International Journal of Heat and Mass Transfer 140 (2019) 807–818

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ijhmt

Quenching power of aqueous salt solution

Luís H. Pizetta Zordão a,⇑, Vinícius A. Oliveira a, George E. Totten b, Lauralice C.F. Canale a
a
Department of Materials Engineering, University of São Paulo, School of Engineering of São Carlos, São Carlos, SP 13563-120, Brazil
b
Department of Mechanical and Metallurgical Engineering, Portland State University, Portland, OR, USA

a r t i c l e i n f o a b s t r a c t

Article history: The mechanical properties of steel components are influenced by the microstructure that is determined
Received 13 March 2019 by the heat treatment cycle. In the quenching of the steel: water, oil, aqueous polymer solutions and
Received in revised form 7 June 2019 aqueous salt solutions (brine) can be used as quenchants that exhibit different characteristic on cooling
Accepted 10 June 2019
mechanisms. For example: when water is used as the cooling media, a stable vapor film is formed around
Available online 19 June 2019
the hot component resulting in non-uniformity of surface heat transfer during the cooling process which
is often responsible for distortion and cracking. Salt addition can reduce or inhibit this vapor film forma-
Keywords:
tion, enhancing the uniformity of heat transfer during the cooling. This work will investigate the cooling
Heat treatment
Quenching
performance of different salt solutions (NaCl, NaNO2, Na2SO4, NaHCO3) varying concentration, tempera-
Aqueous salt solution ture (25 and 45 °C) and agitation (0 and 800 rpm). The description of these medium and the influence of
Low hardenability external factors were made using cooling curves, cooling rates and heat flux. Metallurgical tests were per-
Hardness formed in samples of SAE 1045 steel which were quenched in such solutions and hardness were analyzed.
The use of aqueous salt solutions, in general, has proved to be effective in to inhibit the vapor blanket, to
increase heat extraction and keeping a high heat flux during the most part of the drop of the temperature
that results in a homogeneous cooling able to avoid distortions. NaCl, NaNO2, Na2SO4 solutions were able
to quench SAE 1045 steel.
Ó 2019 Elsevier Ltd. All rights reserved.

1. Introduction quenching is done in a volatile liquid medium where a boiling pro-

cess can be detected, such as in water, aqueous polymer solution or
Steel components when in service are under stress, requiring of oil, at least three stages on cooling mechanism can be seen on its
them different levels of mechanical resistance. So procedures to cooling curve and cooling rate curve. These three cooling zones
promote hardening or even relief of internal stresses are applied are identified as stages A, B and C on Fig. 1, these stages is known
in this class of materials, aiming at their chemical and/or as vapor blanket, nucleate boiling and convective cooling stages
microstructural modification, to guarantee the integrity of the respectively [3].
component and its desired functionality, since the properties of In the vapor blanket stage a slow heat transfer is commonly
steels depend on their structure. detected. Between the stages A and B a wetting front is formed
The so-called quenching heat treatment is one of the most and the contact of the quenchant with the hot surface of the com-
important processes to help improve the performance and applica- ponent occurs through the collapse of the vapor blanket. The vapor
bility of steels by the formation of the martensite metastable blanket depending on the quenchant temperature is very stable,
phase. During quenching, the hot metal is rapidly cooled by a med- but when agitation and additives are included on the quenching
ium (quenchant) from the austenitization temperature, typically in process its stability collapses and the stage B starts sooner [4,5].
the range of 845–870 °C, for carbon steels [1]. Heat transfer and The high stability of vapor blanket is responsible for many soft
wetting are the two important phenomena that occur during surface hardness on quenching steels due to the undesirable
quenching that controls the final metallurgical and mechanical micro-constituents. The process of collapse of vapor blanket is an
properties of the components [2]. important phenomenon that can influences on component distor-
The shape of a cooling curve is indicative of the various cooling tion mainly if it occurs as a non-uniform manner with a variation
mechanisms that occur during the quenching process. When the of the heat transfer coefficient (HTC) around the surface by the
presence of the wetting front, Fig. 2 [3].
The nucleate boiling stage (Stage B) starts when the quenchant
⇑ Corresponding author. reaches the component hot surface and bubbles are formed, this
E-mail address: [email protected] (L.H. Pizetta Zordão).

https://doi.org/10.1016/j.ijheatmasstransfer.2019.06.036
0017-9310/Ó 2019 Elsevier Ltd. All rights reserved.
808 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818

Distilled water typically exhibits much longer film-boiling times

than does tap water [3]. Besides the maximum cooling rate
decrease with increasing water temperature and the temperature
where the maximum cooling rate occurs decrease with increasing
water temperature [8].

