Int. J. Materials and Product Technology, Vol. 24, Nos. 1–4, 2005 101
Copyright © 2005 Inderscience Enterprises Ltd.
Oxidation of vegetable oils and its impact on
Lauralice de C.F. Canale*
Departamento de Materiais,
Automobilística e Aeronáutica,
EESC, Universidade de São Paulo, USP,
São Carlos, SP, Brazil
Fax: 55 16 273 9590 E-mail: email@example.com
Mauro R. Fernandes and
Sylvana C.M. Agustinho
Instituto de Química de São Carlos,
Universidade de São Paulo, USP,
São Carlos, SP, Brazil
E-mail: firstname.lastname@example.org E-mail: email@example.com
George E. Totten
Department of Mechanical and Materials Engineering,
Portland State University, Portland, OR, USA
Alessandro F. Farah
UNIP – Paulista University, Araraquara, SP, Brazil
Abstract: Traditionally, mineral oils have been one of the most important
quenchants used. However, their substantial deficiencies with respect to
environment friendliness and toxicity as well as long-term, low-cost supply
necessitate the search for alternative replacement media. Quenching
performance of petroleum oils is limited by the oxidative degradation
properties, which are composition dependent. Upon repeated exposure to the
relatively high interfacial temperatures between the steel surface and the oil,
petroleum oils undergo thermal and oxidative degradation leading to significant
variation in their quenching performance. Therefore, this is a particularly
important performance parameter that must be examined for any alternative
quenching medium. One class of alternative fluids is vegetable oils, which are
typically biodegradable and non-toxic. However, vegetable oils typically
exhibit relatively poor oxidative stability properties, and therefore it is
important to determine the potential impact of oxidation on quenching
performance. The results reported here are the first step in a larger study. In this
work, uninhibited vegetable oils were studied using a laboratory apparatus and
102 L.C.F. Canale et al.
procedure previously reported to promote premature oxidation of petroleum
oils to approximate actual use conditions. Experimental fluids are examined
and compared to typical petroleum quench oil compositions using viscosity,
infrared spectroscopy, 13 CMR spectroscopy and cooling curve performance
according to ASTM D 6200. The results obtained indicate that vegetable oils
are promising alternatives to petroleum oils as quenchants but that to be
commercially feasible, appropriate antioxidants must be used.
Keywords: vegetable oils; heat treating; quenching; quenchants.
Reference to this paper should be made as follows: Canale, L.C.F.,
Fernandes, M.R., Agustinho, S.C.M., Totten, G.E. and Farah, A.F. (2005)
‘Oxidation of vegetable oils and its impact on quenching performance’,
Int. J. Materials and Product Technology, Vol. 24, Nos. 1–4, pp.101–125.
Biographical notes: Lauralice de C.F. Canale received her MS
(Bioengineering) and PhD (Metallurgy) degrees from University of São Paulo,
Brazil. She did her post-doctoral work at the University of Tennessee. Canale is
currently an Associate Professor in the Departamento de Eng. Materiais,
Aeronautica e Automobilística of Escola de Engenharia de SãoCarlos, at the
University of São Paulo in São Carlos, Brazil, where she is responsible for heat
treatment and surface engineering technology and research. She has
co-authored numerous journal and conference papers in addition to various
book chapters on tribology and heat treating.
Mauro R. Fernandes received his BS, MS, and PhD degrees from Universidade
de São Paulo in São Carlos, SP, Brazil. During his MS and PhD Fernandes
had worked on electroluminescent polymers from PPV – poly (phenylene
vinylene) – class. Atpresent, he is incharge of the Vibrational and Electronic
Spectroscopy Laboratory at the Central de Análises Químicas e Instrumentais
at the Instituto de Química de São Carlos, Universidade de São Paulo, Brazil.
Sylvana C.M. Agustinho, received her BS, MS and PhD degrees from
Universidade de São Paulo in São Carlos, Brazil. She did a post-doctoral stage
at University of Sheffield, England for one year. She is currently responsible
for the Nuclear Magnetic Resonance Laboratory at the Central de Análises
Químicas e Instrumentais at the Instituto de Química de São Carlos,
Universidade de São Paulo, Brazil.
George E. Totten, FASM, received his BS and MS degrees from the Fairleigh
Dickinson University in New Jersey and his PhD degree from the New York
University. Totten is a Fellow of ASM International and SAE. He has
co-authored approximately 20 books and has over 400 patent publications on
various aspects of heat treating, quenching, hydraulic lubrication, and
tribology. He is past president of the International Federation for Heat
Treatment and Surface Engineering (IFHTSE). Currently, Totten is a Research
Professor at the Portland State University in Portland, Oregon, USA.
Alessandro F. Farah is a metallurgist and received his MS from University of
São Paulo, Brazil. He completed his PhD, recently from the same university.
At present, he takes classes on heat treatment at the Universidade Paulista,
Ribeirão Preto, SP, Brazil.
Oxidation of vegetable oils and its impact on quenching performance 103
There have been many investigations on the use of both animal and vegetable oils and
fats as quench ants. One of the earliest studies was conducted by Tagaya and Tamura in
(1954). Although this study did correlate quench severity with fluid source and viscosity
in addition to oxidative stability for various naturally derived fluids, the data are in a
different form than commonly reported today. However, the data reported suggested that
although the Grossman quench severity factors were comparable for both castor oil
(H = 0.199) and soybean oil (H = 0.200), the cooling times from 700 to 300°C were
significantly faster for castor oil (1.8 seconds) than for soybean oil (1.42 seconds) using a
JIS K 2242 silver probe test.
To date, the most commonly cited vegetable oil basestocks used for quenchant
formulation are canola oil (Brennan and Faulkner, 1996), and soybean oil derivatives
(Honary, 1996). Recently, a crambe oil based fluid has been reported as a potential
quenchant (Lazerri et al., 1999). However, in none of these papers was any correlation of
quenching performance, including oxidative stability, with oil structure reported.
Totten et al. (1999) and Prabhu and Prasad (2003) studied the heat flux properties of
soybean oils, crude, partially hydrogenated soybean and coconut, sunflower, and
groundnut oils, respectively. Although quench severities comparable to a conventional,
non-accelerated mineral oil were obtained, correlations between vegetable oil structure
and oxidative stability with corresponding changes in quench severity as a result
of the change in molecular structure of the vegetable oil have not been reported in any
Vegetable oils are of interest as basestocks for quench oil formulations because
they are biodegradable and are derived from renewable sources. Typical fatty acid
ester components of different vegetable oils are shown in Table 1 (Brennan and
Faulkner, 1996; Honary, 1996; Totten et al., 1993, 1999; Moore, 2001; Asadauskas
et al., 1997; Asdauskas and Erhan, 1999; Honary, 2000; Hasson, 1994;
Farah et al., 2000).
