Tropical cyclones formed in the Atlantic Ocean and in the Caribbean Sea and their proximities, constitute a habitual danger for the countries of the Caribbean area. Annually, in this area, about 20 tropical depressions can occur, great part of which develop up to hurricane categories, which reach winds exceeding 250 km /h in the case of category V hurricanes on Saffir-Simpson scale. At present, due to the effects of climate change on the planet, as well as the influence of “El Niño” phenomenon, hurricanes more frequently reach category III. IV or V, and cause severe destruction in economic objectives, among which are the agricultural facilities and the crops themselves.
August, September and October are the months with the highest frequency of cyclonic disturbances and hurricanes in Cuba, formed in the Caribbean Sea, Gulf of Mexico and North Atlantic Ocean.
The damages and losses suffered by Cuban agriculture as result of hurricanes’ onslaught are of great magnitude and demand enormous efforts for its recovery.
According to information from the Department of Defense and Civil Defense of the Ministry of Agriculture cited by Ríos (2017)RÍOS, M.: Agricultura cubana en recuperación progresiva tras huracán Irma, [en línea], Inst. Portal de Radio Habana Cuba., La Habana, Cuba, 2017, Disponible en:http://www.radiohc.cu/noticias/economy/141924-agricultura-cubana-en-recuperacion-progresiva-tras-huracan-irma , [Consulta: 24 de marzo de 2020]., only Hurricane Irma, which hit Cuba in September 2017, affected 50 000 hectares of various crops, of which 26 915 hectares were of banana, a quarter of the existing in the country. Likewise, according to the same source, 1 900 hectares of rice and 4 520 hectares of coffee were damaged, and other crops such as cassava and other food and vegetables were also damaged. In the case of poultry, 383 buildings were damaged, of which 77 were totally destroyed, with numerous deaths of birds, mainly layers. The force of the hurricane also devastated 500 dairy farms and large livestock buildings, causing more than 1000 deaths of calves. The onslaught of this hurricane affected the provinces of Holguín, Las Tunas, Camagüey, Ciego de Ávila, Sancti Spíritus, Havana and Mayabeque.
The president of the Livestock Entrepreneurial Group has declared that due to their characteristics of large open warehouses and light covers with dozens of years, poultry farming "has been practically forced to reborn repeatedly" by the occurrence of different meteorological events Figueredo y Romeo (2019)FIGUEREDO, O.; ROMEO, L.: Producción avícola en Cuba: Realidades y desafíos, [en línea], Cubadebate, 2019, Disponible en:http://www.cubadebate.cu , [Consulta: 24 de marzo de 2020].. According to this same source, Cuban poultry industry has more than 400 base business units of which 289 are production farms, 19 are genetic farms, 64 incubation plants, 75 replacements and raising farms and 106 egg production farms, in addition to 13 feed mills and 5 large warehouses to accumulate protein grains.
Hurricane Gustav, that hit Isla de la Juventud Municipality and Pinar del Río Province in August 2008, affected 63 poultry houses and 3 500 tobacco houses.
The study of the effect of hurricane-force winds on certain objectives has been a complex subject to be approached from an experimental point of view, since it is only possible to reproduce these phenomena and their effects in large-scale wind tunnels and also by applying techniques of the theory of similarity and making models of the objectives to be hit by the wind.
Nowadays, with the advances in computing and the existence of specialized programs in fluid dynamics analysis, based on numerical methods of finite elements (CFD, Computational Fluid Dynamics), it is possible to approach this type of study through numerical simulation, with a high degree of approximation.
The research that is proposed is of great interest to evaluate the effect of hurricane-force winds on different economic objectives, as well as to recommend certain protection measures that can mitigate their effects.
In the case of agriculture, there are numerous examples of targets that can be affected by extreme winds, among which greenhouses, poultry and other livestock farms, milking farms, batteries of photovoltaic panels for the generation of electricity and others, like the crops themselves that are affected by the winds’ action.
The use of Computational Fluid Dynamics tools, applied to the simulation of the action of winds on certain targets, will provide the distribution of pressures and forces on them, under the impact of winds of different magnitude and direction. It will also make possible to assess the mitigation capacity of certain elements, such as windbreaks, improvements in the design of the targets and other means of protection.
