Reading a time temperature transformation curve for metallurgy
Seth Price | August 15, 2024The science of metallurgy is not simply melting a few things together and expecting to get the advantages of both materials. When the liquid metal cools, the atoms are arranging themselves in the most energetically favorable place they can. However, if the cooling rate is fast, they may not have much time to find that place.
When two elements are combined to form an alloy, the first stop for understanding their physical properties is a phase diagram. Phase diagrams assume equilibrium cooling, meaning the metal has as much time as it needs to reach its equilibrium state. Phase diagrams will be covered in a separate article.
The reality is that many times equilibrium cooling is not possible, not economical, or not desirable, depending on the alloy being formed. To determine what will happen to an alloy as it cools under non-equilibrium conditions, metallurgists use a Time Temperature Transformation (TTT) curve.
What is a TTT curve?
A TTT curve plots the temperature on the y-axis and the time on the x-axis. The body of the graph consists of regions that represent different microstructures. As the metal cools over time, it passes through different regions, forming different microstructures. Because microstructure determines physical properties, this cooling curve can help create the best alloy for a given situation.
Purposes
A TTT is used for determining what microstructures are possible with different cooling rates. In steel production, they are used extensively to produce either pearlite or martensite (that will later be tempered) microstructures, depending on the specific application.
Using a TTT
Each TTT is for a specific composition. Therefore, the proper TTT must be selected for the given alloy, as 1076 steel will have a different TTT than 1018 steel, both of which will have a different TTT than 4340 steel, the latter of which has nickel, chromium and molybdenum added to it to increase hardenability.
The first step is to start at time = 0 and some temperature in the austenite region. The “clock” does not start until the temperature has dropped below the eutectoid temperature, marked on the TTT.
From there, drop the temperature in accordance with the cooling rate. The time axis is logarithmic, so keep this in mind when calculating and plotting the temperature at any given time. As the temperature curve passes through different regions on the plot, new microstructures begin to form.
Once austenite has been converted to a different microstructure, it cannot be converted back unless the temperature is raised above the eutectoid temperature again. After the austenite has been completely converted to another microstructure, the process is finished, at least for carbon steels. For some austenitic stainless steels, the austenite region extends all the way down to room temperature (and below).
Confused? Here are a few examples of how this works, using a plain carbon steel TTT curve.
Example #1
A steel sample is held at 800° C for a long period of time to homogenize. Then it is quenched rapidly in a water bath, dropping from 800° C to 100° C in 10 seconds. In this case, the holding temperature was above the eutectoid temperature, so the microstructure is originally 100% austenite. Then, it is rapidly quenched. The clock and the temperature start when the metal reaches the eutectoid temperature, so the graph traces the red line shown below.
Notice that the only line the red line crosses is the Ms, or martensitic transformation temperature. The alloy would have a microstructure that is 100% martensite. “As quenched” martensite is often undesirable, as it is quite brittle, but it can be tempered, making it much more useful as a mill product.
Example #2
Consider another sample of steel that starts the same way, homogenized austenite. However, this time it cools to 550° C in about 5 seconds. Then, it cools all the way down to room temperature in another 5 seconds (for a total of 10 seconds). It will follow the red trace shown below:
In this case, about half of the austenite forms pearlite, as it passes through the first curve and stops at the 50% line. While austenite is not stable at room temperature (for this alloy), pearlite is. As the curve passes back through the same curve, the 50% pearlite is locked into place, but the remaining 50% austenite is converted into martensite.
Example #3
Starting in the same place, a sample’s temperature drops to 550° C in 10 seconds. Then, it is allowed to cool down to around room temperature by a half an hour or so (~1,800 seconds). This cooling condition follows the curve shown below:
In this case, the sample is converted to 100% pearlite. Pearlite is stable, so it remains pearlite, even though it crosses several lines after its initial transformation. This is one of many ways to produce a pearlite structure, which has some desirable physical properties.
Example #4
Suppose instead, there is a need for a sample to have a 100% bainite microstructure. What heating condition could be used to get there? There are an infinite number of right answers, but the key is to cool rapidly at the beginning and then hold the temperature constant for a while. Finally, quench the sample to room temperature. One possible solution is shown below:
This brings out an important point. The temptation may be to reduce the amount of time a part spends in the furnace or to bring down the temperature of the furnace, both in an effort to save energy and reduce emissions. A metallurgist should carefully study the TTT curve to make sure the microstructure will not change, and then perform some energy usage calculations. Notice that the flat line, called the dwell time, in the furnace can be reduced by raising the dwelling or soaking temperature; there is a tradeoff occurring here that must be balanced to ensure the microstructure meets design criteria.
Final thoughts
At first glance, TTTs can be tricky to understand. The first step is to recognize how they relate to the phase diagram and then begin to think in terms of the atomic motion and what is happening at the atomic, then microscopic level. From there, also recognize that the time scale is in logarithmic scale versus linear scale.