Professional Documents
Culture Documents
STSM title: Experimental and numerical (CFD)
assessment of heat stress in dairy barns
STSM applicant Thomas Bartzanas
STSM Host Prof. Thomas Amon
Final Report
Contents
1. Extensive summary ............................................................................................................ 3
2. Step beyond the state of the art ....................................................................................... 4
3. Thermal Stress Index ......................................................................................................... 7
4. Numerical approach (CFD model) ................................................................................... 10
4.1. Governing equations ............................................................................................... 11
4.2. Near wall functions .................................................................................................. 13
4.3. Ventilation rate ........................................................................................................ 14
4.4. Turbulence modelling .............................................................................................. 15
4.5. Modelling the animals ............................................................................................. 17
4.6. CFD modeling of indoor air properties .................................................................... 19
5. Managing Heat Stress ...................................................................................................... 25
5.1. General managing options ...................................................................................... 25
5.2. Influence of barn structural specification ............................................................... 26
6. Future work ..................................................................................................................... 30
7. Literature cited ................................................................................................................ 32
8. Acknowledgements ......................................................................................................... 39
9. Appendix .......................................................................................................................... 40
2
1. Extensive summary
Climate conditions within a livestock building affect animal health and welfare rate.
Heat stress in dairy cattle is one of the leading causes of decreased production and
fertility in cattle during summer months. Cows require additional energy to dissipate
the heat and regulate body temperature.
Heat stress has several serious and economically deleterious effects. Heat stress is
driven by four main climate variables: air temperature, relative humidity, solar
radiation and air velocity. From the above‐mentioned four variables, temperature and
humidity are more readily measured and thus more easily accessible that the other
two. For this reason a common method of measuring heat stress has involved the
development of a temperature‐humidity (THI) index.
There are many ways to decrease the apparent effects of heat on dairy cattle. One
of the most important points to consider is increasing air flow exchanges, a strategy
which theoretically lead to a lower temperature and humidity, thus to a reduced THI.
The majority of large dairy cattle livestock buildings using natural ventilation systems
for air renewal. In that buildings the assessment of THI was realized by unique values
of air temperature and air humidity usually measured in the centre of the building.
However no climatological variable is homogeneous in large naturally ventilated
buildings (NVB) but instead they are numerous exchanges mainly governed by internal
airflow, building structural specifications and cows activity. In NVB, the internal air
flow patterns are primarily influenced by the prevailing flow conditions in the
atmospheric boundary layer, as well as by local side specific factors like structural and
management related conditions of the barn. Numerical techniques, like
Computational Fluid Dynamic (CFD) can render efficiently and accurately the
quantification of the variables compound the microclimate inside livestock buildings.
Actual weather conditions and structural specifications could be simulated and
changed in the CFD model while maintaining stable and intentical boundary
conditions.
3
Aim of the present study is to assess, more efficient and accurate, the heat stress
of dairy cows in large naturally ventilated dairy cattle buildings using computational
fluid dynamic tools. The present report summarises the work carried out from Thomas
Bartzanas during his STSM in ATB‐Potsdam in the framework of DAIRECARE COST
Action, under the supervision of Prof. Thomas Amon and analyse the main issues and
factors involved in the estimation of thermal heat stress based on CFD models.
2. Step beyond the state of the art
Heat stress can be simply defined as a condition that occurs when an animal cannot
dissipate an adequate quantity of heat, whether it is produced or absorbed by the
body, to maintain body thermal balance. This may prompt physiological and
behavioural responses, leading to physiological disorders that negatively affect the
productive and reproductive performance of farm animals (Nardone et al., 2006;
2010). Environmental stress has a severe effect on the productivity of animals and, in
particular, on that of dairy cattle. The prospect of environmental stress is aggravated
by present global warming accompanied by periods of extreme weather. The latter
may have increasingly severe effects on welfare and productivity of cattle, as it is one
of the major factors that can negatively affect milk production, reproduction, and the
health of dairy cows (Jordan, 2003; Bernabucci et al., 2010).
To study heat stress in livestock, the temperature ‐ humidity index (THI) is the most
commonly used bioclimatic index (Hahn et al., 2003), because data of the rest
environmental variables (wind speed, solar radiation) are not measured in a daily
routine basis on livestock farms. Moreover, air temperature and humidity data can be
usually obtained from nearby meteorological stations and these data are usually
provided free of charge.
The effects of the ambient environment on cow performance have been measured
by establishing critical ambient temperatures for the cow (Johnson, 1987), an
equivalent temperature index incorporating temperature, humidity, and air velocity
(Baeta et al., 1987), and temperature humidity index (THI), which incorporates the
4
combined effects of temperature and relative humidity (NOAA, 1976). In classical
work, Johnson et al. (1963) reported that milk yield exhibited significant declines when
maximum THI reached 77. Later research determined that the critical values for
minimum, mean and maximum THI were 64, 72, and 76, respectively (Igono et al.,
1992). It is traditionally believed that milk synthesis begins to decrease when the THI
reaches 72 (Armstrong, 1994). More recent data from the University of Arizona
indicates that high‐yielding dairy cows reduce their milk yield at a THI of approximately
68 (Zimbelman et al., 2009). In general temperature–humidity index values of 70 or
less are considered comfortable, 75–78 stressful, and thermal values greater than 78
cause extreme distress with lactating cows being unable to maintain ther‐
moregulatory mechanisms or normal body temperature
Environmental contributors to heat stress include air temperature (T), relative
humidity (RH), solar radiation, air speed, and their interactions. Lactating dairy cows
prefer ambient temperatures of between 5 and 25 °C, the ‘thermoneutral’ zone
(Roenfeldt, 1998). At ambient temperatures above 26 °C, the cow reaches a point
where she can no longer cool herself adequately and enters heat stress. Body
temperature is usually maintained by in thermoregulatory systems within 1 °C of
normal under ambient conditions that do not impose severe heat stress (Bligh, 1973).
Many indices combining different environmental factors to measure the level of heat
stress have been proposed. However, their use is limited by poor availability of data.
The THI is expressed as a single value representing the combined effects of air
temperature and humidity, which is commonly used to evaluate the degree of heat
stress in dairy cattle (Armstrong, 1994). Bohmanova et al. (2007) compared 7 THI
formulas and concluded that humidity was the limiting heat stress factor in humid
climates, whereas the dry bulb temperature was the limiting factor for heat tolerance
in dry climates.
