[Short title + Author Name - P&H title] xxx (xxxx) xxx

H O S T E D BY Contents lists available at ScienceDirect

Journal of King Saud University - Engineering Sciences

journal homepage: www.sciencedirect.com

Original article

Compression resistance and color evaluation in artisanal red bricks:

Experimental study and numerical simulation
Sergio Alonso Romero a, Luis Ángel Ortiz Lango a, José Eduardo Frías Chimal a,
Roberto Zitzumbo-Guzmán a, Anayansi Estrada-Monje a, *, Leonel Hernández Mena b,
Jorge Del Real-Olvera b
a
Department of Research and Technological Services, Centro de Innovación en Tecnologías Competitivas (CIATEC A.C.), Omega 201, Col. Industrial Delta, León,
Guanajuato CP 37545, Mexico
b
Unidad de Tecnología Ambiental, Centro de Investigación y Asistencia en Tecnología y Diseño del Estado de Jalisco A.C. (CIATEJ), Av. Normalistas 800, Col. Colinas de
la Normal, Guadalajara, Jalisco C.P. 44270, Mexico

A R T I C L E I N F O A B S T R A C T

Keywords: An experimental study is presented, simulating the firing process of artisanal red brick in a laboratory muffle.
Numerical simulation Under controlled heating conditions, the influence of both temperature and time on both the compressive
Red brick strength and surface color of model bricks are evaluated. A production run and numerical simulations of labo­
Compressive resistance
ratory brick heating are used to define the experimental conditions in the laboratory. A statistical model repli­
cates the behavior of the compressive resistance at different temperatures, both in the muffle furnace and in
production. Brick color has a more representative tone of a well-burned brick when created at the burning
temperatures suggested in the literature.

1. Introduction the color of the brick. The former depends on the trust of the client,
while the latter (the color) is determined by the minerals contained in
Artisan bricks are essential for the construction industry in Latin the clay and the temperature (Karaman et al., 2006a). Other authors
America; in fact, it is one of the materials mainly used in the construc­ identified that bricks containing a large amount of iron have a reddish
tion of houses, housing complexes, and other buildings. Bricks are pro­ color (Sandford and Liljegren, 1964), hematite being responsible for this
duced by the molding, drying and firing of mixtures of clay, organic (Cultrone et al., 2020), while those with a high lime content but lacking
matter, and/or similar materials (ONNCCE, 2012). Given the variety of in iron tend to have a yellowish or creamy color. The resulting redness
possible clays and the different firing processes available, the resulting comes from chemical reactions activated by temperature, the amount of
properties of a brick can be highly variable (Arquiñigo, 2011; INECC, iron present and the oxygen available during the process (Karaman
2016). The quality of the final product is mainly affected by the tech­ et al., 2006a).
nique used in the manufacture of the brick (Zea Osorio, 2005) and the For a brick to be considered to be of good quality, it must reach at
firing process (Karaman et al., 2006a, 2006b; Lara-Mireles et al., 2020). least 800 ◦ C (Cultrone et al., 2020; Luján and Guzmán, 2015; Márquez,
As brick quality is multifactorial, a good product is defined as such when 2002). The heating process is carried out in an artisan kiln, where the
it meets the Mexican NMX-C-404-ONNCCE-2012 standard (ONNCCE, heat is usually generated from the bottom up, in the combustion
2012). chamber, and transferred to the bricks above. As the combustion and
Assessing the quality of a brick production ‘lot’ involves sampling heating are usually intermittent and non-uniform, there are differences
and sending selected specimens to the laboratory. In practice, this pro­ in temperature inside the kiln (Guerrero Gómez et al., 2018). Very high
cedure does not occur due to the economic cost related to the charac­ temperatures may yield vitrified bricks (Karaman et al., 2006a);
terization. During the purchase-sale activity, the evaluation of quality is whereas, low temperatures may produce unburned ones. Temperatures
undertaken subjectively, based on the experience of the producer and of around 600 ◦ C in the upper beds of a Marquez kiln may indicate that a

* Corresponding author.
E-mail addresses: [email protected] (S. Alonso Romero), [email protected] (L.Á. Ortiz Lango), [email protected] (J.E. Frías Chimal), [email protected]
(R. Zitzumbo-Guzmán), [email protected] (A. Estrada-Monje), [email protected] (L. Hernández Mena), [email protected] (J. Del Real-Olvera).

https://doi.org/10.1016/j.jksues.2023.10.002
Received 1 November 2022; Accepted 11 October 2023
Available online 16 October 2023
1018-3639/© 2023 The Authors. Production and hosting by Elsevier B.V. on behalf of King Saud University. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: Sergio Alonso Romero et al., [Short title + Author Name - P&H title], https://doi.org/10.1016/j.jksues.2023.10.002
S. Alonso Romero et al. Journal of King Saud University - Engineering Sciences xxx (xxxx) xxx

Fig. 1. Location and identification of thermocouples installed in a traditional field kiln (North face).

