For Dowload This Book Click LINK or Button Below
For Dowload This Book Click LINK or Button Below
For Dowload This Book Click LINK or Button Below
com
https://ebookmeta.com/product/advances-in-design-
simulation-and-manufacturing-iv-proceedings-of-
the-4th-international-conference-on-design-
simulation-manufacturing-the-innovation-exchange-
dsmie-2021-june-8-11-2021-lviv-ukraine-vol/
OR CLICK BUTTON
DOWLOAD EBOOK
https://ebookmeta.com/product/recent-advances-in-manufacturing-
automation-design-and-energy-technologies-proceedings-from-
icoft-2020-1st-edition-sendhil-kumar-natarajan/
Advances in Design,
Simulation and
Manufacturing IV
Proceedings of the 4th International
Conference on Design, Simulation,
Manufacturing: The Innovation
Exchange, DSMIE-2021,
June 8–11, 2021, Lviv, Ukraine –
Volume 1: Manufacturing
and Materials Engineering
Lecture Notes in Mechanical Engineering
Series Editors
Francisco Cavas-Martínez, Departamento de Estructuras, Universidad Politécnica
de Cartagena, Cartagena, Murcia, Spain
Fakher Chaari, National School of Engineers, University of Sfax, Sfax, Tunisia
Francesco Gherardini, Dipartimento di Ingegneria, Università di Modena e Reggio
Emilia, Modena, Italy
Mohamed Haddar, National School of Engineers of Sfax (ENIS), Sfax, Tunisia
Vitalii Ivanov, Department of Manufacturing Engineering Machine and Tools,
Sumy State University, Sumy, Ukraine
Young W. Kwon, Department of Manufacturing Engineering and Aerospace
Engineering, Graduate School of Engineering and Applied Science, Monterey
CA, USA
Justyna Trojanowska, Poznan University of Technology, Poznan, Poland
Francesca di Mare, Institute of Energy Technology, Ruhr-Universität Bochum,
Bochum, Nordrhein-Westfalen, Germany
Lecture Notes in Mechanical Engineering (LNME) publishes the latest develop-
ments in Mechanical Engineering—quickly, informally and with high quality.
Original research reported in proceedings and post-proceedings represents the core of
LNME. Volumes published in LNME embrace all aspects, subfields and new
challenges of mechanical engineering. Topics in the series include:
• Engineering Design
• Machinery and Machine Elements
• Mechanical Structures and Stress Analysis
• Automotive Engineering
• Engine Technology
• Aerospace Technology and Astronautics
• Nanotechnology and Microengineering
• Control, Robotics, Mechatronics
• MEMS
• Theoretical and Applied Mechanics
• Dynamical Systems, Control
• Fluid Mechanics
• Engineering Thermodynamics, Heat and Mass Transfer
• Manufacturing
• Precision Engineering, Instrumentation, Measurement
• Materials Engineering
• Tribology and Surface Technology
To submit a proposal or request further information, please contact the Springer
Editor of your location:
Dragan Peraković
Editors
Advances in Design,
Simulation
and Manufacturing IV
Proceedings of the 4th International
Conference on Design, Simulation,
Manufacturing: The Innovation Exchange,
DSMIE-2021, June 8–11, 2021,
Lviv, Ukraine – Volume 1: Manufacturing
and Materials Engineering
123
Editors
Vitalii Ivanov Justyna Trojanowska
Sumy State University Poznan University of Technology
Sumy, Ukraine Poznan, Poland
Dragan Peraković
University of Zagreb
Zagreb, Croatia
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
v
vi Preface
We would like to thank members of the program committee and invited external
reviewers for their efforts and expertise in contributing to reviewing, without which
it would be impossible to maintain the high standards of peer-reviewed papers.
Program committee members and invited external reviewers devoted their time and
energy to peer-reviewing manuscripts. Our reviewers come from all over the world
and represent 18 countries and are affiliated with more than 70 institutions.
Thank you very much to keynote speakers: Vitalii Pasichnyk (Ukraine),
Katarzyna Antosz (Poland), and Alper Uysal (Turkey) for sharing their knowledge
and experience.
The book “Advances in Design, Simulation and Manufacturing IV” was orga-
nized into two volumes according to the main conference topics: Volume 1—
Manufacturing and Materials Engineering and Volume 2—Mechanical and
Chemical Engineering. Each volume is devoted to research in design, simulation,
and manufacturing in the main conference areas.
The first volume consists of five parts. The first part includes recent develop-
ments in product design and manufacturing processes. Notably, it presents ways for
ensuring the technological parameters of automobile engines, studying the contact
pressure in parts, the development of approaches for recognition cutting parts of
cutters during machining, and fractal analysis of metal structures. This part also
includes the research of spindle units for multi-operational lathes, technological
inheritability and damageability of materials, studies of wear characteristics for
hardened steels using laser–ultrasonic surface treatment, and rational design mod-
eling of an interference fit. Recent developments in designing implants via additive
technologies, dynamic modeling of automatic clamping mechanisms, optimal
designing of automobile pipe adapters, and grinding hard alloys using solid
lubricants are also presented in this part. Finally, the first part includes studies in
ensuring technological parameters of surface shaping, design calculation of elec-
trohydraulic drives and technological equipment, optimizing the interelectrode gap
during electrical discharge grinding, and designing thread joints for thin-walled
shells.
The second part includes studies in the implementation of intelligent solutions
within the Industry 4.0 strategy. Notably, ways to implement blockchain infor-
mation management systems and agile project management for IMS and IT projects
are analyzed. This part also consists of optimization work with a digital human
model, the implementation of Industry 4.0 supported by service robots in produc-
tion processes, intelligent numerical control of profile grinding, and carrier behavior
strategy. Autonomous data-driven integration systems, 3D technology radar models
to evaluate emerging technologies, and ensuring the reliability of transport systems
are also included. Finally, the second part presents an intelligent scheduling system
architecture for manufacturing systems and a new approach for providing internal
logistics according to Industry 4.0 concept and corresponding requirements.
The third part is devoted to contributing to ICT in engineering education.
Mainly, it presents studies aimed at developing a mobile application for test control,
integrated quality assurance at HEI, and Android application for explaining form
deviations based on 3D models. This part also includes recent developments in
Preface vii
Steering Committee
General Chair
Vitalii Ivanov Sumy State University, Ukraine
Co-chair
Ihor Hrytsai Lviv Polytechnic National University, Ukraine
Members
Oleksandr Gusak Sumy State University, Ukraine
Yaroslav Kusyi Lviv Polytechnic National University, Ukraine
Oleksandr Liaposhchenko Sumy State University, Ukraine
Ivan Pavlenko Sumy State University, Ukraine
Vadym Stupnytskyy Lviv Polytechnic National University, Ukraine
Program Committee
Gabriel Abba University of Lorraine, France
Jean-Francois Antoine University of Lorraine, France
Katarzyna Antosz Rzeszow University of Technology, Poland
Jose Manoel Balthazar Institute of Aeronautics and Space, Brazil
Kristina Berladir Sumy State University, Ukraine
Jozef Bocko Technical University of Kosice, Slovak Republic
Ricardo Branco University of Coimbra, Portugal
Dagmar Caganova Slovak University of Technology,
Slovak Republic
Emilia Campean Technical University of Cluj-Napoca, Romania
Robert Cep VSB-Technical University of Ostrava,
Czech Republic
ix
x Organization
DSMIE Team
Kristina Berladir Sumy State University, Ukraine
Oleksandr Gusak Sumy State University, Ukraine
Ihor Hrytsai Lviv Polytechnic National University, Ukraine
Vitalii Ivanov Sumy State University, Ukraine
Yaroslav Kusyi Lviv Polytechnic National University, Ukraine
Oleksandr Liaposhchenko Sumy State University, Ukraine
Ivan Pavlenko Sumy State University, Ukraine
Andrii Slipchuk Lviv Polytechnic National University, Ukraine
Vadym Stupnytskyy Lviv Polytechnic National University, Ukraine
Contents
xv
xvi Contents
Advanced Materials
Strengthening of the NKV Type Centrifugal Pump’s Shaft
by Chemical-Thermocycling Treatment . . . . . . . . . . . . . . . . . . . . . . . . . 525
Kristina Berladir, Tetiana Hovorun, and Oleksandr Gusak
The Influence of Synthesis Modes on Operational Properties
of Oxide Ceramic Coatings on Aluminum Alloys . . . . . . . . . . . . . . . . . . 536
Nataliya Imbirovych, Oleksandr Povstyanoy, Olha Zaleta,
Sergey Shymchuk, and Olga Priadko
Optimal Design of Composite Shelled Sandwich Structures
with a Honeycomb Filler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
Andrii Kondratiev, Olexander Potapov, Anton Tsaritsynskyi,
and Tetyana Nabokina
Preparation and Characterization of a Biocomposite Based
on Casein and Cellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
Mykola Melnychuk, Victoria Malets, Marcin Sosnowski,
Ivanna Mykhaylyuk, and Inna Boyarska
Advanced Technologies of Manufacturing Readily Removable Cores
for Obtaining High-Quality Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
Olga Ponomarenko, Igor Grimzin, Nataliia Yevtushenko, Tatiana Lysenko,
and Dmytro Marynenko
The Use of Ion Bombardment Modeling as a Component
of the Structural Engineering of Nanoperiodic
Composite Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
Oleg Sobol, Nataliia Pinchuk, Andriy Meilekhov, and Mariia Zhadko
Powder Technology and Software Tools for Microstructure Control
of AlCu2 Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
Oleg Zabolotnyi, Viktoriya Pasternak, Nataliia Ilchuk, Nataliia Huliieva,
and Dagmar Cagáňová
Rational Choice of a Material for Orthopedic Insoles Based
on the Mechanical Characteristics and Practical
Application Purposes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
Ruslan Zaloha, Kostiantyn Dyadyura, Viliam Zaloga, and Michal Hatala
Structure and Properties of Surface Bandage Shelves for the Gas
Turbine Engine’s Blades . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
Natalia Zaichuk, Sergii Shymchuk, Anatolii Tkachuk, Yurii Feshchuk,
and Jacek Szczot
1 Introduction
Modern computer-aided design of cast engine parts is a powerful tool for developing new
parts and upgrading existing ones. Current systems for computer-engineering modeling
of production processes and analysis of the thermal and stress-strain state of cast parts of
internal combustion engines are the most effective in design and technological design.
Processes computer-aided design, technological preparation, and production of cast
engine parts are an integral part of a systematic approach inherent in the CALS-
technologies, i.e., a set of the corresponding time sequence of state changes of the part.
In current conditions, the main requirements for manufactured products are high qual-
ity, low cost, and minimum development time for new products, which can be achieved
using computer modeling [1] and parameter identification [2] in technological design.
