Vacuum Heat Treatment of Tool Steel, Process Parameters and Grade Schedules
A vacuum furnace hardens tool steel without the atmosphere that oxidises or decarburises it, and the vacuum level itself becomes a process variable, because the alloying elements in a high chromium grade will evaporate if the pressure is held too low while the part is at temperature. The tables below set the vacuum level against the heating temperature, give the factors used to calculate the soak from the section thickness, compare the cooling gases, and list the published vacuum cycles for thirteen tool and mould steel grades, from the carbon and low alloy tool steels through to the high speed grades.
Reference only. The figures below are a reference summary of published Chinese handbook practice, not an Aobo Steel specification. Confirm the cycle against the mill certificate and the governing drawing before an order or a production run is set against them.
What the vacuum cycle does for tooling
The furnace chamber is held below one atmosphere, so the only gas left around a hot part is a trace of carbon monoxide and hydrogen, both of which are reducing rather than oxidising. The part therefore comes out with its original surface, and any oxide already on it is reduced rather than grown. On a die that has been finish ground before hardening this removes the cleaning step that follows a salt bath or an open air furnace.
Two further effects matter on tooling. The first is distortion. Heating by radiation is even across the load, and a pressurised gas quench can drive the gas at the part from all sides, so a vacuum cycle is quoted as producing roughly seventy per cent less distortion than a salt bath on the same part. The second is degassing, where hydrogen and other dissolved gas leave the steel at temperature, which suits large die blocks and parts that have been forged or welded.
The cost of the process is that the part calls for a longer heat. Radiation heats a load more slowly than a circulating atmosphere at low temperature, so a vacuum soak runs longer than the soak in a salt bath or an air furnace for the same section, and this is the reason the soak factors in the table further down are read on their own rather than taken from a general chart. Oil quenched parts also come out with a light surface layer of residual austenite that a normal tempering temperature will not remove, and a high alloy or high speed grade quenched in oil needs the higher temper to clear it.
Furnace and cooling gas
Three furnace types cover the cycle. A vacuum annealing furnace holds a high vacuum and uses a slow controlled cool. A vacuum hardening furnace is either a gas quenching or an oil quenching chamber, and the gas chamber is generally preferred on tooling because the surface stays clean, the distortion is lower and the chamber does not need the oil cleaning line. A vacuum tempering furnace follows the quench, because tempering in an ordinary furnace after a vacuum quench brings back the poor surface, the uneven hardness and the under tempering the vacuum cycle was chosen to avoid, and it needs a fast cooling facility so that a grade sensitive to temper embrittlement does not sit in the embrittling range.
The gas in the chamber decides how fast the quench runs. The four common gases are ranked below against hydrogen, which transfers heat fastest but is avoided because a hydrogen volume fraction above five per cent in air is an explosion risk and the gas costs about ten times as much as nitrogen. Nitrogen is what most Chinese vacuum furnaces use, and the quench is run either at positive pressure or under vacuum.
| Cooling gas | Relative cooling time |
|---|---|
| Hydrogen (H2) | 1.0 |
| Helium (He) | 1.2 |
| Nitrogen (N2) | 1.5 |
| Argon (Ar) | 1.75 |
Source, section 8.2.6, printed page 477. Cooling time relative to hydrogen, taken as 1. The handbook notes that a mixture of 60 to 70 volume per cent helium with 30 to 40 volume per cent nitrogen is regarded abroad as the best compromise.
A high pressure gas quench at 0.5 to 0.6 MPa will through harden a part of 80 to 110 mm, which covers most die blocks. Positive pressure quenching with high purity nitrogen at 0.2 to 0.6 MPa is used for high speed steel, and negative pressure quenching with high purity nitrogen at 7.9 times ten to the fourth to 9.3 times ten to the fourth pascals is the milder option. Where the gas quench is not fast enough and the part has to go into oil, the note in the source is to use a gas quench first and finish in oil, so that the white layer oil alone would leave on a high chromium or high speed part does not form.