Fig. 2. Geometric pattern of brick and square mesh.

Table 1
Experimental design in the muffle.
Run Block Temperature [◦ C] Time [h]

1 1 0 0
2 1 0 0
3 1 0 0
4 1 0 0
5 1 1 1
6 1 1 − 1
7 1 − 1 1
8 1 − 1 − 1

sufficient temperature will be achieved for correctly firing the entire

load (Marquez, 2002). Karaman et al. (2006a) reported that a sudden
increase in compressive strength is obtained when 1000 ◦ C is reached,
the result being linked to a decrease in porosity and an increase in
product density due to a vitrified product (Karaman et al., 2006a). In
practice, it is well known that the lack of porosity is a problem, since the
absorption properties of a brick decreases when mortar is used. Fig. 3. Specimens for experimentation in the laboratory.
The brick burning process is further investigated to better under­
stand brick quality, with regard to the production process and labora­ 2. Materials and methods
tory variables. For laboratory specimens and production-made bricks,
compressive strength measurements have been determined and linked 2.1. Stage one: Brick burning process in traditional kiln
to both burning temperature and time. External color assessment has
been realized, in order to identify the CIELAB coordinates at which a Internal temperatures were recorded during a brick burning pro­
brick is considered fully burned. The only reference for the work being duction process in an artisanal kiln, by means of dataloggers specially
reported is that of Karaman et al. (2006b), whose results are com­ designed and built at CIATEC. K-type thermocouples were placed with
plemented by those obtained by this study. their tips flush to a brick, each one being evaluated later, with regard to
compression tests. 48 thermocouples were located on the four walls of a

2
S. Alonso Romero et al. Journal of King Saud University - Engineering Sciences xxx (xxxx) xxx

Fig. 4. Laboratory experimental setup.

Fig. 5. North wall temperature profiles of a production kiln. Fig. 7. Thermal profile applied to the burning of samples in a labora­
tory muffle.

Fig. 6. Thermocouple time/temperature plots for the North wall. Fig. 8. Behavior of brick temperature simulation.

