Originally I was told this project had to come in as a park trailer. Once it was found out that it was not factory built, I then had to submit as R3. Then I was told if I had engineer approval, it would be permitted. After I submitted, I was told that "every system would need to be evaluated" first.
If I had proposed a small
house using typical construction methods; with 100 year old technology and relatively
shoddy materials, I understand my permitting would be much easier. If my project
were designed for rapid unskilled construction instead of careful boatbuilding,
I understand that I would have an easier permitting process.
One may think that this project is too complex to understand structurally. Not true. People actually have the conceptual tools already.
First lets look at what does comply with code. A small typically framed house could be easily permitted.
Lets look at a typical piece of wall first. Structurally it is ½” waferboard or CDX sheathing on one side and 2x4 studs on 16” centers, as you know.
Basically I am using SIP
panel construction. Nothing new there. Compared to standard balloon construction,
they have much better insulation and less critical framing. They are waferboard
faces bonded onto 0.9 lb. foam core. The face panels are butt joined to each
other. And they are however still vulnerable to termites, rats, moisture and
SIP panels are stronger than a typical framed wall because they have the face panels both as far away as possible from the centroid of the wall. A framed wall only gets to count the studs and outside sheathing. A typical 16” wide frame wall section with one stud and ½” sheathing has a section modulus of 6.15 in^3. A SIP with two 3/8” faces has for comparison, a SM of 18 in^3
Mine are SIP panels also,
but thinner. Mine are 5mm or almost ¼” thick, with a SM of 11 in^3.
But notice this. Waferboard has a bending strength of about 980 psi. My plywood
is BS1088. It will test at 15,000 psi bending strength. 15 times the bending
strength. Unless you have lived in Port Townsend, you have never seen plywood
of this quality in your entire career.
That’s just the beginning. Every join between my face panels is scarphed. Not a single butt join is allowed in any part. The load transfer in the plane of the face panels is nearly 100%. With regular butt join SIPs it is almost 0%, depending only on the shear strength of the foam. More on that later.
Now consider this. Every piece of plywood, indeed every piece of the fiberglass and foam is completely epoxy encapsulated. Completely. Then every exterior surface and every wear surface next gets a fiberglass cloth coating. With a paint coating to stop UV damage.
The foam I used is the Owens Corning Foamular 400. It has double the density and core shear strength of standard SIP panel foam.
All the parts are joined
with biaxial coves inside and out. The biaxial roving has a bending strength
of 40,000 psi and a shear strength of 3,000 psi. Understand that is for every
inch of the perimeter of every part, inside and out.
I have pounded enough 16d nails into wall plates. Those nails are at the centroid of the wall; the worst place possible against an out of plane load. My biaxial coves are both at the face planes; the best possible place to join parts. In tests we did at UW Department of Architecture, an epoxy laminate shears away in-plane at about 5000 psi. Suppose you want to see how much strength the biaxial gives one panel. If only bonded at the base, with half of a typical 6” biaxial tape, you get 288 sqin. of laminate contact. At 5000 psi, that gives 1,440,000 lbs. For simplicity lets use just 1000 lbs/sqin. That is 288,000 lbs. BUT that is only one side. Every panel is bonded on both sides. Double that 288K. That is 576K lbs. How many pounds are a half dozen 16d nails and a dozen 6d nails good for in comparison?
Now imagine a panel bonded on all 4 sides with 3” wide band. It has 1152 square inches of biaxial contact. At only 1000 psi for ease of calculation, that gives 1,152,000 lbs of load resist. That is over a million lbs. How many 16 penny nails are needed to match that? Again, that is only for one side. We get to count both sides. Double that strength. That is 2,304,000 lbs. Gentlemen.
The cored parts actually
have only minimal loads. They rest on the systems ring where the loads are resolved.
These are the same BS1088 plywood, but in 3/8” thickness. And again, all
parts have biaxial coves inside and out. All fasteners have biaxial local reinforcements
as per standard boatbuilding practice. Those are ½”thick biaxial
All fasteners are 316 stainless steel.
The systems ring components are loaded to push into each other, like an arch. An arch all around. They would probably stay together with minimal fasteners, but I use 7/16” bolts, good for over 55 ksi each.
Besides being bolted, all the systems ring bins are bonded to each other with 3M 5200 adhesive. That product has 885 psi tensile strength. Each bin will have 84 linear inches of 5200 contact with the adjacent bin. If the bead is ½” wide, that is 42 sqin or 37,170 lbs bonding strength at each join.