Table 1 Vegetable oil structure and composition
Canola oil 15.7 – – – 4 2 62 22 10
Coconut oil 0.1 6 47 18 9 3 6 2 –
Corn oil 6.7 – – – 11 2 28 58 1
2.8 – – 1 22 3 19 54 1
Olive oil 4.6 – – – 13 3 71 10 1
Palm oil 1.0 – – 1 45 4 40 10 –
Peanut oil 4.0 – – – 11 2 48 32 –
Soybean oil 5.7 – – – 11 4 24 54 7
Percentages may not add up to 100% due to rounding and other constituents not listed.
104 L.C.F. Canale et al.
Vegetable oils are triglycerides of fatty acids and may be classified as:
Examples of saturated fatty acids include palmitic and stearic acids.
All carbon-hydrogen bonds are saturated, that is, they are methyl (CH
) or methylene
). Saturated hydrocarbons are the most stable biologically in that they do not become
rancid or oxidised. However, because they are saturated, the neighbouring fatty ester
chains form close-packed structures with relatively high melting points, often forming
waxes at room temperature. Generally, increasing the chain length (molecular weight) of
the saturated carboxylic acid increases the melting point as shown in Table 2:
Table 2 Effect of chain length and degree of unsaturation on the melting point of fatty acids
Fatty acid Symbol Empirical structure Melting point (°C)
Lauric acid 12:0 CH
Myristic acid 14:0 CH
Palmitic acid 16:0 CH
Stearic acid 18:0 CH
Oleic acid 18:1
Linoleic acid 18:2
Linolenic acid 18:3
An example of a monounsaturated fatty acid is oleic acid (a common ester component in
olive oil), which possesses one double bond (-CH=CH-). These double bonds are in the
cis configuration, which provides a bend in the molecule thus inhibiting solidification.
(A cis configuration means that the hydrogen atoms on the double bond are planar and on
the same side. Figure 1 illustrates the differences in cis and trans double bond
configurations for octadecenoic acid.) Monounsaturated fatty acids are liquids at room
temperature and are relatively resistant to becoming rancid upon storage. Although
slightly less oxidatively stable than saturated acids, they still find wide utility as
basestocks for lubricating oil formulation.
Figure 1 Illustration of difference between the cis configuration of cis-9-octadecenoic acid
(oleic acid) and the trans configuration of trans-9-octadecenoic acid
(Elaidic acid) – both isomers of octadecenoic acid. Naturally occurring vegetable oils
and fats are composed of fatty acid esters in the cis configuration
Source: Anon (2004)
Oxidation of vegetable oils and its impact on quenching performance 105
Polyunsaturated fatty acids possess two or more double bonds
(-CH=CH-) in conjugation with each other. An example of a fatty acid with two double
bonds is linoleic acid (an omega-6 fatty acid) and a fatty acid with three double bonds is
linolenic acid (an omega-3 fatty acid).
The ‘omega’ terminology refers to the position of the carbon containing the first
double bond. The double bonds in both fatty acids are typically conjugated and are in the
cis conformation. Increasing double bond content results in increased ‘kinks’ in the
molecular structure and the freezing point typically decreases with the number of double
bonds for a homologous series of fatty acids with the same number of total carbons in the
molecular chain. Therefore, polyunsaturated fatty acids are typically liquid even when
refrigerated. However, increasing double bond content results in progressively increasing
biological and oxidative instability. Linolenic acid is especially reactive relative to
linoleic acid. Figure 2 provides an illustration of the linear structures and the cis
configurations of the polyunsaturated C-18 carboxylic acids: linoleic and linolenic, and
illustrates the effect of increased double bond content on the resulting configurations of
the homologous series of fatty acids: stearic, oleic, linoleic and linolenic acid which helps
to explain the melting point data shown in Table 2.
Figure 2 Illustration of the configurations of the homologous series of fatty acids: stearic, oleic,
linoleic and linolenic acid
106 L.C.F. Canale et al.
All plant and animal derived fats and oils are triglycerides with a distribution of
saturated, monounsaturated, and polyunsaturated acids. A triglyceride is a reaction
product of three moles of a fatty acid and one mole of glycerine. Figure 3 illustrates one
triglyceride component of olive oil. Animal fats typically contain 40–60% of saturated
acids and therefore are solid at room temperature. The level and type of unsaturation of
vegetable oils, however, are highly dependent upon the climate in which they are grown.
Vegetable oils from tropical climates are typically highly saturated. Coconut oil, for
example, contains 92% saturated acids whereas vegetable oils grown in more northern
climates contain increasing amounts of polyunsaturated fatty acids.
Figure 3 Illustration of one triglyceride component of olive oil; a reaction product of two moles
of oleic acid and one mole of stearic acid. Naturally oils and fats are composed of a
distribution of various structural triglycerides. They are not one pure triglyceride
Table 3 provides a summary of typical physical properties exhibited by various
commonly encountered vegetable oils. Although vegetable oil basestocks are
biodegradable, they do suffer from a number of significant disadvantages relative to
petroleum oil such as oxidative instability and a relatively narrow useful viscosity range
(Brennan and Faulkner, 1996; Honary, 1996; Totten et al., 1999; Moore, 2001).
Of these, oxidative stability may be the greatest limitation to potential widespread use.
Figure 4 provides a schematic illustration of the general oxidation of a fatty acid.
Table 3 Typical vegetable oil properties
Canola oil 24 238 335–338 50 77
Coconut oil 25 175 288 37.8
Corn oil –5 230 340–343 54.4 28.7
Cottonseed oil 0 215 234 37.8
Olive oil –6 190 225 37.8
Palm oil 35 220 320 37.8
Peanut oil 3 225 340–343 37.8
Soybean oil –16 225 343–346 37.8 35.4
Oxidation of vegetable oils and its impact on quenching performance 107
Figure 4 Simplified scheme of the oxidation of a fatty acid
Source: Brucker Application Note and Barr et al. (n.d.)
The ‘smoke point’ and ‘flash point’ of a vegetable oil is one measure of its thermal
stability when heated in contact with air. This is a very important performance safety
criterion for non-aqueous, hydrocarbon-based quenchants. The ‘smoke point’ is the
temperature at which smoke is first detected in a laboratory apparatus which is protected
from draughts and equipped with illumination. The ‘flash point’ is the temperature at
which the volatile products are evolved at such a rate that they are capable of being
ignited but not capable of supporting combustion.
