ArticlePDF Available

Abstract and Figures

Experiments were conducted to study light propagation in a light waveguide loop consisting of linearly and circularly moving segments. We found that any segment of the loop contributes to the total phase difference between two counterpropagating light beams in the loop. The contribution is proportional to a product of the moving velocity v and the projection of the segment length Delta(l) on the moving direction, Deltaphi=4piv x Delta(l)/c(lambda). It is independent of the type of motion and the refractive index of waveguides. The finding includes the Sagnac effect of rotation as a special case and suggests a new fiber optic sensor for measuring linear motion with nanoscale sensitivity.
Content may be subject to copyright.
Generalized Sagnac Effect
Ruyong Wang, Yi Zheng, and Aiping Yao
St. Cloud State University, St. Cloud, Minnesota 56301, USA
(Received 18 March 2004; published 27 September 2004)
Experiments were conducted to study light propagation in a light waveguide loop consisting of
linearly and circularly moving segments. We found that any segment of the loop contributes to the total
phase difference between two counterpropagating light beams in the loop. The contribution is
proportional to a product of the moving velocity v and the projection of the segment length l on
the moving direction, 4v l=c. It is independent of the type of motion and the refractive
index of waveguides. The finding includes the Sagnac effect of rotation as a special case and suggests a
new fiber optic sensor for measuring linear motion with nanoscale sensitivity.
PACS numbers: 42.25.Bs, 03.30.+p, 42.81.Pa
The Sagnac effect [1] shows that two counterpropagat-
ing light beams take different time intervals to travel a
closed path on a rotating disk, while the light source and
detector are rotating with the disk. When the disk rotates
clockwise, the beam propagating clockwise takes a longer
time interval than the beam propagating counterclock-
wise, while both beams travel the same light path in
opposite directions. The travel-time difference between
them is t 4A=c
, where A is the area enclosed by the
path and is the angular velocity of the rotation. Since
the 1970s, the Sagnac effect has found its crucial appli-
cations in navigation as the fundamental design principle
of fiber optic gyroscopes (FOGs) [2,3]. In a FOG, when a
single mode fiber is wound onto a circular coil with N
turns, the Sagnac effect is enhanced by N times so that
the travel-time difference is 2vNl=c
, where v is the
velocity of the rotating fiber, l is the circumference of
the circle, and Nl is the total length of the fiber. The
travel-time difference in a FOG can be expressed by the
phase difference 2tc=, where is the free
space wavelength of light. Today, FOGs have become
highly sensitive detectors measuring rotational motion
in navigation [4,5].
It is believed that the Sagnac effect exists only in
circular motion. However, we have discovered that any
moving path contributes to the total phase difference
between two counterpropagating light beams in the
loop. In a previous experiment using a fiber optic conveyor
(FOC), we showed our preliminary result that a segment
of linearly moving glass fiber contributes
4vL=c to the phase difference [6]. Here, we generalize
our finding for the Sagnac effect with a more complete
study and a series of new experiments. Our experiments
include different types of motion and light paths with a
glass fiber and an air-core fiber [7]. In this study, we found
that in a light waveguide loop consisting of linearly and
circularly moving segments, any segment of the loop
contributes to the phase difference between two counter-
propagating light beams in the loop. The contribution is
proportional to a product of the moving velocity v and the
projection of the segment length l on the moving direc-
tion, 4v l=c. It is independent of the type of
motion and the refractive index of waveguides. The total
phase difference of the loop is 4
v dl=c.
This general conclusion includes the Sagnac effect for
rotation as a special case. The finding also suggests a new
fiber optic linear motion sensor having nanoscale sensi-
tivity, which is much more sensitive than any existing
linear motion detectors. This linear motion sensor may be
applied to accelerometers in navigation and seismology.
Our experiments are different from the Fizeau-type
experiment [8],‘drag’’ experiment of a moving medium.
In the Fizeau-type experiment, there is relative motion
between the light path and the medium, water or glass.
