No abstract

Vibration testing of packaging is a critical part of the distribution packaging analysis process. The accuracy of simulated vibration is important for packaging optimization. Because of this, several researchers have developed improved simulation methods to produce more realistic vibration tests. Correlation studies are required to verify these methods, ideally using actual packaged products in transit. Unfortunately, cost, time and complexity issues make carrying out studies with actual product difficult.This article uses a specially designed and proven test rig, which simulates the damage mechanism of scuffing, to carry out a correlation study. The study compares the level of damage produced when performing simulations using a range of improved techniques in comparison with the time-history reproduction of a journey (used as a benchmark) and the established method using the average power density spectrum to create a Gaussian simulation signal.The level of scuffing damage produced varied between the different simulation methods, with the modulated root mean square (RMS) technique and the accelerated power density spectrum (with a time compression of 5 and a k equal to 2) best reproducing the level of damage observed from the benchmark time replication test.

Fourteen years ago Hiittner and Tammannl mentioned that it was possible, from the behavior of the temperature during crystallization, to draw a conclusion as t o the amount of impurity. Other writers have since discussed the melting or freezing curve more a t length, but little practical use has yet been made of Hiittner and Tammann's suggestion. The present paper considers under what conditions this method is most likely to be useful, and also discusses the technic. Theoretical BasisThe underlying laws are familiar but may deserve to be summarized. Impurities may form solid solutions, or only eutectics, and the crystallization phenomena have significant differences in the two cases.Where mix crystals only are formed, the temperatures and concentrations change as in Figs. I and 2. (More complicated freezing curves than these are known, but if the amount of impurity is small, so that the extreme right or left end only of the phase rule diagram is used, that end will rarely differ essentially from the end of I or 2 .) As the liquid mixture cools for either figure its concentration and temperature follow the line UV. At temperature V mix crystals of composition W begin to form. The relation of V to W depends on the form of the double pointed figure CD, which is a characteristic of the combination of substances in question. Since the composition W is different from the original melt, the separation of crystals W changes the composition of the liquid, which moves toward composition X as solidification proceeds. A t the same time, if complete equilibrium persists, the composition of the crystals, including those already deposited, changes along WY.K. Hiittner and G. Tammann: "Ueber die Schmelzpunkte und Unwandlungspunkte einiger Salze," 2. anorg. Chem., 43, 218 (1905).

1 As in Fig. 5 (in which, however, the comparison was with a fixed portion of the same wire). This curve corresponds to the integral of any one of the curves given by Hall Campbell and Serviss (loc. cit.). EvaminaTion No. 6.] THE CONSTANCY OF THERMOELEMENTS.