Professional Documents
Culture Documents
net/publication/318339420
CITATIONS READS
0 774
5 authors, including:
Some of the authors of this publication are also working on these related projects:
INTELLIGENT SOFTWARE SYSTEM FOR 3, 4 AND 5-AXIS MACHINE TOOL AND INDUSTRIAL ROBOTS PROGRAMMING BASED UPON STEP-NC LANGUAGE View project
Research and development of new concept for connection between undercarriage frame and revolving upper structure in construction and transportation machines
View project
All content following this page was uploaded by Goran V. Pavlovic on 13 August 2017.
The paper >13@ carried out the analysis of the effects function, i.e. smaller girder mass is obtained for a larger
that have inertial forces on the stresses and deflections, by number of constraints.
FEM, where the derived certain conclusions for the Taking into account the above mentioned results and
optimization of these types of girders. conclusions, the aim of this paper is to define optimum
In addition to FEM application, is increasingly being values of parameters of the geometry of cross-sectional area
applied various analytical and numerical methods for of the rectangular box girder of the double beam bridge
optimization process. Multi-criteria optimization was crane that will lead to the reduction of its mass.
conducted in the paper >14@ for the purpose of reducing the 2. MATEMATHICAL FORMULATION OF THE
crane mass using Ms Excel software, and the solution was OPTIMIZATION PROBLEM
verified using Ansys software. The mass of the girder was
reduced about 10%. The task of optimization is to define geometrical
In the paper >15@, problem of reducing the weight of parameters of the rectangular box section of the girder
the box girder of the bridge crane, at the same time which result is its minimum mass.
increasing the productivity and to improve quality of Matlab Optimization Toolbox can be used to solve
product as required by the Indian standard, is tackled by the problems of linear programming, non-linear
preparing programme in Visual Basic 6.0. Optimal programming and multi-target programming.
dimensions of the box girder of the bridge crane are affected The optimization algorithm of the parameters of the
by choice of standards, as shown in the paper >16@, where cross-sectional area of steel rectangular box girder is to
the results obtained by Serbian standard and European utilizing the fmincon function in the Optimization Toolbox
standard, and compared, on examples of existing solutions to realize the minimization of the cross-sectional area of the
of single-girder bridge cranes. GRG2 algorithm in Ms Excel rectangular box girder, i.e. constrained minimization
software was used for the optimization process. problem.
Application of the centered differential evolution The optimization problem can be defined by the
(CDE) algorithm in the process of optimization of the box following mathematical formulation, >21@:
girder of the bridge crane is shown in >17@, where as min f ( X ) (1)
opposed to the previously mentioned papers, the cross- subject to:
section of the box girder is variable along the span. In this C( X ) d 0 (2)
way, the mass of the girder is decreased by about 23.50%. Ceq ( X ) 0 (3)
Application of Mathcad software in optimization of
A X d b (4)
the box girder of the bridge crane is shown in >18@, where is
graphically presented the dependence of the cross-sectional Aeq X beq (5)
area in relation of capacities of the cranes, for characteristic lb d X d ub (6)
spans. Similarly to previous, in >19@ is graphically presented where:
the dependence of the cross-sectional area in relation of the X , b, beq, lb, ub - vectors
spans, for characteristic capacities of the cranes. In >20@ is A, Aeq - matrices
also applied Mathcad for multi-criteria optimization and
C ( X ), Ceq ( X ) - functions that return to vectors
comparison of the results for the welded box girder, for the
cases when the objective function is the cross-sectional area f ( X ) - the target function (the objective function)
and the cost of the girder manufacturing. Functions f(X), C(X) and Ceq(X) can be nonlinear
The mentioned papers point to the importance of functions.
optimization of the main girder of the bridge crane and In Matlab Optimization Toolbox, the constrained
creation of the model which can allow a more real nonlinear optimization problem is realized by fmincon
description of the crane behaviour in operation. As it can be fuction. The specific forms are as followings:
seen in the mentioned papers, there are different constraint
functions so that it can be concluded that a better objective
> X , fval , exitflag , output , lambda, grad , hessian@ fmincon fun, X 0, A, b, Aeq, beq, lb, ub, nonlcon, options (7)
where: grad - shows the gradient of the target function at X
fval - objective function value at solution (returned as a hessian - shows Hessiab value of the target function at X
real number) X 0 - the initial value
exitflag - shows reason for stop of fmincon function A and b meet the requirements of the inequality
fun - the target function constraint A;ڄb, if there is no such constraint, then A=[ ]
and b=[ ].