Most of the cast parts for transport and particular purpose types of diesel are made
by casting method, which should be based on manufacturing’s technological aspects,
namely various casting defects [3]. These defects occur due to the lack of methods and
tools to influence the technological modes of production. The design documentation and
technical specifications usually specify the size and number of various casting defects
that are not allowed in part, which means that such defects affect its quality and reliability.
The diesel’s central part under consideration is a block-case, the quality of which
will depend on the required characteristics laid down at the design stage. According
to the technical conditions, requirements for the quality and reliability of cast parts of
internal combustion engines are laid and fulfilled at the production stage using methods
for determining technological defects [4].
Obtaining a high-quality block crankcase at the casting stage consists of two main
processes: pouring molten metal into the mold and forming the cast part during the phase
transition during cooling and crystallization.
2 Literature Review
understanding of these processes, and therefore better manage them for reducing scrap
in molding and increase yield.
Full-scale CAD/CAM/CAE systems are complex multifunctional systems that
include a large set of modules (from 40 to 50) for various functional purposes [12].
In practice, various specialized and universal CAD/CAE/CAM systems of various
versions and configurations are in operation in most enterprises. Very often, different
strategies are used in divisions of the same organization. Sometimes this is even the case
at the level of individual developers.
In the application of design of cast parts of modern systems (CAD/CAM/CAE/PDM
and CIM), it is possible to provide [13]:
• Reduce production time and increase competitive products and output. Due to pro-
grams that provide speed for creating 3D models, regardless of the time spent on
producing drawings on paper [14, 15];
• They can also provide significant cost savings. By using a software package that
reduces the number of ongoing changes that accompany any design process, as well
as errors [16];
• Improving the vibration reliability [17] and quality of manufacture. With the help of
effective analysis of the created products, it can be tracked related defects in some
software packages [18];
• It will allow you to design and produce better models [19] continuously;
• Maximum error elimination. Due to the visualization of the designed product, which
allow us to eliminate typical errors in production promptly [20];
• Introduction of newer and more modern methods that allow us to develop productive
strategies of 3D model design, as well as their visualization, consideration of fill
processes, as well as possible defects [21];
• Avoiding large-scale layout products will reduce the time to create a specific part [22].
3 Research Methodology
For achieving this objective, the following sequence of stages was proposed, according
to the method of computer modeling of thermal and hydrodynamic casting processes:
6 O. Akimov et al.
A cast 4DTNA1 diesel crankcase was selected as the studied part for computer-
integrated modeling (Fig. 1).
Using the 3D import module built into the ICS NovaFlow, the block crankcase model
with a gating-feeding system and the equal model were converted into a finite-volume
model.
The optimal cell parameters are set based on the calculation time-adequacy of the
results (Fig. 2). The cell size is 3.7 mm.
Ensuring the Technological Parameters of Cast Block Crankcase 7
Fig. 2. Final-volume model of the casting in the chill with the displayed cells.
The assignment of boundary conditions for the material of casting elements and
technological equipment is made by dividing into separate parts, each of which is given
its color:
4 Results
• passing through the elements of the gating system, the melt does not acquire a turbulent
character of movement;
8 O. Akimov et al.
• the transition from the liquid to the solid phase during cooling of the part in the form
(Fig. 4);
• gas-shrinkage defects expressed by the Niyama criterion [24] (Fig. 5).
Analysis of the dynamics of cooling of the casting, the phase transition, the connec-
tivity of the zones that crystallize last allowed to determine the places of the possible
occurrence of gas-shrinkable defects.
The conclusion is based on the analysis of the location of defects expressed in the
ICS NovaFlow by the Niyama criterion.
The criterion for determining the location of gas-shrinkable defects and their mag-
nitude is Niyama, which we use to predict microporosity and gas-shrink porosity large
enough to be detected by radiographic testing. This criterion is a reliable predictor of
porosity for simple castings. Still, in the case of castings with complex geometries, its
use requires a more thorough analysis of the simulation results since many factors affect
the formation of gas-shrink porosity.
Analysis of the locations of defects showed that the most prone to shrinkage places
are:
Fig. 4. The transition from the liquid to the solid phase during the cooling of the part in the chill.
5 Conclusions
The developed 3D model of block crankcase casting with a technological gating-feeding
system allowed creating a finite-difference model of casting and tooling and performing
engineering modeling of casting processes in the ICS NovaFlow. The analysis of physical
10 O. Akimov et al.
features of the processes of filling and cooling the castings in the form was performed
for the cast block crankcase of 4DTNA1 automobile diesel, the locations and sizes of
gas-shrinkable defects were determined according to the Niyama criterion.
The research results allowed us to form boundary and initial conditions for modeling
the stress-strain state of the block crankcase in the places where gas-shrink porosity is
formed.
Further development of the above studies will be carried out for castings of new
types, configurations, other casting technologies, newly synthesized alloys.
References
1. Mehrabi, M.G., Ulsoy, A.G., Koren, Y.: Reconfigurable manufacturing systems: key to future
manufacturing. J. Intell. Manuf. 11, 403–419 (2000). https://doi.org/10.1023/A:100893040
3506
2. Pavlenko, I., Trojanowska, J., Ivanov, V., Liaposhchenko, O.: Parameter identification of
hydro-mechanical processes using artificial intelligence systems. Int. J. Mechatron. Appl.
Mech. 2019(5), 19–26 (2019)
3. Prasad, R.: Progress in investment castings. In: Science and Technology of Casting Processes.
IntechOpen (2012)
4. Thomas, D.S., Gilbert, S.W.: Costs and cost effectiveness of additive manufacturing. NIST
Spec. Publ. 1176, 12 (2012). https://doi.org/10.6028/NIST.SP.1176
5. Milchev, G., Miltchev, R.: Development of information technologies, planned obsolescence
and modification of the life-cycle of the CAD/CAM/CAE systems. Eur. J. Sustain. Dev. 7(3),
217 (2018). https://doi.org/10.14207/ejsd.2018.v7n3p217
6. Rao, S.S.: The Finite Element Method in Engineering. Butterworth-Heinemann (2017)
7. Sang, Z., Xu, X.: The framework of a cloud-based CNC system. Procedia CIRP 63, 82–88
(2017). https://doi.org/10.1016/j.procir.2017.03.152
8. Kumar, V., Madan, J.: A system for computer-aided gating design for multi-cavity die-casting
dies. Proc. Inst. Mech. Eng. Part B: J. Eng. Manuf. 231(11), 1983–1999 (2017). https://doi.
org/10.1177/0954405415614727
9. Khan, M.A.A., Sheikh, A.K.: A comparative study of simulation software for modelling metal
casting processes. Int. J. Simul. Model. 17(2), 197–209 (2018). https://doi.org/10.2507/IJS
IMM17(2)402
10. Kozłowski, J., Sika, R., Górski, F., Ciszak, O.: Modeling of foundry processes in the era of
Industry 4.0. In: Ivanov, V., et al. (eds.) DSMIE. LNME, pp. 62–71. Springer, Cham (2019).
https://doi.org/10.1007/978-3-319-93587-4_7
11. Bianchi, M.F., Gameros, A.A., Axinte, D.A., Lowth, S., Cendrowicz, A.M., Welch, S.T.:
On the effect of mould temperature on the orientation and packing of particles in ceramic
injection moulding. J. Eur. Ceram. Soc. 39(10), 3194–3207 (2019). https://doi.org/10.1016/
j.jeurceramsoc.2019.03.049
12. Ng, C.B.R.: A concurrent design facility architecture for education and research in multi-
disciplinary design of complex systems (2019)
13. Wang, X., Bi, Z.: New CAD/CAM course framework in digital manufacturing. Comput. Appl.
Eng. Educ. 27(1), 128–144 (2019). https://doi.org/10.1002/cae.22063
14. Akimov, O., Soloshenko, V., Kostyk, K.: Computer-integrated design of cast parts on the cri-
terion of performance on the example of the turbine wheel turbocharger. In: CEUR Workshop
Proceedings, vol. 2353, pp. 324–338 (2019)
Ensuring the Technological Parameters of Cast Block Crankcase 11
15. Akimov, O.V., et al.: Computer engineering and design of cast parts for internal combus-
tion engine crankcase. J. Eng. Sci. 6(2), E24–E30 (2019). https://doi.org/10.21272/jes.2019.
6(2).e4
16. Dhahad, H.A., Alawee, W.H., Marchenko, A., Klets, D., Akimov, O.: Evaluation of power
indicators of the automobile engine. Int. J. Eng. Technol. 7(4.3), 130–134 (2018)
17. Pavlenko, I., Ivanov, V., Kuric, I., Gusak, O., Liaposhchenko, O.: Ensuring vibration reliability
of turbopump units using artificial neural networks. In: Trojanowska, J., Ciszak, O., Machado,
J.M., Pavlenko, I. (eds.) MANUFACTURING 2019. LNME, pp. 165–175. Springer, Cham
(2019). https://doi.org/10.1007/978-3-030-18715-6_14
18. Boboulos, M.A.: CAD-CAM & Rapid Prototyping Application Evaluation. Bookboon (2010)
19. Kumar, S.L.X.: State of the art-intense review on artificial intelligence systems application
in process planning and manufacturing. Eng. Appl. Artif. Intell. 65, 294–329 (2017). https://
doi.org/10.1016/j.engappai.2017.08.005
20. Laurel, B.: Design Research: Methods and Perspectives. MIT Press (2003)
21. Otto, K.N.: Product design: techniques in reverse engineering and new product development
(2003)
22. Khan, M., Khan, S.S.: Data and information visualization methods, and interactive mecha-
nisms: a survey. Int. J. Comput. Appl. 34(1), 1–14 (2011)
23. Parsons, J.L., Campbell, J.R.: Digital apparel design process: placing a new technology into a
framework for the creative design process. Cloth. Text. Res. J. 22(1–2), 88–98 (2004). https://
doi.org/10.1177/0887302X0402200111
24. Ignaszak, Z.: Discussion on usability of the niyama criterion for porosity predicting in cast
iron castings. Arch. Foundry Eng. 17(3), 196–204 (2017). https://doi.org/10.1515/afe-2017-
0115
The Contact Pressure in Drawing Parts Without
Clamping the Workpiece Flange
Abstract. In the paper, dependence has been obtained to calculate the contact
pressures when drawing down the axisymmetric workpiece without a blank flange
collet. The solution is based on the assumptions of the momentless theory of
shells. The adequacy of the mathematical model is confirmed by experimental
data for a narrow interval of forging blanks. The experiments have been carried
out on a specially designed tooling to measure load cell deformation using a
strain gauge. All the equipment used has passed metrological control. To calculate
the meridional and tangential stresses on the torus-shaped portion of the matrix,
dependencies were obtained that contain a term connecting the thickness of the
workpiece with the value of the stresses arising during drawing, which more
accurately describes their distribution on the drawing edge of the matrix. The
expression for calculating the surface contact pressure during the drawing of a
cylindrical part makes it possible to consider the friction stresses at the radius of
the matrix rounding and calculate the drawing force. The obtained dependence
differs from the conventional ones in its simplicity and clarity and can be used
at the preliminary stage of choosing equipment for stamping. It is shown that the
friction stresses between the contacting surfaces can be controlled over a wide
range while achieving a significant change in the stress state and the distribution
of deformations in the volume of the workpiece.
