Featured Heat Treatment of Coin Dies

Discussion in 'Coin Chat' started by NSP, Dec 21, 2020.

  1. NSP

    NSP Well-Known Member

    One of the beneficial attributes of steel is that it can be heat treated in numerous ways to get a wide variety of properties for many different end applications. One of these applications are the dies used to strike coins. I’ve seen terms like “hardening,” “annealing,” and “quenching” in coin-related publications before, and after working with this subject for a year and a half, I’ve gotten to learn quite a bit about how the various heat treating processes work. I thought it would be helpful to share what goes on when heat treating a coin die, like the 1820 half eagle die shown below (I believe this is the ANA’s image).


    To start out, steel is characterized as an iron-carbon alloy. Other alloying elements (e.g., manganese, chromium, nickel, molybdenum, etc.) are added to achieve a wide array of properties (e.g., strength, hardness, toughness, corrosion resistance, machineability, etc.). At room temperature, steel is generally made up of two phases with different amounts of carbon atoms interspersed amongst the iron atoms: ferrite (extremely low carbon content) and cementite (relatively high carbon content). At high temperatures, though, a third phase is formed: austenite. Austenite is not stable at room temperature, but once the steel is heated hot enough to completely transform to austenite (called “austenitizing”), it can be cooled in different ways to get a variety of different microstructures and properties.

    Depending on how quickly or how slowly the steel is cooled will affect how much the carbon can migrate when the steel tries to return to the ferrite and cementite phases. Time is an important factor because the austenite phase is not stable once it cools below a certain point, so once a piece of steel is taken out of the furnace, the clock starts ticking.

    The transformations can be plotted on a temperature versus time (logarithmic) plot to form a “TTT diagram” (“TTT” stands for “time, temperature, transformation”). It should be noted that each steel composition will have a different TTT diagram based on how the alloying elements affect how quickly the austenite reverts to ferrite and cementite. Below is a generic TTT diagram for a carbon steel that is roughly 0.77% carbon.


    While this TTT diagram is relatively simple compared to diagrams for other steels, there are some things that require explanation. First, for this steel, the dashed line at 727°C represents the temperature above which austenite (Greek letter gamma γ) is stable (the “austenitizing temperature”). Below this line, austenite is no longer stable and begins to transform into ferrite (Greek letter alpha α) and cementite (Fe3C). Depending on how quickly the steel is cooled after dropping below 727°C will determine which microstructures (pearlite, bainite, or martensite) form.

    Pearlite, bainite, and martensite have different mechanical properties. Pearlite has relatively low strength and low hardness, but is fairly tough. Martensite has high strength and high hardness, but has low toughness. Bainite’s properties are somewhere in between. For the sake of clarity, strength, hardness, and toughness are defined as follows:

    Strength: ability to resist deformation when a force is applied.

    Hardness: ability to resist friction/wear.

    Toughness: ability to resist fracturing/breaking when a force is applied.

    To get pearlite or bainite, the steel has to cross through the yellow shaded region. The left side of this region marks where the unstable austenite begins to transform into ferrite and cementite. The right side of this region marks where essentially all of the austenite has transformed into ferrite and cementite. The dashed line in between represents where 50% of the austenite has been transformed. Once the steel “passes” entirely through the yellow shaded region, the transformation process is essentially over and it can be cooled to room temperature without much effect on the microstructure.

    If the austenite was transformed through the yellow shaded region above ~500°C, a pearlite microstructure will form. If the austenite was transformed below ~500°C, a bainite microstructure will form. For this steel, it would be easy to get pearlite to form by cooling it somewhat slowly. This could be accomplished by cooling along the green curve, which represents an annealing heat treatment where the steel is heated above 727°C and allowed to slowly cool with the furnace to room temperature.

    To get martensite to form, the steel must be cooled rapidly enough to avoid crossing into the yellow shaded region. This could be accomplished by rapidly cooling along the red curve, which represents quenching (“hardening”), generally in a liquid medium like water or oil (though depending on the steel it could be gas quenched).

    For this particular steel, you would need to cool it extremely quickly because at the “nose” of the yellow shaded region, the austenite will start to transform in less than 1 second around 550°C. Therefore, for the steel represented by this particular TTT diagram, it would be very difficult to completely transform all of the austenite to martensite. However, die steels used to make coin dies include other alloying elements that shift the yellow region further to the right and allow a little more time to get past the “nose.” This makes it possible to quench them and successfully achieve a martensite microstructure.

    Circling back to how dies are heat treated, dies can be quenched (rapidly cooled) or annealed (slowly cooled) after heated above the austenitizing temperature. If the engraver wants to engrave a die, they would need to austenitize and then anneal the die to reduce the hardness and make it easier to work with. Once the engraver is finished with preparing the die, they would need to austenitize and then quench the die to develop the necessary strength and hardness. However, as mentioned previously, the martensite produced by quenching has low toughness, meaning it would be very susceptible to cracking and breaking apart (either due to striking coins or internal stresses). Since this is a very undesirable characteristic for a coin die, an additional heat treatment after quenching is needed: tempering.

    For tempering, the steel is heated well below the austenitizing temperature, held at that temperature, and then cooled. In martensite, the iron and carbon atoms are arranged in a very strained manner (causing internal stresses), and tempering allows for the atoms to move into less-strained configurations. Tempering reduces the hardness and strength, but it increases the toughness. Generally the hotter the tempering temperature, the more negatively impacted strength and hardness become. Since coin dies need to be fairly strong and hard, they would probably be tempered at a lower temperature, with the understanding that their toughness may not be particularly great.

