Part VIII

Mold design was much simplified over previous concepts except that provisions were incorporated for molding the finished sheer line and the forward end of the mold had to be carefully developed so that the build-up of the successive elements of the stem layup, including the transition from sandwich to solid laminate, would result in the desired outer surface contours. Layup procedures developed for use with the female mold were considered applicable to the male mold concept with only minor changes.

In anticipation of possible parting difficulties a removable strip was provided along the mold f from the stem aft to and including the transom. Removal of this strip would allow the mold to be slightly collapsed if such appeared necessary.

A small test mockup had been made in order to find what surface quality could be obtained on a wood mold with one ply of polyester impregnated glass cloth as a surface. Results were entirely satisfactory and mold fabrication was begun.

Mold frames were lofted 3/4" inside the lines of the hull and constructed of 2" nominal dimension lumber. These frames, together with a built-up wooden stem member were erected on 24" centers on a foundation consisting of three 24" square launching timbers. The structure was then carvel planked with 3/4" select grade white pine. After sanding the surface fair with coarse sanding boards, a 1-ply butted skin laminate was laid up on the mold surface followed by a highly activated polyester overlay. Smoothing of the laminate surface was complicated by air inhibition which left a tacky surface over a considerable part of the mold. It was found that washing with acetone, followed by scouring with a proprietary scouring powder and water, would remove the soft outer resin and expose the cured laminate underneath.

After the surface was sufficiently smooth, the locations of the bulkheads, hull E, girders, and a reference water line were scribed in the mold surface in the belief that these lines would be picked up on the inside of the hull. Other marks were made on the mold surface to locate the leading edge of each ply of cloth making up the inner skin shingle layup. When these marks were in place, the mold surface was waxed and polished and the parting film applied by spray gun. Before the layup of the hull was begun, a gel coat was applied by brush to protect the parting film from injury during layup of the inner skin. Adjustable steel scaffolds were provided on both sides of the mold which could be rolled fore and aft as the operations proceeded. See photograph 6.

Layup of the inner skin was begun at the stern and proceeded toward the bow. Here, as in all major fabrications, cloth was machine-impregnated and laid up with additional resin so that in troweling out the cloth, air would be displaced from the weave by the previously applied resin (most bench layups were made using resin and dry cloth in a similar manner; first resin, then cloth). During the operation, ambient temperatures at times exceeded 100°F which made it necessary to adjust activator concentration to prevent premature gelling of the impregnated cloth.

The plan of laying up forward from the stern to about midship and then starting at the bow and working aft till joining the aft layup was carried over from the first boat, although the aft layup could just as well have continued through to the bow, since only one layup crew was used.

The shingle layup, used with average pot life formulations proved ideal for 1-shift operations, since air inhibition acted to the extent that the final ply of one day's layup would carry sufficient tack overnight that the next morning's layup would effect a virtual primary bond.

About half way through the inner skin layup, a delay in cloth delivery made it necessary to suspend full speed operation. For several days one ply only would be added to the previous day's layup. In this way the necessity for bonding to a fully cured surface was avoided and the remainder of the inner skin layup was accomplished without incident.

The curling tendency of the selvage edges of the cloth, such as occurred with the outer skin layup in the female mold, was not encountered here.

Local reinforcements of the hull E, shaft logs, sheer line, and strut and davit clamp were all redesigned using vermiculite-resin instead of solid laminate for bolt-bearing fillers. The C.C.A. cores were first bagged in place, followed by honeycomb that was involved in the location of the bolt-bearing fillers. The remaining honeycomb was then installed, leaving space for the girder foot doublers, the location of which had been marked on the inner skin layup. Each piece of honeycomb as before was set in one ply of impregnated mat. Transverse voids were left in way of the bulkhead locations which were later to be filled with vermiculite-resin to act as water stops and to prevent possible crushing of the honeycomb in way of the bulkheads during docking operations. These voids proved very useful as areas in which to effect vacuum bag sealing. All honeycomb was scored longitudinally both to assist in forming the honeycomb to the shape of the hull and to provide a means for bleeding air from the honeycomb during the vacuum bagging operations. Vacuum bagging operations were much more effective than on the first boat, to the extent that honeycomb bond repairs involved a very small percentage of the total area, about 1/10 of 1%.

