Materials Science · Processing routes and microstructural evolution

Powder consolidation and polymer processing

Useful background: Processing, structure and useful properties; Energy, stability and environmental context. Each worked case explains the quantities it uses.

Back to the chapter

After reading, you can

Compare consolidation and polymer shaping by their mechanisms.
Distinguish particle contact, neck growth and densification without assuming complete melting.
Use mass, solid density and an explicitly supplied envelope volume to calculate porosity and conditional shrinkage.
Distinguish melt shaping, solvent removal and chemical network formation in polymer processing.
Interpret a bounded reaction-heat conversion record and explain what a bulk average can hide.
Select internal-state evidence matching a stated local requirement rather than accepting external shape alone.

Different feedstocks require different routes

Powder consolidation bonds particles and may remove pores through pressure and thermal transport. Polymer shaping can use melt flow, solution routes or chemical curing. These operations have different requirements even when they produce similar external shapes.

A mould can give a powder compact and a polymer part the same outline, while their interiors develop by different routes. A useful comparison follows the feedstock, the mechanism that permits shaping, the process that fixes or bonds the structure, and the evidence needed at the end. Shape is one observation in that chain.

A powder contains separate particles with spaces between them. Pressing can rearrange the particles and deform some contacts, increasing contact area and reducing some void space. The resulting green compact is the shaped body before the later consolidation treatment; its handling strength can include temporary binder as well as particle contacts. Binder removal, particle bonding and pore reduction are distinct changes. Actual routes may combine pressure, heat and a liquid phase; no one sequence applies to every powder.

A larger contact need not mean a denser body

During solid-state sintering, material redistributes between surfaces, contacts and the particle interiors. Reducing the free-surface contribution can favour forming a neck, a bonded bridge where particles meet. Atoms require an accessible transport route and time; a favourable energy change does not guarantee a fast process. Complete melting is not necessary for this route.

Neck growth and densification answer different questions. In a simple pair of particles, surface-sourced transport can enlarge the neck while leaving their centres approximately as far apart as before. Some bonding can improve without a decrease in the whole body’s pore volume. Transport from the interparticle grain-boundary/contact region to the neck can let the centres approach and reduce pore volume. Transport through a particle interior can densify or not, depending on where the transported material originates. Which route supplies material matters, not just whether the word diffusion appears.

Real compacts contain many contacts, sizes and pore shapes. Pressure gradients, friction, agglomerates and remaining binder can make the initial arrangement uneven. As pores close, trapped gas can resist further shrinkage; grain growth and changing transport distances can alter later evolution. A smooth surface or one photographed neck does not measure the density or bonding throughout the interior. These are mechanisms and possible limits, not a temperature/pressure recipe or a rate law.

Keep solid volume separate from the body’s envelope

Use an original fictional single-solid-phase record. The solid has density ρ_s = 6.00 g/cm³ at the common measurement temperature. After any binder removal, mass m = 36.0 g remains constant, with no reaction, phase change, liquid uptake or material loss. The supplied envelope volume V_b includes the solid and every pore inside the outside boundary. It is not a skeletal volume that excludes accessible pores. These are ideal supplied quantities, not an instruction for a particular density apparatus.

The solid occupies V_s = m/ρ_s = 6.00 cm³. Define bulk density ρ_b = m/V_b, relative density D = ρ_b/ρ_s = V_s/V_b, and pore fraction φ = V_pore/V_b = 1 − D. All fractions here are volume fractions. Equal mass is not equal bulk density when the envelope changes.

Initially V_b = 10.0 cm³, so ρ_b = 3.60 g/cm³, D = 0.600 and φ = 0.400. After the stipulated change V_b = 8.00 cm³, so ρ_b = 4.50 g/cm³, D = 0.750 and φ = 0.250. Pore volume fell from 4.00 to 2.00 cm³ while solid volume stayed 6.00 cm³. A 20.0% decrease of total volume is therefore not a 20-percentage-point decrease of porosity.

