Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

Advances in Production, Automation and Transportation Systems

Economic Lot Sizing for a Multi-Operation Machining Process


Allowing for Parts Transportation Cost
MĂDĂLIN CATANĂ, SERGIU TONOIU
Department of Machine Manufacturing Technology
University POLITEHNICA of Bucharest
Splaiul IndependenŃei 313, Bucureşti 060042
ROMANIA
mg_catana@yahoo.com, sergiu_ton@yahoo.com, http://www.upb.ro/en/

Abstract: This paper presents an economic lot sizing model for a serial, multi-operation machining process of
identical parts in which transportation cost of the parts between production stages may be significant. A general
cost function is provided for the model to suit different production flow schemes adopted for machined parts. Cost
function includes production setup cost, opportunity cost of capital investment in work-in-process inventory of parts,
and transportation cost of the parts along manufacturing route. A two-step lot sizing procedure uses model’s cost
function to conclude on optimal and economic sizes of manufacturing and transportation lots for machined parts.

Key-Words: Lot sizing, Cost model, Multi-stage production, Production flow schemes, Transportation cost

1 Introduction parts between machining operations is presented in the


Lot sizing is one of the most important decisions for paper. This model is an expansion of single-item lot
every manufacturer mainly due to its direct impact on sizing model proposed by Neagu et al. [5], to reveal
production cost. Consequently, from the first lot sizing the influence of parts transportation cost on economic
model introduced in 1915 by Harris till now, dozens sizes of both manufacturing and transportation lots.
of models were developed and reported to be fitted for
different production environments [8].
Traditional economic lot sizing models accounted by 2 Description of the model
production management literature [1, 7, 8] respectively The proposed lot sizing model uses a unique aggregate
for uniform demand of manufactured items (e.g., EOQ, cost function for optimization of both manufacturing and
EPQ) and for lumpy demand (e.g., SM, PPB, WW) are, transportation lots. This function includes three relevant
in fact, some very simple cost models that do not depict costs incurred by a serial, multi-operation machining
accurately all inventory costs incurred by a multi-stage process of parts: production setup cost; opportunity cost
manufacturing process, especially inventory holding cost. of capital investment in work-in-process (WIP) inventory
More production flow-based and, thus, more precise of parts; transportation cost for moving the parts along
economic lot sizing models are reported by Romanian manufacturing route. The cost function is devised to be
books [5, 6] for multi-operation machining processes. fitted to three alternative flow (transportation) schemes
However, these models do not take into account parts of the parts within multi-operation machining process,
transportation cost incurred by multi-stage production which are by manufacturing lot, by transportation lot,
environment, despite that several flow (transportation) and by piece production flow schemes respectively.
schemes for machined parts within the process are also Model’s cost function is minimized by the optimal
taken into consideration. size of manufacturing lot of parts, which is rounded after
A number of economic lot sizing models devised that to obtain the economic size of manufacturing lot. A
explicitly for multi-stage production processes [2, 3, 4] similar procedure may be used to derive optimal and
does consider the cost of transportation the items between economic sizes of transportation lot of parts in the case
processing stages. However, these models focus only that a by transportation lot flow scheme is adopted for
on slightly different variants of the same flow scheme multi-operation machining process of the parts.
of manufactured items within the process, which is by
transportation lot (sub-lot of process lot) flow scheme.
Taking into consideration the deficiencies revealed 2.1 Model’s cost function
for the above mentioned models, an improved economic Total cost function (C) to be minimized within the lot
lot sizing model for single-item, multi-operation machining sizing model is expressed as follows:
processes with notable costs for transportation of the C = C S + CO + CT [mu/pc] (1)

ISBN: 978-1-61804-193-7 51
Advances in Production, Automation and Transportation Systems

where: CS is production setup cost for manufacturing


lot of parts, in monetary units per piece (mu/pc); CO -

Capital investment
opportunity cost of capital investment in WIP inventory
of parts, in mu/pc; CT - transportation cost for moving
the parts along manufacturing route, in mu/pc.
In order to use function C for optimization of both
manufacturing and transportation lots, the dependence

