💡 In This Article
- How the measurement purpose of DMA differs from DSC and TMA
- Why Tg values differ between techniques and how this relates to segmental motion
- How the position, height, width, and shape of the Tan δ peak can provide clues about polymer structure and relaxation behavior
- How DMA can be used to estimate crosslink density in a cured epoxy network
In the previous articles, I explained that DSC measures changes in heat flow and TMA measures dimensional changes as temperature changes. So what does DMA, or Dynamic Mechanical Analysis, actually measure? DMA also provides Tg-related information, but it asks a somewhat different question: as temperature rises, when does the material begin losing its ability to mechanically support a load? For an epoxy developer, that’s an extremely practical question.
What Does DMA Measure?
DMA applies a small oscillating load to a specimen and tracks how the material responds as temperature changes — repeatedly deforming it by a very small amount and observing how stiff it remains, how easily it deforms, and how much of the input mechanical energy gets dissipated. DMA results commonly report three related quantities: storage modulus (E′), loss modulus (E″), and tan δ.
Epoxy and other polymeric materials are viscoelastic — they show both elastic and viscous behavior. Unlike a simple static mechanical test, DMA applies a sinusoidal deformation or stress, and because molecular motion within the polymer doesn’t respond instantaneously to the applied load, the resulting strain ends up phase-shifted relative to the applied stress. That phase difference, δ, is what lets the mechanical response split into two components: storage modulus (E′), representing the elastic component and the material’s ability to store mechanical energy, and loss modulus (E″), representing the viscous component — the portion of mechanical energy dissipated, largely as heat. Their ratio is expressed as:
tan δ = E” / E’
Tan δ is commonly used as an indicator of damping or energy-dissipation behavior. These three parameters offer different views of the same molecular response, which is why looking at the whole DMA curve tells you more than any single number would.
Why Can Different Techniques Give Different Tg Values?
Many people are surprised the first time they compare Tg values across analytical techniques — an epoxy might show a Tg of 120°C by DSC but roughly 135°C by DMA. That doesn’t necessarily mean either instrument is wrong. DSC, TMA, and DMA are observing different physical consequences of the glass transition:
| Technique | Main response observed |
|---|---|
| DSC | Heat flow / heat-capacity-related change |
| TMA | Dimensional change / CTE change |
| DMA | Mechanical modulus and viscoelastic response |
The glass transition isn’t a single, sharply defined molecular event at one universal temperature — it’s better understood as a temperature-dependent shift in molecular mobility and relaxation behavior playing out over a range of temperatures. DSC picks up the associated thermodynamic change, like the shift in heat capacity; TMA picks up the change in dimensional response and thermal expansion behavior; DMA is particularly sensitive to the mechanical consequences of increased segmental motion, especially the α-relaxation tied to the glass transition. As segmental mobility rises, the polymer becomes dramatically less stiff and better able to dissipate mechanical energy, and DMA captures that shift directly through the temperature dependence of modulus and tan δ.
The apparent Tg, in other words, can depend not just on the instrument but on the physical property being monitored, the criterion used to define Tg, heating rate, oscillation frequency, sample history, and measurement conditions. That’s why I found it more useful, as a developer, to ask why the numbers differ rather than simply which Tg value is “correct.”
Storage Modulus: The First Curve I Often Looked At
You might expect the first move with a DMA result to be finding the tan δ peak. In actual development work, though, I often looked at the storage modulus curve first.
As temperature rises, the material gradually softens — but not every material loses stiffness the same way. Some formulations show a gradual modulus decrease, while others undergo a much sharper drop around the glass transition, and that difference can matter a great deal for the actual application. An adhesive might have a high room-temperature modulus and still lose a large portion of its stiffness at the intended operating temperature, in which case room-temperature mechanical strength alone tells you little about how stable the adhesive will actually be in use. That’s why the temperature dependence of storage modulus can be more informative than a single room-temperature mechanical value.
What Does Tan δ Tell Us?
Tan δ is often introduced as a convenient way to pin down Tg, and the peak position certainly matters. But in formulation development, I never looked only at where the tan δ peak sat — I paid attention to its height, width, and shape too.
A broad tan δ peak can mean the relaxation process spreads across a wider temperature range, which in some systems may point to structural heterogeneity, variations in local mobility, phase-separated morphology, or a distribution of network environments. A relatively sharp peak means the dominant relaxation happens over a narrower range. That said, peak width or shape shouldn’t be read as a direct measure of structural uniformity on its own — molecular structure, crosslink density distribution, phase morphology, additives, measurement frequency, and other factors can all shape the relaxation curve. This is why the curve’s shape is often more informative than the peak temperature alone.
Never Interpret a DMA Curve in Isolation
The same principle that applies to DSC and TMA applies here. It’s unusual to draw a conclusion from a single DMA curve in development work — instead, we compare curves after changing the formulation, filler content, cure conditions, aging conditions, or other processing variables, and ask how much the storage modulus changed, whether the tan δ peak shifted, whether it broadened or narrowed, whether the modulus dropped more sharply around Tg, and whether the rubbery-region modulus changed.