Computational Fluid Dynamics is a valuable tool for fluid simulation studies and due to that, it has been used by various researchers during the performance of dissimilar tasks related to fluid flow (Hofmann et al., 2001HOFMANN, M.; STOFFEL, B.; COUTIER-DELGHOSA, O.; PATELLA, R.F.; REBOUD, J.-L.: “Experimental and numerical studies on a centrifugal pump with 2D-curved blades in cavitation condition”, En: CAV 2001:session B7.005, 2001.; Balbastro et al., 2004BALBASTRO, G.C.; SONZOGNI, V.E.; FRANCK, G.; STORTI, M.: “Acción del viento sobre cubiertas abovedadas aisladas: simulación numérica.”, Mecánica Computacional , 23, 2004.; Balbastro y Sonzogni, 2007BALBASTRO, G.; SONZOGNI, V.: “Simulación de un ensayo en túnel de viento aplicando CFD”, Mecánica Computacional, 26: 3779-3787, 2007.; Xia et al., 2007XIA, G.; LI, D.; MERKLE, C.L.: “Consistent properties reconstruction on adaptive Cartesian meshes for complex fluids computations”, Journal of Computational Physics, 225(1): 1175-1197, 2007, ISSN: 0021-9991.; Gofran, 2008GOFRAN, C.M.: Experimental validation of CFD model predicting wind effects on inclined-roof mounted photovoltaic modules, KTH Industrial Engineering and Management Department of Energy Technology Division of Heat and Power Technology, Master of Science Thesis, Stockholm, 2008.; Alexandrikova et al., 2011ALEXANDRIKOVA, T.; PAVLOV, A.; STRELTSOV, V.: Hybrid density-and pressure-based splitting scheme for cavitating flows simulation, Ed. A.A. Mammoli, C.A. Brebbia, WIT Transactions on Engineering Sciences WIT Press, vol. 70, 41-56 p., 2011, ISBN: 978-1-84564-518-2.; Herrera et al., 2012HERRERA, P.M.I.; GARCÍA DE LA FIGAL, C.A.E.; RAMOS, C.E.; MARTIN, T.M.: “Simulación mediante la dinámica de fluidos por computadora del efecto de la velocidad del viento en el desempeño de los pulverizadores agrícolas de ventilador”, Revista Ciencias Técnicas Agropecuarias , 21(1): 19-25, 2012, ISSN: 1010-2760, e-ISSN: 2071-0054., 2014HERRERA, P.M.I.; DE LA FIGAL, C.A.E.; DE LAS CUEVAS, M.H.R.; MARTINS, T.M.: “Evaluación mediante la Dinámica de los Fluidos por Computadora (CFD) de la corriente de aire del pulverizador agrícola ASS-800”, Revista Ciencias Técnicas Agropecuarias , 23(2): 5-11, 2014, ISSN: 1010-2760, e-ISSN: 2071-0054.; Martínez et al., 2013MARTÍNEZ, R.A.; LAFFITA, L.A.; LOMBANA, S.M.Y.: “Impacto de vientos extremos en baterías de paneles solares”, En: VII conferencia de Ingeniería Agrícola, AGRING 2013, San Jose de las Lajas, Mayabeque. Cuba, 2013, ISBN: 978-959-16-2185-6.; Herrera et al., 2014HERRERA, P.M.I.; DE LA FIGAL, C.A.E.; DE LAS CUEVAS, M.H.R.; MARTINS, T.M.: “Evaluación mediante la Dinámica de los Fluidos por Computadora (CFD) de la corriente de aire del pulverizador agrícola ASS-800”, Revista Ciencias Técnicas Agropecuarias , 23(2): 5-11, 2014, ISSN: 1010-2760, e-ISSN: 2071-0054., 2015HERRERA, P.M.I.; GARCÍA DE LA FIGAL, C.A.E.; DE LAS CUEVAS, M.H.R.; MARTINS, T.M.: “Efecto del viento en el flujo de aire de un pulverizador”, Revista Ciencias Técnicas Agropecuarias , 24(2): 44-48, 2015, ISSN: 1010-2760, e-ISSN: 2071-0054.; Hsu y Wu, 2017HSU, S.-T.; WU, T.