TJ Memo 835-003_-_Fixed Ballast Installation FEA Memo.docx

TECHNICAL MEMORANDUM

References [A] NOAA Drawing 101-6634860, Structural Module Assembly No. 1 [B] NOAA Drawing 101-6634861, Structural Module Assembly No. 2 [C] NOAA Drawing 101-6634867, Structural Module Assembly No. 8 [D] NOAA Drawing 101-6634868, Structural Module Assembly No. 9 [E] NOAA Drawing 801-6634613, Docking Plan [F] NOAA Drawing TJ-101-001, Fixed Ballast Installation [G] NOAAS Thomas Jefferson Fixed Ballast Addition Weight Change Calculation [H] ABS Rules for Building and Classing Steel Vessels Under 90 Meters (295 Feet) In Length, Part 3, Hull Construction and Equipment, July 2019

List of Figures

Introduction NOAA Ship Thomas Jefferson is scheduled to enter drydock for the total addition of 94LT of fixed ballast and structural modifications. These additions are needed to correct a tendency for the vessel to list to port. The new Fixed Ballast Tank (4-8-1-V) is located below the lower deck from frames 8 to 14. The arrangements of the hold before and after the Fixed Ballast Tank installation are shown in Figure 1 and Figure 2. In Figure 2, the Fixed Ballast Tank extends from the bottom shell to the Lower Deck, 14’-0” above baseline, as is further illustrated in Figure 3. Ref [F] details the modifications necessary to make this tank and to reinforce key areas of the ship’s structure to support the fixed ballast.

As calculated in Appendix A, this modification will result in a net weight change of 1.91 LT in structural steel, manholes, and pipes. Therefore, 92.09 LT of fixed ballast must be added to reach the target 94 LT weight addition.

The fixed ballast selected for installation is Permaballast®, produced and installed by Ballast Technologies, Inc (BTI). Permaballast® is poured into the fixed ballast tank as a liquid and solidifies into a non-shifting substance, the density of which can be calibrated to between 130 to 300 lb/ft3. Assuming the tank has a permeability of 97% and assuming the tank can only be filled to 95% capacity, the minimum fixed ballast density required is 294.4 lb/ft3 to add 92.09 LT with the tank 95% full, as is calculated in Appendix A. The minimum required ballast density increases as permeability decreases; permeability may be as low as 95.2% before the minimum required ballast density reaches 300 lb/ft3, which is the maximum possible ballast density. It is therefore recommended to use the maximum density of 300 lb/ft3 to mitigate the risk that there will not be enough space for the ballast required.

Based on past performance, BTI claims to be able to fill the ballast tank to within 1-2% of the target weight, which is ±1.8 LT. It is noted that, by using 300 lb/ft3 for the fixed ballast density, the ballast tank is estimated to be at 93% capacity when the target weight is reached, leaving enough reserve volume to add 1.74 LT additional before the tank reaches the limiting 95% capacity. This margin mitigates risk of inaccuracies in the tank permeability and the Appendix A weight estimate.

Figure 1 - Plan View of Hold Before Installation

Figure 2 - Plan View of Hold After Installation The purpose of this technical memo is to describe the Finite Element Analysis (FEA) and Finite Element Model (FEM) used to validate the ship’s structure in way of the new fixed ballast load, which is over 4 times greater than the ballast water load which the structure has previously withstood.

The FEA is conducted using FEMAP version 2020.2. The units used in the model are inches and pounds. The origin is located at the intersection of baseline and frame 0, on centerline, where +X is forward, +Y is port, and +Z is up. The model is composed entirely of plate elements.

The analysis performed is a static analysis with dynamic accelerations considered, in accordance with Ref [H], Section 3-1-18/7. The application of dynamic accelerations is further described in the Loads and Constraints section of this document.

Assumptions

1. No ship checks were conducted as part of this analysis.

2. The fixed ballast tank is filled to the top (14’-0” above baseline) with fixed ballast. It is noted that the tank will only be filled to approximately 95% capacity due to limitations in the filling process.

3. The fixed ballast tank has a permeability of at least 95%.

4. The ballast will remain rigid and will not shift in response to vessel motions.

5. The vessel has a bow thruster recess, which is a 5” recess in the shell extending downward and aft from the bow thruster tube and fairing into the shell as shown in Ref [A]. This feature has minimal contact with the fixed ballast and is minimally impactful on the structural integrity in way of the Fixed Ballast Tank. The bow thruster recess is expected to induce complications and errors in the FEM geometry, so it is neglected. Instead, the regular shell is modeled.