N • C1 + L
of C on the sizes of these lots will be proved as follows.
Production setup cost (CS) is inverse proportional

N • Cm
to the number of part pieces within manufacturing lot, Lot 1 Lot 2
according to the following relationship:
L
CS = [mu/pc] (2)
N
where: L is setup cost per manufacturing lot (a fixed
cost), which may be derived in monetary units as shown 0
by Neagu et al. [5]; N - size of manufacturing lot of Tc Tc Time
parts, expressed in units. Tr
Opportunity cost of capital investment in WIP (CO)
represents the loss of the firm by using its money in the Fig. 1 Variation of capital investment through time
production process of a manufacturing lot of parts and for serial production of machined parts
not in a savings account. The larger the manufacturing
lot and the longer its flow time, the higher will be the Lot production flow time Tc in equation (3) depends
cost CO, according to the following relationship: not only on the size of manufacturing lot N, but also on
K ⋅ E ⋅ Tc production flow scheme adopted for the parts in the lot,
CO = [mu/pc] (3) according to the following relationships [5]:
D ⋅ Tr n
where: K is capital investment in the production of a Tcs = N ⋅ ∑T
k =1
uk
[min] (6)
manufacturing lot, in monetary units; D - total demand
n n
(production quantity) of the parts over the production
period, in units; E - interest rate in the capital market
Tcm = N t ⋅ ∑
k =1
Tuk + (N − N t ) ⋅ ∑ (T
k =1
uk − Tuk + 1 )+ [min] (7)

over the production period of the parts; Tr - cycle time n

of manufacturing lots during the production period of Tcp = ∑T


k =1
uk + ( N − 1 ) ⋅ Tu max [min] (8)
the parts, in time units; Tc - production flow time for
where: Tcs, Tcm, and Tcp represent lot production flow
each manufacturing lot of parts, in time units.
times for by manufacturing lot, by transportation lot,
For lot sizing purposes, invested capital K can be
and by piece production flow schemes respectively;
estimated as follows [5]:
Tuk - run time per piece of part at machining operation
N ⋅ ( C m + C1 ) + L
K = [mu] (4) k (k=1÷n), in min/pc; Nt - size of transportation lot of
2 parts, in units (1<Nt<N); (Tuk–Tuk+1)+ - positive run time
where: C1 is manufacturing cost per piece of part (a difference between succeeding operations k and k+1, in
variable cost), which may be calculated in mu/pc as min/pc; Tu max - run time per piece of part at the longest
reported by Neagu et al. [5]; Cm - raw material cost per machining operation in the process, in min/pc.
piece of machined part (a component of cost C1), in By using detailed relationships (5), (6), (7), and (8)
mu/pc. The relationship in equation (4) assumes a linear for terms Tr and Tc in equation (3), the ratio Tc / Tr can
increase of invested capital K during the production flow be expressed generally as follows [5]:
time of manufacturing lot (Tc), from a lower value equal Tc S
to N • Cm to an upper value equal to N • C1 + L, as =Z + (9)
illustrated in fig. 1. Tr N
Manufacturing cycle time Tr in equation (3) depends where terms Z and S do not depend on manufacturing
on the size of manufacturing lot N as follows: lot size N, but on the flow scheme adopted for the parts
T r = N ⋅ R [min] (5) in the lot, according to the following relationships:
n
where R denotes the takt time for parts manufacturing
during the production period, which may be stated, in
∑T
k =1
uk

minutes per piece (min/pc), as shown by Neagu et al. [5]. Zs = (10)