The result of the Fizeau-type experiment is dependent on
the refractive index of the medium, and the drag coeffi-
cient is zero when the refractive index is 1. In our experi-
ments, as in the FOG, there is no relative motion between
the light path and the medium, glass fiber or air-core fiber.
The experimental system in our study consists of a fiber
optic loop, a FOG [9], and a mechanical conveyor, as
shown in Fig. 1(a). A single mode fiber loop with different
configurations was added to the FOG, which was cali-
brated for the added length. The loop includes significant
portions of fiber segments that move linearly. The func-
tion of the FOG in this experiment is to transmit and
receive the counterpropagating light beams and to detect
the phase differences between two light beams. Because
the FOG is not rotating in the experiment, the detected
phase differences are caused by the movement of the
added fiber optic loop.
An air-core fiber was used to verify that the phase
difference is independent of the refractive index and
various types of motion. The new photonic band gap fiber
[7] has a hollow air-guiding core for light with a wave-
length of 1310 nm. The fiber was used to construct a two-
wheel FOC shown in Fig. 1(a) and a three-wheel FOC, not
shown. Shown in Fig. 1(b) are the detected phase differ-
ences caused by the fiber motion with both configurations.
The length of the fiber loop is 4.1 m and the speeds of the
motion are from 0.001 91 to 0:211 m=s. Each phase dif-
ference is an average of eight measurements. Thirteen
143901-1 2004 The American Physical Society 143901-1
Physical Review Letters 93 (2004) 143901
phase differences range from 0.000 246 to 0.0276 rad. The
maximum conveyor velocity error is 1.7%. The least-
squares linear regression of the phase differences is
0:0323vL, which closely matches with
4vL=c 0:0320vL with 1310 nm. It also
matches the result using a glass fiber [6]. Therefore, it is
confirmed that the contribution to the phase difference in
a moving light waveguide is independent of the refractive
index of the waveguide and independent of the loop
To study the relationship between the motion of the
fiber and the fiber orientation, we conducted an experi-
ment in which the fiber zigzags and has an angle with
respect to the direction of fiber motion. Thus, for a fiber
segment having an actual length of l, its effective length
is l cos, which is a projection of the fiber onto the
motion direction. As shown in Fig. 2, our experiment
demonstrates that the effective length contributes the
phase difference, not the actual length; therefore, the
phase difference is not 4vl=c,but
4vl cos=c 4v l=c, which is the dot product
between two vectors.
A further experiment was conducted to study the phase
difference when different segments of the loop moved at
different speeds. Figure 3(a) shows a fiber optic ‘paral-
lelogram where the top arm moves with the conveyor
and the bottom arm is stationary. While moving, the two
sidearms, being flexible, are kept the same shape so that
the phase differences in these two sidearms cancel each
other. There is no phase difference in the bottom sta-
tionary arm. Therefore, the detected phase difference is
contributed solely by the motion of the top arm. The
experiment was conducted using a glass fiber and an
air-core fiber. The detected phase differences are shown
in Fig. 3(b). For the glass fiber configuration, the length of
the top arm is 1.455 m and there are 11 turns. The
measured phase differences range from 0.000 482 to
0.121 rad for speeds from 0.000 917 to 0:233 m=s.In
another configuration, the top arm is air-core fiber with
a length l of 5.23 m. The phase differences range from
0.000 301 to 0.0346 rad for speeds from 0.00186 to
0:207 m=s. The least-squares linear regression of the
phase differences is 0:0317vl, which agrees with
4vl=c 0:0320vl with 1310 nm.
According to our experiments, we can draw a conclu-
sion about the generalized Sagnac effect that in a moving
fiber loop or waveguide, a segment l with a velocity v
contributes 4v l=c to the total phase differ-
ence between two counterpropagating beams in the loop.
The contribution is independent of the refractive
index of the waveguide, and the motion of the segment
FIG. 2 Phase differences of moving fiber segments
that have the same effective length and different actual
lengths. The effective length is 7.20 m and four actual lengths
are 8.31, 10.2, 14.4, and 27.8 m for 30
, 45
, 60
respectively. The velocity is 0:091 m=s. Four phase differences
range from 0.0206 to 0.0215 rad, with standard deviations from
0.000 280 to 0.000 743 rad. In this experiment, 4vl cos=
c 0:032vl cos 0:0211 rad, which is represented by the
dashed line.
