Home-  Sailboat-  Submarines-  ROVs-  Metal Working-  Other Stuff -  About Us

Why Build a Sub
   Submarine 101
   Wet Sub
   Planing Wet Sub
   Dry Ambient Sub
   Submarine Yacht
Design Evolution
Cutting Aluminum
Building the Hull
Kort Nozzles
Battery & Chargers
Engine & Jet Drive
   Engine Tune-Up
   Hull Cooler
   Wet Manifold
   Engine Mounts
   Engine Box
   Jet Pump Valve
Helm Controls
Ballast Sled
Landing Gear
Trim Tanks
Wiring Harnes
Tow Truck
Rules of Thumb



Design Evolution

Design Philosophy

Too many of us are afraid to do something because it's dangerous, or we will likely fail. But what inventions and discoveries are we missing because we teach our children to never take a risk? Is there no place in life for danger and adventure? Isn't it possible to lose your life because you are so worried about protecting it?

If design by experimentation and courage scares you then fine, go on and copy someone else's work, and make sure your children wear their bike helmets. I am certainly not the brightest person walking around, and may very well end up with a Darwin award, but I will be adventurous, learn, invent, and enjoy my life. And damn the obstacles erected by everyone who lack the balls or imagination to do build something from scratch.

Design Approach

My personal preference is to use an experimentation approach to design. If you have two or three ideas, then start following up on all of them. Even build them if needed and then refine and abandon as needed. While it is not an approach for everyone, it has merits. It assures full ownership for success and failure, maintains the possibility for discovering something new, and it makes the endeavor educational, adventurous, and fun.

Being an armature home builder allows you the freedom to design without being encumbered by preconceived notions or industry standards.  And doing it without investors or millions of dollars in a research and development budget insures that you will pursue every concept that enters your head.

You are never on your own. The internet and books are full information and working models, and if you want to make no real decisions you can also find plans and plenty of people that will make the decisions for you.  Be fore warned that there are also many people that will attempt to frighten you into abandoning unconventional ideas. Listen and evaluate their suggestions but keep in mind that there are lots of professional engineers without a creative brain cell to their name.

Front View
Front View
Divers Hatches Open
Divers hatches open. Latter changed the design to have two forward support legs.
Top View
Top view with top hatch open. The top hatch as sense been changed to open inward so that it seals with natural outward  ambient pressure.
Aft View
Aft view. Designed changed to close the rear view port opening and from a single large forward trim tank.  This simplifies the trim tank vent and added more forward buoyancy to offset buoyancy lost as the cabin breaks the surface.
Design before the tail was added
Originally the engine was going to draw air from a pipe that led to the top for the forward trim tank, but a tail fin was used instead as it required significantly less water to be pump out before I was used upon surfacing.
Framing with Ballast Sled
Framing with ballast sled
Plan with dimensions
Plan with dimensions


Form Follows Function

Somewhere along the way after we decided to build a submarine paid a visit to the "Mel Fisher Museum" in Key West, Florida.  There we saw the gold, silver and other artifacts raised from the wreak of the Atocha.  Already advancements in side scan sonar, magnetometers, and underwater ROVs and AUVs  had all but rendered submarines useless for anything other than military applications.  But clear shallow waters of the Florida coast with the numerous coral heads within a few miles of shore presented a niche where a submarine would excel.  Side scan sonar's are not well suited for shallow waters, and the frequent storms through the area has disperse debris trails from wounded vessels over miles of bottom. The targets that are exposed are so small that magnetometers easily miss them. And like most costal areas these waters are filled with and array of trash targets from shipping and pleasure boats. 

With a function of the submarine clearly in mind we decided on the traits that would best serve the purpose and came up with the list below.