-C.: “Simulated wind action on photovoltaic module by non-uniform dynamic mechanical load and mean extended wind load”, Energy Procedia, 130: 94-101, 2017, ISSN: 1876-6102.; Herrera et al., 2018HERRERA, P.M.I.; DE LA FIGAL, C.A.E.; DE LAS CUEVAS, H.; MARTINS, T.M.: “Modelling of the air current in the vertical plane of Hatsuta agricultural sprayer”, Revista Ciencias Técnicas Agropecuarias, 27(2): 5-11, 2018, ISSN: 1010-2760, e-ISSN: 2071-0054.). This computational tool represents a valuable and economical alternative for conducting complex studies that have traditionally been carried out in wind tunnels (Natalini et al., 2001NATALINI, M.B.; CANAVESIO, O.F.; NATALINI, B.; PALUCH, M.J.: “Wind tunnel modelling of mean pressures on curved canopy roofs.”, En: American Conference on Wind Engineering, Clemson, 2001.; Guan et al., 2003GUAN, D.; ZHANG, Y.; ZHU, T.: “A wind-tunnel study of windbreak drag”, Agricultural and forest meteorology, 118(1-2): 75-84, 2003, ISSN: 0168-1923.; Gromke y Ruck, 2008GROMKE, C.; RUCK, B.: “Aerodynamic modelling of trees for small-scale wind tunnel studies”, Forestry, 81(3): 243-258, 2008, ISSN: 1464-3626, DOI: 10.1093/ forestry/cpn027.; Bitog et al., 2011BITOG, J.; LEE, I.-B.; HWANG, H.-S.; SHIN, M.-H.; HONG, S.-W.; SEO, I.-H.; MOSTAFA, E.; PANG, Z.: “A wind tunnel study on aerodynamic porosity and windbreak drag”, Forest Science and technology, 7(1): 8-16, 2011, ISSN: 2158-0103.). the use of windbreaks as a means of protection against the action of winds (Borrelli et al., 1989BORRELLI, J.; GREGORY, J.; ABTEW, W.: “Wind barriers: a reevaluation of height, spacing, and porosity”, Transactions of the ASAE, 32(6): 2023-2027, 1989, ISSN: 2151-0032, e-ISSN: 2151-0040.; Balbastro et al., 2004BALBASTRO, G.C.; SONZOGNI, V.E.; FRANCK, G.; STORTI, M.: “Acción del viento sobre cubiertas abovedadas aisladas: simulación numérica.”, Mecánica Computacional , 23, 2004.; Martínez et al., 2013MARTÍNEZ, R.A.; LAFFITA, L.A.; LOMBANA, S.M.Y.: “Impacto de vientos extremos en baterías de paneles solares”, En: VII conferencia de Ingeniería Agrícola, AGRING 2013, San Jose de las Lajas, Mayabeque. Cuba, 2013, ISBN: 978-959-16-2185-6.); including crop protection (Boldes y Colman, 2003BOLDES, U.; COLMAN, J.: La protección de los cultivos de los efectos del viento, Ed. Viento, Suelo y Plantas, INTA, Golberg A.D.; Kin A.G, BsAs ed., Argentina, 2003, ISBN: 987-521-104-4.).
In general, crops and agricultural facilities are located in open areas and are exposed to the incidence of extreme winds. This situation is aggravated by the geographical situation of Cuba, located in the Caribbean Sea in an area frequently hit by hurricanes of different categories. Consequences of climate change like the increasing, both, frequency and intensity of hurricanes are added to these vulnerability elements.
Given this situation, as well as the economic, social and environmental cost imposed by the occurrence of hurricanes, it is necessary to carry out studies that make it possible to take measures to mitigate or diminish the effects that these extreme winds may cause on certain objectives of the rural areas of the country.
As it is not feasible to carry out conventional experiments, creating artificial hurricanes acting on economic objectives of great extension or magnitude, this research is oriented to the use of numerical simulation methods, supported by computerized tools of fluid dynamics, in order to carry out virtual experiments on digitized models and determine the effect of extreme winds on certain economic objectives subjected to the action of hurricanes.