FIG. 1 Experiment for detecting the phase
difference of two counterpropagating light beams in a moving
fiber loop. (a) Experimental setup. The fiber loop is driven by
the conveyor at a velocity v. The conveyor has a length of 1.5 m
and can move from 0.001 to 0:25 m=s. The diameters of the
wheels are 0.3 m. The FOG consists of a 1310-nm superlumi-
nescent light-emitting diode as the light source and a phase
difference detector that has an output rate of 1162.6 mV per rad.
(b) Phase differences of two counterpropagating beams in a
moving air-core fiber.
143901-2 143901-2
Physical Review Letters 93 (2004) 143901
can be either linear or circular. Thus, for the entire loop,
the total phase difference between two counterpropagat-
ing beams in the loop is
v dl=c:
This general conclusion includes the Sagnac effect of
rotation as a special case. In fact, the Sagnac result can
be derived from this general equation by assuming a
circular motion of the loop and using Stokess theorem,
v dl=c 4
r vdA=c
2k dA=c 8A=c:
This generalization provides a design principle for a
new fiber optic linear motion sensor (FOLMS), which has
a high sensitivity and a high stability. The basic structure
of this sensor can be similar to that shown in Fig. 3(a).
The linear motion of the top arm of the sensor is detected
with a phase difference 4vNl=c. Because two
beams share the same optical path, the sensor is optically
stable. Just as a FOG detects the rotational motion of an
object, a FOLMS can detect the relative linear motion
between two objects fixed on the top and bottom arms of
the parallelogram. The optical technologies developed in
recent decades for the FOG can be utilized for the
FOLMS; therefore, the sensitivities and stability of the
FOLMS are comparable to that of the FOG. The sensi-
tivity of a FOG can be 10
rad of the phase difference
[5]. With Nl 500 m and 10
m, a FOLMS can
detect a linear velocity of
v c=4Nl 4:8nm=s
which is a nanoscale velocity.
This nanoscale sensitivity linear motion sensor can
detect the very small relative motion appearing in an
accelerometer, which is important in navigation and seis-
mology. The common design of an accelerometer is a
spring-mass system. Improving the sensitivity of an ac-
celerometer requires improving the sensitivity of detect-
ing the linear movement of the mass relative to the base.
Utilizing a FOLMS to detect the relative motion between
the mass and the base will greatly increase the sensitivity
of the accelerometer. It can be foreseen that an acceler-
ometer using the FOLMS combined with the fiber optic
gyroscope may be beneficial for navigation because both
use the same technology and both are very stable
We thank Dean Langley for useful discussions and
help. We thank Robert Moeller of the Naval Research
Laboratory for technical assistance and NRL for the
loan of the FOG. The air-core fiber was purchased from
Crystal Fiber, Denmark.
[1] G. Sagnac, C. R. Acad. Sci. Paris 157, 708 (1913).
[2] V. Vali and R. Shorthill, Appl. Opt. 16, 290 (1977).
[3] W. Leeb, G. Schiffner, and E. Scheiterer, Appl. Opt. 18,
1293 (1979).
[4] H. Lefevre, The Fiber-Optic Gyroscope (Artech House,
Boston, 1993).
[5] W. K. Burns, Optical Fiber Rotation Sensing (Academic
Press, Boston, 1994).
[6] R. Wang, Y. Zheng, A. Yao, and D. Langley, Phys. Lett. A
312, 7 (2003).
[7] R. F. Cregan et al., Science 228, 1537 (1999).
[8] H. L. Fizeau, C. R. Acad. Sci. Paris 33, 349 (1851).
[9] R. P. Moeller et al.,inFiber Optical and Laser Sensors
XI, edited by R. P. DePaula, SPIE Proceedings Vol. 2070
(SPIE–The International Society for Optical
Engineering, Boston, MA, 1993), p. 255.
