Vous êtes ici : Accueil > Publications > Chicago Pile(s) > The Monte Carlo model of CP-2

The Monte Carlo model of CP-2

​​​​Due to safety reasons, since CP-1 had no radiation shielding nor cooling system, it was decided to dismantle the pile after about three months of testing, including some experiments conducted at 200 W to probe the temperature effects. The operation of CP-1 was officially terminated on February 28, 1943. The bricks of the pile were transported to Argonne Site A, and a new pile was then built and named CP-2. The dismantling of CP-1 and the building of CP-2 took only 21 days. CP-2 was built in a approximately cubic shape. On March 20, 1943, CP-2 began operation, which continued until 1954, when CP-2 was decommissioned. Thanks to a number of experimental facilities (including a thermal column and a neutron chopper), CP-2 contributed extensively to material testing. We have created a TRIPOLI-4 model of CP-2 based on the original plans of the pile, which has allowed comparing the main features of the pile with respect to CP-1.

cp2.jpg
A photo of CP-2, displaying the concrete shield. On top of the pile an experimental facility (familiarly called the "penthouse") was used to extract neutrons from a thermal column for material irradiation. Photo courtesy: ANL. 
​
cohvrum6.png
 A drawing of CP-2, displaying the inner metal uranium core and the outer oxide core, from reference E. Fermi
et a​l., US Patent 2,708,656​, N​eutronic reactor (1955).

Publié le 23 novembre 2023

​Building a model of CP-2

CP-2 basically used the same kinds of graphite bricks and the same types of fuel lumps as CP-1, the main difference being that the poorer materials used in CP-1 were abandoned, thanks to the availability of better oxide and metal fuel lumps, and better graphite. Overall, CP-2 contained about 470 tons of graphite and about 52 tons of uranium in the form of oxide and metal (roughly 17700 fuel lumps, in the form of cylinders of different sizes and pseudospheres). Using the calibration procedure already applied for CP-1, including minimal modifications due to new fuel/graphite combinations that were not used in CP-1, we have built a database of materials to be used in CP-2. For each combination, we have also computed the infinite multiplication coefficient and the key nuclear parameters.

CP2_layer38.png 

TRIPOLI-4 model of the layer 38 of CP-2, showing the fuel lumps in the graphite holes (metal uranium in the center of the pile, and oxide in the periphery), and the slots of the safety rods.​

CP2_horizontal_section_EW_fuels.png

TRIPOLI-4 model of CP-2: the zoom of this portion of the lattice shows various kinds of fuel lumps (cylinders of different sizes, black oxide pseudospheres and brown oxide pseudospheres).

​

A three-dimensional model of CP-2

Based on the elementary bricks, we have built the full CP-2 model, respecting the constraints on the overall mass and number of fuel lumps.​ The structure of CP-2 contains 54 lattice layers, the last 3 layers being used as a reflector in order to increase the excess reactivity. The base of the model is a thick concrete slab. The pile is surrounded by a concrete shield, and on the top of the pile is laid a wood and lead shield. The full model incudes also the ionization chambers, the thermal column, and the slots for the safety, shim and control rods.​​

​CP2_horizontal_section_ionization_chambers.png

TRIPOLI-4 model of CP-2: radial cut showing the inner and outer core and the ionization chambers. The pile is surrounded by a graphite reflector and by a concrete shield.​


CP2_vertical_section_EW_ionization_chambers.png 

TRIPOLI-4 model of CP-2: axial cut showing the inner and outer core and the ionization chambers. The pile is surrounded by a graphite reflector and by a concrete shield. On top of the pile  is laid a wood and lead shield.​

Testing the CP-2 model

The reactivity of the full CP-2 model has been computed using TRIPOLI-4. For the model with 51 layers, which is supposed to be critical with all the control and safety rods out, we have obtained a k eigenvalue ​k = 0.99723 +/- 10 pcm, which is compatible with the slight discrepancy on reactivity observed for CP-1.​