3
S. Alonso Romero et al. Journal of King Saud University - Engineering Sciences xxx (xxxx) xxx

Table 2 simulate it numerically. This defines the laboratory test conditions at

Results of compressive strength tests under the NMX-C-404-ONNCE-2012 which a modeled brick reaches a homogeneous temperature, thus
standard. establishing the time necessary for said brick to be considered fully fired,
Run Block Temperature Time Compressive strength allowing the laboratory burning time to be set.
[◦ C] [h] [Kg/cm2] The following information was used to model and simulate the
1 1 800 2 68.97 process:
2 1 800 2 68.54
3 1 800 2 89.51 • Material properties (brick): density = 1700 kg/m3, isotropic thermal
4 1 900 3 99.53
conductivity = 0.658 W/m ◦ K, specific heat 2297 J/Kg ◦ K. Thermal
5 1 900 1 97.91
6 1 700 3 56.17 and mechanical properties were obtained from MatWeb (2020) and
7 1 800 2 74.05 were held constant.
8 1 700 1 49.44 • Uniform initial specimen temperature, established at 22 ◦ C.
9 2 800 2 63.16 • Transitory heat transfer regime which allowed monitoring of the
10 2 800 2 59.03
11 2 800 2 65.54
temperature inside the brick by conduction, taking a horizontal
12 2 900 3 80.25 plane in the center of the specimen as a reference.
13 2 900 1 91.15 • A step of 15,600 s, divided into 53 sub-steps, was used.
14 2 700 3 61.82
15 2 800 2 61.51
A three-dimensional geometric model of a brick, fitting the muffle,
16 2 700 1 43.07
17 3 800 2 69.4 was developed (135 × 85 × 65 mm), having been generated using the
18 3 800 2 66.17 ANSYS SpaceClaim 3D modeling software (Fig. 2, left). After a transient
19 3 800 2 82.32 mesh sensitivity analysis, a size of 5 mm was selected, obtaining a very
20 3 900 3 90.43 regular mesh, as shown in Fig. 2 (right).
21 3 900 1 90.94
24 3 700 3 43.65
23 3 800 2 65.67 2.3. Stage three: Experimental design
24 3 700 1 45.32
The prediction of compressive strength, as a function of temperature
and burning time, was obtained from an experimental design; after the
efficient process research strategy had been applied to a variety of
Table 3
themes (El Hajam et al., 2020, Khan et al., 2023). An array of 24 ex­
ANOVA for the compressive strength tests.
periments with four center points (as in Rahman et al., 2021) with two
Variable Sum of squares DF Mean square F p-value replicates (Prasath et al., 2022) was selected to represent the burning
A: Temperature 5239.210 1 5239.2100 108.71 0.0000 process in the laboratory (Table 1). The center points were added to the
B: Time 13.380 1 16.3800 0.34 0.5671 22 design to identify the possible non-linearity of the process, whereas
AB 94.080 1 94.0800 1.95 0.1794 the replicates contributed to better assessing the errors and the curva­
Block 395.971 2 197.9850 4.11 0.0339
Total error 867.513 18 48.1952
ture. The results obtained in the production kiln were used to define an
Total (corr.) 6613.160 23 adequate experimental temperature range.

kiln and identified according to the cardinal orientation of each wall, a

letter for the row and a number for the level (Fig. 1).
Table 4
R-squared parameters for compressive strength.

2.2. Stage two: Numerical simulation of model brick burning R-square 86.88%
R-square (adj. d.f.) 83.23%
Standard error 6.942
Measuring the experimental internal temperature of a brick as it is Mean absolute error 4.424
burned is difficult while it maintains its physical integrity. An available Durbin-Watson statistic 1.90403 (P = 0.2677)
alternative is to model the muffle brick burning process and then

Fig. 9. Graph of the main effects on compressive strength.

4
S. Alonso Romero et al. Journal of King Saud University - Engineering Sciences xxx (xxxx) xxx

Fig. 10. Estimated response surface.

Fig. 11. Contours of the estimated response surface.

2.4. Stage four: Model brick preparation and experimental laboratory set- device.
up
2.5. Stage five: Statistical model development and specimen
The laboratory specimens were prepared with the same brick characterization
formulation as that used in the traditional kiln brick burning process of
stage one. The molding process (Fig. 3) was carried out using a special The experimental design was analyzed in Statgraphics software,
wooden frame and the specimens were left to dry at room temperature complying with normality and independence of the data. ANOVA tables
and under shade for at least three days. were obtained, in order to identify the significant variables affecting the
The experimental set-up is shown in Fig. 4. Temperature monitoring response variable and r2. A response surface plot and a statistical model
inside the muffle was carried out by two thermocouples, one of them in were also obtained, the latter being validated by the experimental lab­
contact with the brick surface (as in the kiln production process) and the oratory results and those obtained from production bricks.
other one focused on the internal ambient temperature of the muffle The evaluation of compressive strength was carried out by following
(factory installed). the Mexican standard NMX-C-404-ONNCCE. This established a mini­
Pre-operative experiments in the laboratory allowed us to stan­ mum value of 70 Kg/cm2 for the brick to be considered of good quality, i.
dardize the firing temperature curve in such a way that all samples e., suitable as a structural element in load-bearing walls.
received the same thermal history, although the standard curve actually The surface color for each experimental laboratory specimen was
changes due to the set point temperature to be reached at each run. characterized by CIELAB coordinates, by means of a DATACOLOR
The specimens were always placed in the same position within the model 45G CT (plane geometry colorimeter spectrophotometer, 10◦
muffle and the measurement thermocouples were placed in their angle, D65 luminance, 45/0 observer, and specularity included). The
respective locations (Fig. 4). Next, the muffle was turned on and pro­ results are reported in L, a, and b coordinates.
grammed from the ambient to the test temperature. Once the set point
was reached, the muffle was left for 30 min, to thermally stabilize. Thus,
the burning time took place and then the system cooled down to room
temperature. During the execution of the tests, temperatures were
manually recorded every five minutes with a Fluke model 54IIB thermal