1 Introduction
Drawing of axisymmetric parts is characterized not only by the loss of flange part stability
but also by the deformation localization in the zone of the transition of the wall to the
bottom and the destruction of the workpiece given location [1, 2]. The destruction and
localization of deformations can be eliminated by reducing the tensile stress value. When
drawing down with a collet, this problem is solved by reducing the collet force, using
highly effective lubricants, and also polishing the die body contact surfaces and the blank
holder [3–5].
2 Literature Review
The drawing down without workpiece flange pressing is limited by the formation of
workpiece uncompressed part. The bottom separation can occur if the process of work-
piece retraction into the die body hole continues, ignoring the folds on the flange. As
shown in the papers [6, 7], it is possible to extend the interval of non-collet forging by
increasing the resistance of the flange, loss of stability, or increasing tensile stresses in
the initial stage of deformation. The blank flange practically does not contact the sur-
face of the die body, since when the punch is lowered, the flange portion rises above
the die body and forms a conical surface with an angle α at the apex [8–11]. There-
fore, the contact of the workpiece with the die body occurs along its toroidal surface
with a certain rounded radius. The stresses in the drawn metal, the force, the drawing
coefficient, and the corrugation and destruction of the blank depending on the radius of
curvature of the drawing edge of the matrix [6, 12]. Therefore, it is chosen very care-
fully in each specifically selected technological process [3, 13]. The existing analytical
dependences for calculating stresses at the radius of curvature do not fully consider all
the minute details and technological factors of the drawing process, and their accuracy
is not satisfactory [14, 15]. The presence of refined formal dependencies describing the
distribution of stresses on the drawing edge of the matrix, taking into account friction,
bending moments, contact pressure, blank thickness, and others, will make it possible
to expand the range of drawn products without pressing the flange and reduce rejects
to a minimum. Therefore, it is of great scientific interest to theoretically determine the
contact stresses on the torus-shaped section of the matrix and the factors that make
it possible to control the value of these stresses to find conditions for expanding the
possibilities of drawing without a folding holder. This will enable finding more accu-
rate theoretical dependences for calculating and analyzing the stress field when drawing
cylindrical parts.
3 Research Methodology
To determine the contact stresses, we start from the equilibrium equations for the torus
[6, 16], but taking into account the surface load
∂R2 N1 ∂((R1 sin θ)T )
⎫
∂φ + ∂θ − N2 ∂R
∂φ + T
2 ∂R1 sin θ
∂θ = 0, ⎪
⎬
∂((R1 sin θ )N2 ) ∂(R1 sin θ)
∂θ + R2 ∂T∂φ − N1 ∂θ + T ∂R ∂φ = 0, ⎪
2
(1)
N1 N2 ⎭
R1 + R2 = q.
Next, we calculate according to the procedure [8, 18, 19] with the only difference that
the pressure on the workpiece when it slides along the edge of the die body is applied
from within half a torus.
where Rd – is the radius of the drawing part;
rm The radius of the die body input edge; R = Rd + a sin θ ; a = rm .
When the torus is affected by a surface load with loss of stability in the form of bulge
formation, the expression for q has been obtained by the authors of [8, 20, 21] as k → 0:
E s 3
q= n2 − 1 , (3)
12 1 − μ2 a
Regarding our drawing conditions, it should be slightly modified. For further calcu-
lation, we use the cylindrical rigidity of the shell corresponding to the secant modulus
3
D ≈ D = Ec9s , we sort out the value of n for which n2 − 1 = 1. Further, since
Ec = σεii , we accept σi = σcp , and εi = 21 εθ [22, 23]. The draw ratio is m = DD0 , hence
εθ = 1 − m = 1 − DD0 , where D and D0 are the diameter of the workpiece and the
diameter of the blank, respectively. Then expression (6) will have the form:
3
2σs s
q= . (4)
9 1− D rm
D0
We substitute this relation into the formulas for the internal forces and, passing to the
stresses, we get:
2
σs s
σρ = , (5)
9 1− D rm
D0
1 + k2 sin θ 2σs s 2
σθ = − . (6)
1 + k sin θ 9 1 − D rm
D0
Formerly [6, 24], formulas were derived from calculating the magnitude of the stress
tensor components on the torus-like part of the die body without considering the surface
load. Using the principle of superposition [16, 25] and carrying out simple transforma-
tions, the final dependencies for calculation of meridional and tangential stresses on the
rounded radius of the die body at θ = 90º will look like:
R0 rm R0 s 2
σρ = σs ln − + 0, 1 , (7)
Rd + rm Rd + rm R0 − Rd rm
R0 rm 1 + k2 R0 s 2
σθ = −σs 1 − ln + + 0, 22 . (8)
Rd + rm Rd + rm 1 + k R0 − Rd rm
The Contact Pressure in Drawing Parts Without Clamping 15
In contrast to correlations [6, 26], to calculate the meridional and tangential stresses
on the torus-like part of the die body, the dependences (7) and (8) contain a term that
relates the thickness of the blank to the value of the stresses arising while drawing, which
describes their distribution on the drawing edge of the die body more accurately.
Below there are stress graphic dependences (Fig. 1) – the blank thickness for R0 =
50 mm, Rd = 25 mm and matrix rounding radius rm = 10 mm, created according to
formulas (7), (8).
Fig. 1. Stress graphs (σ/σs ) – the blank thickness (s), created according to dependences (7), (8).
Using the second system Eq. (2), as well as (10) and (8) and carrying out simple
transformations, we find the surface pressure, which causes the plastic deformation of
the blank during drawing without workpiece collet:
⎛ ⎞
R0 rm Rd
ln +r − +r
q ⎜ Rd m Rd m (Rd +rm ) 2
⎟
= σs ⎝ 2
2 1+ 2
k ⎠. (9)
s + R0R−R
0
d r
s
m 9(R d
1
+r m ) − 9r m 1+k − r
1
m
We analyze the terms of the expression for given deformation parameters – R0 = 50 mm,
s = 2 mm, Rd = 25 mm and rm = 10 mm. The first and second terms are of the same
order of smallness and can be neglected with an error up to the third decimal place. Then
the surface load is the following:
s
q=− σs . (10)
rm
To verify the obtained dependence by determining contact pressures, experiments have
been carried out to measure the load cell deformations caused by frictional forces.
Experimental and theoretical studies demonstrate [13, 18, 27, 28] that the frictional
stresses between the surfaces to be contacted can be controlled within wide limits while
achieving a significant change in the stress state and the distribution of deformations
in the volume of blank, which is of great importance for the production of high-quality
workpieces using drawing down.
Since the flange of the workpiece practically does not contact with the mirror of the
die body while drawing down without the blank holder, and friction occurs mainly along
the rounded radius of the latter, it is of great interest to determine the frictional stresses
and to study their effect on the occurrence of the loss of flange stability.
16 R. Arhat et al.
The experiments were carried out on a tensile machine of UME-10TM type with a
force of 10 tf [14, 24, 29].
The experimental equipment appeared to be a special die body with an inlet of
50 mm and a set of punches with diameters of 49.3 mm; 46.4 mm; 49 mm, providing
the drawing of blanks without thinning the stiffening plate of 08kp steel with a thickness
of 0.15 mm, of aluminum A2 with a thickness of 1.4 mm and copper M4 of 0.25
mm correspondently. The dying body was installed on the lower traverse of the tensile
machine. The exchangeable punches were on the upper traverse. Both traverses were
equipped with clamping elements.
Special equipment appeared to be a dying body, with an insert for the drawing radius,
which can be displaced in the direction of movement of the punch during drawing and
the punch of a conventional design (Fig. 2) [15, 30, 31].
Fig. 2. Equipment for measuring frictional stresses at the rounded radius of the exhaust die body:
1 – punch; 2 – blank; 3 – movable insert; 4 – a fixing bolt; 5 – a connecting ring; 6 – load cell; 7
– strain gage.
The equipment for measuring the load cell deformations is a strain gauge operational
amplifier MCP606-I/P, ADC E14–440, power supply unit AX-1803D and personal com-
puter Pentium 4 CPU 2.40 GHz 1.0 GB RAM. Load cell deformations caused by fric-
tional forces during drawing determined strain gages deformation, which is included
in a quarter-bridge scheme without compensation of temperature stresses [32]. This, in
turn, caused the change in the resistance of the resistor and the current in the circuit.
The current oscillations were amplified by an operational amplifier and transmitted to
the ADC. On the computer monitor, these current readings were recorded for a prede-
termined time. It was KF 5P1–5-200-A-12 strain gage, with an operating resistance of
R = 199.7 ± 0.2 and a base of 5 mm. The label of the resistors was produced with
the “cyacrine” glue according to the technology described in the paper [33, 34].
4 Results
Thus, the contact pressure between the hemispherical surface of the die body and the
workpiece was measured. Measurements were subjected to a minimum of 16 blanks for
The Contact Pressure in Drawing Parts Without Clamping 17
each metal and alloy with different coefficients of drawing. The load cell rounded radius
rm = 4 mm and rm = 1.5 mm, without lubrication of the surfaces of the workpiece
and the tool. Simultaneously, the depth of the punch’s progress was recorded, such as
force increase; a sharp force increase meant the occurrence of corrugations that made
it difficult to draw a flat workpiece into the hole in the die. Based on the measurement
results, typical ADC voltage-time charts are shown (Fig. 3). Also, the experimental data
are used to verify the adequacy of the mathematical model of the distribution of contact
pressures on the die body rounded radius when drawing without collet.
Fig. 3. A typical schedule of measurement of load cell deformations: k1 – the first load cell
(breakage); k2 – the second load cell; k3 – the third load cell; k4 – the fourth load cell.
For greater accuracy of contact pressure measurements, the first lot of blanks was
chosen in such a size that their diameter was equal D0 = D + 2rm , where D is the
diameter of the drawing workpiece. This was to minimize the effect of blank flange
deformation and reduce the impact of the bending moment on the torus-like section. The
remaining two lots of blanks were selected with dimensions D0 = D + 4rm and D0 =
D + 6rm correspondently. The results of the experimental data have been subjected to
statistical processing [12, 35, 36], then averaged and summarized in Table 1. Processing
experimental data on the measurement of contact pressures showed that their distribution
is subjected to normal law at a significance level q = 0.05. The confidence error in
the results of the experiment was assumed to be symmetric and amounted to =
±0.26 MPa.
In Table 1, in the numerator of the values of the diameter, the draw ratio, the stroke
of the punch, the value of the contact pressure, which relate to the drawing option with
the rounding radius of the die body r m = 4.0 mm; in the denominator it is the same for
r m = 1.5 mm; , % is deviations of the experimental data from the theoretical ones,
calculated from the dependence (3).