After the 10-ply girder foot doublers were in place with their accompanying honeycomb, the vermiculite-resin bolt-bearing fillers were applied. The transverse voids in way of bulkheads were not filled at this time. At first, the vermiculite-resin was applied in excessive bulk with the result that high exothermic heat was developed and considerable cracking occurred. While such cracking would be of little importance in ordinary bolt-bossing applications, it was felt that such cracks in the area of thru-hull bolts below the waterline could contribute to leakage into the adjoining honeycomb, so these cracks were carefully filled. Filling the voids in three or more successive applications of vermiculite-resin eliminated the trouble.

When honeycomb installation was complete and all surface irregularities sanded fair, the outer bonding layer was applied. This consisted of one ply of impregnated mat accompanied by one ply of cloth which served to stabilize the mat during impregnation and handling. The bonding layer was vacuum bagged to insure intimate contact with the honeycomb, the transverse voids again being used for bag sealing. When the bonding layer was complete, the transverse voids were filled with vermiculite-resin as was the void along the sheer line. Layup of the doubler plates on the shaft logs and strut and davit clamps completed the layup operations on the core material - internal reinforcement phase. All irregularities in the outer surface were ground fair and the surface roughened by sanding to assure adequate bonding of the outer skin.

Following the fairing and sanding of the outer bonding layer, location marks were applied to locate the position of each ply in the initial 4-ply shingle layup. The two finish plies were to be butted. Application of the shingle layup proceeded in much the same manner as the inner skin and no difficulties were encountered.

On one occasion, enough impregnated cloth to make several plies was unused at the end of the working day. This cloth was unrolled, cut to suitable lengths, and stored flat in individual sheets between films of polyvinyl chloride. The inhibiting effect of the PVC and the low bulk of the single cloth thickness combined to prevent gellation of this material and it was used the following day, interspersing the stored material with freshly impregnated cloth, so that any overnight styrene loss or possible activator migration into the PVC would be offset.

Following the completion of the 4-ply shingle layup, the vermiculite-resin forefoot filler was applied along with the forward section of the keel core. The layup of the 30-ply tapered stem reinforcement was to follow. It was decided to make this layup in two parts, each of 15 plies, allowing one day for each half. At the conclusion of the layup of the first 15 plies, it was observed that the layup was slowly sliding out of position with considerable lumpiness developing. It had been a cold day and the heating system had not been fully completed. This no doubt contributed to retarded gellation of the earlier applications and the accumulating weight finally overcame the shear resistance of the laminating resin. A vacuum blanket was quickly applied and the sliding tendency was arrested. The following day the lumpiness was ground off and the remaining plies were laid up in 5-ply increments. No further trouble was experienced. This stem layup was so designed as to accommodate the forward end of the subsequent keel layup.

When the stem layup had been ground fair, the butted 2-ply finishing layup of the outer skin was applied, followed by a trowelled and brushed overlay of resin.

Prior to actual keel application, the C.C.A. keel core had been prefabricated including 10-ply transverse webs and vermiculite-resin blocks in way of lifting and V-strut bolts. The core prefab was made in several sections in order to simplify handling and to lessen danger of breakage in getting the core into position on the hull.

The prefabricated core assembly was erected on the inverted hull using 1 ply of impregnated mat as an adhesive layer. The keel laminate was then applied, starting aft and working forward, using a shingle layup so applied as to give 3/8" laminates on the keel sides and 3/4" on the bottom, with 3/8" flanges tapering in 9" to a feather edge and bonded secondarily to the bottom of the hull. A single finishing ply was applied over the shingle layup, following grinding of irregularities.

Upon completion of the keel layup, a second resin overlay was applied to the entire hull and keel. The hull shell was now complete and ready for separation from the mold.

The mold structure and completed hull were rolled out of the shop and placed in a suitable location to allow crane handling.

The hull and mold were lifted and turned over as a unit by two locomotive cranes and placed in the cradle provided. The mold was then lifted from inside the hull. No parting difficulty of any kind was experienced.

The hull and its cradle were returned to the shop and the inside of the hull inspected. Surface imperfections were few and slight.

It was found that the reference and location lines had been picked up from the mold surface as desired and the waterline so produced was used in levelling up the hull shell prior to installation of the internal structural members.