R

ISBN: 978-1-61804-193-7 52
Advances in Production, Automation and Transportation Systems

n J s = 0 ; J m = 0 ; J p = T [mu] (23)
∑(T
k =1
uk − Tuk + 1 ) +
Terms Gs, Hs, and Js; Gm, Hm, and Jm; and Gp, Hp, and Jp
Zm = (11) are relevant for by manufacturing lot, by transportation
R
Tumax lot, and by piece production flow schemes respectively.
Zp = (12) By substituting in expression (1) the values of CS, CO,
R and CT from equations (2), (16), and (20), the following
Ss = 0 (13) relationship for model’s cost function C, in monetary units
 n n
 per piece, can be obtained:

 k =1 k =1

N t ⋅  Tuk − (Tuk − Tuk +1 )+ 
 C=
L
+
E  S
⋅  Z +  ⋅ [N ⋅ ( C m + C 1 ) + L ]
Sm = (14) N 2⋅D  N
R (24)
n G H J
+ + +
∑Tk =1
uk − Tu max N Nt 1
Sp = (15) where terms Z, S, G, H, and J have specific expressions
R
according to the flow scheme adopted for the parts in
Terms Zs and Ss, Zm and Sm, and Zp and Sp are significant
the manufacturing lot.
for by manufacturing lot, by transportation lot, and by
piece production flow schemes respectively.
By substituting in expression (3) the values of K and
Tc / Tr from equations (4) and (9), a general formula for
2.2 Lot sizing procedure
calculating opportunity cost CO in monetary units per Cost function C in equation (24) will be minimized for
the optimization of both manufacturing and transportation
piece can be derived, as follows:
lots. For sizing the transportation lot, terms Z, S, G, H,
E  S
CO = ⋅  Z +  ⋅ [N ⋅ ( C m + C1 ) + L ] (16) and J of the cost function have to be particularized for
2⋅D  N by transportation lot production flow scheme and term
Transportation cost (CT) is caused by the movement N will be used as a constant whose value corresponds
of parts in manufacturing lot between the work centers to the adopted size of manufacturing lot.
that run machining operations. Considering a fixed cost
incurred by each transport of parts along manufacturing 2.2.1 Sizing the manufacturing lot
route, it results that transportation cost CT depends on Optimal size of manufacturing lot (No) that minimizes
production flow scheme adopted for the parts in the lot, model’s cost function C with respect to variable N is
according to the following relationships: obtained as follows:
T 2 ⋅ D ⋅( L + G ) + S ⋅ L ⋅ E
CT s = [mu/pc] (17) No = [units] (25)
N ( C m + C1 ) ⋅ Z ⋅ E
T
CT m = [mu/pc] (18) Since the second term of the sum in the numerator of
Nt the radicand in equation (25) is negligible in comparison
T with the first term [5], a more convenient formula for
CT p = [mu/pc] (19) deriving optimal size of manufacturing lot is as follows:
1
where: CTs, CTm, and CTp are transportation costs for by 2 ⋅ D ⋅( L + G )
manufacturing lot, by transportation lot, and by piece No = [units] (26)
( C m + C1 ) ⋅ Z ⋅ E
production flow schemes respectively; T - transportation
cost per transport of machined parts along manufacturing From equation (26), it becomes apparent that, in the
route, in monetary units. case of by manufacturing lot production flow scheme,
Particular expressions for cost CT in equations (17), transportation cost of the parts on manufacturing route
(18), and (19) can be replaced by the following general may decide the optimal size of manufacturing lot when
formula for calculating transportation cost CT: term G dominates term L (i.e., for parts with very high
transportation cost per transport). Moreover, production
G H J
CT = + + [mu/pc] (20) flow scheme adopted for the multi-operation machining
N Nt 1 process of parts affects all the time the optimal size of
where terms G, H, and J are fixed costs dependent on the manufacturing lot due to the term Z in equation (26).
flow scheme adopted for the parts in the lot as follows: If only discrete, equal sized manufacturing lots can
Gs = T ; Gm = 0 ; G p = 0 [mu] (21) be used over the production period of the parts, optimal
H s = 0 ; H m = T ; H p = 0 [mu] (22) value No has to be rounded up or down to the nearest
divisor of total demand D. This divisor represents the