18 R. Arhat et al.
5 Conclusions
The results of experimental studies showed that the greatest coincidence with theory is
6.7% for drawing with k = 1.12 and rm = 1.5 mm for copper and the greatest discrepancy
in calculations is up to 87% for steel with rm = 4.0 mm a rounded radius of the die
body. The discrepancy between the results of experimental studies and calculated data
in general can be explained by the imperfection of the mathematical model, where all
factors of constructive and technological nature are not taken into account, as well as
hardening of the metal during plastic deformation [37, 38]. The best coincidence between
the results of theoretical and experimental studies, which does not exceed 20% (except
copper drawing, k = 1.04, rm = 1.5 mm), showed data for all metals with rm = 1.5
mm. However, as can be seen from Table 1, an increase in the flange width results in
an increase in the contact pressure, and the best coincidence of the results is observed
precisely with the drawing of blanks with an enlarged flange. Therefore, the completion
of the dependence is obvious. A further direction of research on this issue will concern
its refinement in terms of considering the bending moments acting on the radius of the
matrix rounding, which should result in greater adequacy of solutions.
References
1. Golovlev, V.: The initial stage of drawing sheet metal. Forging and stamping production.
Mater. Work. Press. 7, 4–10 (1995)
2. Kalyuzhnyi, O.: Intensification of forming processes of cold sheet stamping. Sik Group
Ukraine LLC, Kyiv (2015)
3. Yan, G., Wang, X., Deng, L.: A study of hole flanging-upsetting process. Adv. Mater. Res.
939, 291–298 (2014). https://www.scientific.net/AMR.939.291
4. Luo, J.: Study on stamping-forging process and experiment of sheet metal parts with non-
uniform thickness. Huazhong Univ. Sci. Technol. 51, 45–54 (2011).
The Contact Pressure in Drawing Parts Without Clamping 19
5. Gavas, M., Izciler, M.: Effect of blank holder gap on deep drawing of square cups. Mater.
Des. 28, 1641–1646 (2007)
6. Puzyr, R., Savelov, D., Argat, R., Chernish, A.: Distribution analysis of stresses across the
stretching edge of die body and bending radius of deforming roll during profiling and drawing
of cylindrical workpiece. Metall. Min. Ind. 1, 27–32 (2015)
7. Popov, E.: Fundamentals of the theory of sheet punching. Engineering, Moscow (1977)
8. Volmir, A.: Stability of Deformable Systems. The Science, Moscow (1967)
9. Ogorodnikov, V.A., Dereven’ko, I.A., Sivak, R.I.: On the influence of curvature of the tra-
jectories of deformation of a volume of the material by pressing on its plasticity under the
conditions of complex loading. Mater. Sci. 54(3), 326–332 (2018). https://doi.org/10.1007/
s11003-018-0188-x
10. Salenko, Y., Puzyr, R., Shevchenko, O., Kulynych, V., Pedun, O.: Numerical simulation of
local plastic deformations of a cylindrical workpiece of a steel wheel rim. In: Ivanov, V., et al.
(eds.) DSMIE 2019. LNME, pp. 442–451. Springer, Cham (2020). https://doi.org/10.1007/
978-3-030-50794-7_43
11. Yang, C., Li, P., Fan, L.: Blank shape design for sheet metal forming based on geometrical
resemblance. Procedia Eng. 81, 1487–1492 (2014)
12. Anishchenko, O., Kukhar, V., Grushko, A., Vishtak, I., Prysiazhnyi, A., Balalayeva, E.: Analy-
sis of the sheet shell’s curvature with lame’s superellipse method during superplastic forming.
Mater. Sci. Forum 945, 531–537 (2019).https://www.scientific.net/MSF.945.531
13. Borisevich, V., Zahirnyak, M., Drabobetsky, V.: Choosing the optimal parameters for
extracting sheet blanks. Forging and stamping production. Mater. Work. Press. 2, 38–41
(2009)
14. Sosenushkin, E., Yanovskaya, E., Sosenushkin, A., Emel’yanov, V.: Mechanics of nonmono-
tonic plastic deformation. Russ. Eng. Res. 35(12), 902–906 (2015). https://doi.org/10.3103/
S1068798X15120199.
15. Guo, Y., Batoz, J., Naceur, H., Bouabdallah, S., Mercier, F., Barlet, O.: Recent developments
on the analysis and optimum design of sheet metal forming parts using a simplified inverse
approach. Comput. Struct. 78, 133–148 (2000). https://doi.org/10.1016/S0045-7949(00)000
95-X
16. Sivak, R.: Evaluation of metal plasticity and research of the mechanics of pressure treatment
processes under complex loading. East.-Eur. J. Enterp. Technol. 6/7(90), 34–41 (2017). https://
doi.org/10.15587/1729-4061.2017.115040
17. Anishchenko, A., Kukhar, V., Artiukh, V., Arkhipova, O.: Application of G. Lame’s and J.
Gielis’ formulas for description of shells superplastic forming. MATEC Web of Conf. 239,
06007 (2018). https://doi.org/10.1051/matecconf/201823906007
18. Ogorodnikov, V., Savchinskij, I., Nakhajchuk, O.: Stressed-strained state during forming the
internal slot section by mandrel reduction. Russ. J. Heavy Mach. 12, 31–33 (2004)
19. Leyu, W., Daxin, E.: Numerical simulation analysis of variable BHF drawing of rectangular
cup on curve blank-holder. Mod. Manuf. Eng. 2, 73–74 (2006)
20. Wang, X., Ouyang, K., Xia, J.: FEM analysis of drawing-thickening technology in stamping-
forging hybrid process. Forg. Stamp. Technol. 34(4), 73–78 (2009). https://doi.org/10.3969/
j.issn.1000-3940.2009.04.017
21. Akbari-Mousavi, S., Barrett, L., Al-Hassani, S.: Explosive welding of metal plates. J. Mater.
Process. Technol. 202(1–3), 224–239 (2008)
22. Del’, G., Tomilov, F., Ogorodnikov, V.: Stressed state during the hot extrusion of steel. Steel
USSR 5(4), 223–224 (1975)
23. Comsa, D., Banabic, D.: Numerical simulation of sheet metal forming processes using a new
yield criterion. Key Eng. Mater. 344, 833–840 (2007)
20 R. Arhat et al.
24. Puzyr, R., Haikova, T., Trotsko, O., Argat, R.: Determining experimentally the stress-strained
state in the radial rotary method of obtaining wheels rims. East.-Eur. J. Enterp. Technol.
4(1(82)), 52–60 (2016). https://doi.org/10.15587/1729-4061.2016.76225
25. Puzyr, R., Savelov, D., Shchetynin, V., Levchenko, R., Haikova, T., Kravchenko, S., Yasko, S.,
Argat, R., Sira, Y., Shchipkovakyi, Y.: Development of a method to determine deformations
in the manufacture of a vehicle wheel rim. East.-Eur. J. Enterp. Technol. 4(1(94)), 55–60
(2018). https://doi.org/10.15587/1729-4061.2018.139534
26. Renne, I., Ogorodnikov, V., Nakhaichuk, V.: Plotting plasticity diagrams by testing cylindrical
samples in combined tension and torsion. Strength Mater. 8(6), 733–737 (1976)
27. Mori, K., Nakano, T.: State-of-the-art of plate forging in Japan. Prod. Eng. Res. Devel. 10(1),
81–91 (2015). https://doi.org/10.1007/s11740-015-0648-1
28. Zagirnyak, M., Drahobetskyi, V.: New methods of obtaining materials and structures for light
armor protection. In: International Conference 2015, ICMT, vol. 1, pp. 705–710 (2015)
29. Ogorodnikov, V.A., Grechanyuk, N.S., Gubanov, A.V.: Energy criterion of the reliability of
structural elements in vehicles. Mater. Sci. 53(5), 645–650 (2018). https://doi.org/10.1007/
s11003-018-0119-x
30. Harpell, E., Worswick, M., Finn, M., Jain, M., Martin, P.: Numerical prediction of the limiting
draw ratio for aluminum alloy sheet. J. Mater. Process. Technol. 100, 131–141 (2000). https://
doi.org/10.1016/S0924-0136(99)00468-9
31. Del, G., Ogorodnikov, V., Nakhaichuk, V.: Criterion of deformability of pressure shaped
metals. Notification High. Educ. Inst. Mech. Eng. 4, 135–140 (1975)
32. Zagirnyak, M., Kalinov, A., Ogar, V., Lotous, V.: Experimental assessment of the accuracy
of the method for determination the power on an induction motor shaft. In: 18th International
Conference 2017, CPEE (2017). https://doi.org/10.1109/CPEE.2017.8093054
33. Markov, O., Zlygoriev, V., Gerasimenko, O., Hrudkina, N., Shevtsov, S.: Improving the quality
of forgings based on upsetting the workpieces with concave facets. East.-Eur. J. Enterp.
Technol. 5/1(95), 16–24 (2018). https://doi.org/10.15587/1729-4061.2018.142674
34. Kukhar, V., Yelistratova, N., Burko, V., Nizhelska, Y., Aksionova, O.: Estimation of occupa-
tional safety risks at energetic sector of Iron and Steel Works. Int. J. Eng. Technol. (UAE)
7(2), 216–220 (2018). https://doi.org/10.14419/ijet.v7i2.23.11922
35. Markov, O.E., Perig, A.V., Zlygoriev, V.N., Markova, M.A., Kosilov, M.S.: Development
of forging processes using intermediate workpiece profiling before drawing: research into
strained state. J. Braz. Soc. Mech. Sci. Eng. 39(11), 4649–4665 (2017). https://doi.org/10.
1007/s40430-017-0812-y
36. Anishchenko, A., Kukhar, V., Artiukh, V., Arkhipova, O.: Superplastic forming of shells
from sheet blanks with thermally unstable coatings. In: MATEC Web of Conference, vol.
239, p. 06006 (2018). https://doi.org/10.1051/matecconf/201823906006
37. Sotiria, H., Spyros, A.: Stability of long transversely-isotropic elastic cylindrical shells under
bending. Int. J. Solids Struct. 47, 10–24 (2010). https://doi.org/10.1016/j.ijsolstr.2009.09.004
38. Logesh, K., Raja, V., Krishnan, D., Azeemudeen, M., Andrew, A., Hariprasath, K.: Experi-
mental investigation and analysis of tensile and forming behaviour of cold rolled sheet metal.