5
S. Alonso Romero et al. Journal of King Saud University - Engineering Sciences xxx (xxxx) xxx

Table 5 Table 6
Estimated results for the compressive strength of modeled bricks. Estimated values of compressive strength with the statistical model obtained for
Run Temperature Time Laboratory Regression model Absolute error
production bricks.
[◦ C] [h] [kgf/cm2] [kgf/cm2] [%] Specimen Temperature Laboratory Regression model Absolute error
[◦ C] [kgf/cm2] [kgf/cm2] [%]
1 800 2 68.9 75.5 6.6
2 800 2 68.5 75.5 7.0 1 787 73.3 65.9 11.2
3 800 2 89.5 75.5 14.0 2 860 74.0 83.2 11.1
4 800 2 74.0 75.5 1.5 3 795 56.7 67.8 16.4
5 900 1 97.9 98.0 0.1 4 743 67.9 55.5 22.3
6 900 3 99.5 94.7 4.8 5 788 64.1 66.1 3.0
7 700 3 56.1 58.5 2.4 6 741 60.7 55.0 10.4
8 700 1 49.4 50.6 1.2 7 842 62.9 78.9 20.3
9 800 2 63.1 65.6 2.5 8 821 58.1 74.0 21.5
10 800 2 59.0 65.6 6.6 9 797 64.5 68.3 5.6
11 800 2 65.5 65.6 0.1 10 668 83.5 37.7 121.5
12 800 2 61.5 65.6 4.1 11 789 90.6 66.4 36.4
13 900 1 91.1 88.2 2.9 12 847 99.0 80.1 23.6
14 900 3 80.2 84.9 4.7 13 832 96.5 76.6 26.0
15 700 3 61.8 48.7 13.1 14 842 100.0 78.9 26.7
16 700 1 43.0 40.8 2.2 15 767 70.6 61.2 15.4
17 800 2 69.4 69.2 0.2 16 650 71.7 33.4 114.7
18 800 2 66.1 69.2 3.1 17 749 65.1 56.9 14.4
19 800 2 82.3 69.2 13.1 18 716 48.1 49.1 2.0
20 800 2 65.6 69.2 3.6 19 780 67.4 64.2 5.0
21 900 1 90.9 91.7 0.8 20 823 47.4 74.4 36.3
22 900 3 90.4 88.5 1.9 21 808 62.7 70.9 11.6
23 700 3 43.6 52.3 8.7 22 766 61.9 60.9 1.6
24 700 1 45.3 44.3 1.0 23 788 68.0 66.1 2.9
average 4.33 24 798 68.2 68.5 0.4
25 819 75.0 73.5 2.0
26 101 78.6 − 96.6 181.4
3. Results and discussion 27 849 85.0 80.6 5.5
28 756 139.7 58.6 138.4
29 712 50.0 48.1 4.0
3.1. Brick burning in a kiln
30 716 84.5 49.1 72.1
31 891 95.9 90.5 6.0
During the brick burning process carried out in a traditional pro­ 32 824 64.2 74.7 14.1
duction kiln, the temperature profiles show an increasing trend from the 33 791 43.8 66.8 34.4
beginning, up to a maximum corresponding to the end of the fuel 34 786 100.4 65.7 52.8
35 787 70.5 65.9 7.0
feeding. Afterwards, the temperature decreases (the cooling process). 36 786 41.9 65.7 36.2
Fig. 5 shows the average temperatures of four thermocouples located at 37 823 75.0 74.4 0.8
each level of the North face, where the thermal rise in the lower part of 38 714 38.0 48.6 21.8
the furnace is oscillatory and associated with intermittent and manual 39 744 65.0 55.7 16.7
40 748 66.7 56.7 17.6
fuel feeding (Level A). The thermal rise in the upper levels (B, C and D)
41 686 36.7 42.0 12.6
increased more slowly and homogeneously (Munguía-Guillen et al., 42 741 57.2 55.0 4.0
2012). When the fuel feed ceases, the temperature drops in the lower 43 817 92.3 73.0 26.4
part of the kiln (Level A) first, when the natural convection process 44 838 46.0 78.0 41.0
transfers heat from the hotter lower bricks to the colder upper ones, 45 819 78.2 73.5 6.4
46 855 87.3 82.0 6.5
which causes the temperature to continue rising at the top. For this 47 878 51.5 87.5 41.1
reason, the maximum temperatures in levels C and D are delayed, even 48 855 68.8 82.0 16.1
when the fuel supply has finished hours before. Several brick burning Average 27.6
processes were carried out in the same kiln because the thermal
behavior was qualitatively reproducible. The other kiln walls (South,
mechanical properties of a brick for firing times longer than 2 h (Kara­
East, and West) followed a very similar trend.
man et al., 2006a; Marquez, 2002), the experimental design in the lab­
The maximum temperatures at the bottom of the kiln are close to
oratory considered only two hours as the maximum firing time in the
1000 ◦ C, while the maximum temperatures are just above 700 ◦ C in the
muffle. It should be noted, however, that the temperatures reported here
upper zones (Fig. 5). At approximately 1000 ◦ C the crystalline structure
and in the literature are not the brick’s internal ones, but those measured
changes, to produce vitreous, higher resistance and lower porosity
on the brick surfaces, or those of the surrounding environment (Kara­
bricks (Karaman et al., 2006a, 2006b). It was not considered convenient
man et al., 2006a). For this reason, a numerical simulation tool was
to carry out the laboratory experiments at such temperatures. The
used, which validated the time it took for a brick to have a homogeneous
maximum average temperatures of the thermocouples on all of the walls
internal temperature.
ranged between 700 ◦ C and 900 ◦ C, which is why the said interval was
chosen for carrying out the experimental design in the muffle.
An alternative approach to Fig. 5 is its redefinition as horizontal
3.2. Numerical simulation of brick heating
isotherms, which are data that allow analyses of the time that any
thermocouple remains above a specific temperature. Fig. 6 (with the
For the numerical solution of the brick heating inside the muffle, a
same information as Fig. 5) shows an alternative point of view. As an
transient thermal analysis was chosen (Ngom et al., 2021). It is essential
example, the NA2 thermocouple remained above 700 ◦ C for 9.33 h,
to know the temperature in the center of the brick as a function of time,
above 800 ◦ C for 7.4 h, and above 900 ◦ C for 4.23 h, considering the
to define its complete burning. The simulation mathematically repre­
temperature range to be within the optimum (Márquez, 2002; Cultrone
sented the tests carried out in the laboratory under the same experi­
et al., 2020). Given that the literature does not show differences in the
mental conditions (Ortíz-Lango, 2020).