Desirable Traits:

    Excellent downward visibility
    Provide relative comfort and shelter for extended bottom time
    Allow the crew of two to participate equally
    Let the crew easily exit the sub for hands on inspections
    Be fast enough to quickly reach search areas when the weather is good

I started out with a very crude plywood and fiberglass dry ambient design and I learned all of the basic lessons about water pressure, ballast. This evolved into a submersible fiberglass skiff that would use an outboard while on the surface.  The complexity, fear of failure, and cost pushed me back to a simpler wet sub design and finally after reading some books on boat building and metal boats I settled on a 6' beam, 20' long aluminum V- hull jet drive boat with a ton of movable ballast in bottom of the hull and able to run on the surface at about 20 mph. Air for submerged operation could be supplied by on-board high pressure tanks or by gas powered compressor or hookah system towed behind on the surface. Below is H8 or Hookah submarine, version #8.  This was the starting point of the build and additional modifications would be make as the process moved forward.

Design prior to starting the build.
To dive, the engine is shutdown, the ballast moved forward and the hull flooded, and then thrust will is provided by electric trolling motors. The major design advantage of an ambient sub is that it circumvents the problems with the hull being crushed by the surrounding water.  As the sub descends,  pressurized air fills the cabin which prevents the water from entering. Just like a glass turned upside down in a sink filled with water.  The down side is that the occupants are exposed to this pressurized environment and therefore are subject to decompression just like a scuba diver so rapid accents can kill you as well as going too deep or running out of air.

The CAD designs above and left represent the level of planning that was done prior to starting the build and the photo captions note the various changes that were made later on. Very few details were set in stone before the build process began. Things like the ballast tanks, helm control, engine air supply, retractable landing gear and the ballast sled were only loosely designed. Each had multiple revisions and often experiments, and each would naturally build on lessons learned when building other components.

The first design was started in September of 2002.  Its been a good 10 months of learning how subs work and now its time to start learning how to weld aluminum.


Major Changes Made After Starting the Build

As the building of H8, what is now "Seeker" progressed the design continued to evolve.

Below is a list of the changes that occurred as we went along in the build.

  • The hatch was designed to open outward after testing in a card board mock up, but now that I can test with the real cabin, it is easy to see that opening it inward is not really a problem. And opening it inward is much easier to seal and safer, since the pressure on the hatch when submerged naturally wants to push the hatch outward.  This is a trait of ambient submarines and opposite of their 1 atm (atmosphere) cousins.
  • The rear view window was removed.  This eliminate complexity and allow for larger soft ballast tanks behind the cabin.  One vent pipe is a lot easier than two, and it allows for a long vent pipe that only bends at one joint.
  • Frames in the hull were made smaller and moved to 10 inch centers in order to better stiffen the hull and to allow the ballast sled to ride as low as possible.  The outside rails for the ballast sled were moved up on top of the side longitudinal frames so the sled can be wider.
  • A large tail fin was added to serve as a snorkel for the engine when partially submerged. It replaces a larger air duct in that would have run from the top of the forward cabin.  The change eliminates a large volume of water that would need to be pumped from the intake or drawn into the engine compartment when surfacing.
  • There will be 4 support legs or landing gear and not 3. A center support leg in the front conflicted too much with the framing, and we really don't need them to extend more than 15 inches.  The forth leg also provides additional load caring ability.

The Evolution of the H8 Design - H1 through H7


The H1 is a low cost, 2-person, dry ambient submersible designed for shallow lakes with 15 foot or greater visibility. Primary Air supply is provided from a towed hookah system. The cabin is a simple rectangular box built from fiberglass covered plywood with large tempered glass view ports.  There are two hatches; one in the bottom that allows the divers to inter and exit while submerged, and another in the back on the cabin that allows for dry entry on the surface. Beneath the cabin are two pontoons.  Each pontoon is built from 6, 55 gallon steel barrels with 3 barrels connected end to end with another 3 barrels underneath. 


H1 - Click to Enlarge

When operational, the lower barrels would be about 2/3 full of sand and gravel weighing 1500 pounds. The upper barrels act as main ballast tanks that can be blown free of water or flooded to submerge. Using sand as ballast allows the sand to be dumped when the dive operation is done. Dumping the sand and blowing the water from the lower barrels would also greatly reduce the draft and make getting the sub on a trailer much easier.  When preparing for dive operations sand would be added to the lower barrels. Of course sand would need to be available at the dive site, but hookah system could supply air to a lift pipe that could dredge up sand and gravel and deposit it into the barrels.