Parting film removal proved troublesome and several methods were tried, including washing with acetone, washing with soap and water, wire brushing, and sanding. The most effective method proved to be blowing the film off with a high pressure air jet, a technique suggested by the action of the metallizing gun as observed during fabrication of the female mold for the first boat.

Bulkhead installation involved making a template of the hull shape at the appropriate location, cutting the prefabricated bulkhead panel to shape being careful to preserve desired positions of vertical stiffeners, etc., positioning bulkhead inside the hull, and then laying up 3/16" x 2½" x 2½" plastic bounding angles to the hull. Bulkheads #25 and #32 were installed first so that girder installation and fuel compartment construction could be started while other bulkhead and girder installations were being made elsewhere in the hull.

Girder installation involved similar operations, including cutting to template, taking care to maintain desired locations of vertical girder stiffeners, routing girder sections at bottom and ends for filling with vermiculite-resin bolt-bearing fillers, positioning sections in the hull, and laying up plastic girder-foot angles to the hull. Aluminum angles (1/4" x 2" x 2") were bolted to the ends of the girder sections at the bulkheads, and bolts through the bulkheads connecting these angles provided some girder continuity through bulkheads.

The secondary bonds of all bulkhead bounding angles and girder-foot angles were augmented with mechanical fastenings, either bolts or self-tapping screws set in epoxy resin.

Transverse floors of 3/16" laminate were provided generally at two-foot intervals and were secured either directly to vertical girder stiffeners or to girder webs by means of 3/16" x 1" – 1/4" aluminum angles and self-tapping screws.

Connection to the hull was by means of plastic bounding angles cured in place and with secondary bonds augmented by self-tapping screws. To further stiffen the floors and to provide a means for securing the walking flats, 3/16" x 1–1/4" aluminum angles were secured along the upper edges of the floors.

Fuel cell compartments were designed to follow closely the structure of the wood MSB-5 class and used wood and aluminum throughout except for studs, longitudinal bulkheads, and flats, which were of reinforced plastic construction.

When installation of internal members including the fuel cell compartments had advanced sufficiently, preparations were made for the fabrication of the main and raised decks.

The first step was the location and installation of the permanent deck beams and longitudinals with their bracketing. Next, the temporary intermediate wood beams were installed to further support the 1/2" plywood deck mold platen. The plywood was so secured to the beams, both permanent and temporary, that fastenings could be removed from the underside and the mold platen freed for removal.

All joints in the plywood mold platen were carefully filled to assure maximum integrity for later vacuum-bagging operations.

Following filling and smoothing, the plywood surface was sealed by the application of two coats of a clear lacquer. After this had dried thoroughly, the usual wax and parting film were applied.

Next, the bottom skin of the main deck consisting of 3 plies of 1044 cloth, butted, was laid up. The deck honeycomb with accompanying mat was bagged in place, leaving voids in way of the bulkheads to assist in the bagging operations. Deck honeycomb used was Douglas Aircraft's 125# paper honeycomb, which had been approved by BuShips for use throughout the second boat.

When all honeycomb was in place, the upper bonding layer was applied, using vacuum bags to assure intimate contact. This consisted of 1 ply of impregnated mat and 1 ply of cloth as a carrier. Several small areas of the bonding layer failed to cure properly and had to be removed and replaced.

This was one difficulty that appeared at intervals throughout the entire project and which constituted almost all of the repair work required. The reason for this behavior was never determined, but it was thought that either there was something in the honeycomb that had an inhibiting effect, or resin mixing had been inadequate.

The application of the bonding layer served to complete the sandwich and to stiffen the panel enough to permit removal of the mold platen. After curing, the partially completed deck was wedged up all around and alternate plywood panels taken out until all plywood had been removed. The temporary wooden beams were left in place until the deck was complete, provisions having been made to allow dismantling and removal through hatches.

Self-tapping screws were run through the bottom skin of the deck layup into the permanent beams, longitudinals, and sheer filler of the hull. Holes through the upper bonding layer and honeycomb, bored to allow running these fastenings, were plugged with dowels of C.C.A. set in resin. The deck edges were reinforced with vermiculite-resin and the transverse voids filled with the same material. The honeycomb edges around hatch openings were routed back about 1/2" and impregnated rovings laid up in the resulting voids to minimize the effects of stress concentrations at corners of the openings. Deck edges were ground flush with the sides of the hull and flush with beams and headers at hatch openings.

