Animation Cel Preservation

RH Cycling, Vapor Pressure, and Why “Microclimate Fatigue” Is Real

This article explains why repeated temperature and humidity swings can cause fatigue-style structural damage in animation cels. The core idea is simple: the cel is a composite (CTA base + paint layers). When moisture content or temperature changes, each layer wants to expand/contract differently. That mismatch creates interfacial shear stress. Do it enough times, and you can drive curling, edge lift, microcracking, and paint delamination — even when no single swing looks “extreme.”


2) The “drivers”: thermal strain vs moisture strain

A useful mental model is total strain as the sum of two components:

Strain type Simple model What it means physically
Thermal strain εT = α · ΔT Expansion/contraction from temperature change (α = CTE).
Moisture strain εM ≈ β · Δ(RH) (conceptual) Expansion/contraction from moisture sorption and water activity changes (β is “hygroexpansion sensitivity”).

The reason RH cycling matters so much: moisture-driven strain can be surprisingly large compared to thermal strain, and it also changes the material properties (softening/plasticization), which increases creep and interfacial damage.


3) Quick thermal expansion example (6°F swing)

Thermal expansion by itself is usually not the dominant driver in typical indoor swings — but it still matters because it sets a clearance requirement whenever the cel is constrained (tight mats, clamped frames, edge binding, compressed sleeves).

Assumptions (order-of-magnitude):
  • Cel length L = 11 in (typical display width scale)
  • CTA coefficient of thermal expansion αCTA ≈ 70 × 10−6 / °F (plastic range)
  • Temperature drift ΔT = +6°F (framework anchor)
Quantity Expression Value Interpretation
Free expansion (CTA) ΔL = L · αCTA · ΔT ≈ 0.0046 in (≈ 4.6 mil) This is the extra “room” the film wants when warmed by 6°F.
Percent of length ΔL / L ≈ 0.042% Small as a % — which is why thermal alone often isn’t catastrophic.
Clearance requirement ΔL (if constrained) ≈ 0.0046 in If edges are constrained, this expansion turns into compressive stress instead of harmless growth.

Why this usually isn’t “the big one”: a few mil of free expansion is often tolerated if the cel is not mechanically constrained and the sleeve/matting geometry has slack. Thermal becomes dangerous when it is repeated and constrained, or when it occurs out of phase with moisture-driven strain.


4) Why mismatch matters (CTA vs paint) — interfacial shear, not total growth

The structural risk is not “the cel grew 4.6 mil.” The risk is that the system is a bonded laminate: a base film (CTA) with paint regions that have different stiffness and different expansion behavior. When CTA tries to move and paint resists, the result is interfacial shear stress — especially at paint edges and thickness transitions (stress concentrators).

Mismatch strain (concept):
εmis ≈ (αCTA − αpaint) · ΔT
Even if total expansion is small, mismatch drives shear at the interface.

Why RH Cycling Is a Cumulative Fatigue Mechanism

RH-driven deformation is not an instantaneous failure mode. It is a cumulative mechanical fatigue process acting on a bonded composite (CTA base + paint layers) over long time horizons.

Each RH cycle induces a small amount of interfacial mismatch strain between the CTA film and the paint. Individually, these cycles are often harmless. However, like mechanical fatigue in metals, the damage is banked.

Induction-period behavior: During the early “flat” portion of the bathtub curve, RH cycling consumes interfacial life without visible damage. Once a threshold is reached, failure appears sudden — edge lift, paint cracking, curl memory — even though the damage accumulated gradually.

5) Why RH changes during temperature drift (even when AH is constant)

This is the subtle trap: at lower temperatures, the air holds less water at saturation — so people assume RH swings should “matter less.” But if the air’s absolute humidity (AH) stays constant, then the actual vapor pressure does not change much, while saturation vapor pressure changes strongly with temperature.

The physics in one line:
RH = e / esat(T)
where e is the actual water vapor partial pressure (set by AH), and esat(T) increases with temperature.

Translation: a temperature drift creates an RH swing even if you did not add/remove any water. That RH swing creates a repeated water activity (aw) swing at surfaces, which drives sorption/desorption cycling.


5.1) Example A — 70°F / 50% RH with +6°F drift (AH constant)

Anchor the drift with reality: this is typical HVAC behavior (day/night load, sun, occupancy, thermostat bands). Start at 70°F and 50% RH, then drift to 76°F without changing the amount of water in the air.

Condition T (°F) RH (%) AH (g/m³) e (hPa) esat(T) (hPa) What changed?
Baseline (home) 70 50 ≈ 9.2 ≈ 12.6 ≈ 25.2 Reference state
Warm drift (HVAC) 76 ≈ 41 ≈ 9.2 ≈ 12.6 ≈ 31.0 AH and e stayed ~constant; only esat rose, so RH dropped.