The draw back it that sand is actually not a very good hard ballast.  Lead weights about 700 pounds per cubic foot, and sand is only about 100 pounds per cubic foot.  Fresh water is 62.4 pounds per cubic foot so the added weight when displacing water with sand is only (100 - 62.4) 37.6 pounds per cubic foot.  Adding 1500 pounds of sand requires 15 cubic feet and only adds (15 * 37.6) equates to only 564 pounds of negative ballast once it is in the water.  The wait of sand actually varies depending on the type of sand.  Finer sand compacts together better and displaces more water. Combining sand and gravel is better as rock is more dense and the sand will fill the voids between the rocks.  However have amble storage for the sand is the practical solution, but the six barrels dedicated to sand storage do take block a significant amount of viewable area.  On the plus side 1500 pounds is a lot of weight and is the difference between towing with a mid-size and a full-size pickup.


Freeboard is the portion of the submarine that is exposed above the waterline and more is better as additional freeboard prevents waves from flooding the sub when the hatch is open.  The cabin in this design represents about 40 cubic feet of displaced water and with water weighing 62.4 pound per cubic foot that is just under 2,500 pounds. The cabin holds about 800 pounds of occupants and the batteries so the buoyant force of the cabin is reduced to (2,500 - 800) 1,700 pounds.  The weight of the cabin, steel barrels, steel frame, thrusters comes to just over 1,700 pounds so without any sand ballast the sub is naturally buoyant.  It would be possible to not use any barrels at all provided about 100 of weight was added to the frame, but the sub would have no freeboard.  It is important to keep in mind that positive buoyant force is provided by the water that is being displaced.  Once the top to the cabin reaches the surface it starts to rise from the water displaces less and less water as it rises.  So if it is nearly naturally buoyant the it would not rise from the water at all.  It is the upper set of barrels that provided the buoyancy needed to lift the cabin from the water.  When surfacing an valve would be opened that allows air from the hookah air compressor to blow air into the barrels.  The water would be forced from holes in the bottom of the barrels and the 6 upper barrels would add 2,800 pounds of positive force. Most 55 gallon barrels actually hold  56.5 gallons, and 1 gallon of water is 8.35 pound so: 56.5 gal * 8.35 lbs/gal * 6 barrels equals 2830.65 pounds.  The cabin, occupants and batteries only weight about 1,200 pounds so they will be completely out of the water.  A quick guess can be made about the actual waterline.  As the deck upper frame and air tanks come out of the water the about 200 pounds is added on bring the total weight about the water line to around 1400 pounds.  That is just under half of the displacement of the 6 upper barrels so the 6 barrels will remain half way submerged.

As shown above, this design could work without the need for the lower 6 barrels or any sand ballast, provided the weight of the existing frame was increased by 100 pounds to compensate for the weight of the 6 lower barrels.  However these barrels provide the needed clearance off the bottom to allow for the occupants to easily exit from the lower hatch when the sub is resting on the bottom. 

The lower 6 barrels also contain 1500 pound of hard ballast in the from of sand and gravel, but as discussed above, that addition of 1500 pounds of sand only adds 564 pounds of negative ballast once in the water.  And it would be possible to operate without the sand entirely.  The reason for caring the sand ballast is two fold.  First the air supply fail or a leak develop the sand could be dumped through gates in the bottom of the lower barrels and the sub could rise to the surface.  Secondly when the occupants exit the sub while the sub is on the bottom water will rush in through the bottom hatch equal to the volume of the occupant that just exited.  The added weight of the water is more than enough to compensate for the diver that left, but the air supply to the cabin must be shut off or the vent location raised in order to keep the water in the cabin.  If the air supply system were allowed to force the water from the cabin the sub would become positive buoyant.  If that happened the divers might turn to find the submarine leaving the bottom without them.  The addition of 564 pounds of negative ballast would prevent this from happening.