FIG. 3 Experiment for studying the phase
difference when different segments of the loop move at differ-
ent speeds. (a) Experimental setup. The light from a source is
split into two beams that counterpropagate in the fiber which is
wound onto a parallelogram. The bottom arm is fixed while the
top arm is moving. The phase difference can be enhanced by
multiple turns of the fiber on the parallelogram. The coupler,
source, and detector are replaced by the FOG in the experiment.
(b) Phase differences caused by the linear motion of the top
arm of the fiber optic loop.
143901-3 143901-3
Physical Review Letters 93 (2004) 143901
... It results from the mathematical infeasibility. In addition, the constancy of the speed of light disagrees with the experimental result of the generalized Sagnac effect [16,17], which clearly indicates the anisotropy of the speed of light not only in rotating frames but also in inertial frames [11,18]. On the contrary, the PFT with a unique isotropic frame is consistent with the experimental results, including the generalized Sagnac effect, for the test of STR and it has no contradictions and no paradoxes. ...
... Though a variety of explanations and analyses on the Sagnac effect are available based on these theories [19,20], even the problem of time gap that multiple times are defined at the same place in the rotating frame has not been resolved. The generalized Sagnac effect [11,[16][17][18]21] that shows the anisotropy of the speed of light in inertial frames as well may be a more perplexing conundrum to the theory of relativity based on its isotropy. We demonstrate, as an example to show the usefulness of LT, that the difference between the travel times of counter-propagating light beams in the experiment of the generalized Sagnac effect can be exactly obtained by exploiting it. ...
... In fact, the correction is needed on account of the anisotropy of the light speed. The Sagnac effect has been observed in inertial frames as well as in rotating frames [16,17]. The generalized Sagnac effect that involves both linear and circular motions can be analyzed based on TCL [18]. ...
According to the postulates of the special theory of relativity (STR), physical quantities such as proper times and Doppler shifts can be obtained from any inertial frame by regarding it as isotropic. Nonetheless many inconsistencies arise from the postulates, as shown in this paper. However, there are numerous experimental results that agree with the predictions of STR. It is explained why they are accurate despite the inconsistencies. The Lorentz transformation (LT), unless subject to the postulates of STR, may be a useful method to approach physics problems. As an example to show the usefulness of LT, the problem of the generalized Sagnac effect is solved by utilizing it.
... Paul Langevin's explanation of the same effect in the sense of special relativity as reviewed by Gianni Pascoli [3] has its opponents including Wolfgang Engelhardt [4]. In 2003, Ruyong Wang and colleagues proved a linear version of this effect [5,6]. ...
... Axiom 1 and 2 foretell the results of Ryong Wang's [5,6], Torr and Kolen's [7] and De Witte's [8] experiments. Let us investigate the essence of these two types of experiments: ...
... They added proof by showing that it is the effective length of the fiber into the direction of motion, rather than the actual length of the fiber, which calculates the size of the effect. Mathematically, the length "l" in equation (1) therefore needs replacing by its projection into the direction of motion at angel to yield *The image is a reproduction from the openly accessible pre-print version of Wang et al. 2004 As importantly, the team devised an experiment depicted in figure 3, to prove the above stated claim, that a linear segment in isolation experiences the measured effect. In fact, in the experiment depicted in figure 3, only linear motion accounted for the measured effect which still obeyed the same formula. ...
Full-text available
The present manuscript aims to lay the groundwork for a research proposal to open the gateway for novel signal precision technology.
... However, in your reply to my critique, you did not say a word on the experimental evidence I cited, which contradict SR and the Lorentz transformations (e.g., Wang et al., 2003, Wang, Zheng, & Yao, 2004Wang, 2005, Suleiman, 2014, 2018a. As examples, the experimental results reported by Wang and his colleagues on the Sagnac effect in rectilinear and in circular motion, are clear-cut falsifications of SR. ...
... Since you agree that theories should bow to empirical evidence, then you should either show flaws in the above-cited experiments, or accept their results. Thus, the best way to put an end to your tormenting efforts to convince us to leave SR alone, is to read one of the papers (e.g., Wang, Zheng, & Yao, 2004), and show that it is flowed in design, measurement procedures, or data analysis; or else, accept its result as solid falsifications of SR. ...
