactive force may indeed
approach zero, tracking RTG output with time (dx/dt = -kx)
with decreasing affect on constant Pioneer deceleration.

Based on ref1&2 data extrapolated to infinity,
the Pioneer 10 constant deceleration is 6.9E-10 m/sec^2

Based on ref1 data extrapolated to infinity,
the Pioneer 11 constant deceleration is 8.2E-10 m/sec^2.

These numbers are generally in line
with Anderson's constant deceleration numbers.

The question remains:
What causes the constant deceleration?

Richard D Saam

Op donderdag 4 juli 2013 09:29:23 UTC+2 schreef Steve Willner het volgende:
In article ,

Nicolaas Vroom writes:

My simulations more or less give identical results. You do not need
hugh amounts to simulate flat galaxy rotation curves. At least much
less mass than is directly visible.

I'm surprised by that result. Where are these simulations published?

The results of my simulations are not published. They are only available at my homepage.
For me astronomy is "only" a hobby, but it keeps me very busy.
To see the results go he
http://users.telenet.be/nicvroom/circ11.xls.htm#Circ16
The simulations are performed using an excel program.
Test 4 shows a simulation of a flat galaxy rotation curve with a flat speed of 250 km/sec.
Test 5 shows a simulation of a galaxy rotation curve which starts at 250 and slowly decreases to 200 km/sec. The amount of matter used is roughly 30% less.
For Test6 the final speed is 150 km/sec and for Test7 100 km/sec
For a more general discussion go he
http://users.telenet.be/nicvroom/dark_mat.htm

Nicolaas Vroom

"Richard D. Saam" writes:
Ref: 1. http://arxiv.org/abs/1107.2886v1
2. http://arxiv.org/abs/1204.2507v1

Visually and mathematically in Ref2 Figure 3, the thermal force has
less contribution to Pioneer 10 deceleration with time with the
Pioneer 10 approaching a constant deceleration (not zero)

The authors of ref 2 don't agree with your math:

"To determine if the remaining 20% represents a statistically
significant acceleration anomaly not accounted for by conventional
forces, we analyzed the various error sources that contribute to the
uncertainties in the acceleration estimates using radio-metric
Doppler and thermal models.
[...]
We therefore conclude that at the present level of our knowledge of
the Pioneer 10 spacecraft and its trajectory, no statistically
significant acceleration anomaly exists."

The primary source of uncertainty is "the unknown change in the
properties of the RTG coating", a systematic effect. Unlike
statistical uncertainties, systematic effects usually shift all the
data in the same direction; shift all the thermal acceleration data up
by around .8 sigma or so, and the visual impression of an unaccounted
constant acceleration disappears. This is easier to see in fig 4, the
plot of the error ellipses, which shows an unaccounted effect of less
than 1 sigma significance.

Based on ref1&2 data extrapolated to infinity,
the Pioneer 10 constant deceleration is 6.9E-10 m/sec^2

Based on ref1 data extrapolated to infinity,
the Pioneer 11 constant deceleration is 8.2E-10 m/sec^2.

Confidence intervals?

-dan

On 7/8/13 2:16 PM, Dan Riley wrote:
"Richard D. Saam" writes:
Ref: 1. http://arxiv.org/abs/1107.2886v1
2. http://arxiv.org/abs/1204.2507v1

Visually and mathematically in Ref2 Figure 3, the thermal force has
less contribution to Pioneer 10 deceleration with time with the
Pioneer 10 approaching a constant deceleration (not zero)

The authors of ref 2 don't agree with your math:
The authors tested the decay hypothesis
but not the decay to constant hypothesis.

"To determine if the remaining 20% represents a statistically
significant acceleration anomaly not accounted for by conventional
forces, we analyzed the various error sources that contribute to the
uncertainties in the acceleration estimates using radio-metric
Doppler and thermal models.
[...]
We therefore conclude that at the present level of our knowledge of
the Pioneer 10 spacecraft and its trajectory, no statistically
significant acceleration anomaly exists."

The primary source of uncertainty is "the unknown change in the
properties of the RTG coating", a systematic effect. Unlike
statistical uncertainties, systematic effects usually shift all the
data in the same direction; shift all the thermal acceleration data up
by around .8 sigma or so, and the visual impression of an unaccounted
constant acceleration disappears. This is easier to see in fig 4, the
plot of the error ellipses, which shows an unaccounted effect of less
than 1 sigma significance.
How can an "the unknown change in the properties of the RTG coating"
be used in any conclusion.
Ref 1 & 2 thermal analysis was based on thousands of finite elements
each with unknown radiation absorptivity and emmisivity factors.
It comes down to choosing the factors to 'curve fit'.
Compare this to the minimal observed doppler uncertainty.

Based on ref1&2 data extrapolated to infinity,
the Pioneer 10 constant deceleration is 6.9E-10 m/sec^2

Based on ref1 data extrapolated to infinity,
the Pioneer 11 constant deceleration is 8.2E-10 m/sec^2.

Confidence intervals?

-dan



I have done a more explicit analysis below
with digitized deceleration(aP) data(ref 1 & 2)
as a function of time(t) (years after launch)
and modeled according to

aP = aPo * exp(t*ln(2)/half_life) + aPinfinity

where decaying thermal (or other) decelerating effects
reduce with time
with Pioneers approaching a constant deceleration(aPinfinity).

*********************
For Pioneer 10 Model,
given initial deceleration aPo 9.82 x 10^-10 m/sec^2
at time 8.79 years from Table 1,
then multivariable regression minimizing
stochastic model Root Mean Square(RMS) data
from Table 1 columns 1 and 2
yields the following modeled constants:

half_life = 5.00 years

aPinfinity = 7.00 x 10^-10 m/sec^2

Root Mean Square (RMS) = .224 x 10^-10 m/sec^2

with modeled aP in column 3.

