2021_silverbullet_48
This boat is hull number 3 - the previous two hulls were 1:10 scale models laser cut TIG welded aluminum - 2020 1:10 scale model
The first two were built in a highrise building on the 17th floor and I was limited by space and what size I could fit into the elevator.
The basic hull specs are:
Length: 4.8m (15 feet 9 inches)
Width: 1.3m (4 feet 3 inches)
Height: 1.26m (4 feet 2 inches)
Weight: 600lbs (272kg)
Length to beam: 3.7:1
This is the first full size boat I've built. It's big enough to sleep one adult and one or two children or maybe two adults but it would be a little bit tight.
Think of a giant canoe that you can sleep in.
My expectation is to be able to do photography and wildlife watching on a river or lake or short expeditions with my kids.
Propulsion is will be electric thrusters in form of pod drive or trawling motor.
Boat will be based on a farm with frequent high wind storms (Columbia Gorge). Sometimes the gusts get up to 50 or 60 knots. Having something that's heavy and tough and aerodynamic is desirable for strong wind - you can point it into the wind. For a long time I've wanted to have a vessel that I can put out in the middle of a field and watch extreme weather or spend the night with the kids. We have a trailerable sailboat that we do this sometimes but after several trips I have a list of things I wish were different and I've wanted to design my own for a while. This build is an iteration (hull number 3) in that journey. I hope to build a larger hull in the future but realize I have a lot to learn and am approaching this incrementally with larger and larger iterations.
One of the design goals is dual purpose on land and on water use - we can use it as a camper on the land and also go onto the river and do short overnight trips in an enclosed vessel on the water.
The longitudinal structural members are aligned perpendicular to the side plating with the exception of the bottom (keel) which is aligned to a vertical plane. This is not exactly perpendicular to the bottom plates but I would like to have an axle that wheels can be attached to as part of the internal structure in a future revision and thought this would be strongest.
Hull will have insulation foam in the interior in the top half for safety floatation.
It's a displacement hull - not designed for high speed planing. There is 20 degrees deadrise on the bottom chines (10 degrees each side). The center of gravity is low.
The design (hull shape and structural members) were generated automatically based on an XML definition file.
The way it constructs the model all the chines are developable. Rapid construction is one of the design goals. It's self jigging - the frame is the jig.
bpyhullgen is not a fully polished product but an opensource project that is under development. You are welcome to try it and modify it or use it as you wish.
This is the first build using the new XML hull definition feature. There is a basic user interface bpyhullgen that allows you alter and modify hull parameters - the underlaying data can be stored and retrieved using XML. This is the XML definition file for this hull:
<?xml version="1.0" ?>
<hull>
<size height="3.600" length="4.800" width="1.300"/>
<materials slicer_overcut_ratio="1.100" slot_gap="0.000" structural_thickness="0.005"/>
<generate bulkheads="1" hide_hull="0" keels="1" longitudinals="1"/>
<bulkheads>
<bulkhead floor_height="-0.400" station="0.000" thickness="0.005" watertight="0"/>
<bulkhead floor_height="-0.200" station="-1.800" thickness="0.005" watertight="0"/>
<bulkhead floor_height="-0.400" station="1.800" thickness="0.005" watertight="0"/>
<bulkhead floor_height="-0.400" station="1.000" thickness="0.005" watertight="0"/>
<bulkhead floor_height="-0.400" station="-1.000" thickness="0.005" watertight="0"/>
</bulkheads>
<keels>
<keel lateral_offset="0.000" station_end="1.200" station_start="-1.200" top_height="-0.400"/>
<keel lateral_offset="0.000" station_end="-1.200" station_start="-1.700" top_height="-0.300"/>