As it has been indicated, data of air temperature and humidity are usually used in
order to estimate THI and accordingly to assess potential heat stress. The climate data
are used either from nearby meteorological station or in an ideal situation from
measurements inside the livestock building. Even in this case a measurement device
is used usually located in the middle of the building, and the values are considered
5
representative from the whole building volume. However this is far from the reality,
especially on large naturally ventilated cattle buildings, which climate distribution is
characterised by strong heterogeneity.
CFD models can render efficiently and accurately the quantification of the variables
compound the microclimate inside livestock buildings. Additionally, parametric
investigation can be conducted studying the impact of the external wind speed and
direction on the distributions of air flow, temperature and ammonia in specific points
inside the building (Norton et al., 2007).
The ubiquitous nature of fluids and their influence on system performance has
caused a widespread take‐up of CFD by many other disciplines. As a developing
science, CFD has received extensive attention throughout the international
community since the advent of the digital computer. As a result, CFD has become an
integral part of the engineering design and analysis environment of many companies
because of its ability to predict the performance of new designs or processes prior to
manufacturing or implementation (Schaldach et al., 2000). Typical applications of the
6
CFD in the agri‐food industry includes food storage, drying and sterilization (Le Page
et al. 2009), specific agricultural applications such as the estimation of losses of
pesticide drift from field crop sprayer (Nuyttens et al. 2009), environmental control of
agriculture buildings (Bartzanas et al. 2004; Bartzanas et al. 2007; Bartzanas et al.
Norton et al. 2010a,b,c, Bartzanas et al. 2013a, Berg et al. 2013a,b,c), post‐harvest
applications (Bartzanas et al. 2010a, 2013b).
The final target of this on‐going research collaboration which has initiative under
the STSM of DairyCare Cost Action, is the evaluation of THI of a dairy cattle barn will
be estimated for a series of experimental data and also using a CFD model. Two
different CFD codes will be used: a) Open Source (ATB) and Fluent – Ansys (CERTH).
The outcomes of the two codes will be compared against experimental data as a
benchmark study. In a second step the CFD models will be used to study the influence
of different design and structural changes to the building envelope (height, length,
ventilation openings) and the potential impact to the THI. In this was design
optimisation strategies can be proposed for alleviating the problem of high THI.
3. Thermal Stress Index
Extreme weather conditions during the summer months can severely affect the
welfare and productivity of many livestock species. Compared to other animals cattle
cannot dissipate their heat load very effectively. Cattle do not sweat effectively and
rely on respiration to cool themselves. A compounding factor on top of climatic
conditions is the fermentation process within the rumen generates additional heat
that cattle need to dissipate. Since cattle do not dissipate heat effectively they
accumulate a heat load during the day and dissipate heat at night when it is cooler.
During extreme weather conditions with insufficient environmental cooling at night
cattle will accumulate heat that they cannot disperse. Estimates of the degree of heat
stress affecting animal regulation of body temperature were made by developing
mathematical formulae that combine one or more meteorological variables.
7
Many indices combining different environmental factors to measure the level of
heat stress have been proposed. However, as it has already been indicated that most
of the studies dealing with the assessment of heat stress in livestock have used indices
mainly based to air temperature and relative humidity. A variety of indices were used
to estimate the degree of heat stress affecting cattle and other animals. The most
common of these, the temperature‐humidity index (THI), uses the dry and wet air
temperatures Tdb and wet bulb temperature Twb to estimate the magnitude of heat
stress (Thom, 1959). Conceptually, it is difficult to ascertain whether THI is the most
appropriate measurement of heat stress in dairy cattle. In the related literature there
was no given explanation for the basis of the relative weighting of Tdb and Twb in the
THI. Other THI were formulated empirically and often without reference to body
temperatures of cattle. Nonetheless, the original THI and several variations of it have
been used extensively to estimate the degree of heat stress in dairy and beef cattle.
It is not obvious that THI are better predictors of body temperature in heat‐stressed
cows than other measurements of environmental conditions. In addition, it is possible
that regression analysis using actual data on environmental conditions and cow body
temperature can result in development of heat stress indices that are better
predictors of heat stress than THI.
Following is a list of temperature humidity indices:
Index Reference
Temperature‐Humidity Index (THI)
0.8 x T + (RH/100) x (T‐14.4) + 46.4 Thom, 1959
(0.15 x Tdb + 0.85 x Twb) x 1.8+32 Bianca,1962
Td = 0.55 x (1 ‐ RH) x (Td ‐ 14.4) Kibler, 1964
(0.55 × Tdb + 0.2 × Tdp) × 1.8 + 32 + 17.5 National Research Council, 1971
Tdb + 0.36 × Tdp + 41.2 Yousef, 1985
Temperature‐Humidity Velocity Index
(0.85xTdb + 0.15 x Twb) x V ‐0,058 Tao and Xin, 2003
Environmental Stress Index
8
0.63 T ‐ 0.03 RH (%) + 0.002 SR + 0.0054 (T x RH Moran et al. 2001
(%))‐0.073 (0.1+ SR)‐1
[4.51+THI‐(1.992 x V) + (0.0068 x SR)] Mader et al., 2006
Black globe Temperature Humidity Index
0.8xTbg + [(RH(%)/100) x (Tbg ‐ 14.3)] + 46.4 Buffington et al. 1981
Tbg + 0.36 x Tdp + 41.5 Buffington et al. 1981
Where:
T, is the ambient air temperature (°C)
RH, is the relative humidity (decimal form)
Tdb, is dry bulb temperature (°C)
Twb, is wet bulb temperature (°C)
Tdp, is the dew‐point temperature (°C)
V, is the wind speed (m/s)
SR, is the solar radiation (Wm‐1)
Tbg, is black globe temperature
Tret, Rectal temperature
Tre0, Rectal temperature (initial)
9
4. Numerical approach (CFD model)
Computational fluid dynamics is a sophisticated design and analysis tool that uses
computers to simulate fluid flow, heat and mass transfer, phase change, chemical
reaction, mechanical movement, and solid and fluid interaction. The technique
enables a computational model of a physical system to be studied under many
different design constraints. The quality of a CFD study is a function of not only the
physics available in the software to model the system, but also the understanding that
the CFD modeller has of both the numerics and physics contained in the software
package. If used correctly CFD can provide an understanding of the physics of a flow
system in detail, and does so through non‐intrusive flow, thermal and concentration
field predictions.