The remaining four plies of the butted top skin were then laid up. The laminate was carried around onto the topsides of the hull to act as an outer closing angle and also down inside hatch openings and bonded to the beams and headers. The forward end of the layup was flanged upward and bonded to the aft face of bulkhead #13.

The raised deck was laid up in an identical manner after joiner work forward had progressed to a sufficient extent.

Additional reinforcing angles were laid up along the sheer lines and around the edges of the raised deck.

Temporary wood beams were then removed and the bulkhead bounding angles to the deck installed. Similar angles were laid up connecting the inside of the hull and the underside of the deck.

While deck fabrication was underway, installation of struts, shafting, lifting pads, skegs and V-struts, fresh water coolers, and overboards for sanitary and bilge systems was accomplished. Most through-bolts for lifting pads, struts, and skegs ran through vermiculite-resin fillers provided in the hull and keel. Plastic shims were provided to assist fitting of strut and skeg palms to the contour of the hull.

The standard procedure for the installation of other through-hull fittings was as follows: 1. Locate and bore the hole (or holes); 2. Route back the honeycomb about 3/4" all around the hole; and 3. Fill the void with vermiculite-resin and allow to cure. The fitting was then installed and set in 3M's EC-1159, a self-curing synthetic rubber based sealer, which was also used with all through-bolts.

The installation of the wooden pilothouse, the engines, the electrical system, and other equipment followed ordinary wood boat procedures and will not be further detailed here.

Bulwarks were made up of 1/2" thick sandwich panels using paper honeycomb and having 3 plies of cloth each face. In bulwarks fabrication mat was not used as a bonding layer.

Flush deck hatches were of reinforced plastic construction (see Photograph 169) while other hatches and coamings were of wood. This was more or less in harmony with the use of a wood pilothouse.

To secure the desired flotation, trim, and stability characteristics of the vessel, it was necessary to provide a considerable amount of fixed ballast to equal the weight of all the specialized minesweeping and electronic equipment which would not now be carried aboard.

Disposition of this ballast was determined by several considerations: First, the preservation of essentially the same stability characteristics as found in the MSB-5 class; second, the available space; and third, the importance of keeping access to the hull laminates unobstructed as an aid to visual inspection during evaluation. An additional requirement was that about 15,000 lbs. of ballast be quickly removable to lighten ship for hoisting.

Concrete was selected as the most suitable material considering magnetic qualities, ease of handling, and cost.

The solution involved the casting of a pair of large concrete blocks on deck amidships with provisions for their removal and subsequent replacement. The addition of a pair of smaller blocks on deck aft and two units below in the ex-generator compartment served to bring about the desired weight distribution, both horizontal and vertical. See photographs 208 and 224.

After all major installations were complete, the vessel and its cradle were rolled out of the shop and placed in position accessible to two locomotive cranes. Launching was accomplished by means of these cranes and the hoisting gear furnished for and built into the vessel. The flotation line was noted and the displacement and longitudinal center of gravity computed using Bon Jean's curves as a check on both weight and weight distribution. Computed launching displacement was 44618 lbs. with LCG 3.41 feet aft of the midship perpendicular. Based on the weight estimate, launching displacement had been estimated at 44611 lbs. and the C.G. at 4.01 ft. aft. The fixed ballast on deck was cast in three steps and displacements were computed after each step was a check on the ballast computations. The smaller blocks aft were cast as a part of the third step and the hold ballast was installed last. At the completion of ballast installation, displacement and trim were considered satisfactory pending final placement of the ships outfit.

Both builder's and preliminary acceptance trials were run in choppy seas typical of relatively shallow and unprotected waters.

While the vessel plunged to the extent that spray was thrown well over the top of the pilothouse, the forward hull panels below the waterline showed no evidence of any excessive flexibility. Actually, local pounding shocks were impossible to detect, the entire hull exhibiting a rigidity far above expectations.

Whether or not continued stresses and strains of rough water operation will induce fatigue failures in the hull will be determined only through extensive evaluation, but judging by the absence during trials of any evidence of local distress, it appears that the monolithic hull will stand up under any reasonable treatment.

The delivery trip from Bay City, Michigan to Panama City, Florida via Lakes Huron and Michigan, the Mississippi River, and the Intercoastal Waterway, gave no further evidence of any inherent hull weakness.