Only if shrinkage is uniform and isotropic, meaning equal in all three perpendicular length directions, can its length factor s be inferred as s³ = V_final/V_initial = 0.800. Then s = 0.800^(1/3) ≈ 0.928318: each length decreases by about 7.17%. The cube root is the positive number whose cube is 0.800. The volume ratio alone does not prove isotropy. A body that shrinks 20% in one direction and not in the other two has the same volume ratio.

This conservation account diagnoses supplied states; it does not predict the time, pressure or temperature needed to reach them. Its calculation fails if the weighed mass includes unknown binder, the solid density changes, or the volume convention changes between observations. Density also leaves pore connectivity, shape, direction and contact quality unresolved.

Original fictional constant-mass, constant-solid-density record at matched temperature. The supplied envelope includes all internal pore volume.
State	m (g)	V_b (cm³)	V_s (cm³)	V_pore (cm³)	φ
Initial	36.0	10.0	6.00	4.00	0.400
Later	36.0	8.00	6.00	2.00	0.250

Original invented accounts, not a micrograph or a sintering trajectory. Upper panel: common volume scale zero to ten cubic centimetres. A constant thirty-six gram body contains six cubic centimetres of solid in both states. Initially four cubic centimetres of pores give an envelope of ten; later two give eight. Filled blue means solid and hatched orange means pores. Lower panel is a separate forty-two gram body with two equal four-cubic-centimetre regions. Outer porosity is five percent, core twenty percent and the combined bulk value twelve point five percent. Bar widths share a zero to twenty-five percent scale. A dashed line at fifteen percent is the invented local maximum; the core exceeds it even though the bulk average does not. The bulk bar combines the regions and is not a third independent specimen. Neither diagram depicts physical layer thickness or certifies a property.

Flow, drying and curing fix structure in different ways

For melt shaping of a suitable thermoplastic, sufficiently mobile chains can move relative to one another and the material flows into or through a shaping boundary. Cooling then slows molecular rearrangement; a semicrystalline material can also develop crystallites. An amorphous thermoplastic does not need a crystalline melting transition to soften. Flow and later drawing can orient chains or other constituents, while relaxation before the structure becomes fixed can reduce some orientation. Temperature and time therefore influence the retained state as well as whether a cavity filled.

Extrusion produces a continuing shape through a die; injection moulding fills a finite cavity and then the part is allowed to become dimensionally stable for removal. The same polymer can retain different orientation, crystallinity, voids or residual stress after different histories. The lesson supplies no universal processing temperature, cooling time or claim that every thermoplastic is hard at room temperature.

In a solution route, a suitable solvent allows polymer molecules to be carried and distributed. Removing solvent can leave a film or other shaped structure. Polymer concentration and mobility change during removal; a surface can become less mobile while solvent remains inside. Drying changes a mass/transport account. It does not by itself establish that a chemical network has formed. Some real formulations combine solvent removal with chemical reactions, so the two accounts must then be tracked separately.

For a thermosetting formulation, curing forms or extends chemical connections into a network. Viscosity can rise and a connected gel can develop before the chosen reactions are complete. Reduced molecular mobility can slow later reaction; temperature gradients and reaction heat can also make the interior’s history differ from the outside. An ordinary permanently crosslinked network does not simply resume the same chain-sliding melt flow on reheating. That statement does not rule out specialised exchangeable networks or other reprocessing routes.

The practical question is which change fixes the intended state: cooling/physical arrangement, removal of a carrier, chemical network formation, or a justified combination. A stiff or nonflowing surface is not a universal measure of complete cure. Heating alone is not a mechanism label; it can promote flow, reaction, solvent removal or degradation depending on material and conditions.

Define what a cure number measures

Consider a separate fictional exothermic formulation with a justified calorimetric reference h_total = 300 J/g for its initially uncured state. Heat is normalised per gram of the same entire formulation, including the same filler fraction. Stipulate that all relevant reaction heat can be collected with a suitable baseline, no other thermal event overlaps it, and reaction progress is proportional to that heat. No material-specific kinetic law or operating schedule is supplied.

