GAS QUENCHING WITH CONTROLLABLE HEAT EXTRACTION

Transcription

1 GAS QUENCHING WITH CONTROLLABLE HEAT EXTRACTION B. LISCIC Faculty for Mechanical Engineering and Naval Architecture University of Zagreb Croatia THERMEC' 2006, July 4-8, 2006, Vancouver

2 Modern vacuum furnaces with high pressure gas quenching, using circulated nitrogen of 5 to 10 and as far as 20 bar pressure, and velocities of 10 to 30 m/s are used to harden different finally machined engineering components in batches having masses of several hundred kilogrammes. Besides many advantages compared to quenching in oil (substantially reduced size change and distortion of workpieces; no need to wash the parts after quenching; ecological acceptability), the main shortcoming of it is: inadequate quenching intensity necessary to obtain the required hardness in the core of workpieces having bigger cross-section size. This shortcoming is particularly evident when quenching low alloyed structural steels of low hardenability, and when batches of great mass are quenched.

3 Fig.1. Single chamber vacuum furnace with high-pressure gas quenching

4 There is big difference in heat transfer between quenching in a vapourable liquid quenchant and gas quenching. Because of the Leidenfrost phenomenon when quenching in liquid quenchants three distinct heat transfer phases occur: vapour blanket or film boiling; nucleate or bubble boiling; and convection having very different heat transfer coefficients. In gas quenching there is no Leidenfrost phenomenon; heat transfer takes place predominantly by convection, producing smaller temperature differences in a part during quenching and thus less distortion and dimensional variation.

5 This convective heat transfer is usually described by the following formula: α = C v p d η C p λ showing that the heat transfer coefficient (α) is dependent on gas velocity (v) and its pressure (p). Because cooling the workpieces in a gas takes much longer time than quenching them in a liquid quenchant, quenching parameters (gas pressure, gas temperature and its circulation velocity can be changed during the quenching process. In this way we can change the heat extraction dynamics.

6 a) b) Fig.2. Heat transfer in liquid quenching (a), and in gas quenching (b)

7 By changing intentionally the heat extraction dynamics we actually introduce a new Controllable Heat Extraction (CHE) technology. This technology can increase the quenching intensity during corresponding period of the quenching process; it can increase the depth of hardening and can eliminate the explained shortcoming i.e. it can achieve higher hardness in the core of workpieces having bigger cross-section size. For theoretical explanation of that one can take the Newton's law of cooling: ( ) q= α T T s 0 showing the heat flux density which depends on the temperature difference between the workpiece's surface and the cooling medium.

8 From the other side continuous-cooling-transformation (CCT) diagram of the quenched steel shows that the structure transformation in different points of the workpiece cross-section starts at different times after beginning of the quenching process, and that the cooling rate in the critical temperature range from A 1 to M s (martensite start temperature) is of paramount importance for hardness attained after quenching.

9 Fig.3. Calculated cooling curves for surface (S), three quarter radius (3/4 R), and centre (C) of a 50 mm dia. bar superimposed on the CCT diagram of the steel grade AISI 4140

10 If we increase gas pressure and its velocity, and simultaneously introduce spraying with liquid nitrogen at a later time, when the cooling curve for core passes through the critical temperature range A 1 - M s, we will substantially increase the temperature difference (T s -T 0 ) and correspondingly the heat flux density (q). Increasing later the quenching intensity in described way means actually to have a delayed quenching with discontinuous change of cooling rate.

11 When discontinuous change of cooling rate is involved, as it is known from the work of Shimizu and Tamura, the pearlitic transformation is different from that given by the CCT- diagram, and is related to the incubation period consumed before changing the cooling rate. In this way Shimizu and Tamura have explained the so called "inverse" hardness distribution (showing higher hardness in the core than at the surface), which they have experimentally found by quenching some specimens in oil.

12 Fig.4. Schematic illustration of discontinuous change of cooling rate causing 'inverse' hardness distribution after quenching, according to Shimizu and Tamura

13 An own experiment using double-cone specimens (20/40 mm dia. 110 mm length) made of low alloyed steel AISI 6150, quenched in circulated nitrogen changing the flow direction (top to bottom and bottom to top) every 10 seconds, has produced the following results: Specimen I. Quenched by conventional process: 4 bar gas pressure and 9 m/s circulation velocity from the beginning until the end of cooling-core hardness HRC. Specimen II. Quenched by a delayed quenching process: 1 bar gas pressure and 9 m/s circulation velocity for the first 40 seconds, followed by 10 bar gas pressure and 20 m/s circulation velocity until the end of cooling-core hardness HRC.