Vacuum level against heating temperature
The vacuum level is set by temperature, not by the furnace, because a high chromium grade held at a low pressure at temperature will lose chromium from the surface. A part is loaded cold and the chamber is pumped to 6.67 pascals before the heat is switched on, and above that the level is allowed to rise with the temperature of the load. Back filling with high purity nitrogen is the usual way of holding the level inside the window.
| Heating temperature, °C | Vacuum level, Pa |
|---|---|
| Up to 900 | ≥ 0.133 |
| Over 900 to 1100 | 13.3 to 1.33 |
| Over 1100 to 1300 | 13.3 to 666.0 |
Source, Table 8-17, printed page 476. The level is allowed to rise as the load temperature rises, to hold back the evaporation of alloying elements.
Preheating follows the same grade logic. A cycle that heats to 1000 to 1100 degrees Celsius uses one preheat at 800 degrees Celsius. Above 1200 degrees Celsius a small simple part is preheated once at 850 degrees Celsius, and a larger or a more complex part twice, the first at 500 to 600 degrees Celsius and the second at about 850 degrees Celsius.
Soak time from the section thickness
The vacuum soak is calculated from the effective thickness of the part rather than read off a chart, because the heating rate depends on the load and the section together. The form is a coefficient applied to the thickness plus a time allowance, and both numbers change with the alloy content of the grade, since a high alloy tool steel holds its heat path through the carbides rather than through the matrix.
| Steel group | Soak coefficient K, min/mm | Time allowance T, min | Preheat noted |
|---|---|---|---|
| Carbon tool steel | 1.9 | 5 to 10 | One preheat at 560 °C |
| Low alloy tool steel | 2.0 | 10 to 20 | One preheat at 560 °C |
| High alloy tool steel | 0.48 | 20 to 40 | Preheat at 560 °C and again at 800 °C |
| High speed tool steel | 0.33 | 15 to 25 | Preheat at 560 °C and again at 850 °C |
Source, Table 8-18, printed page 477. Soak time C = K D + T, where C is the soak in minutes, K the coefficient above, D the effective thickness of the part in millimetres and T the allowance in minutes. The prepared tables differ from a general soak chart and replace it for a vacuum cycle.
The same source puts the vacuum quench soak at five to six times the salt bath soak and twice the air furnace soak on the same part. A vacuum temper runs slightly longer than an air furnace temper and one to two times longer than a nitre salt temper, and a vacuum anneal runs about twice as long as an air furnace anneal.
Vacuum heat treatment schedules by grade
The schedule table below is the published cycle for thirteen mould and tool steel grades. The preheat column collapses a two step preheat into one cell where the source gives one, and the two figures in a cell are the first and the second preheat. The quench vacuum level drops from 0.1 pascals on the low alloy grades to 10 to 1 pascals on the grades that go into oil or into a pressurised gas quench, because the pressure at the part has to be high enough to transfer heat once the quench starts. Where a grade carries two tempering rows the second is the secondary hardening temper used for high speed and high chromium grades.
| Grade | Preheat, °C | Preheat vacuum, Pa | Austenitising, °C | Quench vacuum, Pa | Cooling | Temper, °C | Hardness, HRC |
|---|---|---|---|---|---|---|---|
| 9SiCr | 500 to 600 | 0.1 | 850 to 870 | 0.1 | Oil, above 40 °C | 170 to 190 | 61 to 63 |
| CrWMn | 500 to 600 | 0.1 | 820 to 840 | 0.1 | Oil, above 40 °C | 170 to 185 | 62 to 63 |
| 9Mn2V | 500 to 600 | 0.1 | 780 to 820 | 0.1 | Oil | 180 to 200 | 60 to 62 |
| 5CrNiMo | 500 to 600 | 0.1 | 840 to 860 | 0.1 | Oil or high purity N2 | 480 to 500 | 39 to 44 |
| Cr6WV | 500 to 550, then 800 to 850 | 0.1 | 970 to 1000 | 10 to 1 | Oil or high purity N2 | 160 to 200 | 60 to 62 |
| 3Cr2W8V | 480 to 520, then 800 to 850 | 0.1 | 1050 to 1100 | 10 to 1 | Oil or high purity N2 | 560 to 580, 600 to 640 | 42 to 47, 39 to 44 |
| 4Cr5W2VSi | 480 to 520, then 800 to 850 | 0.1 | 1050 to 1100 | 10 to 1 | Oil or high purity N2 | 600 to 650 | 38 to 44 |
| 7CrSiMnMoV | 500 to 600 | 0.1 | 880 to 900 | 0.1 | Oil or high purity N2 | 450, 200 | 52 to 54, 60 to 62 |
