MSIP for RV CRDM penetrations

CRDM nozzles are usually shrunk fit in reactor head penetrations and then welded to the dome on the inner side by partial penetration J-welds. Since all the nozzles are vertical, the outer radial locations require welding of the nozzle to the dome at angles of up to 47º. Welding produces a complex pattern of residual stresses where the tensile stresses are enhanced by significant ovalization of the nozzle extending inside the dome. The pattern of stress becomes more complex after a hydrotest which introduces local plastic yielding at the hole boundary and weld discontinuity.

The mechanical means of applying loads must be adapted to the complex geometry of the CRDM penetration. Such loads can be imposed using a simple device with a central rod extending through the penetration having a head at one end and a hydraulic cylinder on the other (Figure 4). The imposed axial compressive stresses interact with the as-welded circumferential tensile stresses to  enhance plasticity. The resulting plastic flow redistributes the residual stresses to remove tension from the critical weldment region. After removing the imposed axial loads, both the axial and the circumferential residual stresses are reduced to an almost stress-free condition.

Two and three-dimensional inelastic finite element stress analyses were performed on the central and outermost CRDM penetration weldments of typical RPV head design to verify the concept.

Considering the CRDM penetration pattern in the Reactor Vessel Head (RVH), a three dimensional pie-shaped segment of the RVH dome is modeled. This model extends from the center of the head down to the flange connection. One side of the model bisects the center of the outermost CRDM. The area in the vicinity of this CRDM weldment to the RVH dome is modeled with a fine mesh to accurately determine the resulting stresses. Such a shaped model allows symmetry boundary conditions to be accurately applied to the model. The RVH flange is included in the model to account for the effect of the flange stiffness on the deformation and resulting stresses on the analyzed CRDM (Figure 5).

This model was used in comprehensive elastic-plastic analysis to demonstrate the benefits of the MSIP for reducing residual stresses in the weldment to mitigate stress corrosion cracking.

While analyses show that as-welded tensile stresses are reduced by axial contraction of the nozzle, our results show even better improvement of stress would be achieved if the penetrations are contracted when the RPV head is subject to internal pressure, or when a difference in average temperature exists between the RPV head shell and the Inconel penetration. This average temperature difference should be in the order of 85º to 140º C. A combination of heating and cooling would be applied locally to the RPV head penetration weldment being treated with MSIP. The average temperature difference would eliminate the shrink fit and, along with MSIP, improve stresses on the inner surface of the penetration. The cooling process should take a few minutes to achieve the desired cooldown and would be followed by an axial squeeze.

Initial hoop stresses within a range of 275-345N/mm2 tension are reduced, such that after returning to operation, the tensile stresses remain low. Significant improvement is obtained at all locations around the nozzle, below and above the weld. Application of the process eliminates both the hoop and axial 'as-welded' stresses above the weld as well as in the weld region and beneath it.

Qualification tests have been run at  Equipos Nucleares, S.A (ENSA) factory in Santander (Spain) to verify the removal of the 'as-welded' residual stresses. Full scale mock-ups have been designed and fabricated using the same procedures as those used in an actual RV head. Residual stress measurements were made before and after application of MSIP. Geometry and locations of stress measurements are shown in Figure 6, Figure 7 and Figure 8 show the level and distribution of stresses before and after MSIP application. The attached photograph (Figure 9) shows the stress measuring equipment designed by ENSA for measuring residual stresses inside the penetration.

The following table summarizes the results:

Residual axial weld stresses (N/mm2) before and after MSIP

The results show that the tensile stresses are either removed or reduced to such low levels  as to be below the threshold of corrosion cracking.

It should be noted that the Inconel 600 used in the mock-up was conservatively chosen with a high yield strength of 518 N/mm2. This is considerably higher than the typical yield strength values for Inconel 600 which are in the order of 300 N/mm2. Even with such high yield strength material, the process was able to eliminate or reduce the tensile 'as-welded'  residual stresses to begin levels along the inner surface of the penetration. For the normal yield strength values, it is expected that the tensile stresses will be completely removed.