1.1. Salt addition forming an aqueous solution

By adding solutes into the quenching water, it is possible to

alter either the surface of the cooling piece to disturb the type of
liquid vaporization or directly to the liquid vaporization processes
[9]. Salts can be used as additives on water to decrease or eliminate
the film boiling behavior. Some reasons have been studding to
understand this mechanism.

1.1.1. Thermal resistance and larvate boiling

During the first stage, a strong water vaporization occurs letting
the vapor-liquid interface enriched with salt, until the saturation
point when this salt is adsorbed in the hot metal surface. This pro-
cess forms a salt layer that acts like a thermal barrier creating a
Fig. 1. Cooling stages on cooling curve and cooling rate curve during quenching in a thermal resistance [9]. The thermal resistance is responsable for
volatile liquid. the appearance of the larvate boiling process [10].
When the larvate boiling process starts, a contact fluctuation,
point is the Leidenfrost temperature. This contact promotes the which takes fractions of seconds, between the quenchant and the
highest levels of heat transfers, the highest heat transfer coeffi- test-piece surface occurs. Therefore the presence of a thermal
cients and the highest cooling rate as can be seen on the Fig. 1. resistance at the hot surface acts to destabilize the vaporization
The Leidenfrost temperature has an important influence on the and promoting a contact between the liquid and the solid surface
transition between the first and the second stages, variations in even in high temperatures, increasing the heat transfer [10].
quenchants can affect this point [6]. Huang and Carey [7] shows In practice the cooling process can be accelerated by the surface
that the Leidenfrost temperature increases with the salt addition coating, increasing the thermal resistance, and thus the surfaces
on distilled water. The additives introduced in the quenchants have may have a higher boiling temperature than the exposed surfaces,
the function to raise the maximum cooling rate that must be fast destabilizing the formation of the vapor blanket (Fig. 3). Some
enough to avoid the critic rate of an undesirable micro- studies [10,11] show that a modified surface by a oxidation process
constituent of a steel CCT-diagram or to promote its or a coating can enhance the quenching performance. When a bare
transformation. probe was used for a quenching in water the classical boiling
The convective cooling mechanism starts when the hot compo- regimes were observed (Fig. 3a-c) but for the oxidized and rough-
nent temperature drops below the boiling point of the quenchant ness surfaces an accelerated transition of boiling regimes occurred
and so the nucleate boiling process stops [4]. The final stage is (Fig. 3d-f). Therefore, during the water quenching the surface prop-
the slowest stage, the agitation, viscosity of the quenchant medium erties and parameters of the liquid-solid contact can change and
and bath temperature can influence the heat transfer rates that control the cooling mechanisms [11].
will be responsible for distortions and internal stress inductions
[3]. 1.1.2. Micro explosion
To increase the heat transfer, agitation usually is introduced on The ability of an aqueous salt solution in to collapse the vapor
a quenchant to reduce the length and stability of vapor film, there- blanket can also be related with the deposition of small crystals
fore the Leidenfrost temperature increases with increasing adsorbed on the hot surface on the initial instants of the process.
quenching velocity and it promotes a faster wetting process [4]. Due to the high surface temperature the salt adsorbed on the sur-
Totten et al. [3] showed differences between distilled water and face by the vaporization and saturation causes micro explosions as
tap water on cooling rate when they were used as quanchant. shown in the Fig. 4. The micro explosions break the vapor blanket

Fig. 2. Transitions between the cooling mechanisms and wetting front scheme identifying the variation of the heat transfer coefficient (HTC).
 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818 809

Fig. 5. Schematic illustration of double electrical layer in a hot metal surface

charged (a) positively (b) negatively.

Fig. 3. Schematic illustration of influence of surface properties on the cooling

mechanisms during the contact liquid-solid. Ts corresponds the surface
temperature.

and allow that the nucleate boiling process begins sooner even in
high temperature [12], moreover, the collapse of the vapor blanket
occurs more homogenously across the surface being very interest-
ing to avoid distortions.

