This is some related cut and pastes, Chapter 4 of the wood handbook also has a good discussion of creep under "duration of load".

Wood is considered a viscoelastic material. This viscoelasticity explains the creep phenomenon in which a given load will induce an immediate deformation, and if that load is allowed to remain on that piece, additional secondary deformation (i.e., creep) will continue to occur over long time periods.

Mechanical properties deal with stress and strain relationships that are simply functions of chemical bond strength. At the molecular level, strength is related to both covalent and hydrogen intrapolymer bonds. At the microscopic level, strength is related to both covalent and hydrogen interpolymer bonds and cell wall layer bonds (S1–S2 and S2–S3). At the macroscopic level, strength is related to fiber-to-fiber bonding with the middle lamella acting as the adhesive. Thus, any chemical or environmental agent that affects those bonds also affects strength.

Ordinary climatic variations in temperature and humidity
will cause creep to increase. An increase of about 28oC (50oF)
in temperature can cause a two- to threefold increase in creep.
Green wood may creep four to six times the initial deformation
as it dries under load.

Me here,
Water is the most common environmental agent affecting chemical bond strength. If the interpolymer hydrogen bonds (these are the linking arms between strands of cellulose within the cell wall) are busily occupied with H2O...bound water, then the number of hydrogen bonds holding cellulose molecules to each other is reduced. The cellulose molecules a much more free to slip and slide in relation to one another, like a wet noodle. Green wood is about half as strong as dry for this reason.

Chemically when you bend a stick of wood the polymers of cellulose slip by one another. The hydrogen bonds linking one molecule to the next are drawn tight, break, align with the next bonding site and rebond, when you release the stress the process reverses. Get the water out of the mix and the cellulose is more thoroughly cross linked.

Keep bending the stick. When the carbon-carbon and carbon oxygen bonds begin to break the deformation is unrecoverable.

Tim;

I saw a graph in my reading about creep that seemed like it gained at a steady, very slow pace if the temperature and humidity variations remain consistant. One of the warnings of a "creep rupture" failure is an accceleration of the creep. In other words, if you have a saged rafter, monitor it. If the creep is accelerating, evacuate the building.

Creep also happens to metal and concrete.

Jim

And old glass.

Tim

I think this image both exemplifies Creep and the forces we sometimes ask of our framing greatly -





This Post had been two panel points in from the end of this bridges span, (so the load was not even as great as asked of the Posts farther from midspan) yet 150 yrs of constant load imparted from the Brace (compression diagonal) had slowly created this much deformation. Glare has the ruler unreadable but that string is at 1 ¾” at the Bottom Chord dap, (just above where it's pictured) and the braces CE is only 12 ¾” (½ way up that “joggle” that is the Braces abutment) above the Chords reciprocal dap which will buttress the load. Of course this ones problem is exacerbated by neglect and other extenuating circumstances, so this example is not representative of all Burr or MKP trusses. But it is super creeped out

Sorry about the techno-speak, can't think of a better / simpler way to describe it.