For typical vegetable oils with a free fatty acid content of about 0.05%, smoke and
flash points of about 420º and 620º, respectively, are typically obtained. The degree of
unsaturation exhibits little, or no, effect on the smoke point and flash point of a vegetable
oil. However, vegetable oils containing relatively low molecular weight carboxylic acids,
such as coconut oil, do exhibit lower smoke points and flash points. In addition, after
extended use, vegetable oils will typically exhibit increased free fatty acid content which
results in correspondingly lower smoke points and flash points. Some of the factors that
impact smoke point include:
• temperature to which the oil is heated
• number of times the oil is used
• length of time the oil is heated
• exposure to oxygen.
Therefore, although the smoke point and flash point are important performance
characteristics indicating the feasibility of continued use of an oil, they are not effective
parameters to monitor the oxidative stability of an oil, irrespective of whether it is a
vegetable oil or a mineral oil.
108 L.C.F. Canale et al.
Oxidation resistance and viscosity of mineral oils are usually determined by the
refinement process and final quenchant formulation. However, when exposed to high
temperatures, mineral oils also undergo oxidative degradation which leads to physical
and physical-chemical changes. These changes can lead to substantial loss of the heat
extraction characteristics of the oil (Totten et al., 2003; Bashford and Mills, 1984; Bodin
and Segerberg, 1993). Therefore, quench bath maintenance of both petroleum oils and
vegetable oils using viscosity tests, cooling curve analysis, water contamination
measurements, acid number, and flash point is necessary to ensure continued safety and
performance of the heat treatment process (Totten and Webster, 2003).
There are two vegetable oils of particular interest to Brazil: castor oil and soybean oil.
Castor oil is a naturally occurring product of which Brazil is the second leading producer
in the world (108,500 MT), along with India (810,240 MT) and China (2,396 MT).
Together, Brazil, India, and China produce 90% of the castor oil annually. Castor oil is
produced by crushing castor beans, which yield as high as 50–55% castor oil. Castor oil,
which is produced commercially, contains approximately 90% of a hydroxy, unsaturated
C18 fatty acid-ricinoleic acid (see Figure 5) – which is an 18-carbon acid having a double
bond in the 9–10 position and a hydroxyl group on the 12th carbon. A typical analysis of
castor oil shows: 87% ricinoleic acid, 7.4% oleic acid, 3.1% linoleic acid, and 2.4%
Figure 5 Molecular structure of ricinoleic acid
The combination of hydroxyl group and unsaturation occurs only in castor oil and is an
interesting structure because as the temperature increases one molecule of water is
evolved through a dehydration reaction. Although castor oil has excellent stability at
room temperature, the dehydration reaction shown in Figure 6 will occur at
approximately 75–120°C producing a 60:40 mixture of unconjugated: conjugated dienes
(Bertz et al., 1999). Typically, these diene-containing ester by-products do not
decompose further until approximately 343°C.
Figure 6 Dehydration reaction of ricinoleic acid to produce a diene (conjugated + unconjugated)
At the dehydration temperature of 75–120°C, the resulting diene that is formed
(see Figure 6) may then potentially dimerise (intra- or intermolecularly) by a Diels-Alder
(4 + 2) cycloaddition reaction to form a ‘dimer acid’ (if an intramolecular reaction
occurs) product of which four (4) possible isomers may be formed as shown in Figure 7
for linoleic acid (Berman and Loeb, 1975). This is the predominant mechanism of
molecular growth for dienoic and trienoic functional carboxylate acids or esters. Trans
Oxidation of vegetable oils and its impact on quenching performance 109
dienoic substrates are more reactive than cis dienoic substrates as shown in Table 4
(Berman and Loeb, 1975).
Table 4 Relative thermal polymerisation rates for dienois and trienoic methyl esters
(See Figure 7 for reaction products)
Methyl ester Conjugation Cis trans configuration Relative reaction rate
9,12-Linoleate No Cis, trans 0.74
9,12-Linoleate No Cis, cis 1.0
9,12-Linoleate No Trans, trans 1.2
10,12-Linoleate Yes Trans, trans 26.0
9,11 and 10,12-Linoleate Yes Cis, trans 5.8
9,12,15- Linolenate No Cis, cis, cis 2.4
Figure 7 Possible skeletal structures of linoleic acid Diels-Alder dimer products
Ricinoleic acid should be contrasted to oleic acid, which does not form a Diels-Alder
dimer product when heated. Instead it forms the dimer product shown in Figure 8 as the
principal component of the reaction mixture.
110 L.C.F. Canale et al.
Figure 8 Principle component of methyl oleate thermal polymerisation reaction
At temperatures of ~350°C, ricinoleic acid pyrolyses producing volatile by-products as
shown in Figure 9.
Figure 9 High-temperature pyrolysis of castor oil (rinoleic acid)
There are two technical grades of castor oil. One is a ‘pale pressed’ oil, also known as
‘AA standard’. It is an ‘unrefined’ grade that is obtained from the first pressing of the
castor bean. Pale pressed castor oil is lighter in colour and lower in acidity. The second
grade of castor oil that is commercially available is designated as ‘#1 Imported’ and is
also known as ‘industrial castor oil’. This grade is obtained from a mixture of the first
pressing and the second phase of production, solvent extraction. Industrial grade castor
oils are designated as: ‘1st quality’, ‘2nd quality’ and ‘3rd quality’ oils. The 1st quality
oil is processed without solvent extraction and the 2nd and 3rd quality grades are solvent
extracted after pressing.
The second vegetable oil that is of great interest to Brazil and that is studied here
report is soybean oil. Of the important edible vegetable oils, soybean oil is the world’s
largest – about 31% in 2001/2002 followed by palm oil and rapeseed oil. Brazil is the
world’s second-largest soybean producer. (The USA, the world’s largest soybean oil
producer, generally accounts for about a third of total production, producing nearly twice
that of Brazil). Brazil produces about 41.5 MMT (thousand metric tons)/year and crushes
approximately 24.4 MMT/year. Soybeans make up approximately 95% of all oilseed
production in Brazil.
The composition and selected properties of soybean oil are provided in Tables 1 and 3
Soybeans have many uses including as food, feed, and industrial products. Before use,
soybeans must be processed and more than 95% of the Brazilian production is processed
by solvent-extraction plants to extract the oil and meal. The soybean-crushing process
produces 18 pounds of soybean oil and 80 pounds of soybean meal from 100 pounds of
soybeans. After crushing, the soybean oil is processed by refining, bleaching,
deodorising, degumming, winterising, etc., as required.