    So, to sum it all up, a coin die would would generally have the following heat treatments:

    • Annealing to soften the die prior to engraving.
    • Quenching to harden the die after the engraving was completed.
    • Tempering to improve toughness without sacrificing too much strength and hardness.

    The effort that goes into heat treating a die was probably an important factor in the early days of the US Mint when it came to deciding to overdate an outdated die or just use the die as-is. If the die hadn’t been quenched yet, it would be fairly easy to overdate the die. However, once the die was quenched/hardened, it would be much more challenging to overdate the die. This is likely why there are very few early US coin dies that were overdated after being quenched/hardened and used to make coins.

    Here are some links that cover modern heat treatment processes employed by the Philadelphia and Denver Mints, and the Royal Canadian Mint:



    https://www.mint.ca/store/dyn/PDFs/SM_RT_Leaflet_Heat Treating and Coining Dies.pdf

    For a more thorough technical explanation of steel heat treatments, see the following ASM International publication:

    Brina, Nyatii, PlanoSteve and 16 others like this.
  2. Avatar

    Guest User Guest

    to hide this ad.
  3. Burton Strauss III

    Burton Strauss III Well-Known Member

    NSP, NOS and RonSanderson like this.
  4. alurid

    alurid Well-Known Member

    This is the most educational essay on modern coins dies I have ever seen in one article. Truly a superb compilation of information. Thank you for your efforts.
    Brina, NSP, mlov43 and 1 other person like this.
  5. Conder101

    Conder101 Numismatist

    The early mint workers also had the problem that they didn't know the precise composition of the steel, nor could they accurately judge the heating temperature.

    Just two that we can be sure of because the dies were used both pre and post overdating. A die represented a lot of time and effort to produce and the time it was most likely to fail was during the quenching. So if you had a perfectly good hardened die of the previous year, you were much more inclined to use it as is rather than risk losing it during a second quenching cycle. The two they did do it with had short use lives afterward as well. (1806/5 quarter and half dollar)
    NSP likes this.
  6. gxseries

    gxseries Coin Collector

    I love scientific article like this. Keep up the good work!!!

    As Conder101 mentioned, I am genuinely intrigued over how steel was made back in the 1800s. Today's steel is made by removing impurities and adding carbon. I can only imagine that technology back then wasn't as well refined and hence would have some interesting trace elements, which would have meant that different dies may had slightly different heating properties.
    NSP likes this.
  7. Marshall

    Marshall Junior Member

    Another die used both pre-changed and post changed is the Large Cent Reverse K of 1801 which was used as a 1/000 fraction and then used on the 1803 as Reverse G after an attempt to correct it.

    Since copper is a softer metal, it lasted long enough to make the S-249 an R2 variety indicating a long life. Or it may have been a particularly fortunate softening, engraving and re-hardening process.
    NSP likes this.
  8. NSP

    NSP Well-Known Member

    I believe that two other dies that were annealed, overdated, and hardened again are the 1806/5 $2.5 (per the Dannreuther/Bass book) and the 1827/3/2 25C (per the Tompkins book).
    Marshall likes this.
  9. kaosleeroy108

    kaosleeroy108 The Mahayana Tea Shop & hobby center

    Very comprehensive and we'll put , I honestly have several coin and die sets that I purchased purchased during my time collecting and it has been a pleasure to own each one of them my next plan to purchase is a 19 97 Olympic coin and die set at least one that's been certified I'll pay the $650 for it or $700 however much they charge for it. I definitely say it is worth collecting these because they're unique to the industry and the industry stopped producing them as far as the u.s. mint goes.. now there's a secondary market for them and they tend to go pretty well you can still find them close to the original price if you ever take a search on eBay
    Brina and NSP like this.
  10. hogwash

    hogwash New Member

    Does anyone know what happened to all the Franklin Mint dies?
  11. halfcent1793

    halfcent1793 Well-Known Member

    This is true. In the early days, they used "blister steel," which is made by sticking the end of an iron bar into burning charcoal for long enough for carbon to leach into the end of the bar and then knocking the end of the bar that had become steel off. The quality was obviously variable.
    NSP likes this.
  12. CaptHenway

    CaptHenway Survivor

    Well done! Thank you!
    NSP likes this.
  13. capthank

    capthank Well-Known Member

    Excellent. Thanks
    NSP likes this.
  14. John Wright

    John Wright Well-Known Member

    Just two that we can be sure of because the dies were used both pre and post overdating. The two they did do it with had short use lives afterward as well. (1806/5 quarter and half dollar)[/QUOTE]

    A third example: Large cent 1801 Reverse 'K' (Sheldon NC3, quite scarce) has fraction 1/000 'corrected' in 1803 Reverse 'G' (S-249) which is much more common but develops a reverse 'cud' failure.
    NSP likes this.
  15. John Wright

    John Wright Well-Known Member

    NSP, you have done a *GREAT* presentation that I thank you for *PROFUSELY* !!
    NSP likes this.
  16. Mountain Man

    Mountain Man Well-Known Member

    Great post. Thanks.
    Brina and NSP like this.
Draft saved Draft deleted

Share This Page