6. Densities are as follows:

a. Fixed ballast density is 300 lb/ft3. This is the maximum possible density for this fixed ballast product, and it is expected to be greater than or equal to the density of the ballast installed.

b. Seawater density is 64 lb/ft3.

Criteria The FEA is conducted in accordance with the guidance in the ABS Rules for Building and Classing Steel Vessels Under 90 Meters from July 2019 (Ref [H]), Section 3-2-18. NOAA Ship Thomas Jefferson has a length overall of 208’–2¼”.

Model Extent The model extends vertically from the keel to the Main Deck (23’-2 5/8” above baseline) and longitudinally from Frame 2 to 20, which is three times the length of the Fixed Ballast Tank. Rhino version 6 was used to develop the surface model of this section of the ship. All structural elements are modeled as surfaces, such that they will be analyzed as plate elements in the FEA. An overview of the Rhino model which was used to create the Finite Element Model (FEM) is shown in Figure 3.

Figure 3 - Model Overview Figure 4 shows the lower hold with the Lower Deck and all structure above it hidden for clarity.

Figure 4 - Model Overview Focused on the Hold Loads and Constraints Two conditions are considered, each with different loads and constraints. The conditions are:

Load Case 1: Operating in seawater: The worst case loading scenario the Fixed Ballast Tank will experience during operation is when all adjacent seawater ballast tanks are empty. In this scenario, the fixed ballast will impose static pressures against the tank boundaries as well as vertical, transverse, and longitudinal accelerations identified in Ref [H], 3-2-18/7. The particulars of this load case are as follows:

· A fixed ballast hydrostatic load is vectored outward from within the Fixed Ballast Tank and perpendicular to each of the tank boundaries. The static loads are calculated by the formula [P = 0.1736 lbf/in3 × h]. In accordance with Ref [H], 3-2-18/7, dynamic loads are included by using a multiplier in the static pressure equation. The multipliers used are as follows:

· Dynamic loads against vertical bulkheads: Longitudinal and transverse accelerations will increase the pressure load against vertical bulkheads. The ABS guidance recommends using 0.5g as the lateral and transverse acceleration. Therefore, the static pressures against vertical bulkheads are multiplied by 1.5 to account for dynamic factors. The longitudinal and transverse pressure loads which receive this multiplier are applied as illustrated in Figure 5.

Figure 5 - Fixed Ballast Loads on Vertical Bulkheads (Plan View)

· Dynamic loads against the side shell, bottom shell, and bow thruster tunnel: A combination of vertical and horizontal accelerations will increase the static pressure load against the side shell and bottom shell, which have curvature in way of the tank. The ABS guidance recommends using 1.775g as the vertical acceleration factor for this tank. It is assumed that this includes acceleration due to gravity, so the static pressures with vertical components (i.e. pressures against the hull, shown in Figure 6) are multiplied by 1.775.

To be fully accurate when applying the accelerations, it would be necessary to calculate the vertical and horizontal components of acceleration at each point, which would vary depending on the angle of the hull. The curvature of the side shell complicates this method, so, for expediency’s sake, the model conservatively assumes that the side and bottom shell will experience only vertical acceleration with no horizontal components, similar to the loading of the flat bottom of a similar fixed ballast tank. This assumption is conservative because, for example, if the hull had a 45° angle, the resulting acceleration components would be:

A similar assessment could be carried out for any hull angle to find that the maximum multiplier attainable is 1.775g.

Figure 6 - Fixed Ballast Loads on Non-Vertical Surfaces (Section View)

· A seawater hydrostatic load is vectored inward perpendicular to the shell and bow thruster tunnel up to the waterline, which is 14’-0” above baseline (coincident with the Lower Deck).

· Fixed constraints are placed at the forward and aft extents of the model (frames 2 and 20), as shown in Figure 7. These extents are far away from the load we are concerned with, so the shell is expected to deflect equally on either side of frame 2 and 20, resulting in zero rotation of those nodes. Therefore, a fixed constraint is considered the best approximation.

Figure 7 - Constraints for Load Case 1, in Operation Load Case 2: In drydock: The worst case loading scenario the Fixed Ballast Tank will experience in drydock is when the blocks are nearest the fixed ballast in the forwardmost docking position (No. 2 in reference [E]) and when ballast tanks 4-8-2-W and 4-10-1-W and forepeak ballast tank 4-0-0 are full of seawater. When in dry dock, the ship will not be subject to the dynamic accelerations described in the Load Case 1 above. The particulars of this load case are as follows:

· A fixed ballast static load facing outward from within the Fixed Ballast Tank, similar to Load Case 1.