As a result of the dearth of important and necessary molecular structure information
relating to the impact of molecular structure of vegetable oils during their use as
quenchants and the corresponding variation in cooling properties relative to the ‘new’,
as-received condition, a study was conducted to obtain these data on two vegetable oils of
Oxidation of vegetable oils and its impact on quenching performance 111
particular interest in Brazil: castor oil and soybean oil. The results of this study are
2 Experimental procedures
The vegetable oils used for this work included a pale-pressed castor oil (which was
obtained from Dissoltex Indústria Química Ltda) and a refined soybean oil provided by
Shell Brasil (which is not a commercial Shell product). A conventional mineral oil,
designated as MC1 from Castrol Brasil, was used as the mineral oil for comparison to
vegetable oil performance. Typical physical properties of these oils reported by their
manufacturers are summarised in Table 5.
Table 5 Summary of typical physical properties reported for castor oil, soybean oil and a
formulated mineral oil
Physical property Castor oil
MC1 petroleum oil
Viscosity (cSt at 40°C)
254,53 28.49 31.04
Viscosity (cSt at 100°C)
35,15 7.60 –
Melting Point (°C)
5.5 59-68 –
Boiling Point (°C)
245 153 –
Acid Value (mg KOH/g sample) 0.44. 0.04 0.05
Iodine Number (g I
80–91 128–134 –
Specific Gravity 25/25 0.958 0.9188 Please add data
Open Cup Flash Point (°C)
224 296 183
The castor oil data was supplied by Dissoltex Ind. Com. under the product name
The soybean oil data was supplied by Shell Brasil under the product name Soybeam Oil.
The petroleum oil quenchant data was supplied by Castrol Brasil under the product name
An accelerated laboratory ageing system was built according to the apparatus and
procedure used by Bashford and Mills in their earlier study of the effect of quench oil
oxidation on cooling curve performance. A schematic illustration is provided in Figure 10
(Bashford and Mills, 1984). The oxidation test was conducted as follows: into a 2300 mL
glass vessel (12.5 cm diameter, 19 cm in height) was added 2000 mL of the vegetable oil
which was heated to 150 ± 2°C for 12 minutes with agitation. During this time, the fluid
was aerated using a gas sparge tube at 4 L air/hour. (The agitation was provided by an
electrically driven propeller mixer with speed settings of 0–10 and a setting of 6 was
used.) After 12 minutes, the electrical resistance immersion heater, aeration system and
agitation was automatically shut off for 12 minutes during which time the fluid was
cooled to 125 ± 2°C in 3 minutes. The fluid was reheated again with agitation and air
sparging and these 12 minute cycles were repeated over the test duration of 48 hours.
The fresh and oxidised fluids were characterised by viscosity, iodine number,
112 L.C.F. Canale et al.
Fourier transform infrared analysis and
CMR spectral analysis were performed
(ASTM D 974–93, 1993; ASTM D 445–94, 1994).
Figure 10 Schematic process of accelerated ageing laboratorial system
Source: Farah (2002)
Fluid viscosity was determined at 40°C according to ASTM D 445 (Standard Test
Method for Kinematic Viscosity of Transparent and Opaque Liquids (the Calculation of
Dynamic Viscosity)) (ASTM D 445–94, 1994).
Iodine number is defined as a measure of the unsaturation of fats and oils and is
expressed in terms of the number of centigrams of iodine absorbed per gram of sample
(percentage of weight of iodine absorbed) and was determined by the method Tex-809-B
described by the Texas Department of Transportation (1993). In this test, two 250 mL
iodine flasks were used; one was labeled ‘Blank’ and the other was labeled ‘Sample’.
20 mL of chloroform was added into each flask using a pipette. Approximately 0.3 g of
sample to be analysed was weighed in a one mL micro-beaker to the nearest 0.1 mg and
then placed in the iodine flask labeled ‘Sample’ and dispersed into the chloroform
solvent. 25 mL of Wijs solution (iodine monochloride) was added into both iodine flasks
labelled ‘Blank’ and ‘Sample’ and dispersed. Both flasks were then placed in the dark for
one hour. After one hour, a 50 mL burette was filled with a previously standardised
solution of sodium thiosulphate solution. 20 mL of potassium iodide solution, which was
previously prepared by dissolving 150 g of potassium iodide into deionised water and
then diluted to one L in a volumetric flask, was added to the flask labelled ‘Blank’ after
one hour in the dark. The solution was then titrated to a straw-yellow colour with the
standardised sodium thiosulphate solution while shaking vigorously. A 2 mL aliquot of
starch indicator solution (10 g of soluble starch dissolved into deionised water and diluted
to 1 L in a volumetric flask) was added to the solution in the flask which turned dark
purple in colour. The titration was continued by adding standardised sodium thiosulphate
solution and shaking vigorously until a clear, colourless solution was obtained. Record
the total volume of the standardised sodium thiosulphate solution used in this titration
was recorded and this process was repeated for the iodine flask labelled ‘Sample’.
The iodine number was calculated from:
( ) 104.5
Oxidation of vegetable oils and its impact on quenching performance 113
where: M is the volume in mL of the sodium thiosulphate solution used to titrate the
‘Blank’, S is the volume of the sample used to titrate the ‘Sample’, N is the normality of
the sodium thiosulphate solution, SW is the sample weight, and NVM is the percentage
of total non-volatile matter in the sample used. The chemical reactions involved in this
R-CH=CH-R + ICl
→ R-CHI-CHCl-R + ICl
+ 2KI → KCl + KI + I
+ starch + 2Na
(blue) → 2NaI + starch + Na
Although iodine number is used throughout the vegetable oil industry as a measure of
total unsaturation, the method actually only provides a relative estimate based on a given
analysis procedure. In addition to variations in digestion times; side-reactions, including
the potential reaction of Wijs reagent with the allylic methylene moiety in the
hydrocarbon ester chain, can significantly affect the result. Therefore, the cited method
must be followed precisely in order to obtain reproducible results for comparison
(Berman and Loeb, 1975).
Fourier infrared (FT-IR) spectra were recorded using a Bomem Model MB-102
Fourier transform interferometer system (Quebec City, Canada) with 4 cm
The spectra were run as a capillary film on a silicon disk and a total of sixteen (16) scans
were aquired. A comparison of the FT-IR spectra obtained on the ‘new’ and oxidised
castor oil is shown in Figure 11 and the same comparison for the ‘new’ and oxidised
soybean oil is provided in Figure 12.