Various authors Martínez et al. (2013)MARTÍNEZ, R.A.; LAFFITA, L.A.; LOMBANA, S.M.Y.: “Impacto de vientos extremos en baterías de paneles solares”, En: VII conferencia de Ingeniería Agrícola, AGRING 2013, San Jose de las Lajas, Mayabeque. Cuba, 2013, ISBN: 978-959-16-2185-6.; Báez y Pozos (2017)BÁEZ, D.A.; POZOS, E.A.: Simulación numérica de los efectos del viento sobre un conjunto de paneles solares, Inst. Universidad Nacional Autónoma de México., Cuernavaca, Morelos, México, 2017. have used numerical simulation methods to study the action of hurricane force winds on certain targets, however, they do not report comparative studies with experimental results that allow validating the use of these methods, or at least knowing the level of prediction error of the model with respect to the experimental test.
The present work aims to present the results of a group of experiments carried out in a wind tunnel, aimed at obtaining data that could be compared with the calculations from simulation models using Computational Fluid Dynamics tools and consequently, obtaining some level of validation of the use of these tools in the analysis of high intensity winds.
The experimental runs were carried out in a wind tunnel designed and built in the Center for Agricultural Mechanization (CEMA) of the Faculty of Technical Sciences of the Agrarian University of Havana. Dynamic pressure readings were taken in different sections inside the tunnel with an air flow close to 10 m3 / s, simulating hurricane-force winds of the order of 36 m / s. These measurements were compared with the result of dynamic pressure calculations obtained through simulation in a three-dimensional model of the same wind tunnel, carried out with the “Flow Simulation” tool of the Solidwork 2018 software. The comparison was made with the presence of an obstacle to the air circulation located in the air intake area, in order to provide an aerodynamic situation with a greater degree of possible turbulence.
The wind tunnel used during the experimentation is shown in Figure 1, where its dimensions and 12 access points in different sections can also be seen. They were made for measures taking with a Pitot tube coupled to a differential manometer (Figure 2).
Figure 2 shows the points to be measured in each section of the tunnel, whose readings were processed to determine the mean value of the dynamic pressure in each of the sections, which were compared with the results of the numerical simulation carried out by means of finite element analysis.
To quantify the result of the comparative analysis between the values found experimentally and those obtained through simulation, the prediction error (ΔE:
where:
Quantity observed experimentally;
Magnitude obtained by simulation.
The meshing of the computational domain and the boundary conditions applied to the model are shown in Figure 3, where the mesh refinement carried out can also be seen. From a level 5 of refinement, a total of 223 673 cells were obtained, of which 43 809 cells correspond to fluid, 106 220 cells to solid and 73 644 to partial cells of solid and fluid interface.
The volumetric flow of air that moves the axial fan was made to impinge at the tunnel exit (red arrows) in a direction normal to the X-Y plane with a value equal to 10 m3 / s. At the entrance, the total pressure is declared taking the atmospheric pressure as reference (yellow arrows). On the interior surfaces of the tunnel, the actual wall condition has been declared, specifying the roughness of the steel plate used.
Figure 4 shows the number of iterations performed to satisfy the convergence criterion for the declared engineering goals reached in a time of 524 s, obtaining a satisfactory level of convergence of the results.
The distribution of dynamic pressure inside the tunnel for a volume flow of 10 m3/s, obtained by simulation with C.F.D., is shown in Figure 5 where values obtained by simulation in points of experimental measurement by Pitot tube are pointed out (central points in measurement planes 1, 2 and 3).
The results of the experimental measurements of the dynamic pressure for an average real volumetric flow of 9 763 m3/s are shown in Table 1. There, average values are compared of these measures with the values of the dynamic pressure obtained by simulation.
As it can be seen in Table 1 the prediction error of the simulation ranged between 0,53% and 2,07%, which is highly satisfactory, even when it is necessary to point out that all values simulated were superior to the ones measured experimentally. That is surely due to that, in real conditions, only a level of flow a little inferior to the one selected for the simulation could be reached in the tunnel.
When comparing the results of calculations performed by simulation tools based Computational Fluid Dynamics (CFD) to experimental measurements carried out in the same conditions of simulation, a satisfactory level of prediction was obtained that ranged between 0,53% and 2,07%.
This result confirms the application of numeric methods based in finite elements analysis, “Flow Simulation” del software Solidwork, versión 2018 in this case, during the simulation of air flow of high density.