6
S. Alonso Romero et al. Journal of King Saud University - Engineering Sciences xxx (xxxx) xxx

Fig. 12. Laboratory specimen color variation with burning temperature.

The results of the transient thermal numerical analysis are presented

in Fig. 8. The continuous curve is the simulated experimental average
temperature on the surface of the specimen, the dotted lower curve is the
one at the center of the specimen, and the dashed upper curve is the
ambient temperature inside the muffle. The temperature in the center of
the specimen reaches 800 ◦ C 2.6 h after the start of the test, between 41
and 52 min after the average surface reaches the temperature set point.
This means that two hours is not convenient (Karaman et al., 2006a) but
longer times are required to achieve a fully burned brick. Conversely,
short times should be selected in order to produce unburned bricks, so
that the experimental model is reliable. Thus, according to the numer­
ical simulation, the maximum time expected for a fully burned muffle
brick is 3 h; this was the maximum time selected in the experimental
laboratory design. It must be remembered that numerical simulation is
the mathematical representation of a physical phenomenon, which im­
plies assumptions considered in the model that can oversimplify the real
physical circumstances.

3.3. Experimental design for compressive strength tests

The experimental design for burning bricks in the laboratory was

Fig. 13. Change of CIELAB coordinates for laboratory brick samples processed
defined based on the temperatures provided in the field, from pre­
at different temperatures.
liminary tests of brick burning in the muffle, and on the times calculated
by numerical simulations. The levels of the experimental variables were
The brick temperature boundary conditions were defined according
700 ◦ C, 800 ◦ C, and 900 ◦ C and 1 h, 2 h, and 3 h, respectively. The 22
to preliminary laboratory experimental results. Fig. 7 shows a repre­
design (Table 2) comprised two variables with two levels and a center
sentative temperature brick muffle profile. The muffle is capable of
point to identify curvature, if any. The analysis of the results was carried
reaching a setpoint of 800 ◦ C and keeping it relatively constant in about
out using Statgraphics software, at a confidence level of 95%. Once the
1.25 h. That is the moment when it is considered that the entire surface
experimental time–temperature treatments were carried out in the lab­
of the brick is at the setpoint; with this information the numerical
oratory muffle, the samples were evaluated under the Mexican reference
calculation of internal heating of the brick was carried out. The external
standard NMX-C-404-ONNCCE-2012. Compressive strength tests were
temperature curve regression equation was used as a dynamic boundary
performed by a certified laboratory and the results are shown in Table 2.
condition for the simulation of the temperature profile on the brick
As a reference, unburned specimens were also sent to the laboratory,
surface.
with an average compressive strength of 30.25 kg/cm2. This value puts