The majority of the computational fluid dynamic solvers uses a finite volume
numerical scheme to solve the equations of conservation for the different transported
quantities of flow (mass, momentum, energy, water vapour concentration). The CFD
solver does the flow calculations and produces the results by solving the discretised
form of governing equation. This stage needs clear understanding of flow physics
involved in the problem like phenomena related with heat transfer, mass transfer,
multispecies flow, multiphase flow, reacting flow, turbulence, radiation, etc. In most
of the commercial CFD packages there are different options regarding the numerical
methods that can be used to solve the governing equations. As a solution of all the
governing equations, the flow parameters, like velocity, pressure, density,
temperature, concentration are calculated at each grid point. Special items like
ammonia emissions, heat exchange between cows and surrounding air were
simulated using a customization, i.e. a routine included in a used defined file (UDF)
and built for the determination of the parameters exclusively relevant to the specific
case.
The enormous amount of data generated by CFD solver cannot be analysed by just
looking at the numerical values. The final step in CFD analysis involves the organization
and interpretation of the predicted flow data and the production of CFD images and
animations. Different post‐processing tools like colour plots, contour plots and vector
10
plots are used to go into the problem. The interpretation of these results plays
important role in determining the performance of any system being studied.
4.1. Governing equations
The governing equations of fluid flow and heat transfer can be considered as
mathematical formulations of the conservation laws that govern all fluid flow, heat
transfer and associated phenomena. These conservation laws describe the rate of
change of a desired fluid property as a function of external forces and can be written
as
Continuity equation
(1)
0
Conservation of momentum (Newton’s second law): The sum of the external forces
acting on the fluid particle is equal to its rate of change of line air momentum.
)= (2)
Conservation of energy (the first law of thermodynamics): the rate of change of
energy of a fluid particle is equal to the heat addition and the work done on the
particle.
(ρ (3)
11
CFD enforces these conservation laws over a discredited flow domain in order to
compute the systematic changes in mass, momentum and energy as fluid crosses the
boundaries of each discrete region (Versteeg and Malalsekeera, 1995).
When buoyancy forces occur in the flow, the assumption of fixed physical
properties no longer matches fluid flow behavior. The equation of state relates the
density of the air to thermodynamic state, i.e. its temperature and pressure. Climatic
variables in livestock facilities are a function of varying flow properties caused by the
heating and cooling of air. There are two main methods of modelling the density
variations that occur due to buoyancy. The first is the well known Boussinesq
approximation (Ferziger and Peric, 2002). This has been used successfully in many
greenhouse applications (Bartzanas et al., 2013a):
1 (4)
This relationship only considers dry air as the fluid medium, whereas it is
conceivable in most climatic flows a mixture of dry air and moisture will be involved.
Gan (1994) derived an extended version of the Boussinesq approximation, which
admits a description of the density of moist air as a function of temperature and
moisture concentration. With a reference density, temperature, and water
concentration and by using Taylor’s expansion theorem the density variation can be
expressed as:
(5)
where
1 1
;
Despite the enhanced relation, Eq. (5) is not commonly used in the literature, even
though some CFD software packages allow the front‐end incorporation of this
12
extension (Anon, 2006). The use of Boussinesq approximation offers in many CFD
models faster convergence and solution stability, than considering the density variable
in all equations. However, the Boussinesq approximation is not sufficiently accurate
at large temperature differentials (Ferziger and Peric, 2002). Therefore, another
method of achieving the coupling of the temperature and velocity fields is necessary
when the Boussinesq approximation is invalid. This is done by treating the air as an
ideal gas and expressing the density difference by means of the ideal gas equation:
(6)
This method can be considered as a weakly compressible formulation, which means
that the density of the fluid is dependent on temperature and composition but not
pressure.
4.2. Near wall functions
The presence of a solid boundary (wall) leads to considerable differences in the
flow behaviour and turbulence structure when compared to free turbulent flows
(those not affected by a solid boundary). The effects of a wall boundary are best
explained by defining a Reynolds number, Re, with a length scale based upon the
normal distance, y, from the wall
(7)
where U is the velocity and ν is the kinematic viscosity of the fluid.
The Reynolds number of the flow provides a measure of the relative contribution
of inertia forces to viscous forces. Inertia forces dominate in the flow far from the wall,
however, as the distance decreases towards zero, the Reynolds number also
13
decreases. Just before y becomes zero there is a range of values of y for which the
Reynolds number is of the order of 1. At this distance from the wall, and closer, the
viscous forces are equal in order of magnitude, or larger, than the inertia forces. There
are a number of individual layers with differing flow characteristics in the near wall
region that can be defined. In order to explain these layers it is necessary to introduce
two dimensionless parameters.
A velocity parameter
(8)
and a distance parameter
(9)
where Δyp is the distance from the wall to the nearest point, yp, where velocity is to
be
calculated, uτ is the shear velocity, μ is the dynamic viscosity of the fluid and τ is the
shear stress which, according to the wall law, is related to the turbulent kinetic energy
by
/ (10)
4.3. Ventilation rate
Many CFD studies of greenhouse ventilation have used the step‐down method to
compute ventilation rates. This can be done by including a tracer gas with the same
physical properties as air and of a known uniform initial concentration. Then the tracer
gas can be treated as a passive scalar and can be solved after the flow‐field solution
14
has been obtained, i.e. by decoupling it from the momentum equations. The species
equation used can be written as follows:
(11)
where c represents the concentration of the tracer gas in a cell; t is the time in
seconds; and the R.H.S of the equation represents the tracer gas flux through each
computational cell in the buildings volume. Initially all the cells in the building have a
fixed tracer gas concentration equal to unity and all the external cells equal to zero. In
this way, once an appropriate time‐step, dt, is selected, the decay in concentration of
the tracer can be predicted over time. The tracer gas concentration decreases in the
building at a rate that depends on the local values of the air velocity. Once all the
tracer gas been evacuated from the building, the volumetric average tracer gas
concentration can be tabulated as a function of time and exported to a statistical
package so that an exponential decay of the following form can be fitted
̅ ̅ 0 (12)
where the exponent n is the decay rate of this function and therefore describes the
ventilation rate of the studied volume (air changes per hour).
4.4. Turbulence modelling
As shown by the measurements of turbulent airflows and microclimate patterns in
open field conditions, the airflows are highly turbulent (Monteith & Unsworth, 1990).