Figure 11 Comparison of the FT-IR spectra for: (a) ‘new’ and (b) oxidised castor oil (48 hours)
114 L.C.F. Canale et al.
Figure 12 Comparison of the FT-IR spectra for the ‘new’ and oxidised soybean oil
CMR spectra were acquired in a 5 mm diameter tube with CDCl
as the solvent and
tetramethylsilane as an internal standard using a 200 MHz Brucker AC-200 spectrometer
(Karlsruhe, Germany). Chemical shifts were reported in δ-units. Peak assignments
were made using DEPT 135 (Distortionless Enhancement by Polarization Transfer).
CMR spectra for ‘new’ and oxidised castor oil and soybean oil are shown in
Figures 13 and 14 respectively. Chemical shift assignments were made in accordance
with previously published reference data for these oils (Gunstone, 1993; Bergana and
Lee, 1996; Miyake et al., 1998; Adhvaryu et al., 2000).
CMR spectra of: (a) ‘new’, as-received castor oil and (b) oxidised (48 hours)
Oxidation of vegetable oils and its impact on quenching performance 115
CMR spectra of: (a) ‘new’, as-received castor oil and (b) oxidised (48 hours)
castor oil (continued)
CMR spectra of: (a) ‘new’, as-received castor oil and (b) oxidised (48 hours)
116 L.C.F. Canale et al.
Cooling curves were obtained in triplicate using Wolfson probe according to ASTM D
6200 (2001). The oil was unagitated and the cooling curve analysis was conducted at
60°C ± 1°C. From these curves the critical parameters listed here selected for
characterisation (Tensi and Steffen, 1985):
I A–B transition time(s) (t
II Temperature of A–B transition (ºC) (T
III A-B transition rate (ºC/s) (CR
IV Cooling rate at 700ºC (CR
V Maximum cooling rate (ºC/s) (CR
VI Temperature of the maximum cooling rate (ºC) (T
VII Cooling rate at 300ºC (CR
VIII Time for cooling at 300ºC (t
IX Cooling rate at 200ºC (CR
X Time for cooling at 200ºC (t
Parameters I–III are related to the full-film boiling (vapour blanket cooling) to nucleate
boiling transition time and temperature and the cooling rate at critical temperatures.
Cooling rate at 700ºC, parameter IV, is measured since it is usually desirable to
maximise this cooling rate to avoid the steel pearlite transformation region.
Parameters V and VI are the maximum rate of cooling and the temperature where this
occurs. Generally, it would be desirable to maximise CR
and minimise T
Rate of cooling at temperatures such as 200ºC and 300ºC, parameters VII and IX, are
also determined since they are related to the potential for steel cracking and distortion. To
minimise these problems, it is desirable to minimise cooling rates in this region.
Parameters VIII and X are related to region of the pearlite transformation. It is
generally desirable to minimise these parameters.
Figures 15 and 16 show the position of those parameters in the curves.
Figure 15 Cooling time-temperature curve with some critical parameters
Oxidation of vegetable oils and its impact on quenching performance 117
Figure 16 Cooling rate plot with some critical parameters
3 Results and discussion
Kinematic viscosity test results are shown in Table 6 for each of the oils in the ‘new’,
as-received condition and after ageing. The percentage viscosity increase of the castor oil
is considerably less than that obtained for the soybean oil after the 48 hours. oxidation
test. These results are not unexpected since the castor oil is predominantly a
monounsaturated oil which would be expected to be more stable than the soybean oil
which typically contains greater quantities of the less oxidatively stable polyunsaturated
Table 6 Kinematic viscosity results
Kinematic viscosity (cSt at 40°C)
New Oxidised oil (after 48 hour test)
Percent of change
Castor oil 254.53 364.8 +43.3
Soybean oil 35.15 91.65 +160.7
MC1 petroleum oil 31.04 31.68 +2.1
The expected thermal oxidative degradation of the ricinoleate ester moiety of castor oil
would be expected to occur by a two-step process: dehydration, followed by Diels-Alder
dimerisation by a 4 + 2 cycloaddition process as described earlier (see Figure 7).
The oxidation reaction temperature of 150°C is easily within the reported dehydration
and dimerisation temperature range of 75–120°C over 4 to 24 hours (Bertz et al., 1999).
(If catalysed with a Lewis acid, the dehydration reaction occurs at the lower temperature
The second step is a cross-linking reaction that occurs at higher temperatures by
either inter- or intra-molecular reaction of the double bond functionality of the fatty side
chains of the vegetable oil (Hill, 2000) leading to a fluid viscosity increase and if these
118 L.C.F. Canale et al.
reactions are uninhibited they will ultimately lead to sludge and varnish formation.
Schneider (2002) has reported that the relative rate of oxidation increases as the number
of double bonds in conjugation with each other increases as shown in Table 7.
Table 7 Relative reactivity of fatty alkyl groups
Fatty acid Number of double bonds Relative reactivity
Stearate 0 1
Oleate 1 100
Linoleic 2 1,200
Linolenic 3 2,500
Source: Schneider (2002)
Although these degradation pathways are reasonable, it should be noted that FT-IR
CMR spectral comparisons indicate that if these degradation reactions did
occur, they did not proceed to a sufficient extent to result in significantly different spectra
which could be easily observable visually. Furthermore, since these are not only
degradation reactions but also cross-linking mechanisms, significant extents of reaction
would be expected to result in precipitation and sludge formation. However, this was not
observed although the oxidised fluids were much darker in colour than the ‘new’ fluids
which is common and therefore not unexpected.
FT-IR analysis has been used to measure major structural constituents in various
vegetable oils, even when the compositional variation is not large and the differences
between spectra may not be immediately discernable. In such cases, it is possible to
obtain a quantifiable differentiation by superimposing spectra and comparing the areas
under critical regions of absorption frequencies (SensIR Technologies, 1999). This was
done here for the FT-IR spectra of the ‘new’, as received and used castor oil samples. The
ratio of the area under the hydroxyl (OH) peak (frequency ν = 3660–3126 cm
under the carbonyl (C=O) peak (frequency ν = 1822–1670 cm
) for the ‘new’ and used
fluids was determined by integration and compared. The results are shown in Table 8.
The results showed that the ratio of OH/C=O integral was 5.06% less for the used castor
oil relative to the ‘new’ fluid. This confirms that the dehydration of the castor oil over the
duration of the 48 hours oxidation test occurred only to a relatively limited extent.
