Why Memory Foam and HR Foam Feel Different: The Raw Material Differences Explained

Memory foam and HR foam are both polyurethane. They are made from the same two primary chemical components — polyol and isocyanate — by the same basic reaction described in the previous article. Yet their mechanical behaviours are so different that they feel like entirely different material classes: one moulds slowly to your body and holds the impression for seconds; the other springs back almost instantly. The difference is not in the chemistry category — it is in the specific molecular architecture produced by deliberate raw material and formulation choices. This article explains exactly what those choices are and why they produce such divergent behaviour.

1. The Single Most Important Difference: Glass Transition Temperature

The most fundamental difference between memory foam and HR foam is the position of their glass transition temperature (Tg) relative to the range of use temperatures.

As covered in the Seasonal Foam Behavior article, the Tg is the temperature range over which a polymer transitions from a stiff, glassy state to a soft, rubbery state. Below Tg, polymer chain mobility is limited — the foam is stiff and elastic. Above Tg, chain mobility is sufficient for viscoelastic relaxation — the foam is soft and shows time-dependent deformation.

Memory foam is formulated with a Tg in the range of approximately 10–25°C — deliberately placed within the bedroom temperature range. This means that at typical room temperatures (15–23°C), memory foam is operating in or near its glass transition zone: the foam is partially glassy, partially rubbery, and its properties are strongly sensitive to small temperature changes. The slow recovery, the gradual conformance, the body-warming softening effect — all are consequences of operating in the glass transition region.

HR foam is formulated with a Tg well below room temperature — typically below 0°C, often below −20°C. At all normal use temperatures (10–35°C), HR foam is far above its Tg and in the stable rubbery plateau of its modulus-temperature curve. Its chain mobility is high, its elastic restoring force is strong, and its recovery is fast and nearly complete. Temperature variation within the bedroom range causes only minor property changes — the foam is not near any transition region.

This single difference — where the Tg falls relative to use temperature — accounts for the majority of the behavioural differences between the two foam types. Everything else is detail.

2. Raw Material Differences That Control Tg

Tg is not a single molecular property — it is an emergent property of the polymer network architecture, controlled by several raw material choices made during formulation design.

Polyol molecular weight

The molecular weight of the polyol chain is the primary lever for controlling Tg in flexible polyurethane foam. Higher molecular weight polyols produce longer soft segments with greater chain mobility — more conformational freedom, lower energy barriers for chain rearrangement, lower Tg. Lower molecular weight polyols produce shorter, more constrained soft segments with higher Tg.

Memory foam formulations typically use polyols with molecular weights in the range of 3,000–6,000 g/mol. HR foam formulations typically use higher molecular weight polyols — 4,000–8,000 g/mol — which produce longer, more mobile soft segments and lower Tg. The overlap in the molecular weight ranges is real: the difference is not a simple high vs low molecular weight story, but a combination of molecular weight, polyol type, and functionality.

Polyol functionality

Functionality refers to the number of reactive hydroxyl groups (–OH) per polyol molecule. Higher functionality polyols produce more cross-links per molecule when they react with isocyanate — a denser network with shorter chain segments between cross-link points. Shorter chain segments have lower mobility and higher Tg.

Memory foam formulations use higher functionality polyols (typically trifunctional, f=3) to increase cross-link density and raise Tg into the desired range. HR foam formulations use lower average functionality — a higher proportion of difunctional polyols — to produce longer chain segments between cross-links, lower Tg, and higher chain mobility.

Polyol type: polyether vs special grades

Standard polyether polyols based on propylene oxide (PPO) produce foams with moderate Tg. For memory foam, formulators often use PPO-based polyols with specific modifications — controlled end-group chemistry, blending of different molecular weight fractions, or incorporation of ethylene oxide (EO) end-capping — to achieve the desired Tg positioning and temperature sensitivity profile.