Differential scanning calorimetry compares a sample’s heat response with a reference. For this particular model, let h_rem be the positive residual reaction heat per gram available after the part’s prior history. The reacted portion is h_done = h_total − h_rem. Define α = h_done/h_total = 1 − h_rem/h_total. It is a dimensionless reaction-conversion measure under the stated assumptions, not the fraction of final strength or stiffness. A remaining reaction tail can matter even when a specimen already holds its shape.

The table supplies equal-mass outer and core regions, each 20.0 g, with residual heats 15.0 and 75.0 J/g. Their α values are 0.950 and 0.750. The combined residual heat is (20 × 15 + 20 × 75) J = 1800 J, or 45.0 J/g over the full 40.0 g. Its bulk α is therefore 0.850. This is a mass-weighted average because the heat record uses energy per mass; it is not a volume average borrowed from the powder calculation.

Suppose an invented local screening requirement is α ≥ 0.800 in both named regions. The bulk value passes 0.800, but the core fails the actual requirement. A spatially uniform specimen with α = 0.850 everywhere would have the same bulk value and a different local outcome. The whole-part average cannot choose between those states.

Actual cure reactions can overlap, have several steps and become mobility limited. A shifted peak, a low residual heat or one fitted number is not a universal network or service-property measurement. Check reference formulation, filler and mass basis, sample location, baseline, thermal events and the intended property test before transferring α. The supplied local screen is an editorial exercise, not a standard or a real acceptance limit.

Original fictional residual-heat observations. Each region is 20.0 g of the same formulation; h_total = 300 J/g. These values do not represent a named resin or a cure prescription.
Region	Mass (g)	h_rem (J/g)	α
Outer	20.0	15.0	0.950
Core	20.0	75.0	0.750
Whole-part average	40.0	45.0	0.850

Original invented proportional reaction-heat account for one formulation with reference heat three hundred joules per gram of the entire formulation. Outer and core regions each have twenty grams. Solid blue bars show reacted heat fractions of zero point nine five and zero point seven five; orange hatched remainders are zero point zero five and zero point two five. The forty-gram whole-part average is zero point eight five reacted and zero point one five remaining. All widths use the same zero to one fraction scale; they do not encode length, mass or strength. Dashed threshold at zero point eight is an invented local minimum. The core fails it, although the whole-part average exceeds it. Equal region masses justify this particular arithmetic mean. Baseline, common formulation, complete relevant heat collection and proportionality are supplied assumptions; this is not an experimental resin record or a cure schedule.

The route leaves evidence

Particle boundaries, voids, orientation and cure state can remain in the finished material. Inspect how feedstock quality and processing gradients affect the final structure. A dense external surface does not prove complete internal consolidation or cure.

A second powder body illustrates the corresponding spatial question with a different measured quantity. It has two supplied equal-volume regions, 4.00 cm³ each, of the same 6.00 g/cm³ solid. Their pore fractions are 0.0500 and 0.200. Their solid volumes are 3.80 and 3.20 cm³, so the combined mass is 42.0 g and total envelope 8.00 cm³. The bulk pore fraction is (0.200 + 0.800)/8.00 = 0.125.

An invented requirement φ ≤ 0.150 in each region is failed by the 0.200 core, despite the 0.125 whole-body value. A uniform 0.125 body has the same mass, envelope and solid density, yet meets that particular local screen. Neither bulk value supplies a complete strength, leak or durability assessment. These two supplied powder states are not the 36.0 g shrinkage specimen.

The cure and powder records share an inference lesson, while their mechanisms and weights differ. Pore fraction concerns how volume is partitioned; reaction conversion concerns a defined heat account. Microscopy or suitable three-dimensional pore evidence can test spatial consolidation; representative chemical/thermal and property evidence can test cure. A surface inspection cannot replace every interior measurement, and a small sampled region does not automatically represent the complete component.

Choose evidence for the required internal state

Start a route comparison with the required geometry and material state. Can the feedstock flow, rearrange or be distributed in a carrier? What establishes the final contacts, molecular arrangement or network? What has to leave, such as a temporary binder or solvent, and what can remain? Then identify the corresponding mass, thermal, chemical and structural observations.