The biggest drawback for the H1 is it's speed; both on the surface and submerged.  The cabin and pontoons add up to almost 30 cubic feet of frontal area.  In low visibility speed is dangerous but the H1 would be luck to make half a knot submerged, which is only a crawl.


H2 - Side View
H2 - Side View
H2 - Front View
H2 - Front View
H2 - Top View
H2 - Top View
H2 - Top Font Quarter View
H2 - Top Font Quarter View


The H2 is a 2 person dry ambient with air supplied from a towed hookah compressor that improves on the H1. The seating position lowered the cabin height and sloped the forward view port reducing the cabin's drag and its displacement volume from the 40 cubic feet of H1 down to 34 cubic feet. The reduced drag and frontal cross-section should also result in a 2 mph submerged speed.

The hull material is fiberglass over plywood for the entire hull. A top hatch that provides the ability to enter and exit without getting wet. Total weight approximately 2000 pounds, but using sand for hard ballast lightens the load during road trips by 1,300 pounds.

The aft end of the hull is two pontoons which provide yaw and pitch stability, a location for scuba tanks for emergency air, and deck for storing the hookah system while surfaced.

The cabin is set as far forward as possible and the foot rest will also be a view port allowing the crew to see the bottom between there knees.  Control leavers in the cabin that allow either crew to rotate the thrusters 180 degrees as well as release the sand ballast from the two compartments below the cabin.  Thruster direction and speed will be controlled by joystick which operates relays. 

While the H2 improves on the H1's speed and size, it's design adds complications too. The primary drawback is the diver hatch in the bottom of the cabin.  This hatch opens into a main ballast compartment that has a second hatch which opens forward.  The problem is that the main ballast compartment must be kept water tight in order to surface so the forward hatch must make a good seal.

The top hatch is also designed to open outward.  This is normal for 1 ATM submarine because the external water pressure pushes against the hatch to keep it closed.  In an ambient design however the air inside the cabin is like a bubble of trapped air that is pushing upward against the ceiling of the cabin in an effort to rise to the surface.  The force on the hatch is therefore outward so keeping it sealed if it opens outward requires a stiff hatch with sufficient dogs to latch it in place.

The sloped surfaces of the design will also act as dive planes as the sub moves through the water causing a pitch and downward push that must be counteracted by the pitch of the thrusters and the righting force created by the hard ballast and the displacement of the cabin.  Moving the thrusters forward would help but they will not be able to pivot on single shaft that passes from one side to the other due to conflict with the diver hatches. Having the hatches pass through the main ballast tank also eliminates some of the downward visibility.


The H3 design kept the same cabin as the H2 but replaced the hull with 20 inch diameter fiberglass pipe for pontoons and placing lead ballast and the batteries on sleds inside the pontoons so it could be moved aft to trim the boat when on the surface and reduce the ballast requirements directly under the cabin.


H4 - Side View
H4 - Side View
H4 - Front View
H4 - Front View
H4 - Top Front Quarter
H4 - Top Front Quarter
Intruder - Wet sub, powered on the surface by outboard.
Intruder - Outboard powered wet sub.
Intruder - Aft View
Intruder - Aft View


The H4 design moved all of the hard ballast onto a sled that behind the cabin. This opened up the floor of the cabin for greatly improved downward visibility. 

An outboard motor with a snorkel was added to power the sub when on or just below the surface.  The outboard's housing is supplied with exhaust air from the cabin so that it, like the cabin is maintains ambient pressure with the surrounding water. The "Intruder" is a wet sub design by Herve Jauvert of Seahorse Submarine,, that uses a converted outboard motor for surface propulsion. See photos below left.

