Pick a region and a year (1500 to 1950) to see how the house changed, then watch how its clay, log or brick walls, thatch or tin roof and its big clay stove (pich) hold a winter day's heat. Dates are approximate, regional variants differ.
The science
Heat through a wall. In steady state the heat flow is Q = U · A · ΔT with U = 1 / (Rsi + Σ d/k + Rse): the wall's layers (thickness d, conductivity k) and the two air films add as resistances in series. Log (k ≈ 0.13 W/m·K) is about four times better per metre than clay-straw (≈ 0.55) or brick (≈ 0.75), so a 22 cm log wall loses about as much as a 90 cm clay wall. Thatch (k ≈ 0.09, 30 to 40 cm) is an excellent roof insulator; tin or shingle over a clay-and-straw ceiling is several times leakier.
Thermal mass. A wall stores heat: the layer that follows a daily swing has the penetration depth δ = √(α P / π) with α = k / (ρ c) (about 10 cm for clay, 6 cm for logs). A thick clay wall therefore delays and damps the outdoor swing (hours of lag, a few percent of the amplitude) even when it is not a good insulator; the experiment tab compares the finite-difference result with the exact complex-frequency solution of the heat equation.
The stove (pich). The big clay-brick stove is a heat battery: Cs dTs/dt = η Pfire − h As (Ts − Ta). A firing of a few kilograms of wood in 1.5 hours charges about 2 to 3 tonnes of brick; the stove then gives the heat back with a time constant τ = Cs / (h As) of about 6 hours. The room is a three-node network (air with furnishings, inner wall layer, stove), integrated with RK4 over five spin-up days until the daily cycle repeats.
Fuel. Heat of wood: LHV = 18.5 (1 − w) − 2.44 w MJ/kg (w is the moisture fraction). Stove efficiencies (share of fuel energy that ends up in the room) are 0.3 to 0.6 and depend on the stove type; smoke-heated stoves also lose heat by venting the room while they burn.
Honest simplifications. Only the heated living room (khata) is modelled; the hall (sini) is a buffer at Tout + 0.35 (Ta − Tout); the ground has a fixed temperature; the 3D house is one representative example per region, built from typical ethnographic descriptions, not a specific museum building. Conductivities are typical values with ±30 % spread, the lumped stove surface is one temperature, window U-values are rough for historic glass, and wind, snow load and moisture in walls are ignored. Dates are approximate: change was uneven and a poor household could live in a "1600" house in 1900. Open-air museums such as Pyrohiv in Kyiv and Shevchenkivskyi Hai in Lviv hold real examples.