For the powder case, a larger neck or greater bulk density can be useful evidence without proving complete pore removal or a particular property. For the melt-shaped polymer, cooling and flow history can matter without a curing reaction. For the reactive polymer, a correct outline and a hard skin can coexist with a less converted interior. The original worked case asks exactly why dimensional agreement alone is insufficient.

Compare representative states under matched measurement conditions and retain the uncertainty and sampling extent. Density, conversion and dimensions each answer a specified question. An application needs compatible property and component evidence as well; none of the fictional screens establishes a safe operating range, optimum route or service life.

Worked examples

Work through the case

A polymer part has the right dimensions but an incompletely cured interior. Is shape agreement sufficient acceptance?

Dimensions assess one requirement.
Cure state influences network and response.
Measure relevant internal state and performance.

No. External shape does not establish the intended material state.

Read shrinkage without losing the pore account

Use the 36.0 g powder record. Find initial/final pore volumes and fractions, volume shrinkage and the conditional isotropic length shrinkage. Why is 7.17% not a measured directional result?

Constant solid volume is 36.0/6.00 = 6.00 cm³. Subtract from each envelope: 10.0 − 6.00 = 4.00 cm³ and 8.00 − 6.00 = 2.00 cm³ of pores.
Divide by the matching envelopes: φ_initial = 4/10 = 0.400 and φ_final = 2/8 = 0.250. The 15-percentage-point difference is not the 20% decrease in envelope volume.
If every length scales by s, then s³ = 8/10. The positive cube root gives s≈0.928318 and 100(1 − s)≈7.168%.
Only the volumes were supplied. Unequal directional shrinkage can produce the same product; the isotropic result needs that extra assumption.

Pores fall from 4.00 to 2.00 cm³ and 40.0% to 25.0%; total volume shrinks 20.0%. Equal-direction shrinkage would be about 7.17%, conditional on isotropy.

A passing whole-body value can hide a failing region

Use the separate 42.0 g two-region powder body and the invented local φ ≤ 0.150 screen. Compare it with a uniform body of the same mass, envelope and solid density. Can bulk density distinguish acceptance?

The outer solid volume is 4.00(1 − 0.0500) = 3.80 cm³; the core has 4.00(1 − 0.200) = 3.20 cm³. Their sum 7.00 cm³ gives 42.0 g.
Pore volumes 0.200 and 0.800 cm³ total 1.00 cm³ over 8.00 cm³: φ_bulk = 0.125 and ρ_b = 5.25 g/cm³.
The actual core 0.200 exceeds 0.150. A uniform 0.125 body would meet the supplied local screen although its mass and bulk density are identical.
A complete property requirement would need additional evidence; this comparison only decides the stated local porosity screen.

The supplied gradient fails the local screen and the uniform alternative passes it. Bulk density alone gives the same answer for both and cannot resolve the difference.

Keep the residual-heat reference and the sampling location

For the two 20.0 g regions, h_total = 300 J/g. Calculate each conversion and the whole-part conversion. Does bulk α ≥ 0.800 establish α ≥ 0.800 in each region?

The outer reacted heat is 300 − 15 = 285 J/g, giving α_outer = 0.950. The core has 300 − 75 = 225 J/g, giving α_core = 0.750.
Remaining total heat is 20 × 15 + 20 × 75 = 1800 J. Divide by 40.0 g to obtain 45.0 J/g.
Whole-part α = 1 − 45/300 = 0.850, equivalently the equal-mass mean of 0.950 and 0.750.
The core is below the local 0.800 requirement. The common bulk value cannot establish each region or a mechanical property.

Outer 0.950, core 0.750, bulk 0.850. The local core screen fails despite the passing bulk average.

