Gravitational-wave Events and Jump Phenomenon

[Mordred]

You didn't answer the above questions I had. Did you understand the formula above?

 

There is roughly 6 main equations but they won't do any good unless you understand my last post.

 

that link you posted has the key details on signal processing. The graphs above include the whitening background noise. You need to filter that noise out in order to see the GW Chirp.

 

As mentioned by Imatfaal. You can study that section. I however can detail the GW signal itself. However that won't do any good unless your clear on the polarization differences between GW and electromagnetic interferance.

 

The latter detail is needed to set the filter signal trigger. Think of an oscilloscope that isn't locked into a specific signal. It looks noisy until you filter out the background noise.

 

In the noise dominated graphs above you need those details to trigger (lock onto) the GW signal.

 

Secondly I can't show you the difference in arm movement of the detector without you understanding the above formula (at least conceptually).

 

(I'm positive you don't desire to merely take my word on it, but would rather like the tools to confirm yourself)

Edited September 7, 2016 by Mordred

[worlov]

Basically, general relativity has nothing to do with jump phenomenon. I can well imagine that LIGO detectors were developed so that they would only react to gravitational waves and not to electromagnetic waves. In fact, I would rather prefer an antenna for receiving the programs of the radio stations.

Jump phenomenon is the internal affair of the detectors and has a purely mechanical cause. In the diagram (source: https://losc.ligo.org/s/events/GW150914/GW150914_tutorial.html ) we can see that the oscillations during the crash of the signal have the same period T as the free oscillations before and after event.

Because these oscillations are own oscillations, they are determined attenuated. But this balance can be injured by a disturbance. The system goes out of control, which may end with jump phenomenon.

Such a scenario is not mentioned anywhere. Therefore, I suspect that the operators of LIGO have this potential source of error simply overlooked.

Edited September 7, 2016 by worlov

[Mordred]

Im well aware of of the statistical aspects of jump phenomenon. I can't show you the distinction without detailing the GW wave. Hence my question.

 

If you wish to examine the data you have understand its design including detector design.

 

The detectors are dominated by mechanical noise below 60 hertz. How do you filter out that noise?

 

without understanding the signal you wish to detect and the detector design you cannot answer that question.

 

Nor can you conclude "Jump" without that understanding. It sounds to me like you wish to draw conclusions without understanding...

Edited September 7, 2016 by Mordred

[worlov]

I have suspected the active damping: https://www.ligo.caltech.edu/page/ligo-technology

The source of error may be due to software.The regulation must be quite complicated:

https://www.dspace.com/shared/data/pdf/dspace_magazine/dSPACE-Magazine_LIGO_2008-2_en.pdf

Who looks through? Certainly, PID controller is realized, because it allows fastest and most accurate control. But... from experience I know that the PID controller are unstable when high frequency interference is present. Therefore, I prefer in my practice the PI controller. Yes, it is carrier, but stable.

So, the software of the system for active damping can have a quirk.

[Mordred]

Have you ever looked at the attenuation formula for a GW/electromagnetic interaction? You wouldn't apply PID in the mannerism as you would a transverse dipole. Trust me I have years of experience with PID and motion control filtering.

alright lets detail some differences between transverse dipole, vs quadrupole.

lets look at the behavior or [latex]h^{GW}_{jk}[/latex] under boosts in the z direction. Then compare to EM waves in transverse Lorentz quage...

GW [latex]h^{GW}_{jk},h_+,h_x[/latex] EM wave [latex]A^T_j[/latex] notice we don't have the k subscript in the EM guage??

The same applies when you transform as scalar fields.