Int. J. Mech. Eng. Technol. (IJMET) 8(11), 252–264 (2017)
Complex Recognition Approach for Cutting Part
of Cutters in Finishing Turning
Abstract. In the conditions of finishing and precision turning, the traditional app-
roach to laboratory assessment of the condition of cutters by periodically recording
the parameters of the wear zone along the flank surface and subsequent recognition
is, in the authors’ opinion, insufficiently effective. It does not consider significant
changes (due to wear) in the geometry of the cutting edges and, in particular, in
the forming of their sections, the state of which directly affects the quality of the
processed surface. Therefore, there is a need for complex control and complex
recognition of cutting part states. The article aims to develop an approach to com-
plex recognition of cutters cutting part in finishing turning. The scientific novelty
consists of creating classifiers for complex recognizing the states of cutters for fin-
ishing turning, using the most informative features of the shape of all wear zones,
and analyzing their effectiveness. The research was carried out under conditions
of processing hardened steel 115CrV3 on a lathe model TPC - 125 BH1P. On a
special laboratory stand, equipped with a vision system, comprehensive periodi-
cal monitoring of the condition of the cutting part of the cutters for finishing was
carried out. Practical usefulness consists of developing a method for predicting the
residual life of cutters for finishing turning, using a set of features characterizing
the shape and size of defects and microdefects of all wear surfaces of the cutting
part. Timely replacement of a failed tool with a new one provides a significant
economic effect.
1 Introduction
In the conditions of finishing and precision turning, the traditional approach to laboratory
assessment of the condition of cutters by periodically recording the parameters of the
wear zone along the flank face and subsequent recognition is, in the authors’ opinion,
insufficiently effective.
Nowadays, in cutting tools (CT) technical diagnostics, a trend is being developed
based on smart diagnostic systems. These systems can create, compare, and transform
CT models of a given subject area to make decisions based on accumulated knowledge
and without the operator’s participation in this process. The cutting part (CP) of any CT
is characterized by a certain structure and geometry determined by the tasks required
by the processing results. We can say that the working part of the cutting edge (CE) is a
variable structure system. It does not take into account significant changes (due to wear)
in the geometry of the cutting edges and, in particular, in the forming section of the CE,
the state of which directly affects the quality of the processed surface. Therefore, there
is a need for complex control and complex recognition of CP states.
The connection of the wear zones of the flank and rake face leads to significant
gradual changes in the shape and geometry of the working sections of the CP.
The development carried out was focused on recognizing the states of the worn
flank face of tools. However, in finishing and precision machining conditions, it is also
necessary to recognize the rake surface and CE conditions.
2 Literature Review
The 5G technology in the next future will have a major impact on the industry [1]. This
will enable manufacturers to complete end-to-end automation with the virtual deploy-
ment of new product lines or the entire factory. 5G will enable growth and transformation
in Industry 4.0. In production conditions of Industry 4.0 became an actual new technolog-
ical process. For example, in [2] is proposed progressive manufacturing process, based
on the concept of intensification of machining and application of multiaxis equipment. It
made the possibility to reduce the complexity of the manufacturing process (in drilling,
milling, and boring operations).
One major challenge is designing the strategy for communication between the factory
modules and the identification of adequate communication technologies [3].
Now the tasks of intelligent forecasting of the operability of modern technical systems
are becoming relevant. For example, deep learning model creation is used for predicting
the remaining useful life of the machining tools [4].
An artificial neural network model is developed to predict the main cutting force, and
its ability to predict cutting force was analyzed in [5]. An effort is made to optimize the
cutting parameters to accomplish minimum cutting force using a genetic algorithm. In
work [6] features, obtained by processing the cutting force, vibration signal, and surface
texture of the machined surface in turn, which are found by tool condition monitoring,
are used to estimate cutting tools states.
Cutting tool failure in the manufacturing industry causes damage to the cutting tools;
leads to serious accidents [7]. So, a wear monitoring system is required to mitigate these
negative effects in the metal cutting manufacturing process. More and more diverse
mathematical models are beginning to be used in the metalworking industry when cre-
ating mathematical and software for monitoring systems of machine tools and tools. In
particular, the application of discrete and continuous Markov chains is promising [8].
It is especially noteworthy is the development of new promising strategies in CAD /
CAM systems [9] and their use in diagnostic and monitoring systems for modern auto-
mated manufacturing [10].
An important role in data collections and modern computing technology are playing
cloud computing and cloud manufacturing (CM) [11–13]. CM as a new technological
paradigm has been attracted a large amount of research interest. When solving various
diagnostic problems, such intelligent technologies as pattern recognition [14–16], the
Another random document with
no related content on Scribd:
reported in the afternoon that they had discovered Longa’s column
on the Murguia road, it was decided that a flank guard must be
thrown out in that direction, to cover Vittoria and the high road to
France from possible raids. Wherefore Reille was told that it would
be his duty to provide against this danger. He took out Menne’s
brigade of Sarrut’s division from the line, pushed it over the Zadorra,
and established it in a position level with Aranguiz on the Murguia
road, a mile and more beyond the river. Late at night, apparently on
a report from a deserter that there were British troops behind
Longa[553]—(indeed the man said that Wellington himself was on the
Bilbao road)—Reille took off the other brigade of Sarrut’s division in
the same direction, and sent it with Curto’s regiments of light horse
to join Menne. There remained in front of Vittoria of infantry only
Lamartinière’s division, and the King’s Guards; but there was a very
large body of cavalry in reserve—Tilly’s and Digeon’s dragoon
divisions of the Army of the South, the King’s two Guard regiments of
lancers, and the bulk of the horse of the Army of Portugal, Boyer’s
dragoons and Mermet’s chasseurs: there must have been 5,000
sabres or more in hand.
One further precaution was taken—lest the enemy might be
moving against the high road to France from points even more to the
north than Murguia, a trifling force was sent to cover the exit of the
chaussée from the pass of Salinas; this was the Spanish ‘division’ of
Casapalacios belonging to the Army of the Centre, strengthened by
some scraps of the French troops of the Army of the North, which
had come in from minor garrisons during the recent retreat.
Casapalacios’ Afrancesados, nominally three regiments strong, were
under 2,000 bayonets—they had with them five weak squadrons of
their own nation, a half-battery, also Spanish, and an uncertain (but
small) French auxiliary force, which included a battalion of the 3rd
Line, part of the 15th Chasseurs, and a section of guns equipped
from the artillery dépôt of the Army of the North, which had long
been established at Vittoria and had not been sent to the rear[554].
Casapalacios took post at Durana, covering the bridge of that village,
the most northerly one on the Zadorra: the French battalion was at
Gamarra Menor on the other side of the water.
Wellington spent the 20th in arranging for the general attack,
which he had determined to deliver if Joseph stood his ground. The
plan was ambitious, for the battlefield was far larger than any on
which the Anglo-Portuguese Army had ever fought before, and the
numbers available were 30,000 more than they had been at
Salamanca or Bussaco, and more than double those of Fuentes de
Oñoro. Moreover, he intended to operate with a great detached
turning force against the enemy’s flank and rear, a thing that he had
only once done before—at Salamanca—and then only with a single
division. The battle plan was essentially a time-problem: he had to
arrange for the simultaneous appearance of four separate masses in
front of the French position. All started from parallel points in the
valley of the Bayas, but the obstacles in front of them were of very
varying difficulty, and the distances to be covered were very different.
(1) Hill, with the large 2nd Division (four brigades), Silveira’s
Portuguese division, Morillo’s Spaniards, and V. Alten’s and Fane’s
light dragoons, 20,000 sabres and bayonets in all, was to cross the
Zadorra completely outside the extreme French left wing, to storm
the heights of Puebla, which formed the end of Gazan’s line, and
then, advancing from the defile, to strike at Subijana de Alava with
his main body, while continuing to thrust his flank along the heights
above, which dominated the whole region, and extended far behind
the enemy’s left wing.
(2) Two parallel columns were to march from the camps on the
Bayas; one consisting of the 4th and Light Divisions, R. Hill’s,
Grant’s, and Ponsonby’s cavalry brigades, and D’Urban’s
Portuguese horse, was to advance by the country road from
Subijana-Morillas to the two bridges of Nanclares. Opposite those
passages and the neighbouring ones it was to deploy, and to attack
the French centre, when Hill should have got a footing on the Puebla
heights and in the plain by Subijana. The second column, consisting
of the 3rd and 7th Divisions, was to move from Zuazo and Anda on
the Bayas across the high mountain called Monte Arrato by a
country track, and to descend into the valley of the Zadorra at Las
Guetas, opposite the bridge of Mendoza, up-stream from the ‘hairpin
bend’. In this position it would be almost in the rear of Gazan’s right
wing above Villodas: it was to attack at once on reaching its ground,
if the progress of Hill on the south was seen to be satisfactory. ‘The
movement to be regulated from the right: although these columns
are to make such movements in advance as may be evidently
necessary to favour the progress of the two columns on their right,
they are not to descend into the low ground toward Vittoria or the
great road.’ The total strength of the four divisions and cavalry of this
section of the army was about 30,000 sabres and bayonets.
(3) Graham’s column-head was at Olano, three miles in front of
Murguia, and six from the Zadorra. He had in front Longa’s Spanish
infantry, with whom Anson’s light dragoons were to join up when
operations should begin. Behind were the 1st and 5th Divisions,
Pack’s and Bradford’s Portuguese brigades, and Bock’s heavy
dragoons. The total force was about 20,000 men of all nations. As an
afterthought on the 20th Wellington directed Giron, who was to reach
Orduña that day, to come down to the upper Bayas in Graham’s rear,
where he would act as a support if necessary. The object of this
order is not quite obvious. Giron came down too late for the battle,
arriving at Murguia only in the afternoon. It is known from his own
dispatches that Wellington over-estimated Reille’s force, which
Graham had to fight, not knowing of Maucune’s departure[555], and
Giron may have been intended to add weight to the attack in this
quarter. His troops would, apparently, have been more usefully
placed if they had been sent from Murguia to attack the unguarded
upper fords of the Zadorra.
The orders issued to Graham gave him a rather perplexing
choice of action. He was (like the 3rd and 7th Divisions) to guide
himself by what was going on upon his right: he must get in touch at
once with the centre columns; he might attack if it was obviously
profitable to the main advance, but he was to avoid letting his whole
corps be drawn into close action in front of Vittoria, for his main
object must be to turn the enemy’s position, by getting round its right
wing and cutting the great road to France. The lack of precise
direction in this order is, no doubt, a testimony to Wellington’s
confidence in Graham’s judgement. But it cast a grave responsibility
on him: if he had been told simply that he was to turn the French
right and seize the great chaussée, matters would have been simple.
But he is given leave to attack frontally if circumstances farther down
the line seem to make such a policy desirable: yet he must not attack
so heavily as to make his great turning movement impossible. It
must be confessed that the difficult problem was not well solved that
day by the gallant old general.
The whole day of the 20th was spent in getting the columns into
order, and the arrangements for the attacks synchronized.
Wellington took a survey of all the routes in person, as his wont was.