The RH change is not hand-wavy here: 50% → ~41% purely from a +6°F drift at constant AH. That’s a real cyclic strain input if it repeats daily.


5.2) Example B — 40°F / 50% RH with +6°F drift (AH constant)

Anchor the drift with reality: this is typical compressor cycling in a wine fridge / beverage fridge or small enclosure. We assume the microclimate sits at the low setpoint and drifts upward during compressor cycling. Start at 40°F and 50% RH, drift to 46°F, with no moisture added/removed.

Condition T (°F) RH (%) AH (g/m³) e (hPa) esat(T) (hPa) What changed?
Baseline (cold) 40 50 ≈ 3.2 ≈ 3.7 ≈ 7.4 Reference state
Warm drift (cycling) 46 ≈ 38 ≈ 3.2 ≈ 3.7 ≈ 9.6 AH and e stayed ~constant; esat rose → RH dropped more sharply.

This directly answers the “lower temperature = less available water” intuition: yes, cold air has less AH here (3.2 vs 9.2 g/m³), but a 12% RH swing still happens because the driver is e / esat. The vapor pressure doesn’t have to change for RH — and surface water activity — to cycle.


5.3) Example C — Temperature steady, but moisture is added (cooking/shower/door opening)

Separate the mechanisms: here temperature is steady, and AH increases from life events (cooking, showering, door opening). At constant temperature, increased AH means increased vapor pressure e, which increases RH directly.

Scenario T (°F) RH (%) What changed? Meaning for fatigue
Baseline 70 50 Reference Baseline sorption strain
Moisture added 70 60 AH ↑e ↑ Higher aw → more absorption/swelling
More moisture added 70 70 AH ↑↑e ↑↑ Larger cyclic amplitude; stronger fatigue + creep coupling

Real homes often have both: temperature drift changes RH even at constant AH, and daily life changes AH directly. That’s why “the thermostat says 70” does not mean the cel sees stable water activity at its surface.

6) The “Cold Storage Trap”: Slow Chemistry vs. Fast Physics

Storing cels in a fridge slows hydrolysis (Vinegar Syndrome), but "unintentional" cold storage introduces a mechanical risk: High-Frequency Fatigue.

Hysteresis and Phase Lag in Moisture Cycling

Moisture absorption and desorption do not occur symmetrically. CTA and paint layers exhibit hysteresis: they absorb water and release it at different rates and along different paths.

In environments with frequent cycling (e.g., small refrigerators or unbuffered enclosures), the cel may never reach equilibrium before the next cycle begins.

Engineering translation: The system exists in a state of persistent internal phase mismatch: temperature responds quickly, vapor pressure shifts immediately, but moisture redistribution lags — producing continuous internal stress.

7) Why CTA–paint stress exists (what’s physically happening)

Engineering translation: RH cycling is a repeated strain input into a bonded, non-uniform laminate. That is a textbook setup for interfacial fatigue and creep-set over long time horizons.

8) Anchoring the Framework: The ΔRH Likelihood Bands

In our framework, we anchor risk to Amplitude (ΔRH)—the primary driver of interfacial fatigue.

Lstruct Relative Structural / Mechanical Risk (50 yrs) Examples (handling + RH cycling)
1 – Very Low RH cycling **ΔRH < 5%** per typical cycle (much tighter than a good room). Microclimate buffered so the effective aw band is extremely narrow (e.g., engineered cold box, well-buffered small volumes).
2 – Low RH cycling **5% ≤ ΔRH ≤ ~10%**, slightly better than or equal to a best-case room. Room or microclimate held near a stable setpoint with modest drift, generally as good as (or a bit tighter than) a well-behaved 70 °F / 50 % room.
3 – Neutral RH cycling **≈10–12%** per typical cycle (our reference case). RH swings equivalent to a typical interior room drifting **67–73 °F** around 70 °F with constant AH, giving an aw band of roughly 0.47–0.54 (~47–54% RH equivalent). No extremes, but no deliberate buffering either.
4 – High RH cycling **~12–20%**, but still without sustained condensation. Environment has large day–night RH shifts, seasonal spikes, or repeated excursions above ~60% or below ~30% RH, yet does not routinely hit saturation.
5 – Very High RH cycling **> 20%**, including condensation or very low RH episodes. Environments that push aw toward 1.0 (condensation, damp basements) or toward very low values for long periods, combined with strong repeated T/RH shocks (fridge-box cycling, attic or garage extremes).

9) Mitigation: Dampening the Physics

To move into a safer Likelihood Band, you must use Mass to fight Frequency:

Engineering translation: RH cycling is a repeated strain input into a bonded, non-uniform laminate. To protect cels, solve for both chemical decay (cold) and mechanical fatigue (stability).