Thermal shield cutout

[paulromano]

@nsedayilmaz and I have been comparing the MSRE reference CSG model from IRPhEP (originally a Serpent model that we've converted to OpenMC) to the CAD model here and one thing we've found that seems to have a significant impact on k-effective is how the thermal shield is modeled. In the reference model, the thermal shield is a single cylindrical shell that extends all the way down to the structural base (filled with a homogenized mix of steel and water). The reality is of course more complicated than that and the CAD model here seems to be a bit more faithful in representing the true geometry.

Namely, there is a cutout in the thermal shield that corresponds to a "vertical projection" off the structural base that is designed to fit the inside curvature of the hemispherical containment. Here's what they say about it in ORNL-TM-728:

Attached to, and part of, the base is a vertical section formed to fit the inside curvature of the bottom hemispherical head of the reactor cell vessel. The outer hemispherical plate, the inner semicircular plate, and the inner conical-shaped "skirt" plate of this vertical projection are 1-in-thick 304 stainless steel. The projection is internally reinformed with six 3/4-in-thick radial web plates. ... A large cutout in the bottom of the cylinder fits over the vertical projection on the base, mentioned above.

Here's a corresponding image (from Figure 5.2 in ORNL-TM-728):

They also include an actual photon of the thermal shield sitting on top of the base (Figure 5.20):

I see that in the CAD model, the bend in the insulation is present and there is a cutout in the thermal shield:

The major questions I'm trying to figure out are:

1. Is the vertical projection of the base that fits the hemispherical head (not included in the CAD model) really effectively empty inside? If so, that would allow neutrons to escape. The reference CSG model does not have this feature and instead just has the thermal shield extending all the way down, which seems to increase k_eff significantly and is a major reason why the two models do not agree.
2. In the CAD model, the cutout seems to be 90°, although I don't see this mentioned anywhere in the MSRE documents. Is there a basis for this choice?
3. Does anyone have access to the MSRE design drawings that would shed light on this (see list in Appendix A of this doc)?
   • E-KK-D-40724 Reactor Thermal Shield Assembly and Sections
   • E-KK-D-40725 Reactor Thermal Shield Sections and Details
   • E-KK-D-40726 Reactor Thermal Shield Plugs and Details
   • E-KK-D-40727 Reactor Thermal Shield Removal Cover
   • E-KK-D-40728 Reactor Thermal Shield Miscellaneous Details
   • E-KK-D-40729 Reactor Thermal Shield Elevation
   • E-KK-D-40730 Reactor Thermal Shield Plan

[aslakstubsgaard]

hey paul, all good questions. i had trouble finding good pictures/drawings when i drew the cad, and perhaps i should go back and do a more detailed cad model and include the steel support structure.

1. if you are asking if the space underneath the reactor support is really empty, it doesn't seem that any shielding was provided on the underside of the reactor.
2. if by 90 deg you are referring to the cutout of the thermal shield being level and not slanted or curved in some way, then i think the drawing and picture you added shows that it is level. also from a sheet metal fabrication perspective this would be easiest to build.
3. no, if i had seen the drawing i would have made a detailed design. have now asked ornl library contacts if they have the documents, but with these document numbers there's usually not a good chance.

the best view of the thermal shield design was actually from the color version the 1/8th plastic mock-up they build

note that the mock-up has mock-up steel balls in one compartment of the thermal shield, but not underneath the reactor, but they could have omitted it in the mock-up, but you would still expect to see it in figure 5.2 from ORNL-TM-0728.

something else that i would expect to effect the model significantly is the ratio of steel to water in the thermal shield, i.e. what effective packing factor for the steel balls you assume. also the lid of the thermal shield would have a different ratio than the sides. note that there's many details not included in the the parts external to the shield, e.g. heating elements that was on the inside of the insulation and outside the reactor vessel.