fmincon - function which find minimum of constrained
Aeq and beq meet the requirements of the equality
nonlinear multi-variable function constraint AeqڄX = beq, if there is no such constraint, then
output - function which outputs optimization information Aeq=[ ] and beq=[ ].
nonlcon - function which calculates nonlinear inequality lb and ub meet the requirements OEX XE if there
constraint &; through the accepted vector X and the is no scope, suppose lb=[ ] and ub=[ ].
equality constraint Ceq (X) = 0 is used by estimated C and In the process of optimization, fmincon function
Ceq at the poison of X though the appointed function handle adopts gradient projection method. Gradient projection
lambda - the multiplier of Lagrange (it embodies which method is a method to directly deal with the nonlinear
constraint is valid) programming constraint.
It made a projection matrix according to the initial Ag - the area of the cross section of the rectangular box
points and boundary face where the initial points stand. girder (the objective function)
Repeat this procedure frequently so as to obtain the best X - the design vector made of five design variables
solution.
> x1 x2 x3 x4 x5 @ >b1 h t1 t2 y @
T T
X (11)
3. OBJECTIVE AND CONSTRAINT FUNCTIONS
Other important relations can be written in the
3.1. Objective function following way:
The objective function is represented by the area of b f b f ( X ) b1 t2 bS 4 x1 x4 bS 4 cm (12)
the cross section of the rectangular box girder (Figure 1). H H (X ) h 2 t1 x2 2 x3 (13)
The paper treats five optimization parameters (h, b1, t1, t2, y),
where:
where:
H - height of the girder
h - web height b f - flange width
b1 - distance between webs
The geometrical properties in the specific points of
t1 - flange thickness the rectangular box section (Figure 1) shall be determined
t2 - web thickness by well-known expressions ( I x , Wx1 , Wy1 , Wx 2 , Wy 2 , S x 2 ).
y - value, according to recommendations where:
I x - moment of inertia about x axis
hs
h1
x
than permissible stresses.
h2
bf F F
1 2
R q
Figure 1: The rectangular box section of the main girder of A 1 2 B
the bridge crane I II
0.5(L-e1)
The vector of the given parameters is:
G
x Q, L, mk , bk , ka , f y , bS , hS , Dt ... (8) 0.5e1 0.5L
where: L
Q - the carrying capacity of the crane Figure 2: Model of the main girder of the bridge crane
L - the span of the crane
R q L2
mk - the mass of the trolley M VI ( L e1 ) 2 J (14)
4 L 8
bk - the distance between the wheels of the trolley
bk
ka - the dynamic coefficient of crane load in the horizontal e1 (15)
2
plane, >22@ q 1.1 U g A (16)
f y - the minimum yield stress of the plate material
ª Rh qL º 2
Design Optimization of the Rectangular Box Section of the Double Beam Bridge Crane Using Matlab Optimization Toolbox
A.40 Proceedings of IX Triennial International Conference “Heavy Machinery-HM 2017”, Zlatibor, 28 June - 1 July 2017
f b , R1x N1 x f d (45)
f b , R1 y N1 y f d (46)
e1x 1 N , e1 y
4
1x 1 N 4
1y (47)
h
h1 (48)
y
where:
Design Optimization of the Rectangular Box Section of the Double Beam Bridge Crane Using Matlab Optimization Toolbox
A.42 Proceedings of IX Triennial International Conference “Heavy Machinery-HM 2017”, Zlatibor, 28 June - 1 July 2017
KV 1y - a buckling factor for zone 1 (determined using O2 y - non-dimensional plate slenderness, >23@
Figure 11, >23@) fy
c1r - the width over which the transverse load is distributed O2 y (77)
a
(equivalent to l1r ) KV 2 y V 2e
c2 r
The criterion of stability of the web plate in zone 2 is where:
fulfilled if the following condition is satisfied: KV 2 y - a buckling factor for zone 2 (determined using
e2 y Figure 11, >23@)
§ V S 2x ·
e2 x
§ VS2y · § V S 2 xV S 2 y ·
N 2 xN 2 y c2r - the width over which the transverse load is distributed
6
¨¨ ¸¸ ¨ ¸ ¨ ¸ d 1 (59)
© fb, R 2 x ¹ ¨ fb, R 2 y ¸ ¨ fb, R 2 x fb, R 2 y ¸
© ¹ © ¹ (equivalent to l2r )
§ h / 2 h2 · 3.2.3. The criterion of static deflection
V S 2x Q 1 ¨ V zV 1 V zH 2 ¸ (60)
© h/2 ¹ In order to satisfy this criterion, it is necessary that
static deflection in the vertical plane has the value smaller
Q 1 F1
VS2y (61) than the permissible one.