1.1.3. Double electrical layer

Initially the kinetic energy of the free electrons in a metal is not
enough to allow them escape the metal, however when the metal
reaches high temperatures the kinetic energy is enough to break
the metal boundaries and then the free electrons can escape from
the hot surface, thermionic emission process [13], in this case
metal surface is charged positively (Fig. 5a). But authors [14,15]
consider that the free electrons go to the cold surface, therefore
the metal surface is consist of negative charge distribution Fig. 6. Nonlinear behavior between EC and concentration. Adapted from [13].
(Fig. 5b). In both cases quenchants are constituted of electrolytes
that lead to the formation of a double layer of charges differing
in signal at the metal-liquid interface so the liquid is attracted to 1.1.4. Suppression of bubble coalescence
the metal surface and the formation of a film is impossible or very Quenching of spheres in deionized water and seawater were
difficult to occur. investigated by Hsu et al. [17] and the study reveals that the exis-
The relation between the salt concentration and the number of tence of the coalescence of bubbles is a necessary condition for the
charges carriers is not linear, there is an optimal concentration stabilization of the vapor film on the first stage in deionized water.
which corresponds the maximum critical heat flux [16], therefore But when the seawater was used as a quenchant, the coalescence
if the salt concentration exceeds this point an electrical double of bubble didn’t happen due to the zeta-potential of electrical dou-
layer at the surface metal-liquid can be formed and the number ble layer effect created between the sea water and sphere as well
of charge carriers will decrease. It is also related with electrical as other forces between bubbles in water with dissolved salts,
conductivity (EC) [13], Fig. 6, in such a way that for the greater and the vapor film was not formed. On the zeta-potential of elec-
electrical conductivity corresponds to the greater cooling rates trical double layer the ions in the seawater might be attracted to
and the optimal concentration. the heating surface firmly (Fig. 7) and prevent the coalescence of

Fig. 4. Schematic illustration of vapor blanket collapse by salt micro-explosions.

810 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818

heated surface. Therefore will be necessary raise the surface tem-

perature to producing more rapidly growing bubbles that can over-
come the added resistance of bubble merging consequently. This
behavior explain the increasing of the Leidenfrost temperature by
aqueous salt solution investigated by Huang, C. K. and Carey [7].
The use of salts as additives in aqueous cooling media allows
the physicochemical alteration of the water, and thus may influ-
ence the formation of the bubbles in the nucleation stage. The bub-
bles formed are in general numerous and smaller than that formed
in pure water, avoiding the coalescence and increasing the heat
flux due to the biggest wettability and the increase on the agitation
around the material surface caused by the strong movement of
these bubbles [9].

1.1.5. Balance of vaporization, condensation, and mass transfer

A comparative study among pure water, and solutions of 30%
CaCl2 and 30% H2SO4 used as coolants confirms that the additivities
CaCl2 and H2SO4 were able to form a thinner vapor film than in
water, promoting an increasing in the Leidenfrost temperature
Fig. 7. Schematic illustration of charge distribution due to zeta-potential.
reducing its duration. Using the acid, H2SO4, Arai, Furuya [21] could
concluded that the deposition of salt is not required to collapse the
nucleated bubbles due to stabilizing the liquid film between bub- vapor film. The results indicated that additives in water control the
bles [17]. The presence of charges of same signal can also form a net vaporization of the vapor-liquid interface and decrease the
repulsive force avoiding the coalescence. Other theories and stud- vapor film thickness (Fig. 8). When the thickness of the vapor film
ies have been published however the inhibiting effect of salts is not is reduced, the frequency of contact of solid-liquid is raised and the
yet well understood [18]. duration of the vapor film reduced, therefore the collapse happens
Guan and Carey [19] using theoretical models and molecular in higher temperature [21].
dynamics [20] studied the interface region between liquid and
vapor phase and its influence on the merging bubbles and how 2. Materials and methods
the presence of ions affect the structure of the film and its stability.
Their simulations showed that ions tends to be concentrated in the 2.1. Preparation of quenchants
center of the free liquid between the vapor phases, it can be
responsible to increase the surface tension that increases with Using as additives in distilled water NaCl, Na2SO4, NaNO2 and
the salt addition. They also evaluated the influence of the ions on NaHCO3 formed solutions in concentrations of 2 mass%, 7 mass%,
the critical thickness of stable liquid film, dc, (Eq. (1)) [19]. 12 mass% and 15 mass% except for the NaHCO3 that due to its
low solubility the concentrations were prepared in 2 mass%, 5
!0:25
AH L 2 mass% and 7 mass%. These solutions were analyzed at 25 °C and
dc ¼ ð1Þ 45 °C, without agitation. Distilled water was used as a comparative
4p3 r
quenchant at 25 °C and 45 °C, without and with agitation of
AH is the Hamaker constant, L lateral film size and r is the sur- 800 rpm.
face tension. If the thickness of the separating film is greater than
the critical thickness, film should remain stable and the pair of 2.2. Cooling curves acquisition and analyses
bubbles should remain separate after contact, whereas if film thin
to less than the critical thickness, the resulting rupture of the film Cooling characteristics of the research quenchants were mea-
would lead to the merging of bubbles in their experiment. They sured using a quenching test system, according to the international
related this behavior with the Leidenfrost transition and concluded standards ASTM D6200 and ASTM 6482. The whole quenching sys-
that enhance the stability of the thin film due to the salt addition tem consists of a heating furnace, a test probe (12.5 mm in diame-
hinders bubble merging disturbing the vapor film stability over a ter), a beaker, and a computer with acquisition control system. The

