Abstract: To increasing the production efficiency of nuclear fuel cycle, processing equipment with fissile materials is arranged closer to each other than before. The neutron interaction between units with fissile materials has a great influence on criticality safety of multiple-units. The reactivity effect of neutron interaction between two identical tanks containing 19.75% enriched uranyl nitrate solution with and without isolators was measured on uranyl nitrate solution experiment facility. The tank had 260 mm of thickness and 500 mm of width and distance between those two units was adjustable from 0 to 1 000 mm. Condition of the solution was about 200 g/L in uranium concentration, about 0.8 mol/L in free nitric acidity, in room temperature. One tank was fixed, the other was movable in the direction perpendicular to the interacting surfaces. Different materials and thicknesses of isolators could be placed between two slab cores. In the series of distance effect experiment, the interacting surfaces separations were from 21.66 mm to 510 mm in 6 critical experiment cases. In the series of shielding effect experiment, five materials as stainless steel, ordinary concrete, polyethylene, borated polyethylene, water, with 3 or 4 thicknesses from 10 mm to 200 mm were placed in the center between the interacting surfaces in 18 critical experiment cases. Solution was fed into both tanks by independent metering cylinders, so the levels of two tanks could be same or different. In each experiment where the core distance was predetermined, or the isolator was placed, a critical approach was taken repeating solution fuel feeding of two tanks, solution level measurement and neutron counting. Around each critical solution level, a reactivity measurement was also performed using a digital reactivity meter by period method and inverse kinetics method to estimate a differential reactivity worth. The uncertainty analyses were carried out according to ICSBEP guides. The sources of uncertainty were from solution fuel, slab tanks, distance and isolators. Continuous energy Monte Carlo code MONK10 and CENDL nuclear data libraries were used to build the detailed benchmark models. One-variable-at-a-time strategy was applied to determining total uncertainties for each case. The maximum total uncertainty is 200 pcm. The detailed benchmark models for all critical experiments were built and applied in the applicability validation of two combinations of Monte Carlo code and nuclear data libraries for k_eff calculation in the specific systems. The maximum biases between calculation and benchmark for two combinations are 309.0 pcm and 252.0 pcm, which means both are applicable for the criticality safety design or safety analysis in relevant system. The biases of individual simplifications will be obtained to simplifying the detailed benchmark model, and more critical experiments will be carried out such as solid neutron absorber reactivity measurements.