Temperature, time and structural change

A short hot treatment and a long warm treatment

A treatment record says “150 °C”. A second adds “40 minutes”. Neither tells whether the material spent most of those minutes at 150 °C or cooling away from it. The preceding lesson showed why the order of operations matters. Here we examine the temperature path within one operation and the structural changes that continue while time passes.

Atomic rearrangements require pathways and time. Temperature changes the rates of many diffusion-controlled processes, while thermodynamics determines which states are favoured. A treatment temperature alone therefore cannot specify the final microstructure.

Separate two questions. Under the specified temperature, pressure and composition, would a different arrangement lower the appropriate free energy? Free energy is the thermodynamic quantity used to compare those possible states under the stated conditions. And can atoms reach that arrangement on the available timescale? Atomic mobility describes how readily the required movements occur. An energetically favoured change need not be fast enough to observe.

For example, solute atoms dissolved in a solid may have a tendency to form particles of a different composition, but must redistribute through the surrounding solid. Diffusion is that atomic redistribution through repeated local movements. Many such movements require atoms to overcome energy barriers; raising temperature commonly makes the necessary jumps more frequent. The material need not melt for this to happen.

Temperature also changes which states are favoured and how much solute can remain dissolved. Creating a new particle introduces an interface with the matrix, the surrounding continuous material; this costs free energy and can oppose the initial formation of a particle even when a sufficiently developed new state is favoured. Thus “hotter means more change” is not a complete rule. Both the accessible route and the thermodynamic conditions matter.

Prepare a state that can change during ageing

Use one representative family of precipitation-hardenable alloys. The solubility of the relevant solute is sufficiently greater at a suitable high temperature than at the later ageing temperature. A solution treatment dissolves the intended solute in the matrix. Sufficiently rapid cooling, called quenching, can retain more dissolved solute than would remain in equilibrium at the lower temperature. This supersaturated solid solution is a prepared starting state, not a new overall chemical composition.

Such a state can persist because redistribution is slow, even though a different distribution is favoured. This is an example of metastability: the state is retained because a route to change faces barriers, rather than because it is the lowest-free-energy state available under those conditions. Quenching need not stop every later rearrangement; storage temperature and delay can matter.

During a suitable subsequent ageing treatment, atoms redistribute. Solute-rich clusters and then precipitate particles can develop; early clusters need not already have all the structure and composition of the final equilibrium phase. The particles and surrounding matrix have an evolving arrangement and composition. A precipitate is not simply a small grain: grains distinguish crystal orientations, whereas precipitation here creates chemically and structurally distinct regions within the matrix.

The whole specimen can retain the same bulk composition while solute moves out of the matrix into particles. Starting with undissolved particles, a different prior deformation or a different quench can therefore produce a different response under the same later temperature schedule. The sequence above is a useful mechanism, not a heat-treatment recipe for every metal.

Why ageing can first strengthen and later weaken

Precipitates may form and later coarsen; deformed grains may recover and recrystallise. Cooling can preserve a state or allow further changes. Comparing treatments requires the starting structure and the whole relevant thermal history, not an assumption that every increase in temperature improves a property.

In the representative crystalline alloy, permanent deformation involves dislocation motion. A dislocation is a line defect, not an open crack. Particles can obstruct its movement. In an early regime, dislocations can cut through small particles; changes in the population and particle state can make that cutting harder. The stress required for plastic flow can rise as an effective dispersion develops. This is precipitation strengthening, one contribution to the measured resistance to permanent deformation.

Later, larger particles may be bypassed: a dislocation bows between obstacles and leaves loops around them. When continued ageing produces fewer, larger particles with wider gaps between them, bypassing can become easier. The strengthening contribution can then decrease. A fall after a peak is often called overageing. This describes a state relative to a chosen response; it does not mean that every useful property has become worse or that all strengthening has disappeared.

One route to that later state is coarsening. Smaller particles can dissolve, and their atoms diffuse through the matrix to larger particles. Fewer larger particles can have less total particle–matrix interface for the same total particle volume. Reducing an interface contribution to free energy can drive this redistribution. Whole particles need not travel through the solid and collide.