HR foam more commonly uses “polymer polyols” (also called graft polyols or PHD polyols) — polyether polyols containing dispersed solid polymer particles (typically styrene-acrylonitrile copolymer). The solid filler particles act as physical reinforcement, increasing the foam’s load-bearing capacity and supporting factor without significantly increasing Tg. HR foams made with polymer polyols typically show higher resilience and better support factor than equivalent foams made with unfilled polyols.

Isocyanate selection

TDI (toluene diisocyanate) is the dominant isocyanate for memory foam production. TDI’s reactivity profile and the molecular geometry of the urethane linkages it forms are well-suited to producing the specific hard segment domain morphology associated with high loss factor and viscoelastic behaviour.

MDI (methylene diphenyl diisocyanate) is preferred for HR foam production. MDI-based polyurethane networks have different hard segment geometry — the larger, more symmetric MDI molecule produces more ordered hard segment domains with stronger hydrogen bonding between adjacent hard segments. This stronger, more regular hard segment structure produces a network with higher elastic character and lower loss factor — exactly the properties that define HR foam behaviour.

The difference in hard segment domain structure between TDI-based and MDI-based foams is not just a mechanical curiosity — it also affects the temperature sensitivity of the foam’s properties. TDI-based hard segment domains are less ordered and disrupt more readily with temperature, contributing to the strong temperature dependence of TDI-based memory foam. MDI-based hard segment domains are more stable, contributing to the lower temperature sensitivity of MDI-based HR foam.

3. How the Network Architecture Produces Different Behaviour

With the raw material differences established, we can trace how they produce the macroscopic behaviours that distinguish memory foam from HR foam.

Slow recovery in memory foam

Memory foam’s slow recovery is the direct consequence of operating near the glass transition. At bedroom temperature, many of the soft segment chains do not have sufficient thermal energy for rapid conformational rearrangement — the activation energy barriers for chain rotation are only marginally overcome by the available thermal energy. When the compressive load is removed, the elastic restoring force of the network drives recovery, but the viscous resistance of the partially glassy soft segments slows the process. Recovery occurs over seconds to minutes rather than milliseconds.

In HR foam, operating far above Tg, all soft segment chains have high thermal energy relative to the rotational barriers. Recovery is limited only by the elastic driving force and the foam’s cell structure — not by viscous chain resistance. Recovery is essentially complete within one to two seconds.

Energy dissipation and loss factor

Memory foam’s high loss factor (tan δ) — the property that gives it its motion-isolating character — arises from two molecular mechanisms operating simultaneously near the glass transition.

First, the incomplete relaxation of soft segments under cyclical loading: at frequencies and temperatures near the glass transition, the polymer chains are partially mobile but not fully so. Under cyclical loading, energy is input to drive chain rearrangement during compression, but recovery is incomplete within the cycle period — the unreturned energy is dissipated as heat. This is the primary mechanism of energy dissipation in viscoelastic materials.

Second, the formation and disruption of hydrogen bonds between hard segment domains under load: each hydrogen bond formation and disruption event dissipates a small amount of energy. The cumulative effect across the millions of hard segment interactions in a compressed foam volume is a meaningful contribution to the overall loss factor.

HR foam, operating far above Tg with fully mobile soft segments, undergoes complete elastic recovery within each loading cycle. The energy input during compression is fully recovered on unloading — the loss factor is low and the resilience is high.

Temperature sensitivity

Memory foam’s strong temperature sensitivity near Tg is a direct consequence of the steep slope of the modulus-temperature curve in the glass transition region. A small temperature change moves the operating point significantly along this steep slope, producing a large change in modulus. This is not a design flaw — it is the mechanism behind body-warming softening — but it produces the seasonal firmness variation described in the Seasonal Foam Behavior article.

HR foam, operating on the flat rubbery plateau of the modulus-temperature curve, shows only the gradual decrease in modulus with temperature that is characteristic of rubber elasticity (where higher temperature actually increases the elastic restoring force through entropic effects — the rubber elasticity mechanism). The practical consequence is near-constant feel across seasonal temperature variations.