Table 8 Comparison of the FT-IR OH/C=O integral for ‘new’ vs. used castor oil
Castor oil sample
O-H (ν = 3660–
C=O b (ν = 1822–
Ratio of OH/C=O
‘New’ (as-received) 151,742 45,999 3,299
Used (after 48 hour
148,601 47,451 3,132
Although evidence of substantial degradation of either castor oil or soybean oil was not
observed, the resulting viscosity increases of the oxidised fluids relative to the ‘new’,
as-received fluids was significantly greater than that exhibited by the MC1 petroleum oil
quenchant. This is not unexpected since the MC1 is a fully formulated quench oil that
contains antioxidants to inhibit thermal/oxidative degradation whereas both the castor oil
Oxidation of vegetable oils and its impact on quenching performance 119
and soybean oil do not contain any antioxidant additives. These data clearly indicate that
appropriate antioxidants, those that provide effective inhibition for steel quenching
without adverse impact on biodegradability and exotoxicity, must be identified for these
vegetable oils if their viscosity stability performance is to be competitive with fully
formulated petroleum oil formulations.
Another measure of the degree of oxidation of a vegetable oil is the loss of
double bonds of the fatty alkyl moiety of the triglyceride structure. Double bond content
can be estimated by determining the Iodine Number as described in the Experimental
Procedures Section. The iodine number of ‘new’ and oxidised vegetable oils obtained
after the 48 hour. Bashford and Mills oxidation test is summarised in Table 9. These
results show that castor oil resulted in a slightly greater total loss of double bond content
than did the soybean oil. However, as reported above, substantial differences in the FT-IR
CMR spectra were not observed, hence these results are not readily
explainable with respect to detailed molecular structure from the experimental work
completed to date.
Table 9 Total Iodine value analysis of fresh and used vegetable oils
Vegetable oil designation Iodine value (cg I
Total % loss of double bond content
relative to ‘new’ (as-received) fluid
Castor oil (fresh) 78.14 –
Castor oil (after oxidation) 35.77 54.2
Soybean oil (fresh) 114.69 –
Soybean oil (after oxidation) 70.26 38.7
Although neither direct FT-IR nor
CMR spectral comparison of the oxidised
reaction mixtures provided a molecular structural explanation for the observed
instability of either castor oil or soybean oil with respect to viscosity or unsaturation
changes, this is not unusual. It is well known that fluid oxidation, particularly at the C=C
unsaturation, is very complex and difficult to unambiguously structurally elucidate.
Adhvaryu et al. (2000) addressed this problem more generally using a pressurised
differential scanning calorimetric analysis and subsequent quantitative
analysis of the relative change in C=C functionality of numerous ‘new’ and oxidised
vegetable oils including soybean oil. The results of these analyses showed:
• The complexity of vegetable oil oxidation is due to the involvement of different
structural parameters in the fatty acid chain. It is, therefore, not recommended that
the degree of oxidation be modelled by a single structural parameter.
• The activation energy of the oxidation process decreases with increasing
polyunsaturation and increases with the degree of oleic acid (monounsaturation).
• The presence of allylic, bis-allylic, α-CH
groups attached to a carbonyl (C=O) and
chain length of saturated methylene (CH
) groups exhibit a significant effect on the
oxidation process at different stages.
• Kinetic parameters such as activation energy (E
) and pseudo-first-order rates of
reaction (k) can be used as a quick predictive tool to model vegetable oil oxidation.
(The thermal-kinetic behaviour of castor oil and soybean oil oxidation is not
determined in this work).
120 L.C.F. Canale et al.
Adhvaryu et al. (2000) reported the following
CMR chemical shifts for
quantitatively characterising vegetable oils (including soybean oil) for keys structural
parameters: carbonyl carbons are observed at 173.5–172.0 ppm, and olefin carbons at
130.3–127.5 ppm. The olefin carbons are due to monounsaturation and polyunsaturation
in the fatty acid chain, the CH
group that is in the α-position relative to the ester C=O
occurs at 34.3–33.6 ppm, ω-3 carbons of the saturated and unsaturated fatty acids occur
at 32.0–31.2 ppm; saturated CH
groups in the fatty acid chain are observed at 30.0–28.5
ppm, cis-allylic carbons at 27.5–26.5 ppm, bis-allylic carbons at 26.0–25.0 ppm, three (3)
carbons are observed at 25.0–24.0 ppm, mono- and n-6 polyunsaturated
acids at 23.0–22.0; and the three terminal methyl (CH
) groups are observed at
For castor oil, the analysis is somewhat different owing to the different chemical
structure. In this case, the key differences in structural composition are:
The chemical shifts for these structural elements by carbon number are: C-9
at 123.99 ppm, C-10 at 131.82 ppm, C-11 at 36.95 ppm, C-12 at 68.53 ppm and C-13
at 40.17 ppm (see Figure 13).
By comparing these chemical shifts for ‘new’ and oxidised vegetable oils, a
quantitative determination of percentage of change in the presence of these groups
due to the oxidation process can be obtained. Tables 10 and 11 show the results of the
CMR analytical comparisons for the 48 hour oxidation of the castor
oil and the soybean oil respectively. (A similar quantitative comparison can be performed
Table 10 Adhvaryu quantitive
CMR comparison of ‘new’ and oxidised castor oil
Percent of change
relative to ‘new’
Percent of olefin carbons 8.95 9.30 +3.91
Percent of saturated CH
carbons 28.34 27.82 –1.83
Percent of bis-allylic CH
carbons 4.93 4.87 –1.22
Percent of ω-2 carbons of saturated,
mono, and n-6 polyunsaturated acids
4.99 5.38 +7.82
Percent of the three (3) terminal
5.17 4.60 –11.03
Percent of CH (OH) carbons 4.58 4.31 –5.9
The results of these analyses show that even subtle differences in changes in the
molecular structure of a vegetable oil can be detected by quantitative
spectroscopic analyses of key molecular structural elements of the fatty acid side chain of
the triglycerides in the vegetable oil without chromatographic separation using the
Adhvaryu methodology (2000). The differences observed here are essentially
undetectable without such a comparative analysis and are due to various complex
oxidation reactions. These analyses provide a quantitative ‘picture’ of the castor oils and
soybean oils used for subsequent quenchant cooling curve characterisation work.