A peak in strength is therefore not proof that diffusion has stopped, that equilibrium has been reached, or that the largest possible particle volume has just formed. Different mechanisms can dominate at different stages. Particle size, spacing, composition, structure and interactions all matter; a property curve alone does not uniquely identify which of them changed.

Read the temperature path of the material

The first graph gives invented, piecewise-linear local specimen temperatures. Both paths start at 25 °C, reach 150 °C and return to 25 °C after 40 min. Path A reaches 150 °C at 5 min, holds until 35 min, then cools. Path B reaches 150 °C at 5 min, holds only until 10 min, then cools much more slowly. A hold is the interval maintained at the stated temperature; its duration is the difference between its end and start times.

A therefore has a 30 min hold and a 5 min cooling interval; B has a 5 min hold and a 30 min cooling interval. Equal peak temperature and total elapsed time hide this difference. Structural change can occur during ramps and cooling as well as during the hold. The longer peak hold does not, by itself, establish which final specimen has greater strength.

Do not replace the paths with an uncalibrated “temperature × time” score or simply add minutes spent at different temperatures as if each minute had the same structural effect. The rates and even the active mechanisms depend on temperature and the evolving state. These paths do not supply a numerical final property without additional evidence or a justified model.

A furnace setpoint is also different from the temperature within a workpiece. Heat transfer and heat capacity govern how the specimen responds, and a thick piece can have different surface and interior paths. Specify where temperature was measured, when the hold clock started, and what heating and cooling were included. One surface sensor does not by itself establish an identical path everywhere.

Read a property–time record without inventing a recipe

The following table is a separate, invented ageing record, not measurements of a named alloy and not results calculated for paths A and B. Separate matched coupons begin in the same prepared supersaturated state. They receive the listed isothermal hold, meaning a constant-temperature hold, at either 150 or 200 °C. For this simplified record only, entry into the hold and subsequent rapid cooling are assumed to have negligible structural effect. The 0 h entry represents the common prepared state before ageing.

After cooling, each coupon is tested at the same temperature, direction and rate using the same 0.2% proof-stress definition introduced in the history lesson. That endpoint is the intersection of the tensile curve with a line parallel to the initial elastic slope and offset by strain 0.002; engineering proof stress uses the original cross-sectional area. It is a specified resistance to permanent deformation, not elastic stiffness. Each tensile test sacrifices its coupon, so the table does not repeatedly test one surviving specimen.

Among the listed 150 °C results, the highest proof stress is 320 MPa after 4 h; the 8 h value is lower. Among the listed 200 °C results, the highest is 280 MPa after 1 h. The hotter hold gives the greater value at 1 h, but the cooler hold gives the greater value at 2 h. The ranking depends on both time and temperature. The graph joins the supplied points as a reading aid, not a fitted kinetic law; the true maximum might lie between sampled times.

The rise and fall are consistent with the particle-obstacle explanation, but these property values do not prove a unique particle size, phase fraction or microscopic mechanism. Suitable structural measurements would test that interpretation. For real selection, uncertainty, repeatability, final condition, other required properties and component conditions also matter; the invented exact numbers carry no experimental uncertainty claim.

Carry the structural state into the comparison

For a useful comparison, retain the starting structure, local temperature path, relevant deformation, storage and cooling, then the final test conditions. A coupon after one hour at 150 °C is a changed starting state for any subsequent 200 °C hold. Its response cannot be found by adding two proof stresses from separate single-temperature records.

Choose observations that can distinguish explanations. If the proposed reason for a later decrease is coarsening, compare particle sizes and spacing with a method that can resolve those particles, rather than relying only on an optical image of much larger grains. If the workpiece was nonuniformly heated, compare locations rather than assuming one coupon represents the entire part.

The essential chain remains processing → structure → property → performance. This lesson explains why the first arrow needs a path and a starting state. It supplies neither a universal ageing temperature nor a safe load for a finished component. The later diffusion and transformation lessons develop quantitative rates and process models; here, the task is to make the state and the evidence explicit.