Consequently, turbulent models must be introduced in the Reynolds equations
written to separate the mean flow from its fluctuating components. Several
considerations influence the choice of turbulence model. The most important ones,
15
which guide our selection too, consist in the accuracy and the simplicity of the
turbulence model.
The standard k‐ε model (Launder and Spalding, 1974) assuming isotropic
turbulence was adopted to describe turbulent transport. The k‐ε turbulence model is
an eddy – viscosity model in which the Reynolds stresses are assumed to be
proportional to the mean velocity gradients, with the constant of proportionality
being the turbulent eddy viscosity. The complete set of equations of the k‐ε model can
be found in Mohammadi and Pironneau (1994).
The effect of turbulence on the flow was implemented via the high Re k‐ε model
(standard) model (Launder and Spalding, 1974).
k k t k U i (13)
U j ij
t x j x j x j x j
t 2 (14)
U j C 2 C 1 Pk
t x j x j x j k k
where, the turbulent viscosity is:
k2 (15)
t C
and
U i (16)
Pk ij
x j
and Cμ = 0.09, σκ=1, Cε1 = 1.44, Cε2 = 1.96, and σε = 1.3 are model constants.
16
4.5. Modelling the animals
The presence of animals in a livestock building influences the indoor climate both
because of their heat release and because they physically obstruct the airflow. In most
of the CFD models that have been used to investigate the naturally ventilated livestock
buildings, the presence of animals either is neglected or is simulated as area with
higher resistance to air flow (porous media).
Heat release from the animal surface is dependent on radiation and convection
which are influenced by air velocity, temperatures, body surface of the animal and
thermal insulation between body and tissue layers (CIGR, 2002).
Hoff et al (1992) simulated heat transfer from the floor and Bjerg et al (2000)
assumed that heat was generated from the floor instead of pigs, for some cases. This
method has the advantage of saving design and calculation time but cannot follow
detailed distributions for airflow and air temperature affected by animal occupants.
However, some small animals such as broilers can be regarded as a heat source on the
floor when the animals are raised at high density
Among others researchers, Bjerg et al. (2000) used cylindrical pig simulators, made
from a painted metal tube with a heat generator inside, for investigating air motion
caused by convection around a livestock body, in experiments and CFD simulation. In
another approach the slatted floor and pigs was regarded as porous media with heat
production in the CFD simulation using suitable flow resistance properties for each
element.
However, CFD studies using the real configuration of animals or humans have also
been considered recently. Murakami et al (2000) modelled a mannequin to estimate
heat release from the human body. Total heat transfer was evaluated by integration
of CFD simulation with a thermo‐physiological model for the human body.
For livestock, Gebremedhin and Wu (2001) studied heat and mass transfer from
multiple cows using a transformed coordinate system of body‐fitted geometrical
model, which was an indirect method to simulate the animal in the model. In their
subsequent research, Gebremedhin and Wu (2003) considered the actual dimension
and configuration of randomly distributed multiple occupants in the CFD model.
17
Fig. 1. Flow field in a ventilated room with cows (Gebremedhin & Wu, 2003).
Seo et al (2009) presented a CFD simulation of environmental conditions and
ventilation performance for a full size pig house using simplified pig models. The
simplification process is illustrated in Fig 2 illustrates the distributions of the simplified
pig models in the full size building. The results showed that the presence of pigs in the
house had a significant effect on internal airflow and temperature distribution,
especially in the livestock zone, and the presence of the pig models improved the
accuracy of the CFD results for temperature distributions in the livestock house.
Animal design in the model generally required much time and skill for an optimum
mesh, but did not lead to excessive computational time.
Fig. 2. Simplification process of the pig model to reduce mesh resources (Seo et al.
2009)
Bjerg et al. (2008) suggested using CFD methods to determine the flow resistance
through a small group of geometrically‐modelled animals, and included the values in
subsequent CFD modelling of the entire animal‐occupied zone as a porous media,
within which the assumed animal heat production was released. This method makes
18
inclusion of large numbers of animals in CFD simulation affordable, but needs
validation.
More recently, in a study contacted by Sapounas et al. (2012), is described how a
detail CAD design of a dairy cow can be used to calculate the potential resistance of
the animals to the inside air flow
Fig. 3. 3D model of a virtual dairy cow house with 48 cows (Sapounas et al., 2012)
4.6. CFD modeling of indoor air properties
Early three‐dimensional predictions of airflow in a livestock room found reasonable
agreement with experimental results even though the housed animals were modeled
as a uniformly heated floor (Hoff et al., 1992). Harral and Boon (1997) simulated the
isothermal flow patterns in a three‐dimensional geometrical representation of a
livestock building. Mistriotis et al. (1997) studied the natural ventilation in a broiler
house with a two‐dimensional CFD model. The study focused on addressing the
problem of meeting adequate levels of ventilation on days of high solar radiation
coupled with low wind speeds. The animals were modelled as a constant heat flux,
which was represented by a heated floor. Zhang et al. (1999) studied the buoyant flow
generated by a single simulated pig. Transient simulations were observed to agree
well with measured data. Zhang et al. (1999) recognised that although the simulations
represented a gross simplification of the actual dynamics inside the building, similar
type modelling techniques could be used to develop accurate solutions when more
complex interactions occur. Bjerg et al. (2000) found that their CFD model
inadequately predicted recirculation zones in locations where these occurred
19
experimentally. Nevertheless, Bjerg et al. (2000) were able to form a number of
recommendations for the CFD modelling of similar animal housing airflows.
Bartzanas et al (2007) examined the influence of four commonly used CFD
turbulence models on airflow distribution in livestock buildings and greenhouses.
Comparing their CFD results with experimental data they concluded that k‐ε
renormalization group turbulence model showed the best behavior concerning the
accuracy of the predictions on livestock building ventilation rate (Fig. 4).
0.35
0.30
Air velocity, m/s
0.25
0.20
0.15
0.10
0.05
0.00
0 1 2 3 4 5
Livestock building length, m
Fig. 4. Influence of turbulence model on air velocity distribution along the
livestock building width at a height of 1.75 m above ground, ──── Standard k‐ε model,
─▲─▲─RSM model, ─ ─ ─ Realizable k‐ε model, ─■─■─■─ RNG k‐ε model (Bartzanas
et al. 2007)
Norton et al. (2009) develop and validate a CFD in order to investigate the natural
ventilation of a calf livestock building under different wind incidences (WI). In their
study they used a quantity known as the LMA, was used to define an air change
effectiveness parameter, called the air change ratio. They found that ventilation rates
were not at their highest when wind was blowing normal to the building because a
considerable quantity of the flow left the building via ‘‘short‐circuiting’’. However, the
greatest ventilation homogeneity was experienced when the wind was blowing
normal to the building, owing to the formation of two wind‐driven vortices in the
20
building. Results also showed that the highest level of environmental heterogeneity
occurs at WI of 10–40° because the primary vortex only occupies a portion of the total
building volume.