Fig. 14. Bricks tested by the compressive strength test.

7
S. Alonso Romero et al. Journal of King Saud University - Engineering Sciences xxx (xxxx) xxx

the increment of the mechanical resistance of the bricks (once burned) temperature increases, the color changes towards the characteristic
into context: from 42.3% to 329.0%, depending on the experimental vivid red of a brick considered to be well-fired. Fig. 12 shows specimens
conditions. subjected to different temperatures for 3 h and burned in the same
According to ANOVA, for the compressive strength (Table 3), the muffle as where the experimental design was run; samples that do not
temperature is the only significant factor, i.e., the firing process must be reach 800 ◦ C should not be fully burned nor have a good red color and,
mostly based on the temperature and not on the exposure time of the therefore, should not present a good quality.
brick at that temperature. These results are in accordance with the The colors of the samples in Fig. 12 were analyzed as CIELAB co­
previous literature (Karaman et al., 2006a). On the other hand, the ordinates. All three coordinates increase with temperature, noting a
temperature–time interaction is not representative and so an interaction color plateau at around 600 ◦ C (Fig. 13). A thermogravimetric analysis
of both factors, to reach the maximum values of compressive strength, of the clay revealed a thermal change at 470.3 ◦ C, a change in some
can be ruled out (Fig. 9). Although not shown, the results passed the components of the clay that burns at that temperature and can give rise
normality test with the expected normal distribution carried out in to a ‘peak’ in color change, which suggests that 500 ◦ C may be the point
Minitab 19. where the trend breaks. The authors found no research that interprets
Another parameter that supports the reliability of the method is the what occurs, since the only research which was found only carried out
correlation coefficient r2. For this experiment, its value was 86.88% analyses from 800 ◦ C (Cultrone et al., 2020; Karaman et al., 2006a,
(Table 4), which indicates that the factors included in the experimental 2006b).
model explained 86.88% of the variability of the results. Similarly, the Brick is considered to be of good quality if its external color is the one
value of adjusted r2 corresponded to 83.23%, which is high enough to be expected (Fig. 12). However, a brick may have a good external color but
considered as a correct analysis (Montgomery and Runger, 1996). not be completely burned, which is only detectable through destructive
testing. Fig. 14 shows specimens from production brick firing having an
3.4. Response surface for compressive strength adequate exterior color, but the dark tones inside clearly indicate
incomplete firing. Thus, a good external color is not fully indicative of a
Following the response surface methodology, we sought to optimize well-burned brick; systematic research is still needed to correlate the
the compressive strength, with respect to temperature and burning time external and internal color of a brick with compressive tests.
during the firing process (Fig. 10).
A flat surface without a curvature effect is observed, which implies 4. Conclusions
that the treatment that maximizes the compressive strength is 900 ◦ C
and 1 h of burning time, taken pertinently to increase productivity. Results from the first stage of testing established the test temperature
Fig. 11 shows the time–temperature curves from the statistical ranges for carrying out the controlled laboratory experiments. 700 ◦ C
model, showing the standard of 70 Kg/cm2 being accomplished at was selected as the low level and 900 ◦ C as the high level, for the results
805 ◦ C after one hour of burning; this agrees with what was reported by from the maximum temperature intervals reached in production brick
Martínez González et al. (2002) and Marquez (2002). burning.
The statistical model that represents the compressive strength trend, From stage two, the time necessary for the center of the brick to
with respect to temperature and exposure time, is a regression equation reach the test temperature (homogeneous temperature) was determined
that adjusts the experimental results according to the regression coef­ by numerical simulation. It was possible to replicate the laboratory tests
ficient shown above. This equation is: in a virtual environment, to evaluate the thermal behavior of the brick as
a function of time. It was identified that, although the brick burning test
CompressionResistance = − 144.149 + 0.26495T + 23.5683t − 0.028Tt (1) in the laboratory was carried out between the range 1–3 h, it took
approximately 1 h for the center of the brick to reach the surface tem­
where: T = Temperature [◦ C], t = Exposure time [h]. perature and, therefore, it can reach the minimum resistance values set
Table 5 shows the values generated by the model and the values by the reference standard.
determined experimentally. The comparison between the results ob­ Furthermore, the experimental design turned into a statistical model
tained by the statistical model and the experimental results shows 95.7% that estimated the compressive strength as a function of time and tem­
reliability, and this means that the model is valid for predicting the perature. The model predicted the laboratory results, with an average
compressive strength, with respect to temperature and exposure time, error of 4.3%, and production results, with an error of 27.1%. The
with a 4.33% error. determining factor of the compressive strength is the temperature
Similarly, the predicted estimates of compressive strength from the reached: at least 70 kg/cm2 should be expected at 805 ◦ C for 1 h (the
statistical model, compared to those obtained in the production kiln, are Mexican reference standard).
shown in Table 6. The statistical laboratory burning model can be used The color of the brick (expressed in CIELAB values) is influenced by
to estimate the compressive strength of production bricks, within a temperature, which is well known in practice and can be shown scien­
27.6% error. It should be noted that the brick producer is not quanti­ tifically. Temperature changes the brick from a dark gray (typical of the
tatively controlling the temperature, but it is done manually and from raw clay color) to a bright red, which is representative of a well-fired
experience. Furthermore, the brick formulation changes lot by lot; the product. However, production bricks have shown that the exterior
clay used in production bricks is not usually meshed, so that a quantity color is not enough to guarantee the quality of a brick, since it may not
of small stones may have been presented, which may affect the resis­ have a red uniform color throughout its volume; therefore, further
tance results. research is needed.