Fig. 8. Schematic illustration of reduction of vapor film thickness.

 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818 811

standard test probe was made of a cylindrical nickel–chromium can be observed that the increasing on temperature stabilizes the
alloy (Inconel 600) probe with a K-type thermocouple implanted vapor blanket and reduce the maximum cooling rate, moreover
in the center of it. In testing, the test probe was heated to the agitation can reduce a little the vapor blanket duration that
850 ± 2 °C in the furnace, held for 2 min and immerged into the took 6 s for the 0 rpm and 45 °C condition and 4,75 s for 800 rpm
2000 mL prepared quenchant. All the quenching tests were con- and 45 °C, being absent at 25 °C. Neither the Leidenfrost tempera-
ducted with an acquisition frequency of 8 Hz at least until a varia- ture nor the maximum cooling rates did not suffer a significant
tion no more than 10 °C in maximum cooling rate and its variation considering the same quenchant temperature.
correspondent temperature between two curves. After testing, a
temperature versus time curve and a temperature versus cooling
rate curve were obtained (Fig. 1). The surface phenomenon images 3.1.2. Salts
of the test probe in the different quenchants were captured by a As showed by the Fig. 9 that the temperature stabilized the
Canon EOS-1D Mark III on the high-speed mode of shutter. vapor blanket the further results will show the influence of salt
The cooling mechanism of the aqueous salt solutions were ana- aqueous solution to inhibit the first stage even in high quenchant
lyzed by calculating the heat transfer coefficient and heat flux temperature (45 °C). Figs. 10–15 show that, in general, a common
using a software develop by Cremonini [22], a member of this behavior appears in the NaCl, Na2SO4 and NaNO2 aqueous solution
research group. The software is based on the inverse heat conduc- which increasing the concentration of the salt increase the maxi-
tion problem (IHCP) in order to evaluate quenchants for quenching. mum cooling rates and or its correspondent temperature. The salt
To solve the IHCP Cremonini [22] based his calculations on the addition were able to reduce or eliminate the vapor blanket at
exact solution proposed by Burgraff [23] for cylindrical geometry 45 °C. The optimal concentration resulted in concentrations
and radial unidimensional heat conductivity. between 12 mass% and 15 mass%. However is important notice

2.3. Hardness test

Hardening power of the prepared quenchants was measured by

quenching process of AISI 1045 steel cylindrical samples with 25,4
mm diameter and 100 mm length, which was cut into the median
region of length. The AISI 1045 steel samples were previously nor-
malized and after heated into 850 °C, held for 120 min, and
quenched in the different prepared quenchants at 45 °C without
agitation. The microstructures of the samples after quenching were
observed using optical microscopes and the hardness values of the
samples quenched by different quenchants at 45 °C were measured
by Rockwell C hardness tester.

3. Results and discussion

3.1. Cooling curves and thermal properties

3.1.1. Water
Fig. 9 shows the cooling rates of distilled water at 25 °C and
45 °C with and without agitation (0 and 800 rpm). From the curves
Fig. 10. Cooling rates curves for aqueous salt solution of Na2SO4 at 25 °C.

Fig. 9. Cooling rate curves of distilled water at 25 °C and 45 °C with and without
agitation. Fig. 11. Cooling rates curves for aqueous salt solution of Na2SO4 at 45 °C.
812 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818

Fig. 12. Cooling rates curves for aqueous salt solution of NaCl at 25 °C. Fig. 15. Cooling rates curves for aqueous salt solution of NaNO2 at 45 °C.