Fig. 5. Streamlines illustrating the airflow patterns coloured with LMA distribution in a
calf livestock building for two different wind incidences: 90° WI, and 60° WI.(Norton
et al. 2009)
Bartzanas et al. (2010) analyzed experimentally and numerically the aerosol levels
of three different sizes (PM10, 2.5 &1) and the climatic conditions that prevail in a
typical Greek livestock building that is naturally ventilated and hosts mainly sheep.
Concerning the airflow distribution (Fig. 6) they found that the flow is characterized
by a main core flow along the building and smaller secondary flows in different
positions inside the building. Resulted ventilation rate is strongly reduced to the
windward part of the building due to smaller ventilation openings in that part of the
building.
21
Fig. 6. Air velocity distribution in transversal sections along the length of the
building (Bartzanas et al. 2010b)
Norton et al. (2010a) have been developed a CFD model to predict the ventilation
performance and thermal environment of naturally ventilated calf buildings with
unrestricted, space boarding and ventilated cladding eave opening conditions. The
effect of altering the height of openings on the indoor environment as and ventilation
characteristics of the building has also been investigated. They found that increasing
the eave opening height of unrestricted and space boarding eave openings resulted in
a decrease of the average air velocity at animal level, due to a reduction in size and
change in location of the primary recirculation zone.
(a)
(b)
22
Fig. 7 Streamline visualisation of the flow regime (side‐view) in a building with 0.6m
high unrestricted eave openings condition. The black dashed line highlights the main
flow pattern (Norton et al., 2010a)
The efficiency of air mixing in a naturally‐ventilated livestock building with
different eave opening conditions was studied by Norton et al. (2010b) for wind
dominated and wind and buoyancy‐influenced ventilation. Among other useful results
they found that during wind‐dominated ventilation, the porosity of the eave opening
cladding system was found to influence the mixing of air and a reduced porosity
enhanced the ventilation efficiency of the building.
Fig. 8. Livestock building ventilation rate as a function wind speed for the 6 eave
opening conditions (different effective eave opening area (EffOA) (Norton et al., 2010b)
23
Norton et al. (2010c) tried to assess numerically, the relationship between calf heat
loss and the distributed indoor environment of commercial naturally ventilated calf
structure for a number of climate scenarios. The airflow patterns for each weather
scenario in the are shown in Fig. 9, where it can be seen that different airflow patterns
occur because of the changing calf heat production due to seasonal climates and to
the differences between the thermal boundary conditions of the building surfaces. For
the summer scenario, the temperature difference between the roof and the ambient
air was so great that the heat loss from the animal had a minimal effect on the
predicted indoor airflow patterns. During the spring and winter scenarios the overall
calf heat flux had a greater influence on the indoor climate, and the effect of using
different calf convective heat flux boundary conditions on the predicted airflow
patterns was obvious for only the winter scenario.
Fig. 9 The airflow pattern along a vertical plane of a commercial naturally ventilated
calf structure for (a) the winter scenario; (b) the spring scenario; (c) the summer
scenario. The outdoor wind speed is 0.5ms−1 in all cases (Norton et al. 2010c)
24
5. Managing Heat Stress
5.1. General managing options
The water requirements of cattle increases during heat stress. Cattle lose water
from increased respiration and perspiration. Additionally, consumption of water is the
quickest method for cattle to reduce their core body temperature. Therefore, water
consumption will be greater than typical metabolic requirements. Extra water tanks
should be introduced prior to extreme heat events so that cattle become accustomed
to them. Waterers need to be kept clean to encourage cattle to consume adequate
water.
Heat production from feed intake peaks 4 to 6 hours after feeding. Therefore heat
production in cattle fed in the morning will peak in the middle of the day when
environmental temperatures are also elevated. Cattle should receive a least 70% of
their feed 2 to 4 hours after peak ambient temperature. Changing the ration has been
controversial but research indicates that lowering the energy content of diet will
decrease the heat load. The general recommendation is to reduce the diet energy
content by 5 to 7%.
Shade can be critical in determining whether cattle die during extreme heat events,
especially for black cattle. If the shade structure has an east‐west orientation then
ground under the shade will remain cooler. However, if mud is an issue then a north‐
south orientation will increase drying as the shade moves across the ground during
the day.
Increasing the air flow can help cattle cope with extreme heat events. Wind speed
has been shown to be associated with ability of cattle to regulate their heat load.
Although we cannot influence wind speed, feedlots can increase ability for cattle to
be exposed to air movement. Utilize temporary wind breaks in winter to allow
maximal air movement in summer. Tall earthen mounds will allow cattle more
exposure to air movement. Feedlots should assess their feedlot and know which pens
25
have poor air movement. Avoid using these pens for cattle that will be approaching
slaughter weights in mid to late summer.
Another factor that feedlots can address is to control flies. Biting flies cause cattle
to bunch up which decreases cooling. Minimizing breeding areas for flies and applying
insecticides to decrease fly populations prior to heat stress times is worthwhile.
Sprinklers can be used to cool cattle during times of stress. Sprinklers increase
evaporative cooling and can reduce ground temperature. Sprinklers should
thoroughly wet the animal and not just mist the air in order to cool the animal. Before
installing a system make sure the water supply is adequate to provide drinking water
and sprinklers. Sprinkle intermittently to avoid mud and increased humidity.
Sprinklers should be placed away from feed bunks and waterers. Cattle need to be
introduced to sprinklers prior to extreme heat. Cattle not used to sprinklers will try to
avoid the spray. Additionally, sprinklers need to be used before cattle are in extreme
stress. Thermal shock from cold water can kill cattle that are extremely stressed. Once
sprinklers are utilized they need to be continued until the heat event is over and cattle
can manage on their own.
