Computational and experimental studies to obtain correction factors for heat defect, heat transfer and scattering/perturbation for the water calorimeter in both Co-60 and high-energy x-ray beams are nearing completion.
Modeling of accumulated-dose dependence in the heat defect of H2/O2 exhibits expected peak behavior in qualitative agreement with experiment, and is being further refined for purposes of reducing the associated relative uncertainty from its present value of 0.3 %. Studies to quantify the possible role of convection in room-temperature operation have not yielded a measurable effect except at very elevated dose rates or exposure times. Comparison of results obtained at 3.98 oC and at room temperature exhibited no observable systematic differences, although uncertainties from the noisier 3.98 oC runs could not preclude an effect below the desired level of < 0.2 %. While repeat runs at 3.98 oC are planned, a separate, room-temperature study with the calorimeter in high-energy x-rays yielded a linear response over a 4x variation in the Clinac dose servo, with an uncertainty of < 0.2 %. This is believed to provide an upper bound on convection effects, as the residuals were random and computational modeling of that experiment suggested onset of nonlinearities due to convection only at dose rates in excess of ~10 Gy/min, more than 5x higher than typical operating conditions. Additional measurements and simulations are planned to see whether effects due to convection are detectable below 0.2 %. Monte Carlo simulations of scattering and attenuation by non-water materials in the phantom have been done for both 6 MV and 18 MV x-ray beams and are planned for Co-60. Computational results and measurements conducted with a Farmer chamber were found to be in excellent agreement.