RADIO PULSARS ASP Conference Series, Vol.

**VOLUME***, 2003 Bailes, Nice 9 Thorsett

Improved Bounds on Violation of the Strong Equivalence Principle

1.

Introduction

PSR B0655+64, in a highly circular one-day orbit with a 0.8 M a white-dwarf companion, serves as a control experiment for measurements of orbital decay in the highly relativistic double-neutron-star binaries: General Relativity (GR) has predicted equally well the strong back-reaction to gravitational radiation for PSRs B1913+16 (Taylor 1993; see also Weisberg & Taylor, this volume) and B1534+12 (Stairs et al. 1998), and the absence of detectable orbital evolution for PSR B0655+64 over two decades. The long-term stability of the B0655+64 orbit sets unique bounds on departures from GR that give rise to dipolar gravitational radiation (Arzoumanian 1995, Goldman 1992), the existence of which would represent a violation of the Strong Equivalence Principle (SEP), one of the basic tenets of GR. G6rard & Wiaux (2002) examine the theoretical basis for dipolar gravitational radiation and suggest that bounds from binary pulsars may be competitive with future satellite experiments dedicated to probing SEP violation. A definitive analysis of the available data for PSR B0655+64, and implications for a variety of alternative theories of gravitation, will be published separately (Arzoumanian et al. 2003, in preparation). Following recent observations to extend our long-term monitoring program, I present here preliminary results on the orbital evolution of PSR B0655+64.
1.1.

Binary Pulsars and the Strong Equivalence Principle

The Strong Equivalence Principle posits that the response of a body to an external gravitational field is independent of the gravitational self-energy of the body; it would be violated if, for example, two objects with different gravitational binding energies were observed to fall at different rates. SEP is satisfied by postulate within GR; as a consequence, the lowest allowed multipole order of gravitational radiation is the electric quadrupole. If SEP does not hold,
1

Arzoumanian

however, emission of dipolar gravitational radiation is allowed, and lends itself in principle to indirect detection through decay of a binary orbit at a rate inconsistent with the quadrupole-order prediction of GR. Because self-gravitational effects are important in neutron stars, SEP can be tested with pulsar systems. The most thoroughly studied theory of gravitation that violates SEP, the so-called Brans-Dicke scalar-tensor theory (Brans & Dicke 1961), invokes the existence of a scalar gravitational field that couples to matter through a single dimensionless parameter, WBD. The Brans-Dicke theory becomes indistinguishable from GR as WBD becomes arbitrarily large; solar-system experiments currently place a lower bound WBD 2 500. Other alternatives to GR are also constrained by limits on dipolar gravitational radiation (see, e.g., Damour & Esposito-Farkse 1992): recent observations of distant type Ia supernovae suggesting that the expansion of the universe is accelerating have prompted renewed interest in theories involving scalar fields (e.g., quintessence, Galdwell et al. 1998) or long-range forces that may be responsible for the acceleration. Such fields would radiate predominantly at dipole order. The various contributions to observed changes in the orbital period pb of a binary system (neglecting tidal and mass-transfer effects) can be expressed as

);(

obs

(;)Q+

(;)D+

@ ) G +

(a)

where the subscripts Q, D, G, and a denote the effects of quadrupolar and dipolar gravitational radiation, a change in the gravitational constant with time, and the varying Doppler shift from a relative acceleration of the solar system and pulsar binary. In a large class of metric theories of gravitation, the radiation terms for a circular orbit are given by (Eardley 1975, Will 1981, Goldman 1992)

where n G 2.rr/Pb, ml and m2 are the pulsar and companion masses in solar units, M G ml +m2, and To is the mass of the Sun in units of time. The dimensionless parameters t q , K D , and G describe the strength of quadrupolar and dipolar radiation and the effective gravitational constant in a given theory. In GR, ~1 = 12, K D = 0, and G = 1; in the Brans-Dicke theory, K D G - ~ = 2(2 W B D ) - ~ . At the level of the current constraint on the dipole term D ,the G contribution to i)b in Eq. 1 can be neglected (Arzoumanian 1995, Kaspi et al. 1994). The quantities s 1 and s 2 in Eq. 3 represent the sensitivities of the orbiting objects, the fractional change in binding energy of each star with G,

s=-(-)

alnm alnG N

(4)

To correct for Galactic accelerations and the Shklovskii effect, I follow Damour & Taylor u = 6.8 f 1.1 masyr-. (1991), using a new proper-motion measurement for B0655+64: ,