 

GW

 

[latex]h^{GW}_{jk},h_+,h_x[/latex]

[latex]\acute{h}_{jk}(\acute{t}-\acute{z})=h_{jk}(t-z)=h_{jk}(D(\acute{t}-\acute{z})[/latex]

 

EM

 

[latex]A_x(t-z),A_y(t-z)[/latex] transforms to scalar field [latex]\acute{A}_j(\acute{t}-\acute{z})=A_j(D\acute{t}-\acute{z})[/latex]

 

now in the electromagnetic case each rotation is 90 degrees.. In the GW spin 2 each rotation is 45 degrees.

the GW wave attenuation through matter is

[latex]h_{jk}\sim exp(-z/\ell_{att})[/latex]

the ratio of GW energy to EM wave energy [latex]\frac{T_{GW}}{T_{EM}}=\frac{\dot{h}_+/16\pi}{B_0^2/8\pi}[/latex]

 

If you study the formulas you can draw several conclusions.. which I won't post all the math for...

gravity waves travel without significant attenuation, scattering,dispersion or conversion into EM waves.

If you want further detail on the EM GW interaction I would recommend googling Gertsenshtein effect.

 

By the way mechanical noise doesn't follow spin 2 polarization statistics You can confirm that by performing a Fourier series expansion on various types of vibration. A straight line is one term, a sinusoidal two terms.

 

Though quite frankly you will want a multi-degrees of freedom Langrange's equations can easily derive the equations of motion

 

 

I just detailed key differences between GW and EM...

 

A key note is a spin 2 quadrupole wave has no dipole moment.

 

PS I have little doubt that LIGO examined all the natural resonance frequencies for their locality. After all they probably hired a slew of vibrational analysis studies. I'm fairly confident they know how to critically dampen the PID instructions. Yes PI is better in most cases. I tend to use just PI myself, however there have been sytems where I needed PID.

 

I am aware the standard formulas used to calculate PID tend to lead to a quarterly amplitude decay. As such I further critically dampen depending on the system specifications.

 

You should note sound waves is longitudinal. There is three types of mechanical vibration. Transverse, longitudinal (compression) and torsional.

Edited September 8, 2016 by Mordred

[worlov]

The graphs above include the whitening background noise. You need to filter that noise out in order to see the GW Chirp.

 

 

Back to GW150914. The key to understanding of the data processing provides the following picture:

 

https://physics.aps.org/articles/v9/52

 

The numerical relativity (NR) template is band and notch filtered as well as the measurement data itself. In this way it fit with 5.1-sigma to the measurement data.

 

I have this also checked. We can fetch the data of "matched NR waveform" from the following URL: https://losc.ligo.org/s/events/GW150914/GW150914_4_NR_waveform.txt

 

 

Find you this justified?

[Mordred]

absolutley, have you studied why the differences between the two graphs? Which are both simulated datasets.

 

What precisely is meant by whitened and bandpass to generate the second graph? You have the LIGO links with some of those answers...

 

As far as electromagnetic and vibration is concerned the detector design itself (L shape) can distinquish from a GW wave via its polarity x and y axis transitions.

 

That detail isn't shown on those strain graphs.

 

Have you looked at the specific formulas to generate the first graph ?

 

Lets ask another related question. What frequency will you target with the bandpass filters when you don't know the frequency? You only know an allowable range?

 

Second question What length will you cut the antenna to resonate best to a specific frequency? When you only know the range? Ever wondered why the LIGO arms need to be the length they are? quarter wave matching

The first graph is an idealized perfect detector. That in itself is impossible hence the second graph

Edited September 12, 2016 by Mordred

[worlov]

absolutley...

 

I doubt this. In this way one has not directly confirmed the relativistic simulation. Meanwhile, the relativistic curve is changed significantly by the filtration. There remained only a "zilch". I lack a convincing return from "zilch" to the original function. Then the evidence would have been perfect.

[Mordred]

Well I did ask if you looked at the specific formulas to generate the left hand graph. Have you? Did you understand the different requirements to detect a GW wave vs a dipole wave?

 

I'm really trying to get a handle on your objections as its seemingly oriented toward calibration issues rather than the GW vs dipole differences.

