ArticlePDF Available

Figures

Content may be subject to copyright.
Report on IFHTSE Liquid Quenchant Database
Project
I. Felde*
Background
Originating from discussions held in Rio
de Janeiro in July 2010 and in Glasgow
2011, a proposal was made to
International Federation for Heat
Treatment and Surface Engineering
(IFHTSE) that an international
collaborative project should be launched
with the objective of generating a
comparative database of the cooling
intensities of liquid quenchants for use
industrially as a selection tool in relation
to specific materials, conditions and
engineering requirements. At present,
there is no universally recognised
method or procedure for the
measurement, recording and
comparison of relative cooling intensities
of different quenchants for components
of complex geometry. A globally usable
database is therefore required which
covers a range of selected quenchants
acting under specified conditions.
Goals of LQD project
Globally, for real quenching of real
engineering components, a very wide
range of liquid quenchants of different
provenances and grades are available:
mineral oils, accelerated oils,
martempering oils, polymer solutions,
water, brine and salt baths.Any of these
quenchants may be used in different
conditions (bath temperatures and
agitation rates), thus multiplying the
number of possible combinations. Yet
there is no generally recognised method
or technique for the measurement,
recording and comparison ofthe relative
cooling intensities of different
quenchants. It is also fully
acknowledged that a great deal of
progress has been made over the past
few years, over a wide range of related
but different developments. It will be
generally beneficial to those working on
these activities, as well as to potential
users, to ensure that there is a more
widespread understanding of those
developments and their importance, for
example: intensive quenching, delayed
quenching, spray quenching,
martempering and austempering in salt
baths, for which almost no heat transfer
data exist in practice. Given a database
of the cooling intensities for specific
liquid quenchants in specific conditions,
computer modelling and prediction of
quench hardness, microstructure,
stresses and distortion can become the
normal practice when quenching real
engineering components with complex
geometry.
A first outline of the project was
circulated to potential participants
(around25organisationsin14
countries) and to the IFHTSE Executive
Committee. The reactions from
recipients were almost entirely positive
and the project was generally agreed to
be worthwhile from an industrial, as
well as an intellectual and scientific
viewpoint. The intended output is
believed to constitute a valuable in-
dustrial tool from the viewpoints of:
Nimproved technology
Navoiding non-optimum selection of
quenchants and quenching
conditions
Nsaving of manufacturing costs by
reduction of distortion and scrap.
The two phases of the project are
described as below.
Phase 1
The compilation of a database of cooling
intensities of differentliquid quenchants,
is characterised by laboratory test
methods and evaluation of heat transfer
data in workshop conditions. The
database will cover a wide range of
quenchants for specified conditions. To
achieve this, a group of appropriate
institutions will be invited to take part
in Round-Robin Tests on relevant
methods and selected quenchants.
Each institution will follow a generally
agreed and prescribed test procedure.
The intention is to yield information
about the cooling intensity of a liquid
quenchant in a globally comprehensive
form as distinct from the partial
information available from any given
single source at present. This will mean,
for the most frequently used liquid
quenchants (oils, polymer solutions,
water and inorganic water solutions):
Ndescription of chemical and physical
properties
Nspecification of test conditions
Nincorporate own testing methods
Nlaboratory test ISO 9950 with
resulting cooling and cooling rate
curves, and heat transfer coefficient
as a function of surface
temperature, calculated using a
uniformly predetermined
mathematical procedure
Nworkshop test using the Liscic probe
of 50 mm diameter6
200 mm with three resulting cooling
curves and calculated heat transfer
coefficient as a function of surface
temperature and of time, respectively
Nmaximum heat flux density
determined by e.g. Japanese New
*Corresponding author, email imre.felde@gmail.
com
Table 1 Partners contributing in Phase 1
Organisation Short name Country
Dunau´jva´ ros College DUF Hungary
Burgdorf, Germany BurgD Germany
CNIITMASH, Russia CNIIT Russia
Houghton International, USA HIUSA USA
Idemitsu Kosan, Japan IKOS Japan
Intensive Technologies Ltd, Ukraine IQT Ukraine
IWT, Bremen, Germany IWT Germany
KSHT, Korea BIT Korea
National Institute of Technology, Karnataka, India NIT India
Petrofer Chemie, Germany PTROF Germany
QRC, Zagreb, Croatia QRC Croatia
Swerea IVF, Sweden IVF Sweden
University of Sao Paolo, Brazil USP Brazil
Utsunomiya University, Japan, UtUni Japan
Fuchs Europe Schmierstoffe GmbH Fuchs Germany
ß2014 IHTSE Partnership
Published by Maney on behalf of the Partnership
2International Heat Treatment and Surface Engineering 2014 VOL 8NO 1DOI 10.1179/1749514813Z.000000000103
Silver Probe, according to JIS 2242 –
method B
Nnoise tests to determine transition
film and nucleate boiling processes.