that the increasing not always change significantly the maximum

cooling rate and for an industrial application lower concentrations
can give good quench performance. For the 2 mass% NaNO2 solu-
tions (Figs. 14 and 15), the vapor blanket is absent but the cooling
rate is lower than the distilled water. At 45 °C and for the 2 mass%
of NaCl and Na2SO4 solutions the vapor blanket still exist but with
a shorter duration in comparison the distilled water. For all the
conditions after the maximum cooling rate the values of the cool-
ing rate curves tend to the same behavior. Therefore, aqueous salt
solutions showed a better performance inhibiting or reducing the
vapor blanket and also increasing the cooling rates.
An opposite behavior appeared when the NaHCO3 solutions
were used as quenchants. The performance of these fluids reached
lowest values of cooling rates. The vapor blanket was stable even in
low temperature, which increased the cooling time and the dura-
tion of the vapor film. The thermal decomposition is a possible rea-
son for that. When NaHCO3 is heated over 100 °C two volatiles
compounds are formed H2O and CO2 [24], as indicated by Eq. (2).
Therefore, the NaHCO3 acts contributing for the maintenance and
formation of the vapor blanket by the release of volatile sub-
Fig. 13. Cooling rates curves for aqueous salt solution of NaCl at 45 °C. stances, Figs. 16 and 17.

Fig. 14. Cooling rates curves for aqueous salt solution of NaNO2 at 25 °C. Fig. 16. Cooling rates curves for aqueous salt solution of NaHCO3 at 25 °C.
 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818 813

Fig. 17. Cooling rates curves for aqueous salt solution of NaHCO3 at 45 °C.
Fig. 19. Maximum heat flux vs salt concentration at (a) 25 °C and (b) 45 °C.

D
2NaHCO3 ! Na2 CO3 þ CO2ðgÞ þ H2 OðgÞ ð2Þ
The presence and a long duration of the vapor film is responsi-
ble to limit the cooling process and small variation can be seen on
the maximum cooling rate and cooling rate curves, Figs. 16 and 17.
Heat flux as function of temperature was calculated using For-
ward Finite Differences for the salt solutions, Figs. 18 and 19. A fast
and high value of heat flux is reached and kept for the most part of
cooling duration of all quenchants, except for 2 mass% of NaNO2
and all NaHCO3 solutions. This behavior indicates that a more
homogeneous heat transfer process is happen during quenching.
Although, for the conditions where the vapor blanket appears this
values is reduced and the presence of vapor film can be seen. The
maximum heat flux (qmax) followed the same behavior for those Fig. 20. HTC during quenching in salt solutions of 2% of concentration.
presented in cooling rate curves showing that the increasing salt
concentration increases the heat extraction except for the NaHCO3
solutions and until the optimal concentration, Fig. 19. Fig. 19-b
decays during the cooling and shows a maximum value around
shows that the increasing quenchant temperatures reduces the
200 °C. The maximum HTC is also influenced by the quenchant
values of heat flux.
temperature, when quenchant temperature is raised the HTC is
The heat transfer coefficient (HTC) of the quenchants is pre-
reduced, Fig. 21-b, mainly for lower salt concentrations.
sented on Fig. 20. Differently from the behavior of heat flux, HTC

Fig. 18. Heat flux of salt solutions. Fig. 21. Maximum HTC vs salt concentration at (a) 25 °C and (b) 45 °C.
814 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818

3.2. Hardness test with a duration not so long and the thickness of the metal speci-
men is thick, therefore when the vapor film collapses the surface
To test the quench power of the salt aqueous solutions in a low is already transformed in diffusion phases but the bulk still with
hardenable steel, specimens of AISI 1045 steel were heated at austenite is able to be influenced by the fast cooling promoted
850 °C and cooled in the solutions at 45 °C and without agitation. by the nucleate boiling process and can be transformed in grains
The hardness profile and the predict martensite content is shown with a refined microstructure and in martensite. Fig. 22(b) show
on Fig. 22. Fig. 22(a) shows an uncommon behavior and an internal the lowest values of hardness due to the stable vapor film that
section has a higher hardness than the surface region [25–27]. This had a long duration and weren’t able to quench the steel and pro-
phenomenon appears when the vapor blanket process happens mote the martensite transformation. Fig. 22(c-e) shows the good

Fig. 22. Hardness profile of quenched AISI 1045 specimens with 25 mm of diameter in (a) water (b) NaHCO3 (c) Na2SO4 (d) NaCl (e) NaNO2.
 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818 815

ability to quench a low hardenable steel. The aqueous salt solution 3.3. Cooling process
at 45 °C promoted a fast cooling and quenched the surface of the
specimens reaching more than 90% of martensite in this region. To show the influence of vapor film on quenching process,
The elimination of the vapor blanket by the 2 mass% NaNO2 images during cooling of a probe were recorded and the Figs. 23–
solution, even reaching lower cooling rates than distilled water, 25 show the different cooling stages that happens when distilled
was able to quench the AISI 1045 steel. Therefore, the presence water, NaHCO3 and NaCl were used as quenchants at 45 °C without
of the vapor film is a limitation to quenching. agitation.