Oxidation of vegetable oils and its impact on quenching performance 121
Table 11 Adhvaryu quantitive
CMR comparison of ‘new’ and oxidised soybean oil
relative to ‘new’
Percent of olefin carbons 11.43 10.63 –7.00
Percent of saturated CH
carbons 38.24 37.67 –1.49
Percent of bis-allylic CH
carbons 2.65 3.63 +36.98
Percent of ω-2 carbons of saturated,
mono, and n-6 polyunsaturated acids
4.93 5.67 15.01
Percent of the three terminal methyl
5.41 5.47 1.11
Cooling curve analyses were performed on the ‘new’, as-received and used vegetable oils
in the unagitated condition at 60
C and the quantitative cooling curve parameters are
summarised in Table 12. Figures 17–19 provide a comparison of the cooling curves
Figure 17 Castor oil cooling time (a) and rate (b) curves
Figure 18 Soybean oil cooling time (a) and rate (b) curves
122 L.C.F. Canale et al.
Figure 19 Cooling time (a) and rate (b) curves for MC1, a conventional mineral oil quenchant
Table 12 Comparative cooling curve parameters to sample oils used*
Oil Castor Soybean MC1
Parameters New 48 hour New 48 hour New 48 hour
(s) 4.7 2.7 1.9 1.9 11.0 10.5
(ºC) 736 790 804 804 617.0 621.9
28.2 26.1 34.1 28.1 14.2 16.0
84.26 82.3 96.3 88.3 56.2 60.2
660.8 704.6 652.9 710.6 561.1 562.9
76.2 82.2 92.3 86.2 20.0 22.1
8.1 6.1 12.0 6.0 6.2 8.0
2.0 4.1 2.1 4.0 4.1 4.0
21.0 30.17 15.6 27.0 33.7 30.0
(s) 43.0 55.0 38.7 51.8 56.8 53.3
*The cooling curve data shown were obtain by ASTM D 6200 at 60°C. These values are
the result of single curve determinations.
The cooling parameter data provided in Table 12 shows:
• The time for film-boiling to nucleate-boiling transition (t
) varied considerably
between the ‘new’ and used samples of castor oil and soybean oil. The castor oil
transition time decreased by 42.6% whereas there was no change in the transition
time for the soybean oil. For comparison, the transition time for the mineral oil
decreased by 4.5%. The temperature where this transition occurs did not change
significantly for any of the three basestocks. This phenomenon is related to surface
wetting and/or the presence of significant levels of volatile products. The magnitude
of the difference in fresh vs. used fluid data obtained here suggests a significant
molecular structure change for the castor oil upon oxidation in this test. Oil oxidation
would increase the polar nature of the oil which would be expected to facilitate
interfacial surface wetting which would facilitate faster film-boiling to
nucleate-boiling transition times. Since the maximum cooling rate of the castor oil
actually decreased slightly (see below), it would not seem that the decrease in
transition time is due to increased volatile by-product formation in the used
castor oil fluid.
Oxidation of vegetable oils and its impact on quenching performance 123
• The maximum cooling rate (C
) decreased slightly for the used fluid relative to
the ‘new’ fluid for both the castor and soybean oils. Conversely, the maximum
cooling rate increased slightly for the mineral oil. The decrease in maximum cooling
rate for both vegetable oils is consistent with a decrease in the average volatility and
increase in viscosity of the used relative to ‘new’ fluids. The slight increase in
maximum cooling rate for the used fluid relative to ‘new’ mineral oil fluid was
consistent with expectation (slight increase in the amount of volatile by-products due
to fluid oxidation.)
• The temperature where the maximum cooling rate occurs (TC
) increased slightly
for both the castor and soybean oil but essentially no change was observed for the
mineral oil after oxidation.
• The cooling rate at 300°C (C
) decreased by 24.6% for the used vs. ‘new’ castor
oil fluids and by 50% for the used vs. ‘new’ soybean oil fluids but increased by 32%
for the mineral oil. The data for the mineral oil is expected and is due to increased oil
degradation and a small viscosity decrease of the used fluid relative to new oil.
However, the data for the castor oil and soybean oil are not as obvious. The cooling
rate at this temperature is due to both viscosity and boiling point change of the fluids.
The increase in viscosity of the used vs. ‘new’ condition for the castor oil and
soybean oil would be expected to yield a decrease in cooling rate but if the nucleate
boiling to convective cooling is not yet complete for one of the fluids relative to the
other, then a comparison is more difficult since the cooling mechanism has not
• The cooling rate at 200°C (C
) actually exhibits the opposite trend relative to what
is expected since the viscosity of both the used castor oil (+105%) and soybean oil
(+90.5%) fluids increased quite significantly. Typically, higher viscosities yield
correspondingly lower cooling rates in this region. The 2.4% decrease observed for
the mineral oil is probably not significant and is typical of what would be expected
relative to the changes observed for the other parameters. However, the data for the
vegetable oils are not explainable at this
point of time.
• The time to cool to 300°C (t
) and 200°C (t
) increased for both the castor oil
and the soybean oil. This is consistent with our expectation as are the relatively small
decreases observed for the mineral oil.
Castor oil and soybean oil are among the vegetable oils produced in maximum volume in
the world and Brazil is among the top producers of these oils. Therefore, potential new
uses in applications such as quenchants are of fundamental and great interest.
An overview of vegetable oil chemistry, including soybean and castor oil, has
been provided here to aid in the understanding of the quenching performance of castor
oil, and soybean oil in their ‘new’, as-received condition and after oxidation. Although
molecular structural characterisation in the oxidised state is complex and difficult,
it was shown that comparative variations in key structural fragments in the fatty ester
side-chain using the
CMR spectroscopic methodology previously reported by
124 L.C.F. Canale et al.
Adhvaryu et al. (2000) do provide a useful basis for molecular characterisation of used
Cooling curve analyses of ‘new’ and oxidised vegetable oils show
significantly greater variation in cooling performance relative to conventional petroleum
oil quenchants. These results show that further work will be necessary to identify
effective antioxidants that do not compromise the favourable toxicological and
biodegradation properties of these oils if they are to gain substantial use in the heat
Adhvaryu, A., Erhan, S.Z., Liu, Z.S. and Perez, J.M. (2000) ‘Oxidation kinetic studies of oils
derived from unmodified and genetically modified vegetables using pressurized differential
scanning calorimetry and nuclear magnetic resonance spectroscopy’, Thermochimica Acta,
Vol. 364, pp.87–97.
American Society for Testing and Materials: ASTM D 445–94 (1994) Standard Test Method for
Kinematic Viscosity of Transparent and Opaque Liquids (the Calculation of Dynamic
Viscosity), ASTM International, West Conshohocken, PA.
American Society for Testing and Materials: ASTM D 6200 (2001) Standard Test Method for
Determination of Cooling Characteristics of Quench Oils by Cooling Curve Analysis, ASTM
International, West Conshohocken, PA.
American Society for Testing and Materials: ASTM D 974–93 (1993) Standard Test Method for
Acid and Base Number by Colour-Indicator Titration, ASTM International, West
Anon (2004) Fats, Oils, Fatty Acids, Triglycerides: Chemical Structure, from: http://
www.scientificpsychic.com/fitness/fattyacids.html, February 25.