ThemainoutputfromPhase1isthe
DATABASE on CD, which intended to
be a tool for designers and engineers in
actual practice, and available worldwide
at a realistic price.
Phase 2
It is well known that many academic
institutions have either developed their
own software or use commercially
available software packages for
modelling quenching phenomena and
their consequences, mostly for
‘scientific’ purposes. Larger companies
have also written their own software, or
use commercially available packages,
for prediction of mechanical properties
and distortion in their main production
components. It is also well known that
in the case of quenching a real
workpiece of complex geometry, the
heat flux and consequently, the heat
transfer coefficient, varies over the
surface.
This complex situation usually
means different durations of the
vapour film phase at different points,
which leads to distortion. To calculate
these heat flux variations for a complex
workpiece via 1D (or in the case of
symmetrical bodies 2D) models, the
inverse heat conduction numerical
methods currently in use are not
adequate. A 3D inverse heat
conduction method combined with
calculation of local heat fluxes could
enable computer modelling of complex
cases.
The main objective of Phase 2 is
further development of mathematical
models, and production of adequate
software, for prediction of quench
hardness, microstructure, stresses and
distortion, even in the most difficult
1 Cooling and cooling rate curves of Thermisol QB46: 50, 75 and 100
u
C
Felde Report on IFHTSE Liquid Quenchant Database Project
International Heat Treatment and Surface Engineering 2014 VOL 8NO 13
situations when real workpieces of
complex shape are quenched.
Organisations participating in
Phase 1
After the circulation of the Phase 1
concept, 15 organisations and labo-
ratories expressed their intention
to participate and to perform the
cooling curve acquisitions, and the
HTC estimations are presented in
Table 1.
Quenchants applied in Phase 1
In Phase 1, the cooling curves are to
be obtained for three oil types from
the quenchants producer Fuchs
Europe Schmierstoffe GmbH. The
heat transfer coefficients as a
function of temperature are to be
predicted by performing inverse heat
conduction problem (IHCP)
techniques using the cooling curves
recorded.
Phase 1 (2012 Q1–Q4)
The first phase of the project was
regarding the evaluation of the cooling
curve acquisition method ISO 9950 by
performing cooling curve
measurements and the estimation of
heat transfer coefficient as a function
of time or surface temperature.
The cooling curve acquisitions have
been carried out by eight participants in
eight countries. The partners have
evaluated three types of oils produced
by Fuchs GmbH using three different
temperatures for each coolant (each
measurement has been repeated;
36362518 cooling curves have
therefore been recorded by each
partners). The contributors have applied
instruments produced by Swerea-IVF,
Drayton and some built by their own.
The sample recording frequency was
set from 10 to 100 Hz.
2 Cooling and cooling rate curves of Thermisol QH10: 50, 65 and 80
u
C
Felde Report on IFHTSE Liquid Quenchant Database Project
4International Heat Treatment and Surface Engineering 2014 VOL 8NO 1
The HTC calculations have been
done by six contributors. It is
important to note that four different
approaches have been applied to
calculate the thermal boundary
conditions. These methods are based
on different assumptions to solve the
1D inverse heat transfer problem of
the transient cooling of a cylindrical
rod. The different assumptions involve
different mathematical solutions.
The following parameters have been
calculated from the cooling curves:
Ntime to 600uC, s
Ntime to 400uC, s
Ntime to 200uC, s
NCR
300
,uCs
21
NCR
550
,uCs
21
NT
cp
,uC
NT
vp
,uC
NCR
max
,uCs
21
NT(CR
max
), uC
Nt(CR
max
), s.
