The purpose of the Cosmic Ray Energetics and Mass (CREAM) Experiment is to determine the composition of cosmic rays up to the 1015 eV "knee" in the cosmic ray spectrum. It is suspected that this knee in the cosmic ray spectrum can be explained by the theoretical maximum energy that a supernova can accelerate particles to through Fermi acceleration. This is accomplished with a timing-based charge detector and transition radiation detector at an altitude of at least 110,000 ft with the aid of a high-altitude balloon. After launching from McMurdo Station in Antarctica, the balloon will stay aloft for 60–100 days gathering data on charges and energies of the unimpeded cosmic rays that strike the detectors.
One of the advantages of this type of experiment is that it is possible to identify the original particle that would have caused the air shower detected by ground-based detectors. Maximum detectable energy level is determined by duration of the flight and size of the detector; a difficult barrier to get around for experiments of this type. An accurate measurement of the composition of cosmic rays is necessary in order to understand the origins of the cosmic rays found above the "knee" at 1015 eV. To date, the CREAM balloon experiments have accumulated a total of 161 days of exposure, longer than any other single balloon-borne experiment.
In order to answer these questions, it is of particular interest to investigate cosmic rays in the 1012 to 1015 eV region due to several theories predicting a change in elemental composition just below the knee. To determine the elemental spectrum of cosmic rays, CREAM uses a silicon charge detector, timing charge detector, and scintillating fiber hodoscopes to detect the charge of incident particles up to that of iron (Z = 26). Energies are measured with a transition radiation detector (TRD), along with an ionization calorimeter. Because all detectors are in close proximity of each other, it is a prime concern to minimize the interaction between showers produced in the calorimeter, and the charge-measuring instruments. To dampen this effect, CREAM uses a larger number of pixels with a smaller area, along with a very fast readout time to differentiate between events caused by the primary particle and events caused by back-scattering from the calorimeter.