Disregarded sources of random errors in the nonlinear constants of quartz

Standard errors of the third-order nonlinear material constants of quartz obtained by the least-squares fit reflect only the random errors from a single source. The source is the experimental data set used to determine the nonlinear constants. Other possible sources of errors, namely the linear material constants and the crystal specimen orientation, are reported. To minimize their effect, the most important factor is the accuracy of the linear elastic constants. If these constants have standard errors of 1%, they produce nontrivial errors in the nonlinear constants, about half the size of the errors due to the experimental errors. It appears that the errors caused by the other linear constants and by the orientation angles are of substantially lesser importance. >