<keel lateral_offset="0.000" station_end="-1.700" station_start="-2.400" top_height="-0.200"/>
<keel lateral_offset="0.000" station_end="1.600" station_start="1.200" top_height="-0.300"/>
<keel lateral_offset="0.000" station_end="2.400" station_start="1.600" top_height="-0.200"/>
<keel lateral_offset="0.300" station_end="1.050" station_start="-1.050" top_height="-0.400"/>
<keel lateral_offset="-0.300" station_end="1.050" station_start="-1.050" top_height="-0.400"/>
<keel lateral_offset="0.200" station_end="1.600" station_start="1.200" top_height="-0.300"/>
<keel lateral_offset="0.200" station_end="1.850" station_start="1.600" top_height="-0.200"/>
<keel lateral_offset="0.200" station_end="1.200" station_start="0.950" top_height="-0.400"/>
<keel lateral_offset="-0.200" station_end="1.600" station_start="1.200" top_height="-0.300"/>
<keel lateral_offset="-0.200" station_end="1.850" station_start="1.600" top_height="-0.200"/>
<keel lateral_offset="-0.200" station_end="1.200" station_start="0.950" top_height="-0.400"/>
<keel lateral_offset="-0.200" station_end="-0.950" station_start="-1.200" top_height="-0.400"/>
<keel lateral_offset="-0.200" station_end="-1.200" station_start="-1.700" top_height="-0.300"/>
<keel lateral_offset="-0.200" station_end="-1.700" station_start="-1.850" top_height="-0.200"/>
<keel lateral_offset="0.200" station_end="-0.950" station_start="-1.200" top_height="-0.400"/>
<keel lateral_offset="0.200" station_end="-1.200" station_start="-1.700" top_height="-0.300"/>
<keel lateral_offset="0.200" station_end="-1.700" station_start="-1.850" top_height="-0.200"/>
</keels>
<modshapes>
<modshape mod_mode="intersect" mod_shape="trapezoid" mod_type="cube" name="modshape" symmetrical="1">
<location x="2.300" y="0.000" z="0.000"/>
<rotation x="0.000" y="0.000" z="0.000"/>
<size x="0.400" y="-0.800" z="0.005"/>
<deform p1="0.000" p2="0.000" p3="0.000"/>
</modshape>
</modshapes>
<chines>
<chine name="side" symmetrical="1">
<curve extrude_width="10.800" height="3.600" length="5.200" width="0.650"/>
<offset x="0.000" y="0.000" z="0.000"/>
<rotation x="0.000" y="0.000" z="0.000"/>
<asymmetry a0="0.400" a1="0.000"/>
<longitudinals>
<longitudinal width="-0.150" x_max="2.200" x_min="-2.400" z_offset="-0.100"/>
</longitudinals>
</chine>
<chine name="bottom" symmetrical="1">
<curve extrude_width="10.800" height="3.600" length="5.200" width="0.600"/>
<offset x="0.000" y="0.000" z="0.000"/>
<rotation x="80.000" y="0.000" z="0.000"/>
<asymmetry a0="0.400" a1="0.200"/>
<longitudinals/>
</chine>
<chine name="top" symmetrical="0">
<curve extrude_width="10.800" height="3.600" length="5.400" width="0.650"/>
<offset x="0.000" y="0.000" z="0.000"/>
<rotation x="-90.000" y="0.000" z="0.000"/>
<asymmetry a0="0.400" a1="0.100"/>
<longitudinals/>
</chine>
<chine name="upper_side" symmetrical="1">
<curve extrude_width="10.800" height="3.600" length="5.400" width="0.650"/>
<offset x="0.000" y="0.000" z="0.000"/>
<rotation x="-37.000" y="0.000" z="0.000"/>
<asymmetry a0="0.400" a1="0.000"/>
<longitudinals>
<longitudinal width="-0.100" x_max="2.200" x_min="-2.400" z_offset="0.070"/>
</longitudinals>
</chine>
</chines>
</hull>
You can download the XML File
Inside looking forward showing structural members. There are three keels defined with different lateral offsets. The center keel has lateral offset zero. It would be stronger to have a continuous curved keel instead of a break and shift near the end bulkheads but I decided to break it up so the keel had parallel equal spacing in the tips (bow and stern). I could have designed it as a curved keel so it was one long continuous keel that curved along the longitudinal axis but it would make the interior space hard to utilize - think trying to store batteries containers and other square objects in those spaces if the width is constantly changing. Its a little less strong this way but more functional. Extending the keels parallel all the way to the ends would make it difficult to weld in the tight corner spaces on the tips so this is the design decision I made for this boat.