t 2 l2 r F1, st L3
2 hd 2 tgN 0.2 r 2 hd 2 tgN 0.1 Dt f st ª1 D (1 6 E 2 ) º¼ d f dop (78)
48 E I ¬
l2 r (62)
x
hd 2 hS t1 h1 (63)
F2, st
fb, R 2 x N 2 x f d (64) D 1 (79)
F1, st
fb, R 2 y N 2 y fd (65)
bk
e2 x 1 N , e2 y 4
1 N 4
(66) E (80)
2x 2y L
h ( y 1) f dop kL (81)
h2 (67)
y where:
where: f st - static deflection in the vertical plane,
V S 2 x , V S 2 y - design values of the compressive stresses in f dop - the permissible deflection in the vertical plane, >22@
zone 2 in the x and y directions, respectively D , E - the coefficients, >22@
f b , R 2 x , f b , R 2 y - critical stresses k - the coefficient which depends on the purpose of the
N 2 x , N 2 y - reduction factors for zone 2 in the x and y crane and control condition of the crane, >22@
directions, respectively, >23@ 3.2.4. The criterion of permissible period of oscillation
G11 - the deflection of the girder caused by the action of the adopted for mentioned classification class. Other input datas
unit force are shown in Table 1.
Td - the permissible time of damping of oscillation As initial conditions (vector X0) in the optimization
process are taken the dimensions of the rectangular box
(permissible period of oscillation), which depends on the section of existing solutions of the double beam bridge
purpose of the crane, >22@ cranes (shown in Table 1).
J d - the logarithmic decrement which shows the rate of As additional criteria for the optimization process are
damping of oscillation, which depends on the ratio between taken production feasibility (distance between the webs) and
the height of the girder H and the span L, >22@ the recommended minimum thicknesses of the plates.
Minimum thickness of the web plates is adopted to
4. NUMERICAL REPRESENTATION OF THE RESULTS
be 5 mm and minimum thickness of the bottom and top
OBTAINED
flange is adopted to be 6 mm, which are also the constraint
The optimization process is done by Optimization functions. In addition, as one more constraint function,
Toolbox module of Matlab software. minimum distance between the web plates is taken to be: 20
Existing solutions of the double beam bridge cranes cm, 25 cm and 30 cm. These values are shown by the
that were used as examples for the optimization process are minimum values (lower boundaries) of the optimization
in classification class 2m/M5, so that certain input parameters (vector lb).
parameters necessary for the optimization process are The boundaries for the optimization parameter y are
taken as recommended values (y = 3 y 5).
Table 1: Characteristics of existing solutions of the double beam bridge cranes
Q L mk Dt bk t1 t2 h b1 bf Ag
No. Manufacturer Material
(t) (m) (kg) (mm) (mm) (mm) (mm) (cm) (cm) (cm) (cm2)
1 Colpart 15 20.69 780 170 S355 1000 10 6 98 31 41 199.6
2 Colpart 10 23.3 690 140 S355 1000 10 6 98 30 40 185.6
3 Colpart 2x25 22.5 2x2920 230 S355 1350 12 8 147.6 41.8 52 360.96
Montavar
4 20 18.95 860 170 S235 1000 10 8 98 31 41 238.8
Lola
Table 2: The values of optimum parameters and Savings
t1 t2 h b1 bf Aopt Saving
No. y
(mm) (mm) (cm) (cm) (cm) (cm2) (%)
1 7 6 113.74 20 29.6 4.07 177.93 10.86
1 7 6 113.74 25.04 34.64 4.06 184.98 7.32
1 6 5 111.15 30 39.5 3.96 158.55 20.57
2 6 5 120.91 20 29.5 3.76 156.31 15.78
2 6 5 117.4 25 34.5 3.73 158.8 14.44
3 8 6 136.1 50 60.6 3.84 260.28 27.89
3 8 6 132.28 52.78 63.38 4.02 260.14 27.93
4 6 5 107.26 45.23 55.73 3.89 174.14 27.08
Table 2 shows the results of the optimization for optimization rather than the realization of the specific
four existing solutions of the double beam bridge cranes, calculation. The results of optimization (the main
according to euro codes and material saving at different paremeters of the cross-sectional area of the rectangular
initial conditions (the minimum value of b1). It can be seen box girder) represent the starting point for designers while
that in some cases of existing solutions of the double beam designing box girders.The justification of application of
bridge cranes the optimization process gives different Optimization Toolbox module of Matlab software was
results. The plate thickness values are rounded to whole checked on four solutions of the double beam bridge
numbers. The values with the greatest savings of some cranes which are in operation. The optimization task –
examples are bolded. minimization of the cross-sectional area was successfully
5. CONCLUSION realized, which is seen in the comparison of the obtained
results with the solutions made in practice. The application
Based on the optimization theory and the procedure of this method resulted in significant savings in the
for calculation of steel box girder of the bridge cranes, this material, within the range of 15.78 y 27.93%. Similar
paper combines the optimum design philosophy and the procedure can be carried out for the situation with two
design of the rectangular box girder. The cross-sectional rows of longitudinal stiffeners, which is the case for
area is optimized making full use of the optimization greater spans and load capacities.