Fig. 23. Quenching process in distilled water at 45 °C. Arrows indicate the vapor blanket collapse point and circles show the wetting front propagation.

Fig. 24. Quenching process in 5 mass% NaHCO3 solution at 45 °C. Arrows indicate the vapor blanket collapse point and circle shows the presence of vapor blanket after the
drop surface temperature.

Fig. 25. Quenching process in 12 mass% of NaCl solution at 45 °C. Circles show the bubble cluster burst out from the surface.
816 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818

Fig. 23 shows the quenching process in distilled water. The Table 1

arrows on Fig. 23 show the vapor blanket collapse positions, notice Cooling time for distilled water, 5% NaHCO3 and 12% NaCl solution at 45 °C, obtained
by Tensi probe.
that the first collapse happened on the bottom creating a wetting
front that moves to the top while the probe still remains in high Distilled water Position - Tensi Probe
temperature characterized by the incandescent color of the probe 2 mm 15 mm 30 mm Center
(circle on Fig. 23(e) and (f)). A next position of the collapse of the Temperature [°C] Time [s]
vapor film happened in an upper position. 700 5,5 8,5 11,2 11
Fig. 24 shows the quenching process in 5 mass% of NaHCO3. The 600 6 9,25 13 12,3
vapor blanket collapse happened when the probe was in a low 500 6,5 9,7 13,7 13,1
300 8 11,6 15,3 15,6
temperature and didn’t be able to promote the quench due to the
200 10 14 17,8 18,4
long duration of the first stage that promoted a slow cooling. Dur-
5% NaHCO3 Time [s]
ing the immersion a fluctuation on the quenchant happened and a
700 5,5 8,5 11 11
region on the middle of the probe cooled first but it wasn’t able to 600 6 9,3 13,2 12,3
break the vapor blanket (Fig. 24(a)). The arrows on Fig. 24 show the 500 6,5 9,7 13,7 13,2
vapor blanket collapse positions and two wetting front were 300 8 11,5 15,3 15,6
200 10 14 17,8 18,4
formed that moved one against other (Fig. 24(e)).
Fig. 25 shows the quenching process in 12 mass% of NaCl. No 12% NaCl Time [s]
700 3,3 3,4 3,5 4,5
vapor film were formed and a fast contact between the solution
600 3,6 3,7 3,9 5,1
and the probe happened, the raised Leidenfrost temperature by 500 3,8 4,1 4,4 5,7
the salt addition explained by [19,20] can justify the absence of 300 5,1 5,9 6,1 7,9
vapor blanket. The contact between the solution and the hot sur- 200 7 8,2 8,4 10,5
face promotes a fast evaporation rate due to the high metal tem-
perature and a condensed cloud of vapor is formed around the
surface (Fig. 25(a) and (b)). As reported by S.-H. Hsu et al. [17] dur-
ing the transition boiling a cluster of bubble is formed and burst
mote a non-uniform cooling over the surface that is characterized
out from the surface, pointed by the circles on Fig. 25(b). At this
by the differences on time among the thermocouples positions for
point the metal temperature is low enough to allow the salt
a same temperature. For the 12 mass% NaCl solution at 45 °C the
adsorption on the surface (Fig. 25(c)-(d)). The salt coating con-
cooling time for all thermocouples positions is reduced and a small
tribute to keep the contact between the solution and the hot sur-
variation for a same temperature is observed that represents an
face [10,11] improving the heat extraction. The salt coating is
improvement on the uniformity of the heat extraction through
solubilized (Fig. 25(d)-(f)) and a fast bubble nucleation remains
the cross section of the probe.
on the probe surface (Fig. 25(h) and (i)).
Metallography images of the cylindrical specimens of AISI 1045
Tensi probe (Fig. 26) was used to evaluate the wetting process.
steel quenched are showed on Fig. 27 representing range of
The design of this probe is different due to the position of the ther-
hardness (a) over 55 HRC, (b) 50HRC < HRC < 55HRC, (c)
mocouples that there are 3 of them positioned on the surface
40HRC < HRC < 50HRC and (d) around 30 HRC. Fig. 27(a) show a
region at 2 mm, 15 mm and 30 mm from the bottom of probe
martensitic microstructure characterized by an acicular morphol-
and another one at the geometric center of the probe. The cooling
ogy, this phase is responsible for the highest hardness values.
time is showed on Table 1.
Fig. 27(b) and (c) show the formation of diffusional micro-
The cooling time for distilled water and 5 mass% NaHCO3 solu-
constituents (perlite, bainite and ferrite) in the grain contours
tion at 45 °C show the influence of the wetting front moving to pro-
and acicular martensite inside the grain, the amount of diffusional
micro-constituents on grain contours is higher in Fig. 27(c) and its
reduces the hardness in this region. Fig. 27(d) show a mixture of
diffusional micro-constituents and no martensite phase was
observed due to the thickness of the specimens or the slow cooling
promoted by the quenchant. Variations on the hardness, Fig. 22,
are justified by refining or not of the microstructure.