| 4Cr5MoSiV1 | 500 to 550, then 800 to 820 | 0.1 | 1020 to 1050 | 10 to 1 | Oil or high purity N2 | 560 to 620 | 45 to 50 |
| Cr12 | 500 to 550 | 0.1 | 960 to 980 | 10 to 1 | Oil or high purity N2 | 180 to 240 | 60 to 64 |
| Cr12MoV | 500 to 550, then 800 to 850 | 0.1 | 980 to 1050, 1080 to 1120 | 10 to 1 | Oil or high purity N2 | 180 to 240, 500 to 540 | 60 to 64, 58 to 60 |
| W6Mo5Cr4V2 | 500 to 600, then 800 to 850 | 0.1 | 1100 to 1150, 1150 to 1250 | 10 | Oil or high purity N2 | 200 to 300, 540 to 600 | 58 to 62, 62 to 66 |
| W18Cr4V | 500 to 600, then 800 to 850 | 0.1 | 1000 to 1100, 1240 to 1300 | 10 | Oil or high purity N2 | 180 to 220, 540 to 600 | 58 to 62, 62 to 66 |
Source, Table 8-19, printed page 478. Read from the page image at 300 dpi. A cell with two values carries the first and the second preheat, or the two austenitising temperatures, or the two tempering temperatures, in the order the source gives them. Vacuum levels below 1 Pa are a gauge pressure inside the chamber, and 10 to 1 Pa is the working level used once the quench begins.
Reading the schedules together
Three patterns run through the table. The vacuum level during the quench follows the cooling medium, not the grade, so every grade that can go into oil or into high purity nitrogen is run at 10 to 1 pascals while the grades quenched in oil alone hold 0.1 pascals. The preheat count follows the austenitising temperature. A grade run at 900 degrees Celsius and below takes one preheat, and everything run above 1000 degrees Celsius takes two, one in the low range to even out the block and one just under the austenitising range to hold the section together.
The tempering rows separate the two families of tool steel. The cold work and mould grades are tempered once for hardness, and the grades that carry tungsten or a large chromium content take a second temper in the 500 to 640 degrees Celsius band that produces the secondary hardening peak. The two high speed grades are the clearest case, where the first temper at 180 to 300 degrees Celsius is a stress relief rather than a hardness temper and the second at 540 to 600 degrees Celsius is the one that sets the final hardness.
A vacuum cycle is not the answer for every grade. Where the gas quench cannot reach the hardness the section needs, the part is better moved to a grade designed for a slower quench, such as D2 or H13, than pushed into oil and risked for cracking. Where the same cycle is wanted at a smaller section, the oil hardening grades such as O1 and W1 reach their hardness in a bath that a vacuum chamber cannot match for speed.
Related reference data
The furnace and bath side of the same decision, including the salt mixtures and their working ranges, is on the hardening furnaces and salt bath page. The soak times for a conventional atmosphere cycle are on the tool steel soak time chart, and the full hardening and tempering cycles by grade are on the tool steel heat treatment guide. Where a vacuum cycle is followed by a sub zero hold, the media and the temperatures are on the cryogenic treatment of tool steel page.
Notes on the data
The schedule table is the published reference cycle from a Chinese mould and tool steel handbook, and the page numbers are the printed page numbers of that handbook rather than the sample numbers of this site. The grade names are the Chinese GB designations, so 4Cr5MoSiV1 is the H13 family, Cr12MoV is the D2 family and W6Mo5Cr4V2 is the M2 family, and the equivalents are set out on the grade pages rather than assumed here. The vacuum levels are chamber gauge readings, and a furnace’s own pump-down curve and leak rate will shift the practical figure. The soak coefficients are paired with one manufacturer’s preheat practice, so an operator using a different preheat count should take the coefficient and leave the allowance.
Compiled from a Chinese mould and tool steel heat treatment handbook, section 8.2.6 and Tables 8-17 to 8-19, printed pages 476 to 478. Tables 8-19 was read from the page image at 300 dpi and the rest from the text layer of the same printing. The tables are a reference summary of published practice rather than an Aobo Steel specification, so confirm the cycle against the mill certificate and the governing drawing before an order is placed against it.