ISBN: 978-1-61804-193-7 53
Advances in Production, Automation and Transportation Systems

economic size of manufacturing lot (Ne) that should be the model takes into account the influence on lot sizing
used for the multi-operation machining process of parts. decision not only of production flow scheme adopted for
the parts, but also of the transportation cost of the parts
2.2.2 Sizing the transportation lot along manufacturing route. Consequently, model’s cost
Cost function C in equation (24) is particularized for the function comprises three lot size-dependent production
case that by transportation lot flow scheme and size Ne costs: production setup cost; opportunity cost of capital
of manufacturing lot were adopted for serial production investment in WIP inventory of parts; transportation cost
of parts, as follows: for moving the parts between machining operations.
E  S  On minimizing the model’s cost function, it was
C= ⋅  Z m + m  ⋅ [N e ⋅ ( C m + C1 ) + L ] proved that optimal size of transportation lot and, in
2⋅D  Ne 
(27) specific situations, even optimal size of manufacturing
L T lot heavily depend on parts transportation cost within
+ +
Ne Nt multi-operation machining process.
where terms Zm and Sm are detailed in equations (11)
and (14) respectively.
Optimal size of transportation lot (Nto) that minimizes References:
cost function C in equation (27) with respect to variable [1] S. Axsater, Inventory Control, 2nd Edition, Springer
Nt is obtained as follows: Science + Business Media, New York, 2006.
2 ⋅ D ⋅ Ne ⋅T [2] R.W. Bogaschewsky, U.D. Buscher, and G. Lindner,
N to = [units] (28)
S m′ ⋅ E ⋅ [N e ⋅ ( C m + C1 ) + L ] Optimizing Multistage Production with Constant
Lot Size and Varying Number of Unequal Sized
where Batches, Omega, Vol. 29, No. 2, 2001, pp. 183-191.
n n

∑T − ∑(T
k =1
uk
k =1
uk − Tuk +1 )+ [3] H.N. Chiu and J.H. Chang, Cost Models for Lot
Streaming in a Multistage Flow Shop, Omega, Vol.
S m′ = (29)
R 33, No. 5, 2005, pp. 435-450.
Equation (28) proves the expected influence of the [4] S.K. Goyal and A.Z. Szendrovits, A Constant Lot
cost per transport on the optimal size of transportation Size Model with Equal and Unequal Sized Batch
lot for a multi-operation machining process of parts. Shipments Between Production Stages, Engineering
If only discrete, equal sized transportation lots can be Costs and Production Economics, Vol. 10, No. 1,
used in the production flow of manufacturing lot, optimal 1986, pp. 203–210.
value Nto has to be rounded up or down to the nearest
divisor of the manufacturing lot size Ne. This divisor is [5] C. Neagu, E. NiŃu, L. Melnic, and M. Catană,
the economic size of transportation lot (Nte) that should Ingineria şi managementul producŃiei: Bazele
be used for the multi-operation machining process of parts. teoretice, Editura Didactică şi Pedagogică R.A.,
Bucureşti, 2006.
[6] E. NiŃu, Managementul producŃiei, Editura UniversităŃii
3 Conclusion din Piteşti, 2002.
Economic lot sizing model presented in the paper uses [7] P. Schonsleben, Integral Logistics Management:
a single aggregate cost function for the optimization of Planning and Control of Comprehensive Supply
both manufacturing and transportation lots in the case Chains, 2nd Edition, CRC Press, Boca Raton, 2004.
of a serial, multi-operation machining process of parts. [8] D. Sipper and R.L. Bulfin, Production: Planning,
To be well fitted to diverse practices and environments Control, and Integration, McGraw-Hill, Singapore,
for serial manufacturing of parts, the cost function of 1998.

ISBN: 978-1-61804-193-7 54

You might also like