~~

~~

_____

~~~~

~~~~

PSR B0655+64 and Dipolar Gravitational Radiation

where N is the total number of baryons. The sensitivity of white dwarfs is negligible; neutron stars are thought to have sensitivities in the range 0.15 < s < 0.40 (Will & Zaglauer 1989), with larger values corresponding to softer equations of state. Results of X-ray burst oscillation modeling ( e g 7 Nath et al. 2002), and the recent detection of atmospheric absorption lines from a neutron star (Cottam et al. 2002), suggest that fairly stiff equations of state are appropriate. I therefore adopt a fiducial value s = 0.2 below. It is worth noting that the difference (SI - ~ 2 in) Eq. ~ 3 strongly suppresses dipolar radiation from NS-NS binaries like B1913+16, so that the celebrated agreement of the latters orbital decay rate with the GR prediction does not usefully constrain the existence of gravitational radiation at dipole order. In scalar-tensor theories, gravitational binding energy takes on the role of an effective gravitational charge. Neutron star-white dwarf (NS-WD) binaries can therefore be expected to copiously emit dipolar radiation, if such radiation exists, because of the large difference in gravitational self-energy between the component stars.

1.2.

PSR B0655+64

PSR B0655+64 was discovered during a survey of the Northern sky made with the NRAO 300 Foot telescope (Damashek et al. 1982). Timing observations began soon thereafter, and the companion was optically identified as a massive white dwarf by Kulkarni (1986). If we assume pulsar and companion masses of 1.4 M a and 0.8 M a , Eq. 2 predicts a rate of orbital decay within GR, pFR= -2 x corresponding to a decay timescale for the orbit of TQ 150 Gyr.

2.

Observations and Results

The current data span 20 years, including observations made with the NRAO Green Bank 300 Foot (Taylor & Dewey 1988), 140 Foot (Backus et al. 1982, Arzoumanian et al. 1994b), and GBT telescopes, as well as the NRAO VLA (Thorsett 1991) and Jodrell Bank Love11 (Jones & Lyne 1988) telescopes. We carried out intensive observing campaigns, to obtain good coverage of all orbital phases at a single epoch, every few years with the 140 Foot and recently with the GBT, and these data provide the interesting constraints on orbital evolution. Pulse times-of-arrival (TOAs) were derived from observations following stansoftware package, using the ELL^ dard procedures and analyzed with the TEMPO binary timing formula (Lange et al. 2001). A least-squares fit of the entire dataset for rotational, astrometric, and orbital parameters produces residuals consistent with statist ical fluctuations. Orbital phase residual and period measurements over time are shown in Figure 1. The constancy of p b confirms the result of the global timing solution: the available data bound changes in orbital period at the level IPblobs < 1.5 x (la),a limit 7 times the GR prediction. Accounting for a small correction for relative acceleration (Eq. l), we then have a 2 0 upper limit on orbit a1 evolution,

< 1.0 x
obs

1 0 - l ~ y r - l (20).

(5)

Arzo umanian

\
Q

4. a-

...
I I

........

-5

'

'

'

"

'

"

"

'

'

'

1980

1985

1990
Date

1995

2000

Figure 1. Orbital evolution of PSR B0655+64. Box outlines depict ranges of dates over which phase and period measurements were made. The resulting constraints on WBD are summarized in Fig. 2. Horizontal lines depict the current la and 2a limits on orbital decay in the PSR B0655+64 system. The curving dashed and dotted lines are the expected orbital period changes due to emission of gravitational radiation through quadrupole order, Eqs. 2-3, for NS sensitivities from soft and stiff equations of state respectively, The three curves in each set represent the ( m l , m2) pairs (1.30, 0.7), (1.35, 0.8), and (1.40, 0.9), so that total system mass increases to the upper left. Constraints on WBD then lie at the intersections of the dashed or dotted curves with the measured upper limits on &. From Eqs. 2, 3, and 5, we have in general

and for the Brans-Dicke theory specifically, WBD

2 320 ( ~ / 0 . 2 )(20). ~

3 . Discussion
Prospects for improving bounds on SEP violation and the existence of dipolar gravitational radiation are good. Damour & Taylor (1992) show that, for constant data quality, measurement uncertainty for & scales with data span T as T-5/2; moreover, data quality typically improves with time through improved instrumentation. Also, pulsar surveys continue to discover NS-WD systems similar to B0655+64 but with shorter orbital periods and millisecond pulse periods. Their higher rates of gravitational energy release coupled with higher timing precision suggest that these systems will surpass B0655+64 as laboratories for

PSR B0655+64 and Dipolar Gravitational Radiation

3
2
h
L

- 0.6
w

a- 0.5

\ 0.4

*a- 0.3
0.2

5000

1000 500
%D

100

50

Figure 2. Bounds on glauer 1989).