It was not till late in the afternoon that he became certain from the
dispositions of the enemy that an eleventh hour retreat was not
contemplated by the King[556]. The timing was that Hill, who had a
few miles to march, should attack at eight in the morning, that both
Graham and the column consisting of the 3rd and 7th Divisions
should get into position by the same hour, and make ready to attack,
when it was clear that the flank movement of Hill had already begun
and was making good progress. Meanwhile the other central column,
the Light and 4th Divisions, should cross when Hill had won the
defile of Puebla and room to debouch beyond it, but not before. The
rather late hour fixed, in a month when dawn comes at 4 a.m., was
dictated by the fact that Hill had a river to cross, and the 3rd and 7th
Divisions mountain tracks to follow, which neither could have
negotiated in the dark. Some hours of daylight were needed to get
them into position. Even so the brigades of the left-centre column
were several hours late at their rendezvous.
The French commanders, as we have seen, made no move on
the 20th, save to hurry off convoys and to throw back Reille on the
north-west flank of Vittoria, to protect the royal chaussée and the two
Bilbao roads. Jourdan wrote to Clarke that evening a perfectly
sensible dispatch as to the difficulties of the situation, but one which
showed that he was wholly unaware that he might be forced to a
general action within the next twelve hours. All depended, he said,
on Clausel’s prompt arrival: if he should come up on the 21st there
was no danger: if he delayed, the army might have to choose
between two tiresome alternatives—a retreat on the pass of Salinas
and one on Pampeluna. If Wellington, as he suspected, was making
a great turning movement by Orduña and Durango, the army would
be forced to take the wretched road Salvatierra-Pampeluna. ‘The
King does not yet know what decision he will have to make: if
General Clausel delays his junction much longer, he will be forced to
choose the retreat into Navarre.’ Of any idea that Wellington was
about to attack next morning the dispatch shows no sign.
Jourdan had been indisposed all that day—he was laid up in bed
with a feverish attack. This caused the postponement of a general
reconnaissance of the position, which he and his King had intended
to carry out together. They rode forth, however, at 6 a.m. on the 21st.
‘No intelligence had come to hand,’ writes Jourdan in his memoirs,
‘which could cause us to foresee an instant attack. We arrived at the
position by Zuazo, from which Count Reille had been recently
moved, and stopped to examine it. Its left rested on the mountain-
chain in the direction of Berostigueta, its right came down to the
Zadorra behind La Hermandad and Crispijana. It is dominating, yet
not over steep, and has all along it good artillery emplacement for
many batteries. It connected itself much better than does the Ariñez
position with the ground about Aranguiz now occupied by Reille,
which might be heavily attacked. Struck with the advantages of this
position, the King appeared inclined to bring back the Army of the
South to occupy it, and to place the Army of the Centre between it
and the Army of Portugal. The three armies would have been closer
together, and more able to give each other rapid assistance; and the
eye of the Commander-in-Chief could have swept around the whole
of this more concentrated field in one glance. This change of
dispositions would probably have been carried out on the 20th if
Marshal Jourdan had not been indisposed, and might perhaps have
prevented the catastrophe that was to come. But the officer sent to
call General Gazan to confer with the King came back to say that the
general could not leave his troops, as an attack on him was
developing. It was judged too late to change position. Riding to the
front of Ariñez, to a long hill occupied by Leval’s division, on the right
of the Army of the South, the King saw that in fact the enemy was on
the move. About eight o’clock the posts on the mountain reported
that the Allies had passed the Zadorra above La Puebla: a strong
column was coming up the high road and the defile, a smaller one
had diverged to its right and was climbing the mountain itself.
General Avy, sent on reconnaissance beyond the Zadorra on the
side of Mendoza, reported at the same time that a large corps was
coming in on Tres Puentes behind the rear of General Leval. And he
could see signs in the woods opposite, which seemed to indicate the
march of other troops in the direction of the bridge of Nanclares. No
news had yet come in from Reille, but we prepared ourselves to hear
ere long that he, too, was attacked[557]’.
It is unnecessary to comment on the character of a Head-
quarters where a passing indisposition of the Chief of the Staff
causes all movements to be postponed for eighteen hours, at a
moment when a general action is obviously possible. But it is
necessary to point out that the existing dislocation of the troops left
three miles of the Zadorra front between Reille’s left and Gazan’s
right unguarded when the battle began, and that this had been the
case throughout the preceding day. And during that day who was
responsible for the fact that not a single one of the eleven bridges
between the defile of La Puebla and Durana had been ruined and
only three obstructed[558]? Gazan must take a good deal of
responsibility, no less than Jourdan.
At eight o’clock on the morning of the 21st the action
commenced, on a most beautiful clear day, contrasting wonderfully
with the heavy rain of the 19th and intermittent showers on the 20th.
Every hill-crest road and village for twenty miles was visible with
surprising sharpness, so much so that incidents occurring at a great
distance were easily to be followed by the naked eye, still more
easily by the staff officer’s telescope.
The clash began, as Wellington had intended, on the extreme
right. Here Hill’s columns crossed the Zadorra, where it broadened
out below the defile of La Puebla, far outside the French line. The
2nd Division led up the high road, with Cadogan’s brigade heading
the column; but before they entered the narrows, Morillo’s Spaniards
were pushed forward on their right, through a wood which covered
the lower slopes, to seize the spur of the Puebla heights immediately
above the road. Till this was cleared, it would be impossible to move
the 2nd Division forward. Eye-witnesses describe the deploying of
the Spanish column as clearly visible from the high ground on the
Monte Arrato heights: the first brigade appeared emerging from the
woods below, then came stiffer ground, ‘so steep that while moving
up it they looked as if they were lying on their faces or crawling[559]’
Then the smoke of the French skirmishing fire began to be visible on
the crest, while Morillo’s second brigade, deployed behind the first,
went up the heights in support. The enemy, who was not in great
strength at first, gave way, and the Estremadurans won the sky-line,
formed there, and began to drive up from the first summit to the next
above it. This also they won, but then came to stand—Gazan had
pushed up first one and then the other of the two regiments of
Maransin’s brigade from his left, on the main position, to support the
voltigeur companies which had been the only troops originally placed
on this lofty ground. Hill, seeing the advance on the mountain stop,
sent up, to help Morillo, Colonel Cadogan, with his own regiment the
71st, and all the light companies of Cadogan’s and Byng’s brigades.
This turned the fight, and after a stiff struggle for the second summit,
in which Morillo was severely wounded, Maransin’s brigade was
turned back and flung down hill: it halted and re-formed low down on
the slope, losing the crest completely.
Having his flank now reasonably safe, Hill turned off the other two
battalions of Cadogan’s brigade to follow the 71st, and then pushed
his second British brigade (O’Callaghan’s) up the defile of La Puebla,
and deployed it on the open ground opposite the village of Subijana
de Alava, which lay near the high road, and was the first obstacle to
be carried if the whole corps was to issue forth and attack the French
left. One battery moved up with the brigade, and got into action on a
slope to its right. The rest of the 2nd Division and Silveira’s
Portuguese issued from the defile, ready to act as reserve either to
Morillo on the heights or O’Callaghan in the open ground. Meanwhile
the enemy could be seen dispatching troops from his reserves, to
attack the spurs of the mountain which had been won by Morillo and
Cadogan. For the heights of Puebla commanded all the left flank of
the main French position, and, if Allied troops pushed along them
any further, Gazan’s line would be completely turned. Jourdan says
that his orders were that Maransin’s brigade should have attacked,
with a whole division in support, but that Gazan took upon him the
responsibility of sending in Maransin alone, and only later, when the
latter had been beaten down from the crest, first Rey’s brigade of
Conroux’s division and then St. Pol’s reserve brigade of Daricau’s
division, from the hill on the right behind Leval. It would seem that
Daricau’s two regiments took post on the slopes behind Subijana,
where there was a gap in the line, owing to Maransin’s departure,
while Rey and his brigade went up the Puebla heights, to try to head
off Morillo’s and Cadogan’s attack.
In this it was wholly unsuccessful: after a severe struggle on the
crest, in which Cadogan was mortally wounded[560], the French
reinforcements were checked and routed: the Allied troops began to
push forward again on the heights, and were getting right round the
flank of Gazan’s left wing, while Maransin’s troops on the lower
slopes were being contained by the 92nd and 50th, the rear
regiments of Cadogan’s brigade, and Daricau’s were hotly engaged
with O’Callaghan’s three battalions, which occupied Subijana and
then attacked the hillside above it, but failed for some time to secure
a lodgement there.
Jourdan was by this time growing anxious, and not without
reason, at the rapid progress of the Allies on the Puebla heights. He
ordered Gazan to send up at once his only remaining reserve,
Villatte’s division from the height behind Ariñez, to gain the crest at a
point farther east than any that the enemy could reach, and to attack
in mass along it. In order that he might not be forestalled on the
summit, Villatte was told to march by a long détour, through the
village of Esquivel far to the east, where there was a country road
debouching from the chaussée. Indeed so perturbed was the
Marshal by the threat to his left, that he suspected further turning
movements in this quarter, and sent orders for Tilly’s dragoons, from
the cavalry reserve, to ride out by Berostigueta to the Trevino road,
to see if there were no British columns pressing in from that quarter.
He also directed D’Erlon to move one of his two infantry divisions,
Cassagne’s, in the same direction, to support Tilly if necessary.
Gazan, if we may credit his long and contentious report on the battle,
suggested to Jourdan that it was dangerous to disgarnish his centre,
and to push so many troops to his left, while it was still uncertain
whether Wellington’s column-heads, visible on the other side of the
Zadorra, were not about to move. Might not Hill’s attack be a feint,
intended to draw off the French reserves in an eccentric direction?
The Marshal, says Gazan, then declared in a loud voice, so that all
around could hear ‘that the enemy’s movements opposite our right
are mere demonstrations, to which no attention should be paid, and
that if the battle were lost it would be because the heights oi the left
of Subijana remained in the enemy’s power[561].’ Then lay the real
danger.
Orders were given that when Villatte should have got on the crest
of the mountain, and should be delivering his attack, a simultaneous
move forward was to be made by Conroux, and the brigades of
Maransin and St. Pol, to cast the enemy out of Subijana, as well as
off the heights of Puebla. It is interesting to note that Hill had as yet
only engaged two British brigades, and one small Spanish division,
and had succeeded in attracting against himself a much larger force
—the whole of Conroux’s and Villatte’s divisions, and the two
brigades of Maransin and St. Pol. In the French centre and right
there remained now only Leval’s division and the remaining brigade
of Daricau’s, and in reserve there was only one of D’Erlon’s
divisions, since the other had gone off on a wild goose chase toward
the Trevino road. Wellington could have wished for nothing better.