Try the reasoning

Does powder bonding require complete melting?
Show the answer and reasoning
No.
Solid-state mechanisms can contribute.
Why record flow direction in polymer processing?
Show the answer and reasoning
It can establish orientation.
The finished response may be anisotropic.
A supplied sintering observation shows wider necks but unchanged mass, solid density and envelope volume. Must the two records contradict one another?
Show the answer and reasoning
No. The stated bulk porosity remains unchanged while some contacts can grow.
Neck growth can occur by transport that does not bring particle centres closer. Check the actual transport and spatial evidence; a neck photograph is not a complete density measurement.
A different single-phase compact has m = 24.0 g, ρ_s = 4.00 g/cm³ and a supplied 8.00 cm³ envelope, all at one temperature with no binder or reaction. Find bulk density, pore volume and pore fraction.
Show the answer and reasoning
ρ_b = 3.00 g/cm³; V_s = 6.00 cm³, V_pore = 2.00 cm³ and φ = 0.250.
Use the matching material density before subtracting volumes. Equal pore fraction with another compact does not establish equal contact quality, chemistry or strength.
An invented rectangular body changes from 2.00 × 1.00 × 4.00 cm to 1.60 × 1.00 × 4.00 cm. Is its 20% volume decrease evidence of 7.17% shrinkage in each direction?
Show the answer and reasoning
No. Only the first direction shrinks 20%; the other two are unchanged.
The volume ratio is the product of three length ratios. A cube-root length factor requires equal factors, an assumption not established by volume alone.
A different same-solid body has a 1.00 cm³ region with φ = 0.300 and a 3.00 cm³ region with φ = 0.100. Find bulk φ. Is the unweighted mean 0.200 correct?
Show the answer and reasoning
Bulk φ = (0.300 × 1.00 + 0.100 × 3.00)/4.00 = 0.150. The unweighted mean is incorrect.
Pore volumes add; then divide by total envelope. The 0.300 region still fails any local 0.200 limit even though the whole-body value is 0.150.
A separate formulation is 50.0% reactive resin by mass and 50.0% inert filler. Its resin-only reference heat is 300 J/g_resin. A specimen has 30.0 J/g_formulation residual heat. Under the same proportional-heat assumptions, find α. May 300 J/g be used directly with 30.0 J/g?
Show the answer and reasoning
The matching reference is 150 J/g_formulation, so α = 1 − 30.0/150 = 0.800. Directly mixing the two mass bases would wrongly give 0.900.
One gram of formulation contains 0.500 g of the reactive resin. Normalise input and remaining heat to the same material mass and composition before taking a ratio.
For the same 300 J/g fictional chemistry, 10.0 g outer material has 15.0 J/g remaining and 30.0 g core has 75.0 J/g remaining. Find whole-part α. Does it equal the equal-region mean 0.850?
Show the answer and reasoning
h_rem, bulk = (10 × 15 + 30 × 75)/40 = 60.0 J/g and α_bulk = 0.800; the equal-region mean is inappropriate.
The core supplies three times as much material to the heat account. Averaging region labels instead of mass changes the physical question; the 0.750 core value is not repaired by averaging.
A purely drying record begins with 12.0 g polymer and 8.00 g solvent. Later total mass is 16.0 g; no polymer is lost or reacts. Find remaining solvent mass and its final mass fraction. Has a chemical cure been demonstrated?
Show the answer and reasoning
4.00 g solvent remains, 25.0% of the final mass. No chemical cure follows from this mass record.
Subtract the conserved polymer mass, then divide by the final total. Solvent departure and network formation are different accounts; their coexistence in another formulation would require separate evidence.
Two shaped parts become rigid: one after cooling a mobile thermoplastic, and one while a reactive formulation forms a network. Can that common rigidity justify reheating both as if they had the same melt-flow route?
Show the answer and reasoning
No. Identify the molecular constraints and reaction history before transferring a route.
Cooling can arrest motion without forming permanent chemical links; ordinary permanent network formation changes connectivity. Shape and rigidity alone do not select the mechanism or a processing prescription.

What to carry forward

Identify the bonding or flow mechanism.
Assess internal state as well as shape.
Neck growth and densification are distinct observations; define the volume and mass basis before inferring porosity.
A cube-root length change needs isotropic shrinkage, not only a measured volume ratio.
Cooling, solvent removal and chemical curing can all fix shape through different mechanisms.
A reaction-heat conversion is a defined chemical account, not a direct strength or stiffness fraction.
Use the correct volume or mass weights, and test local requirements with representative spatial evidence.

References

Teaching resources and research supporting this lesson’s explanations and examples.