WBD

from PSR B0655+64 (after Will & Za-

testing theories of gravitation-Table 1 lists the current sample of relativistic NS-WD systems. While B0655+64 remains a valuable object of further study, a potential caveat applies to relativistic studies of all close NS-WD binaries: very small orbital separations can give rise to non-relativistic interactions. Heating of the companion star by irradiation from the pulsar is thought to power tides and mass loss in a handful of low-mass binaries. Outflows and tides cause significant orbital torques (e.g., Arzoumanian et al. 1994a), which would overwhelm any small secular trend due to gravitational radiation. If indeed the closest NS-WD binaries (e.g., PSR B0751+18; see Nice et al., this volume) are clean systems, significant new constraints on SETP violation will emerge in the coming years.

Acknowledgments. I thank my collaborators, R. Dewey, A. G. Lyne, D. J. Nice, J. H. Taylor, and S. E. Thorsett, for their continued interest and participation in this project. The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.

References
Arzoumanian, Arzoumanian, Arzoumanian, 671 Backus, P. R.,

Z. 1995, PhD thesis, Princeton University Z., Fruchter, A. S., & Taylor, J. H. 1994a, ApJ, 426, L85 Z., Nice, D. J., Taylor, J. H., & Thorsett, S. E. 199413, ApJ, 422,
Taylor, J. H., & Damashek, M. 1982, ApJ, 255, L63

Arzoumanian

Table 1. Relativistic NS-WD systems. Spin and orbital periods are shown alongside spin-down luminosity (I?), orbital separation ( A ) and timescales for orbital evolution from quadrupolar ( T Q ) and dipolar ( T D ) radiation. The potential for heating and tidal interactions between the ? and decreasing A. component stars increases rapidly for increasing I PSR J0751+1807 51757-5322 B0655+64 J1012+5307 51435-6100 3.5 0.26 8.9 0.45 195.6 1.03 5.3 0.60 9.3 1.35 33.8 33.2 30.6 33.6 33.1 4.1 6.4 11.3 6.8 14.2 20 20 150 200 250 15 15 60 90 90

Brans, C. & Dicke, R. H. 1961, Phys.Rev., 124, 925 Caldwell, R. R., Dave, R., & Steinhardt, P. J. 1998, Ap&SS, 261, 303 Cottam, J., Paerels, F., & Mendez, M. 2002, Nature, 420, 51 Damashek, M., Backus, P. R., Taylor, J. H., & Burkhardt, R. K. 1982, ApJ, 253, L57 Damour, T. & Esposito-Farkse, G. 1992, Classical Quantum Gravity, 9, 2093 Damour, T. & Taylor, J. H. 1991, ApJ, 366, 501 -. 1992, Phys.Rev.D, 45, 1840 Eardley, D. M. 1975, ApJ, 196, L59 Ghrard, J.-M. & Wiaux, Y. 2002, Phys.Rev.D, 66, 24040 Goldman, I. 1992, ApJ, 390, 494 Jones, A. W. & Lyne, A. G. 1988, MNRAS, 232, 473 Kaspi, V. M., Taylor, J. H., & Ryba, M. 1994, ApJ, 428, 713 Kulkarni, S. R. 1986, ApJ, 306, L85 Lange, C., Camilo, F., Wex, N., Kramer, M., Backer, D. C., Lyne, A. G., & Doroshenko, 0. 2001, MNRAS, 326, 274 Nath, N. R., Strohmayer, T. E., & Swank, J. H. 2002, ApJ, 564, 353 Stairs, I. H., Arzoumanian, Z., Camilo, F., Lyne, A. G., Nice, D. J., Taylor, J. H., Thorsett, S. E., & Wolszczan, A. 1998, ApJ, 505, 352 Taylor, J. H. 1993, Classical Quantum Gravity, 10, S167 Taylor, J. H. & Dewey, R. J. 1988, ApJ, 332, 770 Thorsett, S. E. 1991, PhD thesis, Princeton University Will, C. M. 1981, Theory and Experiment in Gravitational Physics (Cambridge: Cambridge University Press) Will, C. M. & Zaglauer, H. W. 1989, ApJ, 346, 366