The outboard's housing and snorkel provides positive buoyancy like the cabin and the hard ballast sled maintains a balance between the outboard and the cabin.  You can think of the ballast sled as a center (fulcrum) of a teeter totter and the displacement of the cabin as a fat kid sitting close to the fulcrum with a small kid far away on the other side. When submerged the sled is brought forward to compensate for the increase buoyancy of the cabin as it descends into the water.  When surfacing the sled is moved aft to compensate for lost buoyancy of the cabin as it rises out of the water, and give the boat an upward pitch that will help the outboard lift it up onto a plane.  While a flat hull is the easiest hull form to get up on a plane, it is also the bumpiest hull shape in anything but very flat water. 

Retractable skids can be deployed when the sub comes to rest on the bottom.  These allow clearance for the cabins bottom hatch to be opened.  Provided sufficient HP was utilized it would be possible for this design use hydrofoils on fixed skids.

The downward push cause by the dive plane like profile of the cabin is reduced in this design by an acrylic panel that extends the footrest upward.  The H4's design changed the front of the cabin too.

The forward view ports were reduced to small, almost vertical view ports directly in front of the occupants.  The remainder of the forward deck is solid and aligned with the occupants line of site to further increases forward visibility.  This profile also reduces the amount of light that enters the cabin from the surface.  Keeping the cabin dark is essential to maintaining good downward visibility.  If too much light comes in thought upper view ports it will only reflect off the lower view port. 

The new profile for the front deck and view ports also further reduces the displacement of the cabin.  The H3's cabin was 34 cubic feet and the H4 is 25.75 cubic feet.

While this design could be constructed from fiberglass over plywood but plywood will not last very long when subjected to water pressure. Water will soon enter the plywood and delaminate the fiberglass.  Fiberglass laminated plywood is also positive buoyant which will increase the requirement for hard ballast. A better core material composite plastic board. Composite plastic board is made from recycled plastics, it has the same density as water, comes in 4 x 8 sheets as well as larger sizes and is available with fiberglass gauze imbedded in the surface so that it bonds strongly to fiberglass laminate used to bond the seams.


S2 - Wet/Semi-Wet Sub

   S2 - Wet Sub
S2 - Wet Sub

S2 - Wet Sub Details - Click to Enlarge
S2 - Wet Sub Details
S2 with foot rest retracted.
S2 with foot rest retracted.
S2 - Side View
S2 - Side View
S2 - Gimbaled thrusters rotated left
Gimbaled thrusters rotated left
S2 - Gimbaled thrusters rotated up
Gimbaled thrusters rotated up


Design S2 is a complete departure from the previous series. Recall that such departures are part of the design approach. "If you have two or three ideas, then start following up on all of them."  The motivation was also sticker shock when I calculated material cost for the plastic core and fiberglass.

The Sled or S2 is a simpler wet sub design based from Odyssea which is now out-of-business.  The primary departure from the design is the addition of a acrylic bell. It would be easy to transport and should cost under $3000 to build.  The other departure is a retractable foot rest.  The front of the rest sides back, and then 4 rails allow it to retrace up to the bottom of the sub to make for more compact storage. Divers can wear fins while riding by tucking the front edge of their fins under a lip on the front of the foot rest.

The bell can alternately be removed or swung around so that it is out of the way, and lead ballast then removed from the seat.

The thrusters are mount together on a gimbal, allowing them to control direction and pitch.

There are recesses for scuba tanks behind each passengers  seats.  The passengers can use these tanks as their primary source of air or the primary source can be hookah supplied air and the scuba tanks are available for emergencies.





After reading up on boat building, my confidence in the H4 design was up.  Aluminum now looked like the material of choice and the flat hull was abandoned in favor of a V hull.  Steel is another option because it's cheaper than aluminum, however it is heavier too and I am already worried about the weight. Getting a 2.5 ton V-Hull boat to plane will be a difficult enough when it only has a 4 ft beam.  The cost of aluminum is about the same as fiberglass over a dense core but aluminum has a weight savings as a 3/16 sheet of aluminum provides comparable stiffness to fiberglass over a 1/2 inch high density core. 