3.5. Effect of temperature on brick color Declaration of competing interest

Color evaluation is relevant since, in commercial practice, it is the The authors declare that they have no known competing financial
basic selling point with respect to quality. As shown before, for the upper interests or personal relationships that could have appeared to influence
levels of a production kiln the maximum temperatures may not be near the work reported in this paper.
to the expected maximum of 800 ◦ C. This can easily be seen with an
external color assessment (Fig. 12). Color variation means that the brick Acknowledgments
presents physical–chemical changes (Karaman et al., 2006a, 2006b;
Sandford and Liljegren, 1964; Zea Osorio, 2005); as the maximum This work was financially supported by the National Council of

8
S. Alonso Romero et al. Journal of King Saud University - Engineering Sciences xxx (xxxx) xxx

Science and Technology (CONAHCYT Mexico) through the CONAHCYT- behavior. Revista Mexicana de Ingeniería Química 20 (1), 289–304. https://doi.org/
10.24275/rmiq/Fen1749.
SENER-Energy Sustainability Sector Fund, project 234633 and the
Luján, M., Guzmán, D., 2015. Diseño, Construcción y Evaluación de un Horno (MK3)
master scholarship granted to Luis Angel Ortiz Lango. para la Cocción de Ladrillos Artesanales. Acta Nova 7 (2).
Márquez, R.O., 2002. Appropriate Chemistry for The Economically Limited People of The
References Earth. New Mexico State University. PhD Thesis.
Martínez González, G.M., Jiménez Islas, H., Flores Pimentel, J.C., 2002. Estudio
numérico de la operación de un horno cilíndrico para la cocción de ladrillo rojo.
Arquiñigo, T.W.N., 2011. Propuesta para mejorar la calidad estructural de los ladrillos Información Tecnológica 13 (1), 3–12.
artesanales de arcilla cocida en Huanuco. Pontificia Universidad Católica del Perú. MatWeb, 2020. http://www.matweb.com/index.aspx.
M.S. Thesis. Montgomery, D.C., Runger, G.C., 1996. Probabilidad y estadística aplicadas a la
Cultrone, G., Aurrekoetxea, I., Casado, C., Arizzi, A., 2020. Sawdust recycling in the ingeniería, second ed. Limusa Wiley, Mexico.
production of lightweight bricks: how the amount of additive and the firing Munguía-Guillen, J.L., Cárdenas-González, B., Piña-Victoria, J.C., Aréchiga-Viramontes,
temperature influence the physical properties of the bricks. Constr. Build. Mater. J.U., 2012. Análisis energético en hornos ladrilleros MK2. XXXIII Encuentro Nacional
235, 117436 https://doi.org/10.1016/j.conbuildmat.2019.117436. y II Congreso Internacional AMIDIQ, 3799–3804. Los Cabos, Mexico.
El Hajam, M., Noureddine, Kandri I., Plavan, G.I., Harrath, A.H., Mansour, L., Ngom, M., Thiam, A., Balhamri, A., Sambou, V., Raffak, T., Refaey, H.A., 2021. Transient
Boufahja, F., Zerouale, A., 2020. Pb2+ ions adsorption onto raw and chemically study during clay bricks cooking in the traditional kiln; CFD numerical study. Case