Full-text available
A reply to Eric Lord's comments on my critique of his paper titled "Notes on the Meaning of Einstein's Special Theory of Relativity". I focus on the crucial issue of theory testing, which was strangely overlooked in Lord's comments on my critique. I also underscore the importance of considering competitive theories to Einstein's relativity, and demonstrate that Information Relativity Theory (IR) is superior to both SR and GR.
... Now go to my published paper[12] and read the derivations. Actually, IRT is the only theory on the table, which I am aware of, that makes no presumptions and has no free parameters.9. You ended your awkward, full of disinformation "review" by justifying your scientific committee's bad review. ...
Full-text available
A reply to a review by Dr. Peter Hess of my paper titled: "A Measurement-Independent and Parameter-Free Alternative to ΛCDM and General Relativity Predicts the Cosmological Parameters of the Universe and Proposes a (Peaceful) Resolution of the Hubble Constant Tension". The review by Dr. Hess is a post-hoc justification after IWARA 2020 Video Conference (which he co-chairs with Dr. Cesar Zen) rejected the paper. The original review was the following: "We have analyzed the submitted contribution of Ramzi Suleiman. In a unanimous and final decision of the Committee, we concluded that its content does not contain elements worthy of acceptance at the IWARA 2020 event".
... But such experiments have already been performed in rectilinear and circular motion by Wang and his colleagues (Wang et al., 2003, Wang, Zheng, & Yao, 2004Wang, 2005), and we have shown here that the two types of motion are completely equivalent. ...
... Propagation of light in non-inertial frames provides possibilities of testing general relativity referred to non-inertial reference frames in laboratory. The Sagnac effect states that in rotating reference frame, counter-propagating rays, which propagate around a closed path, would take different time intervals [7,8]. It can be described by Born metric known as a relativistic effect [9][10][11]. ...
We construct an alternative uniformly accelerated reference frame based on 3+1 formalism in adapted coordinate. It is distinguished with Rindler coordinate that there is time-dependent redshift drift between co-moving observers. The experimentally falsifiable distinguishment might promote our understanding of non-inertial frame in laboratory.
Full-text available
The existence of the electromagnetic aether is argued from two standpoints. Conceptual, based on the nature of physical waves. And practical: the various experiments that demonstrate it. Possible reasons for the strange nullification of the positive 1887 Michelson-Morley aether-wind result are discussed.
The role of preferred frames for light propagation and time dilation in the region of a massive, spherical, gravitating bodies, where according to general relativity, space–time curvature is described by the Schwarzschild metric equation, is discussed in the context of the Sagnac effect (for light propagation) and the Hafele–Keating experiment (for time dilation). Predictions for both translational and rotational motion relative to the preferred frame are calculated up to order [Formula: see text]. Different published theoretical calculations of the Sagnac effect are critically reviewed. The conflation in the literature of measured time differences in Sagnac experiments (a classical order [Formula: see text] effect) and time dilation (a relativistic order [Formula: see text] effect) are also discussed.
A world system is composed of the world lines of the rest observers in the system. We present a relativistic coordinate transformation, termed the transformation under constant light speed with the same angle (TCL-SA), between a rotating world system and the isotropic system. In TCL-SA, the constancy of the two-way speed of light holds, and the angles of rotation before and after the transformation are the same. Additionally a transformation for inertial world systems is derived from it through the limit operation of circular motion to linear motion. The generalized Sagnac effect involves linear motion, as well as circular motion. We deal with the generalized effect via TCL-SA and via the framework of Mansouri and Sexl (MS), analyzing the speeds of light. Their analysis results correspond to each other and are in agreement with the experimental results. Within the framework of special and general relativity (SGR), traditionally the Sagnac effect has been dealt with by using the Galilean transformation (GT) in cylindrical coordinates together with the invariant line element. Applying the same traditional methods to an inertial frame in place of the rotating one, we show that the speed of light with respect to proper time is anisotropic in the inertial frame, even if the Lorentz transformation, instead of GT, is employed. The local speeds of light obtained via the traditional methods within SGR correspond to those derived from TCL-SA and from the MS framework.
