PHASE INSTABILITY AND NONLINEARITY IN THE DISTORTED DIPOLE PULSATION MODE OF THE RAPIDLY OSCILLATING AP STAR HR-3831 (HD-83368)

HR 3831 is a rapidly oscillating A7p SrCrEu star with T(eff) = 8000 K, R = 1.9 R., M(V) = 2.0, V = 6.168, d almost-equal-to 60 pc, a magnetic field which ranges from about - 720 G less-than-or-equal-to B(eff) less-than-or-equal-to + 780 G with the rotation period of 2.851982 d, and a pulsation period near 700 s. In this paper, new high-speed photometric observations obtained during 101 h in 1990 and 167 h in 1991 are presented. The 1991 observations were obtained contemporaneously from the South African Astronomical Observatory and Cerro Tololo Inter-American Observatory over a time-span of 17 d with a 41 per cent duty cycle. A frequency analysis of the 1991 data shows that HR 3831 pulsates in a single mode which is rotationally modulated, giving rise to a fundamental frequency septuplet, a first harmonic quintuplet, a second harmonic triplet, and a probable detection of a single frequency at the third harmonic. The fundamental frequency septuplet is shown to be adequately described by a sum of axisymmetric spherical harmonics with l = 0, 1, 2 and 3. By far, the dominant component of the sum is the dipole, which is why we call the mode a distorted dipole. By comparing the new 1990 and 1991 data with previous data sets obtained in 1981, 1985 and 1986, it is shown that the form of this distorted dipole probably changes slightly from year to year. It is also shown that the phase of the pulsation is variable from year to year, and that this variability cannot be modelled simply by a periodic Doppler shift, caused by an orbital companion, plus a quadratic evolutionary term. The first harmonic quintuplet has changed substantially in form between the 1986 and 1990 data sets. HR 3831 shows a pi-radian reversal of its pulsation phase at the two magnetic quadratures. This reversal is not well defined for one quadrature; whether this is because of errors in the observations, or is a physical effect in the star, is not yet known. The relative amplitudes of the second harmonic frequencies remain problematic.