activated Dibetou sawdust: application of experimental designs. J. King Saud Univ. Stud. Therm. Eng. 28 https://doi.org/10.1016/j.csite.2021.101672.
Sci. 32, 2176–2189. https://doi.org/10.1016/j.jksus.2020.02.027. ONNCCE, 2012. Norma Mexicana NMX-C-404-ONNCCE-2012. In Piezas para uso
Guerrero Gómez, G., Espinel Blanco, E., Escobar Mora, N., 2018. Curva de cocción de la estructural.
arcilla en la ladrillera el recreo. Revista colombiana de tecnologías de avanzada Ortíz-Lango, L.A., 2020. Predicción de la resistencia a la compresión del ladrillo rojo
(RCTA) 31, 35–41. https://doi.org/10.24054/16927257.v31.n31.2018.2762. artesanal en la zona ladrillera “El Refugio”, a través de la metodología de superficie
INECC, 2016. Análisis de Mercado del Sector de la Construcción y Proyecto Piloto a Nivel de respuesta. MSc Thesis. CIATEC.
Región, Basado en un Portafolio de Políticas Públicas, con el Objetivo de Reducir los Prasath, B.B., Zahir, M., Elsawah, A.M., Raza, M., Lecong, C., Chutian, S., Poon, K., 2022.
Contaminantes Climáticos de Vida Corta (CCVC) de Ladrilleras Artesanales en Statistical approaches in modeling of the interaction between bacteria and diatom
México. under a dual-species co-cultivation system. J. King Saud Univ. Sci. 34 https://doi.
Karaman, S., Ersahin, S., Gunal, H., 2006a. Firing temperature and firing time influence org/10.1016/j.jksus.2021.101743.
on mechanical and physical properties of clay bricks. J. Sci. Ind. Res. 65, 153–159. Rahman, N., Nasir, M., Alothman, A.A., Al-Enizi, A.M., Ubaidullah, M., Shaikh, S.F.,
Karaman, S., Gunal, H., Ersahin, S., 2006b. Assessment of clay bricks compressive 2021. Synthesis of 2-mercaptopropionic acid/hydrous zirconium oxide composite
strength using quantitative values of colour components. Constr. Build. Mater. 20, and its application for removal of Pb(II) from water samples: central composite
348–354. https://doi.org/10.1016/j.conbuildmat.2004.11.003. design for optimization. J. King Saud Univ. Sci. 33 https://doi.org/10.1016/j.
Khan, Z., Shafique, M., Jabeen, N., Naz, S.A., Yasmeen, K., Ejaz, U., Sohail, M., 2023. jksus.2020.101280.
Protease from Bacillus subtilis ZMS-2: evaluation of production dynamics through Sandford, F., Liljegren, B., 1964. La formación de color en ladrillos rojos y amarillos.
Response Surface Methodology and application in leather tannery. J. King Saud Mater. Constr. 14 (113), 40–48. https://doi.org/10.3989/mc.1964.v14.i113.1792.
Univ. Sci. 35 https://doi.org/10.1016/j.jksus.2023.102643. Zea Osorio, N.L., 2005. Caracterización de las arcillas para la fabricación de ladrillos
Lara-Mireles, J., Alonso-Romero, S., Fías-Chimal, E., Zitzumbo-Guzmán, R., del Real Artesanales. Universidad de San Carlos de Guatemala. L’esprit Ingénieux 9 (1),
Olvera, J., Hernández-Mena, L., 2020. Analysis of the MK brick kiln thermal 83–92.