Side profile with model of person to show scale. The intention is to add a small low profile pilothouse in the future high enough to raise your head up and view around you around while remaining protected from the elements. I decided to construct the hull without any openings and cut them out later. The previous models I built hull #1 and #2 I cut out hatches and pilot house holes in the design files - The welds were much more prone to heat warping on the sides around the edge of the pilot house because it was a thin narrow strip and bending the plates was less uniform. The best solution would be to have tabs cut so it's a continuous sheet with dashed edges that can be cut out later after the welding but this is not designed into the software yet. A future software revision should automate this but for this phase of the development cycle this is what it does for now.
The design is just big enough to sit up in to read a book or work on a computer.
Cutaway profile view looking down from top side.
The keel (floor) structural members are level on top to allow placement of a bed or flat surface for sitting on. The floor steps up slightly near the ends like a shelf - this allows a flat floor in the very extreme ends of the hull.
Aft view looking forward showing transom. The hull is not exactly a double ender - the aft end is slightly wider than the bow. This is done by creating non Symmetrical chine curves in the bpyhullgen hull definition.
The top of the structural members for the keel are stepped to allow easier interior placement of flat surfaces on floor instead of curved floor. This feature is determined by the floor height of the keel structural element definition in bpyhullgen.
I sliced the bulkheads into halves so they would nest easier an optimize material usage to minimize waste. The first step of the build process is to weld the halves of the bulkheads together into a single piece.
Two bulkheads welded together - side plating behind it.
Bulkhead clamped to table for welding. I wasn't sure how much heat I could put into it so I welded edges to half center instead of a continuous weld across the joint. The material is 3/16 thick aluminum (4.8mm) looking back I probably could have welded all the way across the seam but was being extra cautions as this was the start of the build and I didn't want to have any heat induced warping.
The previous two hulls I built out of aluminum were TIG welded. This is my first time using MIG process with spool gun. I used ESAB Rebel EMP 215ic with Tweco 160 Amp Spool Gun. 0.35" 4043 filler wire. 100% argon, 30cfh.
Bulkheads welded together.
The material is 3/16 inch (4.8mm) 5052 Aluminum. It was CNC waterjet cut. I know many builders prefer 508x series for boats because it's tougher but it's more expensive and supposedly a little bit harder to weld. For this size boat I think it's plenty strong.
So my idea was to make a self jigging frame. It all fit together perfectly and I used clamps to hold it from moving around for the welding. There is no jig or additional braces - just the frame.
The order of assembly was sides first then upper sides. I used clamps to hold the sides onto the bulkheads once they were aligned prior to welding.
View from inside showing side and upper side plates tack welded. I got a little bit of black soot on the beginning and end of welds. I spent almost a day and a half trying to debug this problem with test welds on scrap. Some of the things I tried to solve this problem:
- Check for air leaks in lines (soapy water)
- Disassemble and reassemble spool gun (two times)
- Alternate argon gas bottle from different supplier
- Different filler wire size
- All different spool gun angles
- Different flow rates
Clamp holding aft side plates to frame prior to tack welding. Bricks were nice and stable to prevent the clamp from sliding off the back.
Top plate tack welded on. If you look very carefully at the top chine on the side plate the edge is not 100% perfect. This is because I exported the DXF file as a series of line segments. Is it perfect? No. Does it work? Yes. For a long curve this big you can tell if you look closely and know what to look for but it fit up really nicely and welded great. After welding you can't tell.