  • R F Cregan
R. F. Cregan et al., Science 228, 1537 (1999).
  • G Sagnac
G. Sagnac, C. R. Acad. Sci. Paris 157, 708 (1913).
  • H L Fizeau
H. L. Fizeau, C. R. Acad. Sci. Paris 33, 349 (1851).
  • R Wang
  • Y Zheng
  • A Yao
  • D Langley
R. Wang, Y. Zheng, A. Yao, and D. Langley, Phys. Lett. A 312, 7 (2003).
Full-text available
Frequency distributions of parasitic helminths within human communities are invariably highly aggregated, the majority of worms occurring in relatively small fractions of the host populations. It has been suggested that the heavily infected individuals are predisposed to this state, not by chance, but by as yet undefined genetic, ecological, behavioral, or social factors. Analyses of individual post-treatment patterns of hookworm reinfection among 112 villagers in an endemic area of West Bengal provide quantitative evidence of predisposition to heavy infection. This observation has implications for the design of control programs based on chemotherapy because of the potential economic advantage of selective or targeted treatment as opposed to mass or blanket treatment.
Inertial guidance and navigation systems are known to be very useful for aircrafts, missiles, land vehicles, robots… For the moment, they make use of mechanical gyroscopes or gas laser gyroscopes. Now the development of low-loss optical fibers and compact and reliable semi-conductor light sources for the telecommunication industry has made possible a new approach: the fiber-optic gyroscope [l, 2, 3]. This device is only composed of low-mass solid-state components-and does not require any mechanical movement (this is still the case with laser gyros which need dithering to solve their lock-in problem about zero rotation).
The fundamental operating principles and technological implementation of single-pass interferometric fiber-optic gyroscopes (FOGs) are reviewed. Topics discussed include the Sagnac effect, sensitivity and single-mode reciprocity, biasing modulation-demodulation schemes, noise and drift, multipath and transient parasitic effects, and truly nonreciprocal effects (the magnetooptic Faraday and nonlinear Kerr effects). Consideration is given to the design of source and sensing fiber coils, the FOG interferometer core (two beamsplitters, a polarizer, a spatial single-mode filter, and modulators), detectors, FOG open-loop sensitivity, and scale-factor accuracy. It is shown that interferometric FOGs are significantly more practical for guidance and navigation applications than FOGs based on a passive ring resonator.
Optical fiber gyroscopes consist of a multiturn coil of a single-mode fiber in an interferometer arrangement. The fringe shift (scale factor) depends on the effective index of refraction and on the dispersion of the fiber. The paper is intended to show that the scale factor is independent of the fiber properties and that the fringe shift obeys the Sagnac formula. The Sagnac effect is well documented and understood, and it has some relation to the Fizeau effect and the Fresnel drag coefficient. Two versions of the fiber ring-interferometer are considered. The first consists of a light source, a beam splitter and a fiber coil, and all elements are firmly mounted on a common frame which can be rotated. The second consists of an interferometer as above, but only the fiber coil is rotated; the connection with the beam splitter is established by the flexible fiber ends.
The feasibility of constructing a sensitive optical gyroscope was tested by using a 950-m single mode optical fiber ring interferometer. Observed fringes were obtained at a wavelength of 6328 A, where several fiber modes were present. The central fringes were not at intensity maximum and minimum positions, indicating that there is a slight (less than lambda/2) optical path difference between the oppositely traveling beams. This occurs when the beams are focused on the fiber ends at an angle with respect to the fiber axes, or when the fiber end is not cut perpendicular to its axis.
The confinement of light within a hollow core (a large air hole) in a silica-air photonic crystal fiber is demonstrated. Only certain wavelength bands are confined and guided down the fiber, each band corresponding to the presence of a full two-dimensional band gap in the photonic crystal cladding. Single-mode vacuum waveguides have a multitude of potential applications from ultrahigh-power transmission to the guiding of cold atoms.