[worlov]

Well I did ask if you looked at the specific formulas to generate the left hand graph. Have you? Did you understand the different requirements to detect a GW wave vs a dipole wave?

 

I'm sorry, I did not try. I am ignorant... but for good reason. I accept that if a gravitational wave actually comes, the detectors will notice it. And when an electromagnetic wave comes, the detectors will ignore them.

Now we look at the noise of the detectors:

What have the "Power line harmonic" lost here? Do they have anything to do with the gravitational waves? - Nothing! Nevertheless, they are in the measurement data. And what a coincidence: the signal reached the maximum amplitude at about 180Hz, that is at the third harmonic.

If the detectors are working properly, then it may have been the lightning strike on the overhead power line, what the detectors have seen.

I'm really trying to get a handle on your objections as its seemingly oriented toward calibration issues rather than the GW vs dipole differences.

 

Thanks for your patience. Actually, I'm to blame, because I immediately addressed several problems. Hence it is a mess. But we can also conclude the discussion.

[Mordred]

Ok you obviously aren't seeing the key difference. Lets try an antenna analogy.

 

Visualize th Ligo detector as having two seperate antennas. One vertical, one horizontal. Now if you ever worked with antennas. You will know that the vertical antenna will not detect an electromagnetic signal but the horizontal antenna will.

 

Try it take a transmitter and an antenna

 

place the antenna horizontal then vertical. In one position you can detect the electromagnetic signal in the other position the signal is significantly weakened.

 

An electromagnetic wave only polarizes in 1 axis. A GW wave has TWO simultaneously polarizations.

 

IT IS COMPLETELY distinct from the polarization of an electromagnetic signal....

Have you ever looked at the attenuation formula for a GW/electromagnetic interaction? You wouldn't apply PID in the mannerism as you would a transverse dipole. Trust me I have years of experience with PID and motion control filtering.

alright lets detail some differences between transverse dipole, vs quadrupole.

lets look at the behavior or [latex]h^{GW}_{jk}[/latex] under boosts in the z direction. Then compare to EM waves in transverse Lorentz quage...

GW [latex]h^{GW}_{jk},h_+,h_x[/latex] EM wave [latex]A^T_j[/latex] notice we don't have the k subscript in the EM guage??

The same applies when you transform as scalar fields.

 

GW

 

[latex]h^{GW}_{jk},h_+,h_x[/latex]

[latex]\acute{h}_{jk}(\acute{t}-\acute{z})=h_{jk}(t-z)=h_{jk}(D(\acute{t}-\acute{z})[/latex]

 

EM

 

[latex]A_x(t-z),A_y(t-z)[/latex] transforms to scalar field [latex]\acute{A}_j(\acute{t}-\acute{z})=A_j(D\acute{t}-\acute{z})[/latex]

 

now in the electromagnetic case each rotation is 90 degrees.. In the GW spin 2 each rotation is 45 degrees.

the GW wave attenuation through matter is

[latex]h_{jk}\sim exp(-z/\ell_{att})[/latex]

the ratio of GW energy to EM wave energy [latex]\frac{T_{GW}}{T_{EM}}=\frac{\dot{h}_+/16\pi}{B_0^2/8\pi}[/latex]

 

If you study the formulas you can draw several conclusions.. which I won't post all the math for...

gravity waves travel without significant attenuation, scattering,dispersion or conversion into EM waves.

If you want further detail on the EM GW interaction I would recommend googling Gertsenshtein effect.

 

By the way mechanical noise doesn't follow spin 2 polarization statistics You can confirm that by performing a Fourier series expansion on various types of vibration. A straight line is one term, a sinusoidal two terms.

 

Though quite frankly you will want a multi-degrees of freedom Langrange's equations can easily derive the equations of motion

 

 

I just detailed key differences between GW and EM...

 

A key note is a spin 2 quadrupole wave has no dipole moment.

