ΔT
DoughKinetics
Dough fermentation simulator · Free, no account

See your next bake before you make it.

Follow your dough from countertop to fridge to oven. Find the yeast amount for your schedule — with real cooling time in the equation.

t = 20.0h • Final proof
20.0h

Which dough will you bake?

Example result · Big batch trap

Your dough at bake time
⚠️ +35% Overproofed

Too much yeast for this schedule

You'd reach 135% at bake. The core stays warm for 8.4h in the fridge.

Rise at bake
135%
Gluten at bake
74%
Fresh yeast for your schedule
15.75gor 5.25g Dry yeast
Flour 1724gWater 1207gSalt 48.3gCurrent yeast 21.20g

Your recipe, step by step

Open a step, answer, then tap Next. Closed steps read like a recipe.

  1. 3000 g
    260 g → 12
    Recommended fresh yeast15.8 g · 0.91%
    1.23% · 21.2 g
  2. → Bake · 20.0h
Every hour tells a story
t 20.0h · 23.1°C · 135% · 74%
Dough core °C Ambient (naive) Rise (%) Naive rise Gluten strength (%)
■ Green band = ideal bake

The hours your fridge can't skip

Normal app says
21.13g
fresh yeast
Thermal model says
15.75g
fresh yeast
Core stays above 8°C in the fridge for
8.4h
+492% in fridge

On the chart, dashed lines are what a naive calculator assumes (dough instantly at fridge temperature). Solid lines follow the real dough core, which cools slowly and keeps fermenting.

⚠️ +35% OverproofedUse 15.75g

Cold fridge. Still-warm dough.

Cooling, fermentation and gluten strength — modeled together.

1. Newton's law of cooling, scaled by mass

Fermenting dough is a poor thermal conductor (k ≈ 0.42 W/m·K) because trapped CO₂ micro-bubbles insulate it. Core temperature T(t) lags behind ambient T_env.

dT/dt = -k_vessel · (1000 / Mass)^(2/3) · (T(t) - T_env(t))

Surface-to-volume scales with M^(-1/3) and conduction path with M^(1/3), so doubling dough mass stretches cooling time by roughly 1.58×.

2. Yeast vs. protease Q₁₀ divergence

Leavening (yeast CO₂) and maturation (protease and amylase breaking down proteins and starch) follow different temperature sensitivities.

R_yeast(T) ∝ Yeast% · 2.28^((T - 20)/10)
R_protease(T) ∝ Hyd_factor · 1.65^((T - 20)/10)

At 4°C yeast drops to ~6% of its 20°C rate while proteases stay at ~15–18%. Over 48h with high hydration and weak flour, enzymes tear the gluten before the dough even looks overinflated.

3. Two-stage refresh

In the first dough (no salt) yeast multiplies logistically. At refresh, temperature jumps by heat balance of first dough, new flour and water (plus mixer friction), 88% of the gas is knocked out, and gluten is averaged between matured and fresh flour.

T⁺ = (C_pre·T_pre + C_flour₂·T_room + C_water₂·T_water) / C_tot + ΔT_mixer
Gluten⁺ = f_pre · Gluten⁻ + (1 − f_pre) · 100

It also runs backwards: given the first dough, the simulator computes the water temperature to close at 24°C. First-dough yeast is found by bisection.

4. Flour blends & shaping

Blend strength is the share-weighted W of your flours; with a pre-ferment it becomes fPre·W₁ + (1−fPre)·W₂ after refresh. Shaping knocks out part of the gas (72/55/35% kept), briefly tightens gluten and, from then on, each ball cools or warms with its own small mass.

W_blend = Σ pᵢ · Wᵢ
Gas⁺ = η_shape · Gas⁻  (η = 0.72 / 0.55 / 0.35)

Water balance: refresh water = F_tot·(H_tot − f_pre·H_pre). If H_tot < f_pre·H_pre the combination is impossible and the simulator flags it.

About & limitations

  • • This is a simplified numerical model, not lab-grade truth. Constants were calibrated against a handful of probe measurements in home fridges and plastic containers.
  • • Real fridges cycle, doors open, and flours vary a lot. Use the output as a starting point, then adjust to your own bakes.
  • • Your settings stay in your browser. No account, no tracking of your recipes.

Guide: cold fermentation yeast and the fridge →