If the ships structure, equipment, and cargo are distributed unevenly there may be large point loads into the structure, and if they are distributed differently than the distribution of buoyancy from displaced water then there are bending forces on the hull.
When ships are drydocked, and when they are being built, they are supported on regularly spaced posts on their bottoms.
The primary strength, loads, and bending of a ship's hull are the loads that affect the whole hull, viewed from front to back and top to bottom. Though this could be considered to include overall transverse loads (from side to side within the ship), generally it is applied to Longitudinal loads (from end to end) only. The hull, viewed as a single beam, can bend
This can be due to:
Primary hull bending loads are generally highest near the middle of the ship, and usually very minor past halfway to the bow or stern.
Primary strength calculations generally consider the midships cross section of the ship. These calculations treat the whole ships structure as a single beam, using the simplified Euler-Bernoulli beam equation to calculate the strength of the beam in longitudinal bending. The moment of inertia (technically, Second moment of area) of the hull section is calculated by finding the neutral or central axis of the beam and then totalling up the quantity for each section of plate or girder making up the hull, with being the moment of inertia of that section of material, being the width (horizontal dimesion) of the section, being the height of the section (vertical dimension), being the area of the section and being the vertical distance of the center of that section from the neutral axis.
Primary strength loads calculations usually total up the ships weight and buoyancy along the hull, dividing the hull into manageable lengthwise sections such as one compartment, arbitrary ten foot segments, or some such manageable subdivision. For each loading condition, the displaced water weight or buoyancy is calculated for that hull section based on the displaced volume of water within that hull section. The weight of the hull is similarly calculated for that length, and the weight of equipment and systems. Cargo weight is then added in to that section depending on the loading conditions being checked.
The total still water bending moment is then calculated by integrating the difference between buoyancy and total weight along the length of the ship.
For a ship in motion, additional bending moment is added to that value to account for waves it may encounter. Standard formulas for wave height and length are used, which take ship size into account. The worst possible waves are, as noted above, where either a wave crest or trough is located exactly amidships.
Those total bending loads, including still water bending moment and wave loads, are the forces that the overall hull primary beam has to be capable of withstanding.
Secondary loads, strength, and bending are calculated similarly to primary loads: you determine the point and distributed loads due to displacement and weight, and determine local total forces on each unit area of the panel. Those loads then cause the composite panel to deform, usually bending inwards between bulkheads as most loads are compressive and directed inwards. Stress in the structure is calculated from the loads and bending.
Tertiary strength and loads are the forces, strength, and bending response of individual sections of hull plate between stiffeners , and the behaviour of individual stiffener sections. Usually the tertiary loading is simpler to calculate: for most sections, there is a simple, maximum hydrostatic load or hydrostatic plus slamming load to calculate. The plate is supported against those loads at its edges by stiffeners and beams. The deflection of the plate (or stiffener), and additional stresses, are simply calculated from those loads and the theory of plates and shells.
This diagram shows the key structural elements of a ship's main hull (excluding the bow, stern, and deckhouse).
The depicted hull is a sample small double bottom (but not double hull) oil tanker.
Shipbuilders today use steels which have good corrosion resistance when exposed to seawater, and which do not get brittle at low temperatures (below freezing) since many ships are at sea during cold storms in wintertime, and some older ship steels which were not tough enough at low temperature caused ships to crack in half and sink during WW II in the Atlantic.
The benchmark steel grade is ABS A, specified by the American Bureau of Shipping. This steel has a yield strength of at least 34,000 PSI, ultimate tensile strength of 58 to 71,000 PSI, must elongate at least 19% in an 8 inch long specimen before fracturing and 22% in a 2 inch long specimen.
A safety factor above the yield strength has to be applied, since steel regularly pushed to its yield strength will suffer from metal fatigue. Steels typically have a fatigue limit, below which any quantity of stress load cycles will not cause metal fatigue and cracks / failures. Ship design criteria generally assume that all normal loads on the ship, times a moderate safety factor, should be below the fatigue limit for the steel used in their construction. It is wise to assume that the ship will regularly operate fully loaded, in heavy weather and strong waves, and that it will encounter its maximum normal design operating conditions many times over its lifetime.
Designing underneath the fatigue limit coincidentally and beneficially gives large (factor of up to 6 or more) total safety factors from normal maximum operating loads to ultimate tensile failure of the structure. But those large ultimate safety margins are not the intent: the intent is that the basic operational stress and strain on the ship, throughout its intended service life, should not cause serious fatigue cracks in the structure. Very few ships ever see ultimate load conditions anywhere near their gross failure limits. It is likely that, without fatigue concerns, ship strength requirements would be somewhat lower.
However, it is still important to be able to manually calculate rough behaviour of ship hulls. Engineers do not trust the output of computer programs without some general reality checking that the results are within the expected order of magnitude. And preliminary designs may be started before enough information on a structure is available to perform a computer analysis.