The cooling curves recorded as well as
the related cooling rate curves are
shown in Figs. 1–3. The calculated
parameters related to each cooling
curve are shown in Fig. 4. The
following observations can be made by
analysing the charts:
Nthe cooling curves and the cooling
rate curves performed by DUF,
FUCHS, IVF, BIT and QRC seem to be
similar and show only a small
(acceptable) difference. A similar
regime of cooling can be observed on
the curves recorded by NIT; however,
a certain time delay occurred in every
measurement
Nthe curves acquired by USP are
very much similar to the curves
made by the other mentioned
partners, but due to the fact that
the Brazilian participant started the
quenching from 900uC, the cooling
curves have been shifted. The
cooling rate curves of USP show a
good agreement with curves
3 Cooling and cooling rate curves of Thermisol QH120: 60, 110 and 170
u
C
Felde Report on IFHTSE Liquid Quenchant Database Project
International Heat Treatment and Surface Engineering 2014 VOL 8NO 15
4 Values for quenchants evaluated of mean and standard variation of time to 600, 400 and 200
u
C, CR300, CR550,
t(CRmax), CRmax and T(CRmax), plus range of these variables and of tfor T
cp
,T
vp
and T(CRmax): see Table 2 for key
Felde Report on IFHTSE Liquid Quenchant Database Project
6International Heat Treatment and Surface Engineering 2014 VOL 8NO 1
collected by the other six
contributors
Nthe curves produced by CNIIT are
significantly different from the
temperatures measured by other
partners using the same quenchants
with the given conditions
Nthe range of the calculated
quantities representing the cooling
characteristics of the quenchants is
unexpectedly huge. For example,
the difference between the
minimum and maximum value of
CR
max
could reach 100uCs
21
, or the
T(CR
max
) could exceed 60uC. The
huge range of the quantities is
mainly driven by the CNIIT curves
Nthe HTC(T) functions generated by
using the cooling curves show
similar trends including the
existence of the three different
stages of oil quenching. The
deviation between the heat transfer
coefficient functions predicted by
using the cooling curves recorded in
the same type of cooling medium
seems unacceptably large.
The reason of the discrepancies could
be originated to the following facts
and/or factors:
Neven though the ISO 9950 method
defines exactly the temperature of
the probe before cooling (850uC),
some participants started the
sampling from higher temperatures
Nsome instruments controlled
temperature recording with a trigger
function. This option starts the data
collection process from the moment
the temperature dropped below the
set temperature threshold (i.e. 850uC).
Owing to starting the quenching from
higher temperatures and applying the
triggering function, part of the cooling
process is not recorded
Nsome participants used 2 L of liquid
to cool the probes while others used
1 L. The difference of quantity
definitely does affect cooling
characteristics
Nthere are contributors who used their
own installed (and developed)
ISO 9950 system, while others
are using industrial products (provided
by Swerea-IVF or Drayton). There is
no information as to if the ‘house
made’ equipment have undergone
the same quality controls as the
industrial ones. The origin of the
discrepancy could be here as well
Nthe determination of thermal
boundary conditions requires the
application of inverse mathematical
(IHCP) approaches. The inverse
heat conduction problems to be
solved by these procedures belong
to highly nonlinear ill posed
problems. Therefore, no exact
heuristics are available or in other
words, several types of solution
technique could provide the proper
solution. The partners are using at
least four different type of inverse
methods based on different
mathematical backgrounds.
Unfortunately, no comparative
study has been carried out to
evaluate the applicability of these
approaches. Consequently, the big
scatter of HTC(T) predicted could be
related to, on the one hand, the
cooling curves altering significantly
and on the other hand, the IHCP
techniques based on different
mathematical methods.
Conclusions and further tasks
The experiments carried out during
Phase 1 showed that acceptable
agreement between the temperatures
recorded in laboratories at four
continents by using the ISO 9950
method could be achieved. However, a
certain deviation between the cooling
curves can originate from the
application of different
instrumentations as well as the
inaccurate attention to the conditions
used during the measurements.
Therefore, the following tasks are
proposed:
Nmore accurately defined and strictly
complied with conditions must be
followed during the cooling curve
acquisition of Petrofer oil
quenchants. In this case, three
different types of oils will be
analysed at three given
temperatures. Better or more similar
results are expected by performing
the sampling using strictly applied
conditions
Nperforming a benchmark test of the
inverse heat conduction algorithms
applied by the contributors would
be beneficial in order to indicate
their performance. The test will be
carried out on five cooling curves
generated by a finite element
method model of an ISO 9950
probe using five different
hypothetical HTC(T) functions. The
task of the partners will be to
generate the thermal boundary
condition functions [to reconstruct
the HTC(T) functions]. The
evaluation of inverse techniques
will be obtained from the predicted
heat transfer coefficients.
Acknowledgement
The contributors would like to thank
the support of Fuchs Schmierstoffe
GmbH and Petrofer Chemie H.R.
Fischer GmbHzCo. KG.