function in the Matlab Optimization Toolbox. The result The conclusion is that further research should be
shows that the Matlab Optimization Toolbox is reliable, directed toward a multi-criteria analysis where it is
convenient and rapid. Therefore, it can be widely utilized necessary to include additional constraint functions, such
in the optimization of the cross-sectional area of steel box as: influence of manufacturing technology, types of
girders for similar constructions. material, conditions of crane control and operation,
Application of Matlab Optimizaiton Toolbox on material fatigue and economy. Previously mentioned is the
sections of the rectangular box girders of the bridge cranes basis for further research in order to save material, and
allows designers to concentrate their attention on the
Design Optimization of the Rectangular Box Section of the Double Beam Bridge Crane Using Matlab Optimization Toolbox
A.44 Proceedings of IX Triennial International Conference “Heavy Machinery-HM 2017”, Zlatibor, 28 June - 1 July 2017
also to minimize the cost of the girder manufacturing, [12] ù. Yildirim and E. Esim, “A New Approach for
>20@. The obtained results can be of great importance both Dynamic Analysis of Overhead Crane Systems Under
for crane designers and for the researchers dealing with Moving Loads”, Proceedings of the 12th Portuguese
similar optimization problems. Conference on Automatic Control “CONTROLO
2016”, Guimarães (Portugal), 14-16 September 2016,
ACKNOWLEDGEMENTS
pp. 471-481, (2016)
A part of this work is a contribution to the Ministry
[13] Y. He, Y. Zhang, B. Yang, S. Liu and D. Chen,
of Science and Technological Development of Serbia
“Finite element analysis in dynamic conditions of
funded Project TR 35038.
bridge crane beam”, Applied Mechanics and Materials,
REFERENCES Vol. 331, pp. 70-73, (2013)
[1] P.F. Liu, L.J. Xing, Y. L. Liu and J.Y. Zheng, [14] L.B. Gayatrimani and A.D. Lagad, “Design of
“Strength Analysis and Optimal Design for Main optimum cross section for EOT Crane Girder”,
Girder of Double-Trolley Overhead Traveling Crane International Engineering Research Journal (IERJ),
Using Finite Element Method”, J Fail. Anal. and Special Issue 2, pp. 4233-4237, (2015)
Preven, Vol. 14, pp. 76-86, (2014)
[15] A.M. Gohil and Y.D. Vora, “Bridge Girder Design of
[2] Q. Wang, X. Zhang and C. Sun, “Lightweight Design an EOT Crane Structure-A Cad Approach”, National
and Research of Box Girder for Double Girder Bridge conference on current trends in technology “NUCONE
Crane”, Applied Mechanics and Materials, Vols. 229- 2007”, Ahmedabad (Gujarat, India), 29 November-1
231, pp. 478-481, (2012) December 2007, pp. 256-259, (2007)
[3] J. Wang, W. Jan and X. Gui, “Application of Matlab [16] G. 3DYORYLü9 .YUJLü6 0LWURYLü0 6DYNRYLü and
Optimization Tool Box on Sections of Prestressed N. =GUDYNRYLü, “Optimization of the box section of the
Steel-concrete Composite Box Girder”, Advanced single-girder bridge crane by GRG algorithm