Table 1 (digitized stochastic data from ref 1 and 2)
Pioneer 10
Time(t) Stochastic aP Modeled aP
(years) x 10^-10 m/s^2 x 10^-10 m/s^2
to 8.79 aPo 9.82 9.91
10.78 9.33 9.20
12.79 8.78 8.65
14.82 8.21 8.24
16.81 8.21 7.94
18.80 7.39 7.71
20.81 7.34 7.53
22.82 7.23 7.40
24.85 7.72 7.30

*********************
For Pioneer 11 Model,
given initial deceleration aPo 9.274 x 10^-10
at time 10.632 years from Table 2,
then multivariable regression minimizing
stochastic – model Root Mean Square(RMS) data
from Table 2 columns 1 and 2
yields the following modeled constants:

half_life = 3.30 years

aPinfinity = 8.20 x 10^-10 m/sec^2

Root Mean Square (RMS) = .160 x 10^-10 m/sec^2

with modeled aP in column 3.

Table 2 (digitized stochastic data from ref 1 and 2)
Pioneer 11
Time(t) Stochastic aP Modeled aP
(years) x 10^-10 m/s^2 x 10^-10 m/s^2
to 10.632 aPo 9.274 9.187
12.623 8.840 8.832
14.631 8.358 8.604
16.602 8.635 8.460

*********************
Assuming aPinfinity = 0 (exponential decay to 0)
then
for Pioneer 10
half_life = 61 years and RMS = .478 x 10^-10 m/sec^2
and
for Pioneer 11
half_life = 160 years and RMS = .273 x 10^-10 m/sec^2

Conclusion:

A model such as:

aP = aPo * exp(t*ln(2)/half_life) + aPinfinity

is more representative of the physical phenomenon
than straight decay. dx/dt = -kx

In article ,
Nicolaas Vroom writes:
For a more general discussion go he
http://users.telenet.be/nicvroom/dark_mat.htm

There are several misconceptions on that page, but I am not sure we
disagree on the overall result: increasing the mass of a galaxy disk
but keeping the same mass distribution cannot produce a flat rotation
curve. Do we agree on that?

If we do, there are two ways out: modified gravity law ("MOND") or
extra matter that is not distributed the same way as the visible
stars. We call the latter "dark matter" (DM). Whether it's baryonic
or not is a separate question. People are working on MOND but so far
without much success. That is to say, one can always "tune" a
gravity law to explain one or a few galaxies, but no single
alternative model explains all the relevant cases.

Leaving MOND aside, I think there is a small amount of parameter
space that would let the DM needed to explain galaxy rotation curves
be baryonic. Presumably the DM is mostly made out of hydrogen (else
where is the hydrogen that should go with it?), and that requires
fairly large objects (say Saturn-size or larger). Lensing
observations put upper limits on the number of such objects, and I'm
not sure whether they are yet sensitive enough to rule out such
objects as major contributors to galaxy mass. Larger objects such as
white dwarfs and super-Jupiters are, I think, ruled out.

Another limit is the overall baryon budget. We know the average
density of baryons in the Universe (from CMB observations and from
Big Bang nucleosynthesis), and we know where a lot of the baryons
reside. One example is at
http://inspirehep.net/record/1081235/plots
(Look all the way at the bottom.) Are there enough left over to be
associated with galaxies at all? The "missing 29%" on the plot might
be enough, but lately I've heard something about more than expected
very hot gas detected in X-rays. That would leave less than 29%,
maybe zero, to be associated with galaxies and explain their rotation
curves.

As I say, I'm not sure baryonic dark matter is entirely ruled out
here, but the available parameter space for it is shrinking. And
even if you can manage the galaxy rotation curves, there is no way to
have enough baryonic matter to explain the galaxy cluster velocity
dispersions.

People interested in playing with galaxy rotation curves may like
http://burro.astr.cwru.edu/JavaLab/RotcurveWeb/

--
Help keep our newsgroup healthy; please don't feed the trolls.
Steve Willner Phone 617-495-7123
Cambridge, MA 02138 USA

"Richard D. Saam" writes:
The authors tested the decay hypothesis
but not the decay to constant hypothesis.

They tested whether any additional parameters were needed to model
the observations, and found no compelling evidence that anything
more was necessary.

How can an "the unknown change in the properties of the RTG coating"
be used in any conclusion.
Ref 1 & 2 thermal analysis was based on thousands of finite elements
each with unknown radiation absorptivity and emmisivity factors.
It comes down to choosing the factors to 'curve fit'.

The thermal model isn't a fit to the doppler data. It was developed
independent of the doppler data using measured material properties fit
to boundary conditions supplied by the thermal telemetry.

Changes in the RTG coating play a big role because they don't have
good measurements of space environment effects on the coating, they
don't have a way to measure it from the thermal telemetry, and they
aren't fitting to the doppler data.

Compare this to the minimal observed doppler uncertainty.

The top graph of figure 3 doesn't plot the doppler uncertainty at all,
but you can see what it is from the residuals in the lower part of
figure 3, or from the error ellipse in figure 4, or from figure 4 of
gr-qc/0507052. The doppler uncertainty isn't minimal--it's about the
same size as the thermal model uncertainty.

I have done a more explicit analysis below

Still no confidence intervals (the dominant errors are not stochastic,
so minimizing stochastic model RMS isn't appropriate).

-dan