A formulation change, for instance, might produce almost no difference in Tg while significantly shifting the storage modulus above Tg — a meaningful result in its own right, since it tells you the network structure or mechanical response changed without much movement in the apparent glass-transition temperature. That’s one reason I treated DMA as a comparative development tool rather than simply a Tg measurement instrument.
In Practice, DMA Was Often More Useful for Predicting Adhesive Stability
An adhesive isn’t necessarily good just because it has high strength — it also needs to hold onto sufficient stiffness under the actual operating conditions, which matters especially for cured epoxy adhesives since the cured material forms a three-dimensional crosslinked network.
Say an adhesive is meant to operate at 80°C. Knowing its room-temperature strength is useful, but knowing how much storage modulus remains at 80°C can be even more relevant to its actual mechanical stability. A material with a high room-temperature modulus but a dramatic modulus loss near the operating temperature can behave very differently from a formulation that holds its stiffness more effectively. This is where DMA becomes especially useful — instead of asking only “what is the Tg?”, we can also ask how much mechanical stiffness the material retains at the temperature where it will actually be used. For an adhesive developer, that’s a much more practical question.
Why Is DMA So Important for Epoxy Formulation?
This gets even more interesting once you look at the structure of cured epoxy. A thermoplastic polymer and a thermoset epoxy respond differently above Tg. A thermoplastic generally grows increasingly mobile as temperature climbs past Tg and, depending on its chemical structure, may eventually reach a highly softened or molten state. A cured epoxy is different — its polymer chains are tied together through a three-dimensional crosslinked network. Above Tg, the molecular segments become much more mobile, but the covalent network keeps the material from simply flowing like a low-viscosity melt. Instead, it enters a rubbery plateau region, where its modulus stays substantially above that of a freely flowing polymer melt. This rubbery-region modulus carries valuable information about the network structure.
The Rubbery Plateau and Crosslink Density
In the rubbery region, an ideal polymer network’s mechanical response can be tied to the density of elastically active network chains. Under classical rubber-elasticity assumptions, storage modulus can be approximated as:
E’rubbery ≈ 3νeRT
where E’rubbery is the storage modulus in the rubbery region, νe is the concentration of elastically active network chains, R is the gas constant, and T is absolute temperature.
Network-chain concentration can also be expressed approximately in terms of the molecular weight between crosslinks, Mc:
νe ≈ ρ / Mc
which gives:
E’rubbery ≈ 3ρRT / Mc
where ρ is polymer density and Mc is molecular weight between crosslinks.
This gives us an important relationship: under the appropriate rubber-elasticity assumptions, a higher rubbery-region storage modulus corresponds to a higher concentration of elastically effective network chains, or a lower molecular weight between crosslinks. That’s one reason DMA is so valuable in epoxy formulation research — by measuring storage modulus at a defined temperature well above Tg, we can use the rubbery modulus to estimate the cured network’s effective crosslink density.
It’s worth remembering, though, that this is an estimate based on a theoretical model. Real epoxy networks can deviate from ideal rubber elasticity because of network defects, dangling chains, loops, physical interactions, phase separation, fillers, and other structural factors, so the calculated value should be treated as an effective network parameter rather than a direct microscopic count of every chemical crosslink.
What Does This Tell an Epoxy Developer?
This is where DMA gets particularly interesting. DSC can use the heat released during curing to estimate degree of cure. TMA can reveal dimensional changes tied to thermal expansion and, in some cases, post-curing. DMA offers a different angle: it shows how the resulting mechanical network structure responds to temperature.
Two formulations might share similar Tg values but show significantly different rubbery-region moduli — a sign that their effective network structures differ. Likewise, a change in equivalent ratio, cure temperature, cure time, or modifier content might barely move Tg while producing a substantial shift in rubbery modulus. In formulation development, that can be an important clue that the network structure has changed even though Tg hasn’t moved much — one of the reasons DMA became such a valuable tool for understanding cured epoxy systems.
A Developer’s Perspective: Look Beyond Tg
After years working with epoxy formulations, I learned that Tg is useful but rarely the whole story. A formulation can carry a high Tg but poor modulus retention at the operating temperature; a similar Tg but a different rubbery modulus; a broad tan δ peak pointing to a broader relaxation process; or a completely different storage modulus profile despite a similar Tg — and these differences can matter a great deal in an actual adhesive application.
So when I looked at a DMA graph, I never asked only “what is the Tg?” I also asked how the material loses stiffness, how much stiffness remains at the operating temperature, what the rubbery plateau tells me about the network, and what changed compared with the previous formulation. Those questions often gave me more useful development information than the Tg value itself.
DMA Is More Than a Tg Measurement
DSC tells us how the material responds thermally. TMA tells us how its dimensions respond to temperature. DMA tells us how its mechanical response changes as molecular mobility increases. For epoxy formulation development, that makes DMA particularly valuable, since the measurement ties together molecular relaxation, mechanical stiffness, damping behavior, and network structure in a single temperature-dependent experiment. That’s why, in practical development work, I came to rely on DMA curves so heavily.
In the next article, Reading a DMA Curve ②: What Does Storage Modulus Really Tell Us?, I’ll take a closer look at storage modulus and what the shape and temperature dependence of the curve can tell us about an epoxy adhesive.