Worked examples

Try the reasoning

1. Two records both say “150 °C, 40 min”, and both specimens begin and end at 25 °C. Does identical peak temperature imply an identical treatment? What would you request?

   Show the answer and reasoning

   No. Request the local specimen temperature–time paths, starting states, hold definitions and cooling records.

   The A/B example meets those endpoints but has 30 versus 5 min at the peak and different cooling intervals. Heating and cooling can also change the structure. A furnace setpoint is not proof of every specimen location’s temperature.

2. A batch has the same bulk alloy composition as the ageing-table coupons, but it entered ageing with substantial undissolved particles rather than the common supersaturated state. Why identify that starting state? Can the table be assigned to it unchanged?

   Show the answer and reasoning

   The starting state determines which redistribution and transformations are available. Matching composition does not justify assigning the table unchanged.

   Already formed particles change how solute is distributed and what can nucleate or grow. The later schedule acts on that actual structure. Establish a comparable preparation and supporting evidence rather than treating an alloy label as a complete initial condition.

3. For the invented record, a screening task requires proof stress of at least 275 MPa and an ageing hold no longer than 2 h. Among the listed temperature–time combinations, which qualifies? Would this settle a real component selection?

   Show the answer and reasoning

   Only the listed 200 °C, 1 h condition qualifies, at 280 MPa. This arithmetic screening does not settle a real selection.

   At 150 °C the eligible 0, 1 and 2 h values are 100, 180 and 260 MPa; all are below 275. At 200 °C they are 100, 280 and 250 MPa; only 280 passes. The 320 MPa point at 4 h violates the time limit. A real 5 MPa margin would need uncertainty and repeatability assessment, and other properties and service requirements remain unspecified. No unmeasured intermediate point is admitted.

4. A coupon receives 1 h at 150 °C followed by 1 h at 200 °C. A technician adds 180 and 280 MPa from the table to report 460 MPa. Another reads 250 MPa at 2 h and 200 °C. Are either of these predictions justified?

   Show the answer and reasoning

   Neither follows from the supplied record; no final proof stress is determined for that two-stage path.

   The table gives separate single-temperature holds starting from a common state. The first hour changes the state that enters the second. Proof stresses are responses, not additive increments, and the two-stage path is not a 2 h isothermal hold at 200 °C. A suitable path-dependent model or matching test data would be needed.

5. In another idealised dispersion of identical spheres, the radius becomes three times larger while total precipitate volume stays constant. Find the new particle number and total interface area relative to the old values. Does the area ratio give the strength ratio?

   Show the answer and reasoning

   The number becomes 1/27 as large and the total interface area becomes 1/3 as large. No strength ratio follows.

   Fixed N r³ requires N_new = N_old/3³. Area then changes by (1/27) × 3² = 1/3. For a finite count, take an initial number divisible by 27. The geometry supports an interface-area argument under the stated model; obstacle resistance also depends on spacing, particle state and the active dislocation mechanism, so it is not proportional to this area by any supplied rule.

6. The largest listed 150 °C result is 320 MPa at 4 h. A report calls 4.000 h “the exact optimum treatment” and declares the structure to be at equilibrium. What does the table actually establish, and what evidence is missing?

   Show the answer and reasoning

   It establishes only the highest of the sampled values for this response and preparation. Neither an exact optimum nor equilibrium has been established.

   Sample more closely around the apparent peak and assess repeatability and uncertainty to locate a practical maximum. Define the other objectives before calling a treatment optimal. Structural observations and a thermodynamic assessment, rather than the top of a property curve, would be needed to evaluate equilibrium. Continued ageing can still change the particle dispersion beyond a strength peak.

What to carry forward

• Thermodynamics identifies a tendency to change; atomic pathways and mobility determine whether change can proceed on the available timescale.
• A representative precipitation sequence can raise resistance to plastic flow, then reduce that contribution as the dispersion changes; continued time need not improve a chosen property.
• Compare local temperature paths, starting states and final test conditions. A peak temperature, an elapsed time or a furnace setpoint alone is incomplete.
• A property–time curve is evidence for a specified preparation and test. It is neither a unique microstructure decoder nor a universal heat-treatment recipe.