Stern Flask Storage System (SFSF)

This project was one of the most rewarding projects I worked on during my internship at Ontario Power Generation (OPG). I was able to apply all of the engineering principles I learned from school to building a product with a significant impact.

Project Background:

OPG removes and temporarily stores radioactive In-Core Flux Detectors from their CANDU reactors inside the Stern Flask Container (see Figure 1).

The containers are transported and stored using a fully welded structural storage Frame. The Frames are shipped first to an underground storage trench followed by a multi-purpose storage building (MPSB).

The original Frame designed in 2011 can store a single Stern Flask Container (see Figure 2). Since storage space is expensive, and more containers are required to be shipped. A new Frame design named the Stern Flask Storage Frame (SFSF) is built to increase space capacity by stacking the Frames (see Figure 3).

 Design Process:

I was responsible for designing new structural concepts on paper and translating them onto CAD. I also developed all the engineering calculations to support the material selection and sizing of the Frame by building excel sheets and writing the calculation report.

There are four new features incorporated that improved the Frame design. The structure is build using low carbon steel (CSA G40.21) square Hollow-Structural Section (HSS). The cone-like features referred to as a "guide pin" allow the Frames to stack up to three high. The guide pins are located on top of the four support columns to secure the Frame stacked above it. The base pate, forklift pockets, and columns are reduced in size to maximize space efficiency and decrease the assembly weight. Finally, the lifting lugs are relocated and supported using a corner brace HSS.

Structural Analysis:

The final design report illustrates how the columns and welds satisfy the structural and strength requirements for handling and storing the waste under the worst-case loading conditions. 

Columns: 

The columns used to support the stacking of the Frames are slender, meaning that the cross-section is relatively small relative to its length. Two critical failure modes can lead to the failure of the columns: 

1. Structural Instability

2. Material Failure

Structural instabilities are dictated by the stiffness of the component. Two factors are analyzed to design this Frame, dynamic and buckling instabilities. 

Dynamic Instabilities occur when an external load does work on a system and inputs kinetic energy into it. The key to studying instabilities is to see how much of the energy is converted to kinetic energy at the operating load and then dissipating this energy in a safe and controlled fashion to stabilize the system. The SFSF will be stored underground, and the only possible external dynamic loading that could influence the behaviour of the system is earthquakes. During an earthquake, the earth slightly rearranges its mass to lower its moment of inertia, increases its angular velocity and converts the gravitational potential energy into kinetic energy. The effects of kinetic energy are analyzed using newton's second law (F = ma) to study the lateral forces exerted on the structure by the ground vibrations. The equivalent load, referred to as the base shear, is the maximum lateral force exerted on the Frame during the earthquake. It is affected by several factors including, the soil condition at the site, the probability of significant seismic ground motion, and the proximity to potential sources of seismic activity. The base shear load was calculated by following the National Building Code of Canada based on the geographical data in Tiverton, Ontario. These inertia forces are transferred downwards through the horizontal and vertical structural elements to the soil underneath. The way the inertia forces move throughout the structure to the ground is called the load path. Earthquake performances of a Frame are evaluated by the soundness of their load paths, regardless of the material used to construct it. The new SFSF columns directly touch the ground, allowing the load to transmit through a continuous and direct load path without being bent or interrupted. This design change will dissipate the kinetic energy gained from the earthquake more effectively. Also, since the earthquake will sway uniformly in two horizontal directions, the Frame is designed to be symmetrical (50 x 50 inches). Otherwise, it may twist about a vertical axis, which is detrimental to its earthquake performance.

Since the soil foundation is shallow, the resistance of the foundation to sliding, global overturning, and exceeding the soil bearing capacity in the trench and MPSB is also investigated using the seismic forces.  

Material Failure:

The compressive deadweight and lateral seismic load combinations acting on the Frame are analyzed using a 2D in-plane rigid structure with fixed lower ends. The reaction forces at the support ends are calculated to ensure that the maximum stresses experienced in the components during a seismic event are within the allowable stresses for the respective material selected.

Buckling Instabilities occur when a structure undergoes rapid sideward deformation due to sudden loss of stiffness. The behaviour of buckling is illustrated using a force-displacement diagram (see Figure 5, left). Initially, the behaviour of the material is elastic because the stiffness is linear (K = stiffness i.e. slope). At the onset of buckling, there is a sudden loss of stiffness followed by a rapid drop to negative stiffness. This behaviour is referred to as local buckling, which may cause permanent deformation and function failure but not a catastrophic failure. The column will regain its stiffness and continue to counter the applied load until it permanently fails globally leading, to a catastrophic failure. There are two methods used in this design analysis to assess buckling. 

1. Linear Eigenvalue analysis

2. Second-Order Analysis (nonlinear buckling)

The Eigenvalue Analysis is a linear analysis that solves for the maximum load at which the structure buckles globally. It idealizes the force-displacement curve (see Figure 5, right). The Canadian Institute of Steel Construction (CISC) Handbook provides an easy method of calculating the critical load based on the square HSS columns used. This method however does not evaluate the numerical values of the displacements or stresses.

In a real structure, imperfections and nonlinear behaviour keep the system from achieving this theoretical buckling strength, leading Eigenvalue analysis to over-predict buckling load.

The second order nonlinear buckling analysis provides greater accuracy than elastic formulation. As the Frame sways laterally, the vertical loads acting through the deflected shape cause additional moments and deflections that are not analyzed in the linear analysis but may be critical to study. During nonlinear-static buckling analysis, the total load is applied incrementally to account for the stiffness change at each displacement step. In contrast to linear-buckling analysis, the nonlinear analysis calculates actual displacements and stresses of the columns.

Welds: 

The Frame assembly is fully welded between the structural elements to provide robust support. FEA licensing is limited for design use at OPG, STAAD.Pro software is instead used to calculate the internal forces in the members during the lifting and stacking/seismic events. As discussed earlier, the vertical columns will direct the loads to the ground when the Frames are stacked, not to the welded joints. The lifting results are used as inputs to evaluate the joints, as they are larger than the stacking/seismic joint loads. A Free-Body Diagram illustrates how the STAAD.Pro forces and moments are applied to the welds (see Figure 5). The shear and bending stresses are calculated for all weld configurations to ensure that the size, material, and weld type are safe under the worst-case loading condition.  

Lifting Components:

The lifting/handling features use rigging, cranes, or forklifts to manoeuvre the Frame assembly during the various stages of the waste management process. The bearing and tear-out stresses on the lifting lugs are analyzed to ensure that the stresses experienced in the components during lifting are within the allowable stresses for the respective materials and specified working load limits. A conservative two-point lift was assumed for the welding and lifting calculations, meaning that the load is carried entirely by the two legs only (i.e. the working lifting load is divided by two instead of four). Due to the symmetry of the Frame, only one side of the welds is investigated (see Figure 6).

Sizing and Testing:

These calculations were the driving factor in sizing the Frame. The Frame design was iteratively enhanced using FEA/CAD software, 3D printed models and required extensive research before completion. The new SFSF design improves space capacity by 200% per 2500 square inches of space and reduced the structure size by 20%. I was very fortunate and happy to lead this project from start to finish. OPG manufactured 75 frames and tested them, as demonstrated in Figure 7. The Frame design saved OPG 3 million dollars on new building new storage infrastructure.