4. Conclusions

The cooling mechanisms are affected by the salt addition. The

increase on salt concentrations increases the cooling rates and
thermal properties until the optimal concentration except for the
NaHCO3 solutions. NaHCO3 suffers a thermal decomposition that
released water and carbon dioxide in gaseous form that con-
tributed for the stabilization and maintenance of the vapor film.
The increase on bath temperature reduces the severity of quen-
chant and stabilizes the vapor film, but for the salt solutions, NaCl,
Na2SO4 and NaNO2, the inhibition of the vapor blanket were able to
quench the surface of a low hardenable steel. The increasing on
hardness occurs due to the formation of martensite phase. Adding
salt forming solutions promoted a more homogeneous cooling that
kept high values of heat flux during the most part of the drop of
Fig. 26. Illustration of the Three Near-Surface Probes (Tensi probe). T01-03 are the temperature. Homogeneous cooling is desired to avoid distortions
surface thermocouples positions, Tz is the centered thermocouple. and cracking during quenching process.
 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818 817

Fig. 27. Metallography images with 200xzoom of AISI 1045 samples quenched and chemically attacked with 2% nital showing range of hardness (a) over 55 HRC, (b)
50HRC < HRC < 55HRC, (c) 40HRC < HRC < 50HRC and (d) around 30 HRC.