5.2. Influence of barn structural specification
Designers of dairy barns should seek to provide an optimal environment, resulting
in excellent welfare, to enable cows to maximize their milk production Dairy barns are
designed for 15 to 30 years of use, and therefore their design should take into account
the increase in extreme environmental events expected to accompany anticipated
climate changes. Thus, achievement of an optimal environment, in particular for the
high‐yield cow, is a challenging goal.
26
Their work and analysis was based in the development of a heat‐stress model that
simulates the ambient threshold temperature (THRT). In the latter model, a cow would
increase its respiratory rate (RR) in response to the development of heat‐stress
conditions, and calculation of THRT took into account wind speed, relative humidity,
ambient temperature, milk yield, milk content and fur thickness. The higher the THRT,
the better the animal can tolerate a harsh environment before suffering stress. Overall goal
of their study was to determine optimum barn characteristics, under the hypothesis
that these affect the THRT, which, in turn, takes into account ambient temperature,
relative humidity, and wind speed.
The daily mean THRT in free‐stall structures (31.68 °C) was found to be lower than
that in loose housing (32.18 °C). The difference in favor of loose housing was evident
throughout most hours of the day.
Fig. 10. Diurnal average changes in threshold temperature during the day in barn
types: loose housing and free stalls, (Shoshani and Hetzroni (2013).
Barns oriented north‐south (perpendicular to prevailing wind direction) were found
to have higher daily mean THRT than east‐west oriented ones and this difference was
consistent throughout most of the day except around midday. The advantage of the
north‐south orientation was evident in both loose‐housing and free‐stall barns, but
was greater for the free‐stall barns.
27
Single slope was least effective with regard to its effect on THTR compared with
open ridge and pagoda. However, no statistical difference was observed between
open ridge and pagoda. Barn orientation had no significant effect on THRT for any
given ridge type.
Roof slopes below 12% were found inferior to all others among which no significant
difference was observed.
Fig. 11. Threshold temperature with various roof slopes. Bars denote standard
error of the mean, (Shoshani and Hetzroni (2013).
A similar work but based on CFD model have be done more recently by Bartzanas
et al (2015). The numerical (CFD) model was successfully validated by exploiting
measurements recorded in a naturally ventilated livestock building for small
ruminants. The analysis reveals that the effective natural ventilation of the building
strongly depends to the external wind direction, while the temperature increases
excessively in the recirculation areas inside the building. Internal climate indexes like
THI and Twc were calculated in order to investigate the existence of areas where cold
or heat stress conditions hold. For the examined wind and ventilation regimes the
building is proved to be well ventilated, since no areas with excessive ammonia
concentration exist neither areas with heat or cold stress climate conditions are
28
developed. The THI is calculated in every point of the computational domain through
the following relationship (Kibler, 1964) THI = Td ‐ 0.55 x(1 ‐ RH) x (Td ‐ 14.4). The THI
was calculated via the CFD model for two cases: a) strong convection case (entering
air stream velocity equal to 2.6 m/s) and b) weak convection case (entering air stream
velocity equal to 0.75 m/s).
In the follow figure cross sections of THI distribution are given for both examined
cases. In the case of strong convection the average THI is 17.1 oC while the whole area
remains below the limit of 28 °C. In the case of weak convection the average THI is
17.61, with the whole area remaining below the upper limit too. The conclusion is that
in both cases there are no areas inside the building where the animals would suffer
heat stress under the examined conditions.
Fig. 12. THI distribution in cross sections for both cases (Bartzanas et al. 2015)
29
6. Future work
The present report summarizes the work carried out in the framework of the STSM of
Thomas Bartzanas in ATB‐Potsdam. During his visit Thomas Bartzanas, apart from the
collaboration of Prof. Thomas Amon has fruitful and useful discussions with other
members of his research team, and will together decide to further continue this work
in order:
to estimate the thermal heat stress index of dairy cows in large naturally
ventilated buildings using as input the spatial distribution values of air
temperature and air humidity inside the building and not only certain values
from the central of the building.
To proposes, based on CFD simulations, alternative building and natural
ventilation system in order to reduce the THI
Based on the above objectives the following work‐plan was agreed:
Definition of thermal stress index: In the literature one can find numerous
relationships for expressing thermal stress in relation to climate variables
(see also section 3 of the present report). The main relationships will be
reviewed and the team will conclude to three (3) different thermal stress
indexes. Two THI will based on the same climate variables (i.e air
temperature and air humidity) whereas in the third one more climate
variables will be added.
Experimental data: Experimental data will be used from the experimental
farms of ATB (see appendix). The team will decide for the quantity (number
of years) and frequency of data.
Estimation of THI based on experimental data. The THI as it was proposed
from task (1) will be estimated for the series of data that will be selected
on task (2)
Estimation of THI with CFD. The THI as it was proposed from task (1) will be
estimated for the series of data that will be selected on task (2) using CFD.
Two different CFD codes will be used: a) Open Source (ATB) and Fluent –
30
Ansys (CERTH). The outcomes of the two codes will be compared against
experimental data as a benchmark study.
Optimisation of livestock building in terms of THI. With the aim of CFD
different design and structural changes will be proposed to the building
envelope (height, length, ventilation openings) an their influence to the THI
will be again assessed using CFD
31
7. Literature cited
Anon., 2006. ANSYS CFX release 10.0 technical specifications. Public notice, ANSYS,
Inc., Southpointe, Canonburg, PA, USA.
Armstrong, D. V. 1994. Heat stress interaction with shade and cooling. J. Dairy Sci.
77:2044–2050.
Bartzanas T., Kittas C., Boulard T., 2004. Effect of vent arrangement on windward
ventilation of a tunnel greenhouse. Biosystems Engineering, 88 (4): 479‐490.
Bartzanas T., Kittas C., Sapounas A A., Nikita‐Martzopoulou Ch., 2007. Analysis of
Airflow through Experimental Rural Buildings: Sensibility to Turbulence Models.