14 CHE technology is not only described high pressure gas quenching in vacuum furnaces, but every quenching process with exactly controlled heat extraction dynamics. To explain how it can influence surface residual stresses let us analyse the Intensive Quenching method developed by N.I. Kobasko. This quenching method is used mostly for workpieces made of non-alloyed or low-alloyed steels, quenching them in pure water flowing with very high velocity.

15 The main characteristic of Intensive Quenching method is that at very beginning of the quenching process the workpiece is cooled very rapidly and uniformly forming martensite simultaneously over the entire surface, creating a strong hardened "shell" with compressive stresses. A schematic presentation of development of surface stresses during Intensive Quenching by means of "segments" and "springs" is shown on the next figure. The main feature of Intensive Quenching method is to interrupt rapid cooling of the workpiece at the moment when compressive stresses on the surface are at maximum.

16 Fig.5. Surface stress conditions during intensive quenching. Source: Brochure of the IQ Technologies Inc., Akron, OH

17 Fig.6. Development of surface stresses vs. time when intensive quenching is applied

18 Controllable Heat Extraction (CHE) technology is a combination of the main stream of gaseous nitrogen (of adequate pressure and velocity) and a temporary spraying of liquid nitrogen, causing instantly very low ambient temperatures so that big temperature differences between the surface of workpieces and the cooling medium cause high heat fluxes and high heat transfer coefficients respectively. Contemporary, vacuum furnaces with a cold high pressure gas quenching chamber are generally suitable for the CHE technology, provided that an adequate nozzle-field system for spraying liquid nitrogen, and a relevant control system are built in. Dispersed liquid nitrogen is sprayed transiently, only during those periods, when high quenching intensity is needed.

19 During these periods the three following heat transfer mechanisms are used: 1. Convection by the circulated gas 2. Radiation from the hot surface of the workpieces to the cold walls of the quenching chamber 3. Vaporisation of the liquid nitrogen. A special feature of spraying liquid nitrogen is that it can instantly change the temperature of the surrounding medium, and can be stopped at once.

20 Automatic heat extraction control The necessary control system consists of three main parts: 1. An instrumented probe to measure its own temperature at the reference points 2. Control devices for every quenching parameter 3. Software package with auxiliary files and relevant programs for calculation of the heat transfer coefficient α(t), and cooling curves respectively. The instrumented probe having dimensions of real workpieces has to be of similar shape as the quenched workpieces, and has to be put in relevant position (standing or lying) within the batch of workpieces.

21 Fig.7. Instrumented probe for measuring temperatures and calculation of the heat transfer coefficient

22 The user has to put in the required hardness value at the specified cross-section location. The relevant CCT-diagram for the steel grade in question is taken from the file of CCT diagrams, and the cooling curve which assures attaining the required hardness at the chosen cross-section location, is drawn onto the diagram. The heat transfer coefficient as a function of the surface temperature α(t), which satisfies the drawn cooling curve, is calculated. With this calculated function α(t) for the workpiece, having different cross-section size than the probe, the target cooling curve at the reference point of the instrumented probe, is calculated.

23 Temperature at the reference point of the probe is measured during the whole quenching cycle and compared with the calculated target cooling curve. If during this comparison discrepances arise between the calculated target cooling curve and the measured cooling curve, a signal is automatically given to the device of relevant cooling parameter in order to increase or decrease the quenching intensity. The action of the relevant cooling parameter lasts only for the period until both cooling curves become equal. In this way it is assured that in concrete case the quenching will in the specified cross-section point realize the cooling curve which will yield the required hardness.

24 Fig.8. Algorithm of automatic control of the heat exctraction during quenching of a batch of workpieces in vacuum furnace (Source: Liscic 2001.)

25 Conclusion Metal's quenching was a skill already in ancient civilizations, but through centuries it was a "black-box" operation. After introduction of computer-assisted calculations, people started to look at the quenching process as a heat transfer problem, what every quenching process actually is. When calculating heat transfer coefficients it became clear that besides the quenching medium itself, its temperature and pressure, its circulation velocity and the characteristics of the workpiece, as well as loading arrangement and mass of a batch, are factors having smaller or bigger influence on the heat transfer.

26 It became evident that in real quenching operations not only the specific quenching intensity of the quenching medium (as it is measured in laboratory tests), but the heat extraction dynamics is the main factor on which structural transformations and resulting hardness distribution on a workpiece's cross-section depends. With introduction of the CHE technology it became possible to intentionally influence the heat extraction dynamics by changing several parameters during the quenching process. So it become possible to change the quenching process from a "black-box" operation to a controllable heat extraction process adapted specifically to the concrete batch of workpieces, which assures repeatable results of hardness distribution, as well as of residual stresses and distortion respectively.