· A seawater ballast hydrostatic load is facing outward from the forepeak tank (4-0-0-W) and the port and starboard ballast tanks (4-8-2-W and 4-10-1-W), which are assumed have a 29.05 ft head in accordance with the design head specified in Ref [H], 3-2-8/5.1, with the overflow taken to be 35 ft above baseline.

· Fixed constraints are placed at the aft extent of the model. The forward extent of the model is left free, as there are no blocks forward of frame 16. Additional fixed constraints are placed where the keel blocks touch the keel (Block Nos. 21 and 22) and in the forwardmost position of side block No. 10 (port and starboard). The aft end and block constraints are shown in Figure 8. Full blocking arrangements are detailed in Ref [F].

Figure 8 - Constraints for Load Case 2, in Drydock The densities of the fixed ballast and seawater are noted in the Assumptions section on page 5.

Materials and Element Properties Material Properties The material properties used are detailed in Table 1. The allowable stresses are determined in accordance with 3-2-18/9 Table 2, where it is noted the maximum mesh size is 2 inches, and the stiffener spacing is 24 inches.

Table 1 - Material Properties

36,288 in way of discontinuities

40,320 in way of discontinuities

Element Properties Plate Element Properties The linear plate elements are meshed using a variety of plate properties. Each plate property is defined by a separate plate thickness ranging from 3/16" through 2".

Table 2 - Plate Element Properties

Beam Element properties No Beam elements are used to model the FEM. All stanchions and stiffeners were modeled using surface and plate elements.

Mass Element Properties No mass elements are used in FEM. The fixed ballast is modeled as a hydrostatic pressure. Where accelerations must be considered, they are applied as a multiplier to the pressure, as described in the Loads and Constraints section above.

Rigid Links No rigid links are used in this FEM.

Model Mesh The default mesh size was set at 2 inches in order to obtain sufficient granularity on the flanges of angles and flange plates. The brackets added at frames 9 and 11 have their meshes refined to 1 inch to get a better idea of how the brackets distribute corner stress concentrations. The standard mesh with a refined bracket is shown in Figure 9.

Figure 9 - Mesh Refinement in way of Brackets

Results The results of the FEA are shown in the figures below. The models show the Von Mises stress remains below the allowable stresses shown in Table 1, as is required by Ref [H] Section 3-2-18/9.1. Ref [H] does not require other stress types to be analyzed. Additionally, the deflections in each model are reviewed to ensure the model deflects in the direction expected (i.e. the loads are applied in the proper direction).

The figures in this section show elevation views looking to port, plan views looking down, and section views looking aft to detail the full results from the models. Key features in each view are as follows:

Elevation views:

Plan views:

Section views:

Load Case 1 Results Two locations in the Load Case 1 model exceed the 32,400 psi, which is the lowest of the allowable stresses in Table 1. However, these locations are in way of discontinuities, so the allowable stress for them is higher, as is noted in Table 1. These two locations are at frames 8 and 9, and are illustrated in Figure 27 and Figure 29.

The following figures show the stress levels seen in the model for Load Case 1, the operational condition. The maximum limit of the stress scale coincides with the minimum allowable stress from Table 1. Stresses which approach or exceed this limit are discussed in further detail.

Figure 10 - Load Case 1: Outboard Profile Figure 10 shows an external profile of the vessel, where the fixed ballast load can clearly be seen. The low stress patterns at the forward and aft ends of the model are due to hydrostatic pressure of water external to the ship. A few spots on the shell have high stress concentrations. As shown in Figure 11 below, none of these stresses are above the allowable stress.

Figure 11 - Load Case 1: Outboard Profile, Peak Stresses Highlighted

Figure 12 - Load Case 1: Elevation Cut at 8'-0" Off Centerline

Figure 13 - Load Case 1: Elevation Cut at 8'-0" Off Centerline, Loaded Region

Figure 14 - Load Case 1: Elevation Cut at 6'-0" Off Centerline

Figure 15 - Load Case 1: Elevation Cut at 6'-0" Off Centerline, Loaded Region

Figure 16 - Load Case 1: Elevation Cut at 4'-0" Off Centerline

Figure 17 - Load Case 1: Elevation Cut at 4'-0" Off Centerline, Loaded Region

Figure 18 - Load Case 1: Elevation Cut at 3'-0" Off Centerline, Loaded Region Figure 18 shows a stress riser at the bottom of a stiffener attached to Frame 8. Closer inspection (see shows the peak stress here is 26,873 psi, which is below allowable.