Benefits of Aluminum:
1) Strength and durability. Sooner or later we're going to hit something hard, like a rock. So I wanted something that would take that kind of punishment. Something like a jet drive river boat. Fiberglass is a great material, but it does not fair well when dashed against rocks.
2) Aluminum is about the same cost as epoxy glass over a core.
3) You can take your time and build aluminum as slow as you like.
4) Fiberglass strength is dependent on skill of the installer, age and condition of the materials, environment, etc; and aluminum is only dependent on getting the weld right.
5) To give a fiberglass hull the required strength it would need frames and longitudinals or it would need a core. The core is more common and seemed much easier than structural members but most core material is light or medium density foam or wood and that makes for a hull that is lighter than water; a problem for a submarine. And high density core material would increase the overall weight of the boat.
6) Finally, having worked a little with fiberglass and metal, I had enjoyed being burned more that having my fingers glued together.

The actual design for H5 was very fluid and mainly sketches in a notebook leaving little to show for it, but it collimated into H6. 


H6 - Front View
H6 - Front View



H6 - Top Front Quarter


H6 - Font Quarter


The hull was now 3/16 aluminum with a 12 degree dead rise angle from the keel to a hard chine and the straight up to form the sides of the hull.  Longitudinals in the form of 3/4 inch angles were added to the outside bottom of the hull as well as a couple down the side to stiffen the flat surface.  Boats normally have curved surfaces which add to the stiffness of the hull but with the cabin up front it just made since to keep the lines straight and not open up a bigger hole than needed when pushing through the water.

The engine is a converted outboard with a tall tail fin that acts as snorkel for the engine when on the surface or with the all but the snorkel under the water.  It also acts as a rudder when running submerged.






H7 - Details
H7 - Details Jet Pump
H7 - Front Quarter H7 - Side View
H7 - Font Quarter H7 - Side View
H7 - Framing
H7 - Framing
H7 - Bottom View, showing acrylic hatches and foot rest.
H7 - Acrylic hatches & foot rest
H7 - Aft Quarter, showing wet storage are behind the cabin
H7 - Aft view of cabin showing wet storage area and view port.
H7 - Resting on the bottom
H7 - Resting on the bottom


On the H7 design the outboard engine gave way to a jet drive. The biggest reason for switching is that a jet drive provided a solution for pumping hundreds of gallons of water out of the hull when resurfacing but converting the cleanout port to suck water from inside the hull.  It also provides a easy means to flood the hull.  Simply shutdown the engine and open the cleanout port.  There also the benefit of of removing a prop from a boat that will frequently operate with simmers in the water.

The design of the bow can be improved. Bringing the hard chines together to form a point is easy to draw and build, but it done not help the boat get up on a plane.  A better bow will bring the sides forward without curving them into the center.  Instead the bottom hull plates will curve up. 

More upward push from the lower portion of the is needed to help overcome the downward push from the cabins deck and forward view port.  Even if the surface area between the surfaces pushing up and the surfaces pushing down, the down surfaces will still win the war because the cabin deck is further forward that the bottom of the hull.  That distance gives it an advantage over the bottom hull just like having a 3 foot pry bar.

The engine snorkel for the H7 design runs forward inside the hull and comes up to the top of the cabin just behind the occupants.  This eliminates a potential snag but was latter dropped because of the overall volume of the snorkel.  Sealing the snorkel with air in it produced too much displacement and allowing it to flood required too much time to pump it dry and added too much flood volume to the engine compartment if it was not first pumped dry.


Front View
H8 - Side extensions made to
the lower section of the hull.

H8 - Bow curved up to provide
more upward push to balance
the slope of the cabin deck.


There were only two real changes made to the H7 to move it up to H8 which is where we left the drawing board and started welding a hull.

The lower section of hull was widened. The will provide more surface area that will help get the boat up onto a plane. The slated front profiles of the side extensions will provide more upward push like dive planes in order to help offset the down push that will be generated by the top front deck of the cabin.  And the side extensions provide a convenient location for storing compressed air cylinders that will backup the hookah supplied air or allow for operation without the surface supplied air.

The other change was to the profile of the bow.  This too will improve the about of upward push.