Biosystems Engineering, 97 (2), 229‐239
Bartzanas T., Bochtis D.D, Sørensen C.G, Sapounas A.A, 2010a. A Numerical Modelling
Approach on Biomass Field Drying. Biosystems Engineering, 106: 458‐469
Bartzanas T., Kacira M., Zhu H., Karmakar S., Tamimi E., Katsoulas N., In Bok Lee, Kittas
C., 2013a. Computational fluid dynamics applications to improve crop
production systems. Computers and Electronics in Agriculture, 98: 151‐167
Bartzanas T., Bochtis D.D Green P., Sørensen C.G., Fidaros D., 2013b. Prediction of
quality parameters for biomass silage: a CFD approach, Computers and
Electronics in Agriculture, 98: 209‐216
Bartzanas T., Baxevanou C. Fidaros D., Papanastasiou D., Kittas C. 2015 Numerical
simulation of flow and transport phenomena in a livestock building, Report of
the project “Sustainable design of livestock buildigns”
32
Baeta, F. C., N. F. Meador, M. D. Shanklin, and H. D. Johnson., 1987. Equivalent
temperature index at temperatures above the thermoneutral for lactating dairy
cows. ASAE Paper 87–4015. Amer. Soc. Agric. Engr., St. Joseph, MI.
Bernabucci, U., N. Lacetera, L. H. Baumgard, R. P. Rhoads, B. Ronchi, and A. Nardone.
2010. Metabolic and hormonal acclimation to heat stress in domesticated
ruminants. Animal 4:1167–1183
Bianca, W. 1962. Relative importance of dry‐ and wet‐bulb temperatures in causing
heat stress in cattle. Nature 195:251–252.
Bjerg, B., Svidt, K., Zhang, G., & Morsing, S., 2000. The effect of pen partitions and
thermal pig simulators on airflow in a livestock test room. Journal of Agricultural
Engineering Research, 77, 317e326.
Bjerg, B., Zhang, G., & Kai, P., 2008. Porous media as boundary condition for air inlet,
slatted floor and animal occupied zone in numerical simulation of air flow in a
pig unit. In AgEng2008, international conference on agricultural engineering.
Hersonissos, Crete‐Greece.
Bjerg B., Norton T., Banhazi T., Zhang G., Bartzanas T., Liberati P., Cascone G., I.‐B. Lee,
Marucci A., 2013. Modelling of ammonia emissions from naturally ventilated
livestock buildings – Part one: Ammonia release modelling. Biosystems
Engineering, 116(3), 232‐245
Bjerg B., Liberati P., Marucci A., Zhang G., Banhazi T., Bartzanas T., Cascone G., I.‐B.
Lee, Norton T., 2013. Modelling of ammonia emissions from naturally ventilated
livestock buildings – Part two: Air change modelling. Biosystems Engineering,
116(3): 246‐258
Bjerg B., Cascone G., I.‐B. Lee, Bartzanas T., Norton T., Hong S‐W., I‐H Seo, Banhazi T.,
Liberati P., Marucci A., Zhang G, 2013. Modelling of ammonia emissions from
naturally ventilated livestock buildings – Part three: CFD modelling. Biosystems
Engineering, 116(3): 259‐275
Bligh, J., 1973. In: Bligh, J. (Ed.), Temperature Regulation in Mammals and Other
Vertebrates. North Holland, Amsterdam, J. pp. 351–354.
33
Bohmanova, J., I. Misztal, and J. B. Cole. 2007. Temperature‐humidity indices as
indicators of milk production losses due to heat stress. J. Dairy Sci. 90:1947–
1956
Buffington, D. E., A. Collazo‐Arocho, G. H. Canton, D. Pitt, W. W. Thatcher, and R. J.
Collier. 1981. Black globe‐humidity index (BGHI) as comfort equation for dairy
cows. Trans. ASAE 24:711–714.
CIGR, 2002. Heat and moisture production at animal and house levels. In S. Pedersen,
& K. Sa¨ llvik (Eds.), Fourth report of CIGR working group on climatization of
animal houses
Cook, N.J., 1990. The Designers Guide to Wind Loading of Building Structures, Part 2:
Static Structures. Butterworths, London, UK.
Ferziger J H., & Peric M., 1996. Computational Methods for Fluid Dynamics, Springer,
London.
Gan G., 1994. Numerical method for a full assessment of indoor thermal comfort.
Indoor Air 4, 154–168.
Gebremedhin, K.G., Wu, B.X., 2001. Numerical simulation of flow field around a cow
using 3‐D body‐fitted coordinate system. Journal of Thermal Biology 26, 563–
573.
Gebremedhin, K. G., & Wu, B. 2003. Characterization of flow field in a ventilated space
and simulation of heat exchange between cows and their environment. Journal
of Thermal Biology, 28, 301‐319.
Hahn, G. L., T. L. Mader, and R. A. Eigenberg. 2003. Perspective on development of
thermal indices for animal studies and management. EAAP Technical Series
7:31–44.
Harral B B., Boon C R., 1997. Comparison of predicted and measured air flow patterns
in a mechanically ventilated livestock building without animals. Journal of
Agricultural Engineering Research 66, 221–228.
34
Hoff, S. J., Janni, K. A., & Jacobson, L. D., 1992. Three‐dimensional buoyant turbulent
flows in a scaled model, slot‐ventilated, livestock confinement facility.
Transactions of the American Society of Agricultural Engineers, 35, 671‐686.
Igono, M. O., G. Bjotvedt, and H. T. Sanford‐Crane. 1992. Environmental profile and
critical temperature effects on milk production of Holstein cows in desert
climate. Int. J. Biometeorol. 36:77–87.
Johnson, H. D., A. C. Ragsdale, I. L. Berry, and M. D. Shanklin. 1963. Temperature‐
humidity effects including influence of acclimation in feed and water
consumption of Holstein cattle. Missouri Agr. Exp. Sta. Res. Bul. 846
Johnson, H. D. 1987. Bioclimates and livestock. Bioclimatology and the Adaptation of
Livestock. World Animal Science. (H. D. Johnson, ed.) Elsevier Science Publ. Co.,
New York
Jordan, E. R. 2003. Effects of heat stress on reproduction. J. Dairy Sci. 86: E. Suppl.:
E104‐E114
Kibler, H.H., 1964. Thermal effects of various temperature – humidity combinations
on Holstein cattle as measured by eight physiological responses, Research
Bulletin of Missouri Agricultural Experiment Station 16, 862‐878.
Le Page J F., Chevarin C., Kondjoyan A., Daudin J D., Mirade P S., 2009. Development
of an approximate empirical‐CFD model estimating coupled heat and water
transfers of stacked food products placed in airflow. Journal of Food
Engineering, 92 (2), 208‐216.
Launder B.E., and Spaiding D.B. 1974. The numerical computation of turbulent flows.