Asadauskas, S., Perez, J.M. and Duda, J.L. (1997) ‘Lubrication properties of castor oil: potential
basestock for biodegradable lubricants’, Journal of the Society of Tribologists and Lubrication
Engineers, December, pp.35–41.
Asdauskas, S. and Erhan, S.Z. (1999) ‘Depression of pour points of vegetables oils by blending
with diluents used for biodegradable lubricants’, Journal of the American Oil Chemists
Society, March, Vol. 76, No. 3, pp.313–316.
Barr, D., Reynhout, G. and Guzinski, J. (n.d.) EPR Spin Trapping as a QC Measurement for
Vegetable Oil Oxidation, Brucker Application Note, Brucker Instruments Inc, Billerica, MA,
Bashford, A. and Mills, A.J. (1984) ‘The development of improved additives for quenching oils
using Laboratory simulations’, Heat Treatment of Metals, No. 1, pp.9–14.
Bergana, M.M. and Lee, T.W. (1996) ‘Structure determination of long-chain polyunsaturated
trigylcerols by high-resolution
C magnetic resonance’, J. Am. Oil Chem. Soc., Vol. 73,
No. 5, pp.551–556.
Berman, L.U. and Loeb, M.L. (1975) ‘Structure and properties of dimer acids’, in Leonard, E.C.
(Ed.): The Dimer Acids, Humko Sheffield Chemical Company, Memphis, TN, Chapter 2,
Bertz, S.H., Miksza, F.M. and Zucker, E. (1999) High Purity adduct of Castor Oil and Maleic
Anhydride, US Patent, 6,225,485, June 29.
Bodin, J. and Segerberg, S. (1993) ‘Measurement and evaluation of the power of quenching media
for hardening’, Heat Treatment of Metals, January, Vol. 20, No. 1, pp.15–23.
Brennan R.J., and Faulkner, C.H. (1996) ‘A new quenching alternative’, in Totten, G.E., Funatani,
K., Howes, M.A.H. and Sjostrom, S. (Eds.): Conf. Proceed. 2nd International Conference on
Quenching and Control of Distortion, ASM International, Materials OH, pp.423–428.
Oxidation of vegetable oils and its impact on quenching performance 125
Farah, A.F. (2002) Characterisation of Vegetable Oils as an alternative Quenchants to Heat
Treatment, Thesis (Doctor): Área de Interunidades de Ciência e Engenharia de Materiais dos
Institutos de Física e Química de São Carlos – Universidade de São Paulo, São Carlos,
São Paulo, Brasil, p.140.
Farah, A.F., Crnkovic, O.R., Canale, L.C.F. and Totten, G.E. (2000) ‘Castor oil: studies for heat
treatment applications’, Proceedings of the 12th International Federation of Heat Treatment
and Surface Engineering Congress, Melbourne, Australia, October, Vol. 1, pp.83–88.
Guner, F.S. (1997) ‘Castor oil dehydration kinetics’, J. Am. Oil Chem. Soc., Vol. 74, No. 4,
Gunstone, F.D. (1993) ‘Information on the composition of fats from their high-resolution
nuclear magnetic resonance spectra’, J. Am. Oil Chem. Soc., Vol. 70, pp.361–366.
Hasson, J. (1994) ‘Modern quenching oils: lower costs and environmental impact’, Proceedings of
International Heat Treating Conference: Equipment and Processes, April, pp.317–322.
Hill, K. (2000) ‘Fats and oils as oleochemical raw materials’, Pure Appl. Chem., Vol. 72, No. 7,
Honary, A.T.L. (2000) ‘Vegetable-base hydraulic oils’, in Totten, G.E. (Ed.): Handbook of
Hydraulic Fluid Technology, Marcel Dekker, Inc., USA, Chapter 21, pp.1095–1151.
Honary, L.A.T. (1996) ‘Performance of vegetable oils as a heat treat quenchant’, in Totten, G.E.,
Funatani, K., Howes, M.A.H. and Sjostrom, S. (Eds.): Conf. Proceed. 2nd International
Conference on Quenching and Control of Distortion, ASM International, Materials OH,
Lazerri, L., de Mattei, F., Bucelli, F. and Palmieri, S. (1999) ‘Crambe oil: a potentially new
hydraulic oil and quenchant’, Ind. Lubr. Tribol., Vol. 49, No. 2, pp.71–77.
Miyake, Y., Yokomizo, K. and Matsuzaki, N. (1998) ‘Determination of unsaturated fatty acid
composition by high-resolution nuclear magnetic resonance spectroscopy’, J. Am. Oil Chem.
Soc., Vol. 75, No. 9, pp.1091–1094.
Moore, D. (2001) ‘Safer liquid quenchants’, Heat Treatment Progress, June–July, pp.29–33.
Prabhu, K.N. and Prasad, A. (2003) ‘Metal/quenchant interfacial heat flux transients during
quenching in conventional quench media and vegetable oils’, J. Mat. Eng. and Perf., Vol. 12,
No. 1, pp.48–55.
Schneider, M. (2002) Government Industry Forum on Non-food Uses of Crops (GIFNFG 7/7) –
Case Study Plant Oil Based Lubricants in Total Loss and Potential Loss Applications,
Final Report prepared for P. Smith Network Manager, University of York, England.
SensIR Technologies (1999) ‘The analysis of natural oils by FT-IR with a Micro-ATR’,
Applications Brief AB-026, SensIR Technologies, 15 Great Pasture Road, Danbury,
CT 06810-8153, USA.
Tagaya, M. and Tamura, I. (1954) ‘Studies on the quenching media, 3rd report: the cooling ability
of oils’, Technol. Rep., Osaka Univ., Vol. 4, pp.305–319.
Tensi, H.M. and Steffen, E. (1985) ‘Measuring of the quenching effect of liquid hardening agents
on the basis of synthetics’, Steel Research, Vol. 56, pp.489–496.
Texas Department of Transportation (1993) ‘TxDOT Manual System, Coatings and Traffic
Manuals, Tex-809-B’, Iodine Number of Oils and Fatty Acids, Chapter 8, Section 10.
Totten, G.E., Webster, G.M. and Wachter, D.A. (2003) ‘Quenchant fundamentals: condition
monitoring of quench oils’, Practicing Oil Analysis, January–February, pp.26–31.
Totten, G.E., Bates, O.E. and Clinton, N.A. (1993) ‘Chapter 4: quenching oils’, Handbook of
Quenchants and Quenching Technology, ASM International, Materials Park, OH, pp.129–159.
Totten, G.E., Tensi, H.M. and Lanier, K. (1999) ‘Performance of vegetable oils as a cooling
medium in comparison to a standard mineral oil’, J. Mat. Eng. and Perf., Vol. 8, No. 4,