Table 2 Key to Fig. 4
Quenchant Temp./uC
Thermisol QB46
150
275
3 100
Thermisol QH10
450
565
680
Thermisol QH120
760
8 110
9 170
Felde Report on IFHTSE Liquid Quenchant Database Project
International Heat Treatment and Surface Engineering 2014 VOL 8NO 17
... This would be expected since the fluid viscosity also increases with time. However, there was no discernible correlation of T S or t hmax with respect to aging time, which may be due to the overall variability of the ASTM D6200 test procedure used for cooling curve data acquisition [214]. ...
Article
There is an ongoing interest in the development and use of renewable base stocks to formulate quenchants. The most common criterion of vegetable oils as renewable base stocks is their biodegradability and that they be non-toxic. A comprehensive overview of all aspects of vegetable oils that impacts their potential for commercial use is provided. Topics discussed include: vegetable oil structure, processing, physical properties, classification, biodegradation and toxicity; oxidation and inhibition; wetting and wetting kinematics; and applications. As a class, vegetable oil-based quenchant formulations reported in the literature to date exhibit a number of disadvantages, the most notable being their relatively poor thermal-oxidative stability in comparison with petroleum oil-based quenchants in use. Potential pathways to vegetable oilbased fluid compositions that may rival the thermal-oxidative stability of many petroleum oil-based quenchants were introduced.
... Originating from discussions held in Rio de Janeiro in July 2010 and in Glasgow 2011, a proposal was made to the International Federation for Heat Treatment and Surface Engineering (IFHTSE) that an International collaborative project should be launched with the objective of generating a comparative database of the cooling intensities of liquid quenchants, for use industrially, as a selection tool in relation to specific materials, conditions and engineering requirements (Felde, 2014). The aim of Phase 1 was to validate the quenchant evaluation method ISO 9950 for providing reliable and repeatable results. ...
Article
The estimation of thermal boundary conditions occurring during the immersion quenching processes is an essential requirement for characterisation of the heat transfer phenomena. In this work, six numerical techniques that are used to calculate the heat transfer coefficient (HTC), HTC(t) functions during immersion quenching of the ISO 9950 probe are compared. The estimation of the HTC(t) functions has been performed by seven contributors (from three continents) using the six different numerical techniques. First, hypothetical HTC(t) function generated and then these function applied to calculate cooling curves by using 1D axis symmetrical heat transfer model. The contributors performed inverse heat conduction problem (IHCP) calculations to recover the original HTC(t) functions using the cooling curves. The accuracy and the relevancy of the estimation approaches applied are evaluated by using error functions. In this paper, the activities obtained at the Step 2 of the project 'Liquid Quenchant Database' initiated by the International Federation for Heat Treatment and Surface Engineering are summarised.
... Moreover, detailed data on liquid quenchant characteristics are also needed. For this reason a global heat treatment (HT) workshop was initiated by Felde to compose a heat treatment database based on the results obtained from industry and research institutes [21,22]. ...
Article
Full-text available
The current study focuses on an industrial hardening process where the steel parts are austempered in a molten salt bath. The aim of the study was to compose a predictive tool for more qualified process adjustment and control in a hardening plant. The process approach considers specific 3.0 to 4.0 mm thick strip steel products and steel grades with carbon content in the range from 0.27 to 0.62 wt%. Austempering in salt produces high hardness and ductility of steel as the final microstructure is bainitic. Salt bath cooling is not so severe as cooling in oil, polymers, or water and has an advantage of a uniform cooling rate. There are several approaches to determining hardenability of steels and interrelating it to the quenching process to predict the final hardness of the parts. The best known are the Jominy hardenability test and the Grossman hardenability number where hardness is presented as the function of the specimen geometry. To determine the quenchant properties the Grossman number, nickel ball, hot wire, cooling curve, and quench factor methods are used. These methods are guiding tools to the engineer for choosing the steel grade and the hardening process. The latest approach is quench factor analysis, which interrelates the cooling curve of the quenchant and material hardenability data with a single number and allows prediction of the properties of the hardened material. The imperfection of the quench factor is limited availability of mathematical constants for different steel grades to make calculations. As the cooling process and phase transformations are nonlinear, the computation accuracy does not seem to satisfy the heat treaters to utilize it in the hardening plant. The types of heat treatment equipment in use have different characteristics, which are related to the technological process used and the design of the machine. The available engineering diagrams for steels and processes are useful to have a general view of the process but still each workshop has to specify the approach to be used in its particular case.
ResearchGate has not been able to resolve any references for this publication.