4. Intermediate Cases: The Spectrum Between Memory and HR Foam

Memory foam and HR foam are not binary categories — they are the extremes of a continuous spectrum of polyurethane foam behaviour controlled by Tg position and network architecture. Commercial foam products span this spectrum, with different formulations positioned at different points to achieve specific combinations of properties.

“Responsive” or “fast-recovery” memory foam

Some products marketed as memory foam — or as “adaptive foam,” “responsive foam,” or similar terms — use formulations with Tg positioned slightly lower than conventional memory foam, or with modified network architecture that reduces the viscous component of the response. These foams show slower recovery than HR foam but faster recovery than conventional memory foam — trading some of the pressure redistribution advantage of true slow-recovery memory foam for better responsiveness and reduced heat retention.

From a formulation perspective, this is achieved by using slightly higher molecular weight polyols (lower Tg), lower cross-link density (longer chain segments between cross-links), or blending conventional memory foam formulations with HR foam components. The resulting material is a genuine intermediate — not better or worse than the extremes, but different, and suited to sleepers who find true memory foam’s recovery too slow but want more conformance than HR foam provides.

Temperature-neutral memory foam

Some formulations attempt to reduce memory foam’s temperature sensitivity while retaining its pressure redistribution performance. This requires shifting the Tg lower — below the bedroom temperature range — while maintaining sufficient viscous character at sleep temperatures to provide slow recovery. This is accomplished through careful polyol molecular weight selection and the use of modified polyol chemistries that broaden the glass transition region, reducing the slope of the modulus-temperature curve in the relevant temperature range.

Truly temperature-neutral memory foam — foam that maintains consistent properties across the 10–30°C bedroom temperature range while still providing meaningful pressure redistribution — is a genuine formulation challenge. Products that claim this property vary significantly in how well they deliver it, and the absence of published DMA (dynamic mechanical analysis) data for most consumer products makes independent verification difficult.

5. Practical Implications: Reading Foam Behaviour from Chemistry

The raw material framework established in this article and the previous one allows a materials-informed interpretation of foam behaviour that goes beyond the ILD specification:

• Slow recovery at room temperature indicates Tg near or above room temperature — a TDI-based, higher cross-link density formulation. This foam will be more temperature-sensitive and will show significant seasonal firmness variation.
• Fast recovery with some conformance indicates Tg below room temperature but with some viscous character — an intermediate formulation or a blend. Less temperature-sensitive than conventional memory foam.
• Near-instant recovery with consistent feel across temperatures indicates Tg well below room temperature — an MDI-based HR foam or a polyol blend with low cross-link density. This foam will show minimal seasonal variation and high resilience.
• Firm feel that softens significantly over the first 10–15 minutes of contact indicates operation near the lower end of the glass transition — the body’s warmth is driving the foam above its Tg at the contact surface. This is conventional memory foam behaviour and is predictable from the formulation strategy.

Summary

The difference between memory foam and HR foam is not mystical — it is a direct consequence of deliberate raw material choices that position the polymer network’s glass transition temperature relative to the range of use temperatures. Memory foam’s Tg is placed within the bedroom temperature range, producing strong temperature sensitivity, slow recovery, high energy dissipation, and progressive body-warming conformance. HR foam’s Tg is placed well below use temperatures, producing temperature-stable, fast-recovering, low-energy-dissipating behaviour.

The specific raw material levers that control Tg are polyol molecular weight, polyol functionality, polyol type, and isocyanate selection. The next article in this series examines how these same levers — combined with density-controlling variables — are used to engineer specific combinations of ILD, density, and loss factor in commercial foam formulations.

The Sleep Mechanic is a materials engineer with hands-on R&D experience in cushioning materials and viscoelastic polymers. Sleep Science Lab applies materials engineering analysis to sleep surfaces — because “it feels comfortable” is not an explanation.