 

PS I have little doubt that LIGO examined all the natural resonance frequencies for their locality. After all they probably hired a slew of vibrational analysis studies. I'm fairly confident they know how to critically dampen the PID instructions. Yes PI is better in most cases. I tend to use just PI myself, however there have been sytems where I needed PID.

 

I am aware the standard formulas used to calculate PID tend to lead to a quarterly amplitude decay. As such I further critically dampen depending on the system specifications.

 

You should note sound waves is longitudinal. There is three types of mechanical vibration. Transverse, longitudinal (compression) and torsional.

That was what this post describes...

 

On point of detail using H× and H+ polarization (quadrupole) we can also extrapolate the angle the signal originates from...

 

WE CAN TELL it originates from space... as opposed to somewhere in the Earths atmosphere....

 

GW wave polarization is easy to distinquish from an electromagnetic wave.

 

POLARIZATION....

These attempts to describe the signal as the same as electromagnetic is 100 percent wrong. Doesn't make any difference what you set your filtering at or what internal errors are involved.

 

The two types of polarizations are DIFFERENT.

 

GW polarization visual aid. Take a soft ball squeeze the sides along the x axis. The Y axis expands at precisely the same time. The L shape of the LIGO arms are designed to detect this motion. That motion DOES NOT occur on dipole polarization. Only the x axis changes regardless of signal strength.

 

That is a fundamental detail on the strain being measured. The physical design regardless of filtering can distinquish a GW quadrupole wave from a dipole wave.

 

this link has some decent animations showing the squeezing ball analogy I used. The other detail is the polarization is 45 degrees but in an electromagnetic wave it is 90 degrees. Again this deals with the different spin statistics.

Edited September 13, 2016 by Mordred

[worlov]

POLARIZATION....

These attempts to describe the signal as the same as electromagnetic is 100 percent wrong. Doesn't make any difference what you set your filtering at or what internal errors are involved.

 

 

It's not about the electromagnetic influence. The "power line harmonic" coming from the electrical outlet. With 60Hz, 120Hz, 180Hz etc. pulsating supply voltage entire electronics. To suppress these pulsations entirely, should LIGO switch to battery supply.

[Mordred]

How would two detectors 300 km apart on two different power supply networks recieve the same power harmonic disruption? at the same scale of disruption on both systems.

 

I honestly cannot see that, however it isn't a bad idea to consider. Harmonic distortions can readily make finding the data rather tricky. I believe LIGO released several papers dealing with the problems of harmonics.

 

here is one out of several dozen...

 

https://www.google.ca/url?sa=t&source=web&rct=j&url=http://www.phys.ufl.edu/LIGO/stochastic/S1caliblines.pdf&ved=0ahUKEwjjl9a354zPAhUC82MKHf46ASIQFggsMAA&usg=AFQjCNHod0VYe9_jRO_13jBCIJttYDRcIg&sig2=gRaRVOtTz4Ttq5ZCRtgQVA

 

seems to be 1 out of a series of crosschecks. S1 to S5 at first quick glance. From what I've read they use a calibration laser and inject simulated signals to hunt for improper filtering.

 

Using known signals is certainly practical... I do a similar technique when calibrating equipment.

 

This seems to be routinely done. A battery power system wouldn't resolve internal harmonic issues. If only harmonics were that easily corrected.

From a quick initial research they looked at every major component.

 

Here is LIGOs catalog calibration list. Though I'm sure this isn't the only listings.

 

https://losc.ligo.org/speclines/

 

As this contains the S5 instrument lines.

 

For power grid harmonics, some corrective techniques is to use the right transformers. For example Hammond has a series of harmonic mitigating transformers.

 

As an aside. One plant I worked at had the wrong transformer type. The harmonics were easily detectable even with a Digital multimeter by comparing the true neutral line to the neutral conductor. You could tell harmonics was present. Not the frequency with a DMM

Edited September 13, 2016 by Mordred