Some of the side plating we were a bit hurried with the alignment - the top plate is slightly too far aft. There is a function on the CNC waterjet machine and also most CNC laser machines to do a low power etching pass to add some alignment marks. I don't have alignment marking capability added to bpyhullgen yet but this highlights the importance of that feature. I'm working on adding this feature. On my next build I'll spend more time and care on alignment. It's not perfect but it welded up nicely and I'm quite confident it will be watertight.
Hull turned upside down for welding inside. I welded inside seams before closing it up while good ventilation was available.
I'm wearing a Miller LPR-100 respirator underneath the welding helmet. It worked quite good but I could smell a bit of argon.
Doing full welds on the keel to bulkhead seams.
Trying to do as much welding as possible on the interior seams before closing it up. In this view you can see the curve on the side... You can't really tell it's a series of line segments.
For the nose I tried something different - angles clamped to the side plates to give the long red clamp something to grab onto. It worked quite well... Some of the stuff on the ends I did with a tig torch.
View of nose tack welded together with big clamp removed. You can see some stitch welding on the side chines...
View of interior welds. I could have done a better job on the corners. I know from some boat building books people sometimes cut out the corners of the bulkheads so you can do continuous welds on the interior chine seams. I was worried because of the small size of the boat and the size of the bulkheads there wasn't enough material to cut away to give me full access to the inside seams. It is what it is... any suggestions welcome.
Nose welded up a bit more and hook plates installed... These plates could be thicker but I used the same sheet of material as the rest of the build... I can always reenforce them later. I guess you could argue that if the force exceeds these eyehooks it will bend the eyehook plates rather than bend the hull. For vertical loads it's plenty. For side loading it is a weak point. They worked great for holding onto with your hands and we could roll the boat over easily with two people - one person on each end of the boat.
I rolled the hull around to make inside welding easier... down or side is easier than upside down.
Looking at the bottom of the hull... There are some longitudinal supports missing, I left them out so it's easier to get in and out for welding. They were added later. You can see all the extra weight on the bottom compared to top.
View of chine edge fitup. Not perfect but worked well and welded nicely.
Fitting up bottom plates - getting ready to weld bottom.
I used a battery powered circular saw for back chipping so seams would get continuous good penetration and be watertight. Sometimes I would have a bead at the start end of my weld line (cold start). The welder I have doesn't have hot start feature or I've heard of some other features to prevent this. I always got a bump at the beginning of weld lines and I don't have a variable feed control on the spool gun. You set it on the machine and what you set is what you get.
The majority of the structural members are placed low to provide a low center of gravity. Previous two hull designs were very similar and demonstrated good static and dynamic stability in real world tests.
No extra ballast was used in these tests - but additional ballast could lower the center of gravity even further if desired.
In this bathtub test - when I pushed it down further so water came in through the window it did sink quickly and was a bit scary how fast it went under. It hit the bottom with a clunk. This got my attention enough to have flotation foam on the list of things to do before it hits the water.
5cm foam installed on the top half of the interior would provide some insulation as well as safety in the event of an accidental rollover event.
This is what 5cm foam would look like. It's rendered on the outside so you can see the approximate location and size in relation to the hull but would be installed on the inside. The 5cm foam in these renderings has a total volume of 0.29 cubic meters - enough to displace water slightly more than the dry weight of the hull. In the event of a rollover it should float on the surface with the hatches facing upright so it's not a deathtrap. You might get wet but won't be trapped inside.
Cutouts in the top around the approximate location of hatch areas.
A displacement simulation using bpyhullsim shows the waterline location with various loading conditions between dry empty weight and 2000kg. 2000kg is way beyond any expected operating load - I just wanted to see what it looked like and where the waterline would be at 2000kg displacement. I anticipate the payload to be less than 250kg for this build under normal operation.
If you want you can download the Blend file
Or the DXF file