Computer Methods in Applied Mechanics and Engineering, 3, 269‐289
Mader, T. L.; Davis, M. S.; Brown‐Brandl, T., 2006. Environmental factors influencing
heat stress in feedlot cattle. J. Anim. Sci. 84, pp. 712–719
Mistriotis A., de Jong T., Wagemans M J M., Bot G P A., 1997. Computational fluid
dynamics as a tool for the analysis of ventilation and indoor microclimate in
agricultural buildings. Netherlands Journal of Agricultural Science 45, 81–96.
35
Mohammadi B., and Pironneau O. 1994. Analysis pf the k‐epsilon Turbulence Model.
Research in Applied Mathematics, Wiley, New York, Masson, Paris.
Monteith J L., and Unsworth M. 1990. Principles of Environmental Physics, 2nd Edition
Edward Arnold.
Moran DS., Pandolf KB., Shapiro Y., Heled Y., Shani Y, Matthew WT., Gonzales RR.,
2001. An environmental stress index (ESI) as a substitute for the wet bulb
temperature (WBGT). Journal of Thermal Biology 26, 427‐431
Murakami, S., Kato, S., & Zeng, J., 2000. Combined simulation of airflow, radiation and
moisture transport for heat release from a human body. Building and
Environment, 35, 489e500.
Nardone, A., B. Ronchi, N. Lacetera, and U. Bernabucci., 2006. Climatic effects on
productive traits in livestock. Vet. Res. Commun. 30(Suppl. 1):75–81.
Nardone, A., B. Ronchi, N. Lacetera, M. S. Ranieri, and U. Bernabucci., 2010. Effects of
climate changes on animal production and sustainability of livestock systems.
Livest. Sci. 130:57–69.
NOAA, 1976. Livestock hot weather stress. United States Dept. of Commerce, Natl.
Oceanic and Atmospheric Admin., Natl. Weather Service Central Region.
Regional Operations Manual Letter C‐31‐76
Norton T., Da‐Wen Sun, Grant J., Fallon R., Dodd V., 2007. Applications of
computational fluid dynamics (CFD) in the modelling and design of ventilation
systems in the agricultural industry: A review. Bioresource Technology 98, 2386–
2414
Norton T., Grant J., Fallon R., Da‐Wen Sun., 2009. Assessing the ventilation
effectiveness of naturally ventilated livestock buildings under wind dominant
conditions using computational fluid dynamics. Biosystems Engineering, 103(1),
78‐99
36
Norton T., Grant J., Fallon R., Da‐Wen Sun., 2010a. Assessing the ventilation
performance of a naturally ventilated livestock building with different eave
opening conditions. Computers and Electronics in Agriculture, 71(1), 7‐21
Norton T., Grant J., Fallon R., Da‐Wen Sun., 2010b. A computational fluid dynamics
study of air mixing in a naturally ventilated livestock building with different
porous eave opening conditions. Biosystems Engineering, 106: 125‐137
Norton T., Grant J., Fallon R., Da‐Wen Sun., 2010c. Improving the representation of
thermal boundary conditions of livestock during CFD modelling of the indoor
environment. Computers and Electronics in Agriculture, 73, 17‐36
National Research Council, 1971. A Guide to Environmental Research on Animals. Natl.
Acad. Sci., Washington, DC.
Nuyttens D., De Schampheleire M., Verboven P., Sonck B., 2009. Comparison between
indirect and direct spray drift assessment methods. Biosystems Engineering,
105(1): 2‐12
Roenfeldt, S., 1998. You can’t afford to ignore heat stress. Dairy Manage. 35 (5), 6–12.
Sapounas, A. A.; H. J. C. van Dooren and M. C. J. Smits., 2012. Natural ventilation of
commercial dairy cow houses: simulating the effect of roof shape using CFD.
International Conference of Agricultural Engineering. AGENG‐2012, Valencia,
Spain.
Seo, I.‐H., Lee, I.‐B., Kwon, S.‐H., Hwang, H.‐S., Hong, S.‐W., Bitog, J. P., et al. 2009.
Numerical investigation of the pig house to prevent pig respiratory disease
according to ventilation system. In The 2009 CIGR international symposium of
the Australian society for engineering in agriculture, 14 September 2009. The
Chifley at Lennons, Brisbane, Queensland, Australia.
Schaldach G., Berger L., Razilov I., Berndt H., 2000. Computer simulation for
fundamental studies and optimisation of ICP spray chambers. ISAS (Institute of
Spectrochemistry and Applied Spectroscopy) Current Research Reports, Berlin,
Germany.
37
Shoshani E., and Hetzroni A., 2013. Optimal barn characteristics for high‐yielding
Holstein cows as derived by a new heat‐stress model. Animal, 7 (1), 176–182
Tao X and Xin X. 2003., Temperature‐Humidity‐Velocity Index for Market size Broilers.
2003 ASAE Annual International Meeting, Las Vegas, Nevada, USA, July 27‐30,
2003 (Paper Number: 034037)
Thom, E. C. 1959. The discomfort index. Weatherwise 12:57–59.
Versteeg H K. & Malalsekeera W., 1995. An Introduction to Computational Fluid
Dynamics, Longman Group Ltd, Harlow, UK
Yousef, M. K., 1985. Stress Physiology in Livestock. CRC Press, Boca Raton, FL.
Zhang G., Svidt K., Bjerg B., Morsing S., 1999. Buoyant flow generated by thermal
convection of a simulated pig. Transactions of the ASAE 42, 1113–1120.
Zimbelman, R. B., R. P. Rhoads, M. L. Rhoads, G. C. Duff, L. H. Baumgard, and R. J.
Collier,. 2009. A re‐evaluation of the impact of temperature humidity index (THI)
and black globe humidity index (BGHI) on milk production in high producing
dairy cows. Pages 158–169 in Proceedings of the Southwest Nutrition
Conference. R. J. Collier, ed. Accessed Feb. 2, 2009.
38
8. Acknowledgements
This work has been financially supported of the DairyCare, COST action (FA1308).
A special thanks to the host Prof. Thomas Amon, Head of Department of Engineering
for Livestock Management at the Leibniz Institute for Agricultural Engineering (ATB)
for accepting my STSM and for his hospitality and collaboration during my stay on ATB.
I would also to thank G. Hoffmann, D. Janke; S. Hempel, W. Pauw and B. Amon and
also the rest staff of the Department of Engineering for Livestock Management for the
fruitful and interest discussion I had with them but also for their effort and willingness
to guide me to the different experimental facilities of the Institute.
39
9. Appendix
The experimental farm
40
41