Materials Research, Vol. 366, pp. 177-180, (2011) according to domestic standards and eurocodes”, 7th
[4] A. Kumar and R.R. Arakerimath, “Numerical and International Scientific Conference On Defensive
experimental buckling analysis of crane girder”, Technologies “OTEH 2016”, Belgrade (Serbia), 6–7
IJRET: International Journal of Research in October 2016, pp. 211-217, (2016)
Engineering and Technology, Vol. 6, pp. 199-197, [17] G.N. Xu, X.F. Wang, J.K. Zhang and W.J. Qin,
(2016) “Optimization of the Multi-Sectioned Box Girder of
[5] '*ąVNDDQG&3\SQR³6WUHQJWKDQGHODVWLFVWDELOLW\ the Bridge Cranes”, Proceedings of the 2015
of cranes in aspect of new and old design standards”, International Conference on Energy and Mechanical
Mechanika, Vol. 17, No. 3, pp. 226-231, (2011) Engineering, Wuhan, Hubei (China), 17-18 October
2015, pp. 927-935, (2015)
[6] T. Yifei, T. Zhaohui, Y. Wei and Y. Zhen, “Research
on Energy-saving Optimization Design of Bridge [18] V. Antsev, A. Tolokonnikov and P. Kalabin,
Crane”, Eksploatacja i Niezawodnosc – Maintenance “Optimization of bridge type hoisting machines' metal
and Reliability, Vol. 15, pp. 449-457, (2013) structure”, ɂɡɜɟɫɬɢɹ ɌɭɥȽɍ Ɍɟɯɧɢɱɟɫɤɢɟ ɧɚɭɤɢ,
Vol. 4, (2009)
[7] D. Qin, Y. Wang, Q. Zhu, J. Chen and Z. Liu,
“Research and Practice for Structural Optimization [19] A. Tolokonnikov and P. Kalabin, “Optimization of
Design of Large-Scale Continuum”, 9th World geometrical parametres of main beams of bridge cranes
Congress on Structural and Multidisciplinary with belts of a different thickness”, ɂɡɜɟɫɬɢɹɌɭɥȽɍ
Optimization “WCSMO-9”, Shizuoka (Japan), 13-17 Ɍɟɯɧɢɱɟɫɤɢɟɧɚɭɤɢ, Vol. 2, pp. 53-57, (2009)
June 2011, (2011) [20] J. Farkas and K. Jármai, “Optimum Design of Steel
[8] J. Gailius “Tiltinio krano konstrukcijos mases Structures”, Springer, Heidelberg (Germany), (2003)
racionalizavimas”, Proceedings of the 17th Conference [21] T.F. Coleman and Y. Zhang, “Optimization Toolbox
for Junior Researchers “Science – Future of Lithuania” for use with Matlab - User’s Guide”, The MathWorks,
TRANSPORT ENGINEERING AND Natick (Massachusetts, USA), (2005)
MANAGEMENT, Vilnius (Lithuania), 8 May 2014,
[22] ' 2VWULü DQG 6 7RãLü, “Dizalice”, Institute for
pp. 243-2461, (2014)
Mechanization of the Faculty of Mechanical
[9] Z. Ning, “Structural Optimization Research on Girder Engineering of the University in Belgrade, Belgrade
of 200t Bridge Crane Based on Ansys”, Advanced (Serbia), (2005)
Materials Research, Vols. 430-432, pp. 1708-1711,
[23] prEN 13001-3-1, “Cranes - General Design - Part 3-1:
(2012)
Limit States and proof competence of steel structure”,
[10] ɂȺ Ʌɚɝɟɪɟɜ ³ȼɥɢɹɧɢɟ ɪɚɡɦɟɪɚ ɤɨɧɟɱɧɨɝɨ (2010)
ɷɥɟɦɟɧɬɚ ɧɚ ɬɨɱɧɨɫɬɶ ɫɬɚɬɢɱɟɫɤɨɝɨ ɪɚɫɱɟɬɚ
[24] D. 2VWULü, “Dinamika mosnih dizalica”, Faculty of
ɧɟɫɭɳɟɣ ɦɟɬɚɥɥɨɤɨɧɫɬɪɭɤɰɢɢ ɦɨɫɬɨɜɨɝɨ ɤɪɚɧɚ´
Mechanical Engineering of the University in Belgrade,
ɇɚɭɤɚɢɫɨɜɪɟɦɟɧɧɨɫɬɶ– 2014, pp. 36-40, (2014)
Belgrade (Serbia), (1998)
[11] T. Haniszewski, “Strength analysis of overhead
traveling crane with use fem”, Transport problems,
Vol. 9, pp. 19-26, (2014)