But by the time that the French counter-attack on Hill’s corps was
developing, matters were beginning to look lively all along the line of
the Zadorra, and the combat on the heights was growing into a
general action. It was now about eleven o’clock, and Wellington had
been for some time established on a high bank above the river,
facing the centre of the French position, to the left of the village and
bridges of Nanclares, from which he could sweep with his glass the
whole landscape from the heights of La Puebla to the bridge of
Mendoza. To his left and right the Light and 4th Divisions lay in two
masses, a mile or more back from the river, and hidden very
carefully in folds of the Monte Arrato, the battalions in contiguous
close column lying down in hollow roads or behind outcrops of rock,
and showing as little as possible. The great mass of cavalry in
reserve, four brigades, had not been brought over the sky-line yet,
and lay some distance to the rear. Only Grant’s hussars were near
the Light Division, dismounted and standing to their horses in
covered ground. The French lines were perfectly visible, ‘unmasked,
without a bush to prevent the sweeping of their artillery, the charging
of their cavalry, or the fire of their musketry acting with full effect on
those who should attempt to cross the bridges in their front, which it
was necessary to carry before we could begin the attack on their
centre[562],’ King Joseph and his staff were conspicuous on the round
hill before Ariñez.
Hill’s advance having begun, and made good progress,
Wellington was watching for the two other movements which ought
to have coincided with the attempt to cross the river at Nanclares for
the frontal attack. They had neither of them developed as yet: at
least nothing was to be seen of the 3rd and 7th Divisions on the
down-slope of the lofty Monte Arrato, and there had not yet been any
heavy burst of cannonading from the upper Zadorra (which was not
visible from the spot where the Commander-in-Chief had placed
himself), to tell that Graham was engaged.
The reason for the comparative silence in this quarter was
undoubtedly the wording of Wellington’s orders to Graham, which left
so much to the judgement of the commander of the great turning
column. He had been placed, almost from the start of his march, in
the presence of a secondary problem. Reille, as we have seen, had
been given on the 20th the charge of the upper Zadorra and the
great road to France. Hoping that he had only Longa’s Spaniards in
front of him, and judging that it would be well to keep them as far
from the road and the river as possible, the commander of the Army
of Portugal had placed Sarrut’s division and a brigade of Mermet’s
light horse in an advanced position a mile and more in front of the
river, on a ridge flanking the main Bilbao road, above the village of
Aranguiz. Lamartinière’s division and Boyer’s and Digeon’s dragoons
had been left on the nearer bank. When therefore Graham, marching
down from Olano, neared Aranguiz, he found a considerable French
force blocking the way. Remembering his orders to look to the right,
to adapt his movements to those of the troops in that direction, and
not to be drawn into unnecessary fighting, he halted for some time,
to see how matters were going on the critical wing, and meanwhile
deployed alongside of Longa’s men in his front, Pack’s Portuguese
brigade and Anson’s light dragoons, with the 5th Division in support,
before Sarrut’s position. He also detached Bradford’s Portuguese to
his right, with the idea of getting in touch with the troops in that
direction. Reille, who was present in person with his advanced
guard, saw with dismay the depth of the column descending upon
him, and recognized that he must fall back and hold the line of the
Zadorra, the only possible front on which he could oppose such an
enemy. About midday or a little later Graham ‘thought himself
justified in advancing, in order to draw the enemy’s attention to his
right, and so assist the progress of the army from the side of Miranda
(i. e. Hill’s column), where the enemy seemed to be making an
obstinate resistance in the successive strong positions which the
country afforded[563].’ The very moment that Graham sent forth Longa
and Pack to advance, the French retired by order, not without some
skirmishing, in which the 4th Caçadores stormed the hill just above
Aranguiz. But neither side had any appreciable losses.
Graham could now advance to within a mile of the Zadorra, and
was in command of the plain-ground as far as the villages of
Abechuco, and Gamarra Mayor and Menor. All three are on the
northern or right bank of the river, and Reille had determined to hold
them as têtes de pont covering the bridges. They had been hastily
barricaded, and the artillery of the Army of Portugal had been placed
on the opposite bank in a line of batteries, ready to sweep the open
ground over which an assault on the villages must be launched.
Graham had therefore to deploy for a formal attack on the new
position. He sent Longa up-stream and over the hills, to attack
Gamarra Menor and Durana, placed Oswald and the 5th Division,
with a section of Lawson’s battery, over against Gamarra Mayor, and
drew out the 1st Division and the two Portuguese brigades opposite
Abechuco, which would have to be taken before the bridge of
Arriaga could be attacked. Keeping in mind Wellington’s main
purpose, indicated in his orders, of cutting the great road to France,
he told Longa and Oswald on the flank that they might push hard,
while he seems to have acted in a much more leisurely way in front
of Abechuco, where no attack was launched till actual orders had
been received from head-quarters bidding him press harder. But
meanwhile Longa took Gamarra Menor from the French battalion of
the 3rd Line, and then, pushing on, came into collision at Durana
bridge with his renegade compatriots the Franco-Spanish division of
Casapalacios. As the great road to Bayonne actually passes through
Durana, and was now under fire from Longa’s skirmishers, it may be
said to have been blocked for all practical purposes at this early
stage of the battle. It was not till the afternoon, however, that Longa
succeeded in storming the bridge and occupying the village, thus
formally breaking the enemy’s main line of communication with
France—to save which King Joseph had risked his all. Apparently he
was hampered by having no artillery, while the Franco-Spaniards
had some four or five guns with them, bearing on the bridge.
Meanwhile Robinson’s brigade of the 5th Division had stormed
Gamarra Mayor, defended by the French 118th and 119th—
Gauthier’s brigade of Lamartinière’s division. This was a brilliant and
costly affair—it being no light matter to attack in column of battalions
the barricaded streets of a compact village. The British, however,
burst in—Colonel Brooke with the 1/4th being the first to force an
entrance: the French abandoned three guns which had been placed
in the barricades, and fell back in disorder across the bridge.
General Robinson endeavoured to improve the success by instant
pursuit, but the French had guns bearing on the bridge, which swept
away the first platoons that tried to cross it. Very few men reached
the other side, and they were shot down before they could establish
a lodgement on the farther bank. It was necessary to halt, re-form,
and bring up more artillery before the attack could be repeated.
It was now past two o’clock, and the noise of Graham’s attack
was sufficiently audible all down the British line, and was carrying
dismay to French head-quarters. But what of the other column, that
of the 3rd and 7th Divisions, which was to appear on the middle
Zadorra opposite Gazan’s almost unguarded flank, in the direction of
the bridge of Mendoza? It was overdue, now that both the large
flanking corps were seriously engaged, and the attention of the
enemy attracted toward them. But before the missing column came
into action, there had been an unexpected modification of the
position in the right centre. At 11.30 Wellington appeared before the
Light Division, and told Alten to move it more to the left, so as to be
over and above the bridge of Villodas, which it would have to attack
at the moment of general advance, leaving the two bridges by
Nanclares to the 4th Division and the cavalry. So rough and wooded
was the ground that the division, moving in a hollow way, was
established less than 300 yards from the brink of the Zadorra and
the French line, without attracting the notice of the enemy. An officer
of the 43rd writes: ‘I felt anxious to obtain a view, and walking
leisurely between the trees found myself at the edge of the wood, in
clear sight of the enemy’s cannon, planted with lighted matches and
ready to apply them[564]. Had our attack begun here, the French
could never have stood to their guns, so near were they to the
thicket—our Riflemen would have annihilated them.’ The British bank
of the Zadorra here completely commanded the bridge and the
French bank, which accounts for the fact that the enemy’s artillery
did not detect the approach of such a large body of troops as the
Light Division. But after a time a bickering fire across the river
between skirmishers on both sides broke out at several points, and
some voltigeurs even pressed across Villodas bridge, and had to be
cast back again by the skirmishers of the 2/95th.
There was to be, however, no attempt to pass this bridge as yet
by the British. While Wellington was still with the Light Division, a
peasant came up to him with the astounding intelligence that the
bridge of Tres Puentes, the one at the extreme point of the ‘hairpin-
bend’ of the Zadorra, was not only unoccupied but unwatched by the
enemy: he offered to guide any troops sent to it. The Commander-in-
Chief made up his mind at once to seize this crossing, which would
outflank the French position at Villodas, and told the peasant to lead
Kempt’s brigade of the Light Division to the unwatched point, about a
mile and a half to the left. ‘The brigade moved off by threes at a rapid
pace, along a very uneven and circuitous path, concealed from the
observation of the French by high rocks, and reached the narrow
bridge, which crossed the river to the hamlet of Yruna (part of the
scattered village of Tres Puentes). The 1st Rifles led the way, and
the whole brigade following passed at a run, with firelocks and rifles
ready cocked, and ascended a steep road of fifty yards, at the top of
which was an old chapel. We had no sooner cleared it than we
observed a heavy column of the French on the principal hill, and
commanding a bird’s eye view of us. However, fortunately, a convex
bank formed a sort of tête de pont, behind which the regiments
formed at full speed, without any word of command. Two round shots
now came among us: the second severed the head from the body of
our bold guide, the Spanish peasant. The brigade was so well
covered that the enemy soon ceased firing. Our post was
extraordinary—we were in the elbow of the French position, and
isolated from the rest of the army, within 100 yards of the enemy[565]
and absolutely occupying part of his position, without any attempt
having been made to dislodge us.... Sir James Kempt expressed
much wonder at our critical position, without being molested, and
sent his aide-de-camp at speed across the river for the 15th
Hussars, who came up singly and at a gallop along the steep path,
and dismounted in rear of our centre. Some French dragoons, coolly
and at a slow pace, came up to within 50 yards of us, to examine, if
possible, our strength, but a few shots from the Rifles caused them
to decamp. We could see three bridges within a quarter of a mile of
each other, in the elbow of the enemy’s position. We had crossed the
centre one (Tres Puentes), while the other two, right and left
(Villodas and Mendoza), were still covered by French artillery[566].’
Expecting to be instantly attacked, and to have to fight hard for
the chapel-knoll on which they had aligned themselves, Kempt’s
brigade spent ‘half an hour of awkward suspense.’ The immunity
with which they had been allowed to hold their position was suddenly
explained by a movement which they had not been able to observe.
The missing column of Wellington’s army had at last come up, and
was plunging with headlong speed into the rear of the troops which
were facing Kempt, so that the French had no attention to spare for
the side-issue in the hairpin-bend of the Zadorra. Instead of being
attacked the brigade was about to become at once an attacking
force.
A word is necessary as to the leading of this column. It had been
placed by Wellington under Lord Dalhousie, now commanding the
7th Division. This was an extraordinary choice, as this officer had
been only a few months in the Peninsula, and had no experience of
the higher responsibilities—though he had commanded a brigade in
the Walcheren expedition[567]. But being as Lieutenant-General
slightly senior to Picton (though they had been gazetted major-
generals on the same day in 1808), he was entitled to take the
command over the head of the war-worn and experienced leader of
the 3rd Division. The latter had been directing one of the great
marching columns during the early stages of the advance, and was
not unnaturally sulky at being displaced. Common report in the army
held that he was in disfavour at Head-Quarters, for intemperate
letters complaining of the starving of his division during the recent
march beyond the Ebro[568]. Be this as it may, Picton was during the
forenoon hours of June 21st in one of his not infrequent rages. For
though his column had started early, and the 3rd Division had
reached Las Guetas, the villages on the south side of the Monte
Arrato, which were to be its starting-point for the attack on the line of
the Zadorra, Lord Dalhousie refused to advance farther than the
edge of the hills, using apparently his discretion in interpreting the
orders given him ‘to regulate his action from what was going on to
his right and only to move when it should be ‘evidently necessary’ to
favour the progress of the columns in that direction. He was
obviously worried by the fact that the two rear brigades of his own
division, Barnes’s and Le Cor’s, had been hindered by an artillery
breakdown in the steep road behind, and were not yet up, though
Cairnes’s battery, which had delayed them, ultimately overtook the
leading brigade[569]. Hence he used his discretion to wait for formal
orders from Head-Quarters, and to do nothing. Picton, who could
note the advance of Hill’s column, and could see that the French
were utterly unprepared for an attack on the middle Zadorra, chafed
bitterly at the delay.