Acknowledgments [11] H. Yeom, H. Jo, G. Johnson, K. Sridharan, M. Corradini, Transient pool boiling
heat transfer of oxidized and roughened Zircaloy-4 surfaces during water
quenching, Int. J. Heat Mass Transf. 120 (2018) 435–446, https://doi.org/
This study was financed in part by the Coordenação de Aper- 10.1016/j.ijheatmasstransfer.2017.12.060.
feiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance [12] W. Luty, Cooling media and their properties, in: B. Liscic, H.M. Tensi, L.C.F.
Canale, G.E. Totten (Eds.), Quenching Theory and Technology, second ed.,
Code 001 Taylor & Francis, FL, 2010, pp. 359–443.
[13] L.V. Petrash, Some physical phenomena during quenching, Translated from
Metallovedenie i Termicheskaya Obradotka Metallov, No 6, 1973, UDC
Appendix A. Supplementary material
621.785.6.08:531.7.
[14] N.I. Kobasko, SH.E. Guseynov, An explanation of the nature of thermal waves
Supplementary data to this article can be found online at ‘‘a poker effect” on the basis of hyperbolic heat conductivity equation analysis
https://doi.org/10.1016/j.ijheatmasstransfer.2019.06.036. and existence of free electrons in metals, in: Muhammet Koksal Balas (Ed.),
Valentina E, WSEAS Press, Recent Researches in Circus and Systems, 2014,
ISBN 978-1-61804-108-1, pp. 167–172.
References [15] N.I. Kobasko, Effect of free electrons in steel on its quenching process in water
and water salt solutions, EUREKA: Phys. Eng. 1 (2018) 39–46, https://doi.org/
[1] G. Krauss, Principles of Heat Treatment of Steel, American Society for Metals, 10.21303/2461-4262.2018.00529.
Ohio, EUA, 1980. [16] N.I. Kobasko, Intensive steel quenching methods, in: B. Liscic, H.M. Tensi, L.C.F.
[2] G. Ramesh, K.N. Prabhu, Spatial dependence of heat flux transients and wetting Canale, G.E. Totten (Eds.), Quenching Theory and Technology, second ed.,
behavior during immersion quenching of inconel 600 probe in brine and Taylor & Francis, FL, 2010, pp. 509–567.
polymer media, Metall. Mater. Trans. B 45B (2014) 1355–1369, https://doi.org/ [17] S.-H. Hsu, Y.-H. Ho, M.-X. Ho, J.-C. Wang, C. Pan, On the formation of vapor film
10.1007/s11663-014-0038-7. during quenching in de-ionized water and elimination of film boiling during
[3] G.E. Totten, C.E. Bates, N.A. Clinton, Handbook of Quenchants and Quenching quenching in natural sea water, Int. J. Heat Mass Transf. 86 (2015) 65–71,
Technology, ASM International, Materials Park, OH, 1993. https://doi.org/10.1016/j.ijheatmasstransfer.2015.02.049.
[4] H.M. Tensi, K. Lanier, G.E. Totten, G.M. Webster, Quenching uniformity and [18] M. Firouzi, T. Howes, A.V. Nguyen, A quantitative review of the transition salt
surface cooling mechanisms, in: Heat Treating: Proceedings of the 16th concentration for inhibiting bubble coalescence, Adv. Colloid Interface Sci. 222
Conference, Ohio, 1996, pp. 3–8, doi: 10.1361/cp1996ht003. (2015) 305–318, https://doi.org/10.1016/j.cis.2014.07.005.
[5] P. Fernandes, K.N. Prabhu, Effect of section size and agitation on heat transfer [19] Y. Gan, V.P. Carey, An exploration of the effects of dissolved ionic solids on
during quenching of AISI 1040 steel, J. Mater. Process. Technol. 183 (2007) 1–5, bubble merging in water and its impact on the Leidenfrost transition, paper
https://doi.org/10.1016/j.jmatprotec.2006.08.028. IHTC14-23330, in: Proc. 14th International Heat Transfer Conference, IHTC14,
[6] K.H.M. Abdalrahman, Sabariman, E. Specht, Influence of salt mixture on the Washington, DC, USA, August 8–13, 2010.
heat transfer during spray cooling of hot metals, Int. J. Heat Mass Transf. 78 [20] Y. Gan, V.P. Carey, Hybrid modeling of interfacial region thermophysics and
(2014) 76–83, https://doi.org/10.1016/j.ijheatmasstransfer.2014.06.070. intrinsic stability of thin free liquid films, Int. J. Heat Mass Transfer 53 (2010)
[7] C.-K. Huang, V.P. Carey, The effects of dissolved salt on the Leidenfrost 2169–2182.
transition, Int. J. Heat Mass Transf. 50 (2007) 269–282, https://doi.org/ [21] T. Arai, M. Furuya, Effect of salt additives on film boiling heat transfer and
10.1016/j.ijheatmasstransfer.2006.06.031. mechanism of quenching temperature rise, Heat Transfer—Asian Res. 40 (2)
[8] C.E. Bates, G.E. Totten, R.J. Brennan, ASM Handbook, vol 4, Heat Treating, ASM (2011) 101–113, https://doi.org/10.1002/htj.20327.
International (1991) 160–290. [22] G.E.S.O. Cremonini, Aplicação do método inverso de condução de calor na
[9] F. Moreaux, G. Beck, P. Archambault, Effect of workpiece surface properties on avaliação de fluidos de resfriamento para têmpera, 2014, pp.173, Dissertação
cooling behavior, second ed., Taylor & Francis, FL, 2010, pp. 289–313. Mestrado, Escola de Engenharia de São Carlos, Universidade de São Paulo, São
[10] F. Moreaux, J.C. Chevrier, G. Beck, Destabilization of film boiling by means of a Carlos, 2014. (portuguese), http://www.teses.usp.br/teses/disponiveis/18/
thermal resistence, Int. J. Multiphase 2 (1975), Elsevier. 18158/tde-11052015-125002/pt-br.php.
818 L.H. Pizetta Zordão et al. / International Journal of Heat and Mass Transfer 140 (2019) 807–818

[23] O.R. Burggraf, An exact solution of the inverse problem in heat conduction [26] N. Shimizu, I. Tamura, Effect of the discontinuous change in cooling rate during
theory and applications, ASME J. Heat Transfer 86C (1964) 373–382. continuous cooling on perlitic transformation behavior of steel, Trans. ISIJ 17
[24] A.S. Maia, V.K.L. Osorio, Decomposição Térmica do Bicarbonato de Sódio – Do (1978) 7.
Processo Solvay ao Diagrama tipo Ellingham, Quim. Nova 26 (4) (2003) 595– [27] B. Liscic, V. Grubisic, G.E. Totten, Controllable delayed quenching. ‘‘Equipment
601. and processes, in: Proceedings of the Second Interl. Conference on Quenching
[25] N. Shimizu, I. Tamura, Inverse quench-hardening of steel, Trans. ISIJ 18 (1978) and the Control of Distortion, ASM International, 1996, p. 8.
5.