When choosing a new furnace, we have to consider a number of criteria that I tried to describe in the previous chapters. In this last section, I would mention three more that may influence our choice of tempering and annealing furnace
Fig. 1 – Factors influencing the energy and economic functionality of the furnace
Temperature control
Temperature regulators are a particular problem. They can be two-position (Fig. 2), i.e. on/off status, or PID (Fig. 3).
Fig. 2 – On/Off temperature regulation [1]
The first form of regulation is not suitable for our purposes because the actual temperature value will always oscillate around the set point.
PID controllers are practically the only solution, however, they have the disadvantage that we can usually set only one variant of constants for whole cycle. However, in the case of a variable load, charge, we can get into a situation where the regulator settings do not allow the necessary values to be reached. This will result in increased energy consumption, longer cycle time, or failure to achieve quality due to underheating or overheating of the furnace (Fig. 4)
Fig. 3 – Temperature regulation with PID controller [2]
Fig. 4 – Critical states of temperature control [2]
Fig. 5 and 6 – The result of temperature regulation on a retort furnace for small and large batches
But we need to reach the state according to Fig. 3. That is, reaching the set temperature without underheating or overheating, with zero deviation. A practical example is in Fig. 5 and 6, where there are records from a retort furnace with a small and a large batch. Using identical PID constants for the entire temperature range leads to irreversible overheating and destruction of parts with a small batch, or low hardness in the case of tempering. This is unacceptable.
Fig. 7 – Precise temperature control with variable PID values
Fig. 8 – 6-zone temperature programmer PRC-8000-6 from Libratherm Ltd., India
In order to achieve an acceptable state, we must have a temperature programmer with the ability to set a different set of PID constants in each heating step. For example, the temperature programmer PRC-8000-6 (Fig. 8) corresponds to this specification, where the furnace program can be set in up to six zones, while each zone can have 16 program steps. All parameters, including PID constants, are adjustable in each step. Such a programmer ensures that we reach the desired value in the optimal time, without deviations from the desired value.
Control of the tempering or annealing process from the batch thermocouple
In vacuum quenching furnaces it has already become standard to use Ts and Tc thermocouples for process control. However, this is far from the case with tempering or annealing. These operations are still performed in furnaces where the temperature of the piece is not measured and the time is determined empirically based on the result of reference tests. That it is inaccurate is quite obvious. That’s also why there is the most non-conformities in every heat treatment operation about this title.
The resulting state of the material, such as hardness or microstructure, is always a combination of time and temperature used. We already know how to control the temperature in the furnace, but we cannot measure the moment when the parts processed by us have reached this temperature. That’s why Tc is important. It measures the temperature of the piece, and at the moment when we have reached the desired value, it starts the set soaking time at the temperature. This is a so-called conditional dwell, and temperature controllers or furnace programs can already work with this data. Therefore, if we care about the repeatability of the results, and we have diverse batches, temperature sensing at Tc is a condition. The repeatability of such a process is more than 95%.
Fig. 8 – Temperatures during tempering or annealing cycle
There are a number of annealing and tempering furnaces where Tc is already the standard. These are mainly vacuum furnaces, chamber furnaces with a muffle or retort, bell furnaces and others. Usually, this solution is applied wherever there is an individual loading of the furnace. An example of the use of a batch thermocouple in a bell tempering furnace with nitrogen is shown in Fig. 9. High-end furnaces allow the connection of up to 12 batch thermocouples (Fig. 11), so we can fully control the entire process in full volume.
Fig. 9 – Possible arrangement of the bell furnace with the addition of a bushing in the bottom of the furnace for a thermocouple Tc
Fig. 10 – Vacuum tempering furnace with the possibility of up to 12 thermocouples
Fig. 11 – Datalogger from PhoenixTM
However, this is not possible in lines where the baskets with batch moves. In this case, a solution with a datalogger (Fig. 11) can be used, which also moves with the fixture and transmits temperature data wirelessly outside the oven via WiFi or Bluetooth. Although the thermal protection of the device will ensure the function of measuring the temperature of parts in the entire temperature range required by us up to 750 °C, the thermal barrier will take up a lot of space. So, we will have to wait for further miniaturization and until then use this method of measurement only for sampling the process.
Furnace contamination
In the last part, I would like to mention the influence of furnace contamination. Especially in the case of forgings or deep drawing parts, we cannot completely avoid the problem of hot zone contamination with graphite powder from lubricant. Therefore, we should protect the ceramic insulation of the electrical connections of the heating elements inside the furnace. Since graphite is electrically conductive, the electrical bushing would short out and the furnace would destroy very soon.
An example is in Fig. 12, where deep-drawn parts produced on a progressive press are contaminated with graphite grease during cold forming. Although the parts go through a washing machine after forming, but because these are deep shapes, it is very difficult to remove this graphite lubricant from these places.
After forming, the parts must be recrystallized annealed at temperatures above 500 C. If a vacuum furnace with heating elements inside the work zone according to Fig. 13 was used for testing, then after one week of operation the electrical bushings on all phases were short-circuited.
Fig. 12 – Example of batch for recrystallization annealing of cold-formed parts
Fig. 13 – Connection of resistors inside the heating chamber
For this type of application, the perfect solution is, for example, the tempering furnace from TAV Vacuum Furnaces, where we have direct heating of the load with heating elements placed inside the working space of the furnace. The heating elements are made of Ni-Cr wire placed in an Inconel tube, and the connection between the individual elements and the power supply is on the outside of the furnace, in the atmosphere (Fig. 14 and 15). Therefore, there can be no short-circuiting of the heating elements with regard to contamination of the interior space. In combination with perfect insulation, it is a device that can satisfy all demanding applications of annealing or tempering processes.
Fig. 14 and 15 – Connection of heating elements on the furnace TAV H8-T
Of course, this problem can also be solved with a gas-tight retort furnace where the heating elements are also outside the working chamber, but as was said, our energy consumption will be 30 to 40% higher due to the heating and cooling of the retort.
Conclusion
There are several reasons why I started writing this tetralogy about tempering and annealing furnaces. The first reason is that this area of heat treatment and related furnace equipment is still significantly undervalued. At the same time, as I described here [3], the need for low-temperature operations, e.g. for the production of HPDC (High Pressure Die Casting), is 7 times higher than the need for quenching furnaces.
The second reason is that many heat treaters invest in quenching only in the last line, so they have no choice but to temper in quenching furnaces. To clarify, I am talking about vacuum heat treatment here. As a result, the tempering or annealing process is completely uneconomical, because it is done in the most expensive furnace they have, and at the same time of dubious quality, because it is not possible to optimally temper or anneal in a furnace designed not for low, but for high temperatures.
The third reason is that many manufacturers of low-pressure carburizing or quenching furnaces do not produce tempering and annealing furnaces at all. And if, for example, an automatic line is to be built for the processing of thousands of parts using the LPC process, tempering and annealing furnaces from third parties are integrated into it. It is therefore necessary to give certain recommendations to the investor so that his equipment is of high quality and at the same time economical.
Fig. 16 – typical layout of ECM line with third party temper furnaces
For the above reasons, this rough guide was created on how such a furnace should look. And what should the FERRARI furnace mentioned in the introduction look like?
And is there such a furnace? Yes, in my opinion the pictures below from TAV Vacuum Furnaces are the promised Ferrari of annealing and tempering furnaces. And reliability? According to the reference from Galvamet, excellent.
[1] – https://www.eurotherm.com/temperature-control/principles-of-pid-control-and-tuning/
[3] – https://www.jstconsultancy.cz/post-processing-forem-pro-tlakove-liti/
Jiří Stanislav
August 29, 2024