We have an interesting picture of him on that morning from eye-
witnesses. He was a strange figure—suffering from inflammation of
the eyes, he had put on not his cocked hat but a broad-brimmed and
tall civilian top-hat—the same that may be seen to-day in the United
Service Museum. ‘During the struggle on the right the centre was
inactive. General Picton was impatient, he inquired of every aide-de-
camp whether they had any orders for him. As the day wore on, and
the fight waxed louder on the right, he became furious, and observed
to the communicator of these particulars, ‘D—n it! Lord Wellington
must have forgotten us.’ It was near noon, and the men were getting
discontented. Picton’s blood was boiling, his stick was beating with
rapid strokes upon the mane of his cob. He rode backward and
forward looking in every direction for the arrival of an aide-de-camp,
until at last one galloped up from Lord Wellington. He was looking for
Lord Dalhousie—the 7th Division had not yet arrived, having to move
over difficult ground. The aide-de-camp checked his horse and
asked the general whether he had seen Lord Dalhousie. Picton was
disappointed; he had expected that he might at least move now, and
in a voice which did not gain softness from his feelings, answered in
a sharp tone, ‘No, Sir: I have not seen his Lordship, but have you
any orders for me.’ ‘None,’ replied the aide-de-camp. ‘Then, pray Sir,
what are the orders that you do bring?’ ‘Why,’ answered the officer,
‘that as soon as Lord D. shall commence an attack on that bridge,’
pointing to the one on the left (Mendoza), ‘the 4th and Light are to
support him.’ Picton could not understand the idea of any other
division fighting in his front, and drawing himself up to his full height
said to the astonished aide-de-camp, ‘You may tell Lord Wellington
from me, Sir, that the 3rd Division, under my command, shall in less
than ten minutes attack that bridge and carry it, and the 4th and Light
may support if they choose.’ Having thus expressed his intention, he
turned from the aide-de-camp and put himself at the head of his
men, who were quickly in motion toward the bridge, encouraging
them with the bland appellation of ‘Come on, ye rascals! Come on,
ye fighting villains[570].’
Ten minutes as the time required to plunge down from the hillside
to a bridge two miles away seems a short estimate. But there is no
doubt that the advance of the 3rd Division was fast and furious—an
eye-witness describes it as shooting like a meteor across the front of
the still-halted column-head of the 7th Division. The military purist
may opine that Picton should have waited till he got formal orders via
Lord Dalhousie to advance. But the moments were precious—Kempt
was across the Zadorra close by, in an obviously dangerous state of
isolation: the French in a few minutes might be sending infantry to
block the bridge of Mendoza, which they had so strangely neglected.
The 7th Division was short of two brigades, and not ready to attack.
Wellington’s orders were known, and the situation on the right was
now such as to justify the permissible advance which they
authorized. Neither Wellington nor Dalhousie in their dispatches give
any hint that Picton’s action was disapproved—complete success
justified it.
Picton had directed Brisbane’s brigade of the 3rd Division straight
upon the bridge of Mendoza, Colville’s upon a ford 300 yards farther
up-stream. Both crossed safely and almost unopposed. The only
French troops watching the stream here were Avy’s weak brigade of
cavalry—under 500 sabres—and their three horse-artillery guns
commanding the bridge. But the latter hardly got into action, for on
Picton’s rapid approach becoming visible, General Kempt threw out
some companies of the 1/95th under Andrew Barnard, from his point
of vantage on the knoll of Yruna, who opened such a biting fire upon
the half battery that the officer in command limbered up and galloped
off. Avy’s chasseurs hovered about in an undecided way—but were
not capable either of defending a bridge or of attacking a brigade in
position upon a steep hill. Wherefore Picton got across with small
loss, and formed his two British brigades on the south side of the
river. Power’s Portuguese rapidly followed Brisbane, as did a little
later Grant’s brigade of the 7th Division—Lord Dalhousie’s other two
brigades (as we have already noted) were not yet on the ground. On
seeing Picton safely established on the left bank Kempt advanced
from his knoll, and formed on the right-rear of the 3rd Division. The
trifling French detachment at the bridge of Villodas—only a voltigeur
company—wisely absconded at full speed on seeing Kempt on the
move. The passage there was left completely free for Vandeleur’s
brigade of the Light Division, who had long been waiting on the
opposite steep bank.
The British were across the Zadorra in force, and the critical
stage of the action was about to commence: the hour being between
2 and 3 in the afternoon.
SECTION XXXVI: CHAPTER VIII
BATTLE OF VITTORIA. ROUT OF THE FRENCH
When Picton and the 3rd Division, followed by the one available
brigade of the 7th Division, came pouring across the Zadorra on the
side of Mendoza, while Kempt debouched from the knoll of Yruna,
and Vandeleur crossed the bridge of Villodas, the position of Leval’s
division became desperate. It was about to be attacked in flank by
four brigades and in front by two more, and being one of the weaker
divisions of the Army of the South, only 4,500 bayonets, was
outnumbered threefold. Its original reserve (half Daricau’s division)
had gone off to the Puebla heights hours before: the general army-
reserve (Villatte’s division) had been sent away in the same direction
by Jourdan’s last orders. The nearest disposable and intact French
troops were Darmagnac’s division of the Army of the Centre—two
miles to the rear, in position by Zuazo: the other divisions of that army
had (as we have seen) gone off on a wholly unnecessary excursion
to watch the Trevino road. The left wing and centre of the Army of the
South was absorbed in the task of keeping back Hill, and had just
begun the counter-attack upon him which Jourdan had ordered an
hour before.
The sudden change in the situation, caused by the very rapid
advance of Picton and the brigades that helped him, was all too
evident to King Joseph and his chief of the staff, as they stood on the
hill of Ariñez. The whole force of the 3rd division struck diagonally
across the short space between the river and Leval’s position—
Brisbane’s brigade and Power’s Portuguese making for the French
flank, while Colville, higher up the stream, made for the rear of the
hill, in the direction of the village of Margarita. Kempt followed
Brisbane in second line, Grant’s brigade of the 7th Division, when it
crossed at Mendoza, came on behind Colville. So did Vandeleur, from
Villodas, after he had pulled down the obstructions and got his men
over the narrow bridge. Nor was this all—the 4th Division, so long
halted on the scrubby hillside opposite the left-hand of the two
Nanclares bridges, suddenly started to descend the slope at the
double-quick, Stubbs’s Portuguese brigade leading.
Jourdan had to make a ‘lightning change’ in all his dispositions.
Leval, obviously doomed if he did not retire quickly, was told to
evacuate his hill and fall back past Ariñez, into which he threw a
regiment to cover his retreat, on to the heights behind it. The two
brigades of Daricau and Conroux, which had stood on the other side
of the high road from Leval, were to make a parallel movement back
to the same line of heights. The other brigades of Daricau and
Conroux, with Maransin—now deeply engaged some with
O’Callaghan and others with the 50th and 92nd—were to abandon
the attack which they had just begun, and which had somewhat
pushed back the British advance. They too must go back to the
slopes behind Ariñez. It would take longer to recall Villatte, who was
now far up on the crest of the mountain to the left, engaged with
Morillo’s Spaniards and the 71st. But this attack also must be broken
off. Lastly, to fill the gap between Leval’s new position and the
Zadorra, the Army of the Centre must come forward and hold
Margarita, or if that was impossible, the hill and village of La
Hermandad behind it. But only Darmagnac’s division was immediately
available for this task, Cassagne’s having to be brought back from the
eccentric counter-march toward the Trevino road, to which it had
been committed an hour before. In this way a new line of battle would
be formed, reaching from the Zadorra near Margarita across the high
road at Gomecha, to the heights above Zumelzu on the left. It was at
best a hazardous business to order a fighting-line more than two
miles long, and bitterly engaged with the enemy at several points, to
withdraw to an unsurveyed position a mile in its rear, where there was
practically no reserve waiting to receive it. For on the slopes above
Ariñez there was at that moment nothing but Pierre Soult’s light
horse, and Treillard’s dragoons, with two batteries of artillery[571]. The
new front had to be constructed from troops falling back in haste and
closely pursued by the enemy, combining with other troops coming in
from various directions, viz. the two divisions of the Army of the
Centre. And when the line should be re-formed, what was to prevent
its left from being turned once more by the Allied troops on the
Puebla heights, or its right by columns crossing the middle Zadorra
behind it[572]. For there would still be a gap of two miles between
Margarita village and the nearest troops of the Army of Portugal, who
were now engaged with Graham at Abechuco and Arriaga.
To speak plainly, the second French line was never properly
formed, especially on its left; but a better front was made, and a
stronger stand, than might perhaps have been expected, though the
confusion caused by hasty and imperfect alignment was destined in
the end to be fatal.
On the extreme left Villatte had been caught by the order of recall
at the moment when he was delivering his attack on Morillo and the
British 71st. He had reached the crest as ordered, had formed up
across it, and then had marched on a narrow front against the Allies.
Both British and Spanish were in some disorder when he came in
upon them—they had now been fighting for four hours, and in
successive engagements had driven first Maransin’s and then St.
Pol’s brigades for two miles over very steep and rocky ground. At the
moment when Villatte came upon the scene, the Allied advance had
just reached a broad dip in the crest, which it would have to cross if
its progress were to be continued. The Spaniards were on the right
and the 71st and light companies on the left, or northern, part of the
ridge. It would have been a suitable moment to halt, and re-form the
line before continuing to press forward over dangerous ground. But
the officer who had succeeded Cadogan in command[573] was set on
‘keeping the French upon the run,’ and recklessly ordered the tired
troops to plunge down the steep side of the declivity and carry the
opposite slope. He was apparently ignorant that fresh French troops
were just coming to the front—several eye-witnesses say that a
column in light-coloured overcoats with white shako-covers, which
had been noticed on the right, was taken for a Spanish
detachment[574]. At any rate, the 71st crossed the dip—four
companies in its centre, the remainder at its upper end—and was
suddenly met not only by a charging column in front, but by an attack
in flank and almost in rear. The first volley brought down 200 men—
the shattered battalion recoiled, and remounted its own slope in utter