Figure 19 - Load Case 1: Elevation Cut at 3'-0" Off Centerline, Peak Stresses Highlighted

Figure 20 - Load Case 1: Elevation Cut at 2'-0" Off Centerline

Figure 21 - Load Case 1: Elevation Cut at Centerline

Figure 22 - Load Case 1: Plan View at Lower Deck, 14'-0" Off Centerline

Figure 23 - Load Case 1: Plan View Cut Below Lower Deck, 13'-4" Above Baseline

Figure 24 - Load Case 1: Plan View 13’-4” Above Baseline, Loaded Region

Figure 25 - Load Case 1: Plan View 9’-0” Above Baseline, Loaded Region

Figure 26 - Load Case 1: Section at Frame 8, Watertight Bulkhead, Looking Aft It is noted that Frame 8 has a maximum stress of 32,909 psi where it meets the centerline bulkhead. As noted in Table 1, the allowable stress in way of discontinuities is for structure which is not longitudinally effective is 40,320 psi, which is still greater than the stresses observed. This peak stress is shown in Figure 27, which omits any stresses below 32,400 psi, which is the lowest allowable stress from Table 1.

Figure 27 - Load Case 1: Section at Frame 8, Watertight Bulkhead, Looking Aft, Highlighting Peak Stresses

Figure 28 - Load Case 1: Section at Frame 9, Looking Aft The radiused bracket shown in Figure 28 near the base of the centerline bulkhead has some elevated stresses. These are explored in more detail in Figure 29 below, which hides all stresses below 32,400 psi, which is the lowest of the allowable stresses in Table 1.

Figure 29 - Load Case 1: Section at Frame 9, Looking Aft, Peak Stresses Highlighted Similar to the example of Frame 8 (Figure 26 and Figure 27), these stresses are below the allowable stress of 40,320 psi in way of discontinuities on structure which is not longitudinally effective.

Figure 30 - Load Case 1: Section at Frame 10, Partial Watertight Bulkhead, Looking Aft Although Figure 30 appears to show some questionable spots, both have stresses which remain below the allowable stress, as shown in Figure 31 below.

Figure 31 - Load Case 1: Section at Frame 10, Partial Watertight Bulkhead, Looking Aft, Peak Stresses Highlighted

Figure 32 - Load Case 1: Section at Frame 11, Looking Aft

Figure 33 - Load Case 1: Section at Frame 12, Partial Watertight Bulkhead, Looking Aft As shown in Figure 33, Frame 12 has a stress concentration at one point where it meets the bottom shell. As shown in Figure 34 below, this remains below the allowable stress.

Figure 34 - Load Case 1: Section at Frame 12, Partial Watertight Bulkhead, Looking Aft, Peak Stresses Highlighted

Figure 35 - Load Case 1: Section at Frame 13, Center of Bow Thruster Tunnel, Looking Aft

Figure 36 - Load Case 1: Section at Frame 14, Watertight Bulkhead, Looking Aft

Figure 37 - Load Case 1: Section at Frame 14, Watertight Bulkhead, Looking Forward

Load Case 2 Results Load Case 2, the most severe condition in drydock, is described similarly to Load Case 1. All stresses in Load Case 2 remain below the allowable stresses identified in Table 1. The peak stresses for this load case are on the frame 14 bulkhead (see Figure 55).

Figure 38 - Load Case 2: Outboard Profile The high stresses near the forward end of the model, in way of frames 2 and 3, are erroneous. They are caused by the seawater ballast pressure load against the shell near the end of the model, which is unconstrained in order to simulate actual global deflections and stresses imposed by the fixed ballast and seawater ballast against the keel blocks in a drydock. Because the model does not include the structure forward of frame 2, this portion of the model is seen as unsupported and free to deflect outwards. The stress anomaly at frames 2 and 3 does not match actual ship conditions, and it is far enough away from our area of interest, which is the fixed ballast tank, that it is disregarded.

Figure 39 - Load Case 2: Elevation Cut at 8'-0" Off Centerline

Figure 40 - Load Case 2: Elevation Cut at 6'-0" Off Centerline

Figure 41 - Load Case 2: Elevation Cut at 4'-0" Off Centerline

Figure 42 - Load Case 2: Elevation Cut at 3'-0" Off Centerline

Figure 43 - Load Case 2: Elevation Cut at 2'-0" Off Centerline, Peak Stresses Highlighted

Figure 44 - Load Case 2: Elevation Cut at Centerline

Figure 45 - Load Case 2: Plan View at Lower Deck, 14'-0" Above Baseline

Figure 46 - Load Case 2: Plan View 13'-4" Above Baseline

Figure 47 - Load Case 2: Plan View 9’-0" Above Baseline The high stress shown in Figure 47 is on the horizontal girder on the transverse bulkhead at frame 16, which is a watertight ballast tank boundary. This stress does not propagate to any elements on the longitudinal bulkhead. These stresses are therefore below the 40,320 psi allowable stress in way of discontinuities for structure which is not longitudinally effective (see Table 1).

It is also noted that the stress riser shown in Figure 47 is two frames (4 ft) aft of the fixed ballast installation, and it is outside the scope of this modification. The load causing this stress is from the seawater in ballast tank 4-10-1-W, which is already an existing load on the ship. It is assumed that this part of the structure was approved by ABS during reviews for initial construction of the vessel.

Figure 48 - Load Case 2: Plan View 6’-0” Above Baseline, Bottom Shell and Blocks Shown

Figure 49 - Load Case 2: Section at Frame 8, Watertight Bulkhead, Looking Aft It is noted that the stresses shown on the starboard side of frame 8 in Figure 49, which is the forward bulkhead of the new Fixed Ballast Tank, has much lower stress compared to that of Load Case 1, shown in Figure 26.

Figure 50 - Load Case 2: Section at Frame 9, Looking Aft

Figure 51 - Load Case 2: Section at Frame 10, New Bulkhead from CL to , Looking Aft It is noted that the new bulkhead installed by Ref [F] at frame 10 is much less loaded than in Load Case 1 (Figure 30). As previously discussed, this is partially because the seawater ballast load from tank 4-10-1-W is counteracting the fixed ballast pressure in the new Fixed Ballast Tank. Another reason for the lower stress is that fixed ballast accelerations are not considered in the drydock model, as the ship is not expected to encounter appreciable accelerations while in drydock.

Figure 52 - Load Case 2: Section at Frame 11, Looking Aft

Figure 53 - Load Case 2: Section at Frame 12, Partial Watertight Bulkhead, Looking Aft Some of the stress patterns shown in Figure 53 are results from the hydrostatic seawater ballast pressures in ballast tanks 4-8-2-W and 4-10-1-W. The stresses shown on the starboard cluster of the stress pattern (the left hand side of the figure) are the result of the fixed ballast load.

Figure 54 - Load Case 2: Section at Frame 13, Center of Bow Thruster Tunnel, Looking Aft

Figure 55 - Load Case 2: Section at Frame 14, Partial Watertight Bulkhead, Looking Aft Two stress risers are noted here. Both of these risers are in way of discontinuities on structure which is not longitudinally effective. Therefore, the allowable stress is 40,320 psi, as noted in Table 1.

Conclusion The results show the new structure meets the criteria of Ref [H] Section 3-2-18 in both load cases considered. Load Case 1 is the most adverse load condition overall for two reasons. Load Case 1 includes dynamic load factors, whereas Load Case 2 does not, and Load Case 1 does not include seawater ballast, which acts against the fixed ballast load on the fixed ballast tank boundaries, reducing the net pressure. These factors result in greater overall Von Mises stresses for Load Case 1. In accordance with Ref [H], 3-2-18/9.1, the Von Mises stresses are verified to remain below the allowable stress limits defined in Table 1. No other stress types are analyzed.

830 Pacific Avenue, Bremerton, WA 98337 | 360-479-5600 | www.artanderson.com -2-image3.png image4.png image5.png image6.png image7.png image8.png image9.png image10.png image11.png image12.png image13.png image14.png image15.png image16.png image17.png image18.png image19.png image20.png image21.png image22.png image23.png image24.png image25.png image26.png image27.png image28.png image29.png image30.png image31.png image32.png image33.png image34.png image35.png image36.png image37.png image38.png image39.png image40.png image41.png image42.png image43.png image44.png image45.png image46.png image47.png image48.png image49.png image50.png image51.png image52.png image53.png image54.png image55.png image56.png image1.jpeg image2.png image57.jpg