"fi'v- —-—~

 

REDIREC‘TING HIGH QUALITY
PALLET DECK BOARDS '
INTO FURNITURE DRAWER SIDES

Thesis for the: Degree of M. S.
MICHEGAfl STA'EE UNIVERSITY
EDWARD KiNG PEPKE
1975

IHESIS

  
   

LIB RA R Y
Michigan Stave
Um‘vcrsicy

 

   
   
  
 

,. amSmG av .-
, ”W "BAG '& SIJNS'
i g; 800K BINDERY INC.

. 7"?»‘QY BINDERS
n -- -:— "0.9L uncanny

 
 
    

 

 

 

 

%\ J ABSTRACT

The desire for more efficient utilization of wood led to research
on redirecting higher quality pallet deck boards into furniture drawer
sides. Eight pallet companies provided data collection sites for
conducting production studies and for estimating the quantity of high
quality, clear wood being produced by Michigan pallet mills in 1974.
Further studies were conducted under laboratory conditions at the
Michigan State University Forestry Department's dry kiln to provide
information on packaged handling and drying of pallet lumber. Finally
the dried pallet stock was processed into drawer sides at three
furniture companies' rough mills.

Analysis of the data collected allowed comparisons of the
relative costs of pallet deck boards and drawer sides. Cost reductions
will vary between the drawer side dimension stock producing companies
depending on their efficiency and lumber costs. The sale of grade
pallet deck boards would open a large profitable market for the pallet
sawmills. While the demand for drawer side dimension stock is
unpredictable because of variations in furniture styles and general
economic trends the pallet companies have flexible production
operations which could meet a fluctuating demand by shifting emphasis

between market options.

REDIRECTING HIGH QUALITY PALLET DECK BOARDS

INTO FURNITURE DRAWER SIDES

By

Edward King Pepke

A THESIS

Submitted to
Michigan State university
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

Department of Forestry

1975

ACKNOWLEDGMENTS

The completion of this thesis was made possible by the help and
guidance from Dr. Henry A. Huber, Extension Specialist at Michigan State
University. Professors Alan W. Sliker and Eldon A. Behr's expertise was
also appreciated in this study.

Funding for the study was provided by the United States
Department of Agriculture Forest Service through Mr. Alan Boelter, State
and Private Forestry, and the Michigan Department of Natural Resources.
Coupled with the federal grant was a graduate study assistantship
awarded by the Forestry Department at Michigan State University. The
cooperation of the owners, supervisors, and employees at the sample

pallet mills and the furniture companies is appreciated.

ii

TABLE OF CONTENTS
Page
LIST OF TABLES o o o o o o o o o o o o o o o o o o o o o o o o o o 0 V
LIST OF FIGURES. o o o o o o o o o o o o o o o o o o o o o o o o o 0 Vi
INTRODUCTION 0 o o o o o o o o o o o o o o o o o o o o o o o o o o o 1

PALLET INDUSTRY 0 O O O O I O O O O O I O O O O O O O O O O O O
FIJRNITLIRE INDUSTRY 0 O O O O O O O O O O O O O O O O O O O C O O

UIN

OBJECTIVES O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 7

PROCEDURE 0 O O O O O O O O O O O O O O O O O O O O O O O O C O O O O

m

PALLET MILL SELECTION . . . . . . . . . . . . . . . . . . . .
GRADING SYSTEM FOR SELECTION OF HIGH QUALITY WOOD . . . . . . .
DATA COLLECTED AT PALLET MILLS. . . . . . . . . . . . . . .
DATA COLLECTED IN THE LABORATORY ON DRYING LUMBER . . . . . . .
DATA COLLECTED AT FURNITURE ROUGH MILLS . . . . . . . . . . . .

F‘F‘
+ara~o<n<w

RESIJIJTS AND ANALYSIS 0 O O O O C O O O O C O C O O O O O O O O O O O 13
DATA TABULATIONS FROM PALLET MILLS. . . . . . . . . . . . . . . 13

EQUIPMENT. . . . . . . . . . . . . . . . . . . . . . . . . 13
TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
PRODUCTS . . . . . . . . . . . . . . . . . . . . . . . . . l4
SPECIES. . . . . . . . . . . . . . . . . . . . . . . . . . 15
INPUT. . . . . . . . . . . . . . . . . . . . . . . . . . . l7
CLEAR CANT INPUT . . . . . . . . . . . . . . . . . . . . . l9
OUTPUT . . . . . . . . . . . . . . . . . . . . . . . . . . 21
WASTE. . . . . . . . . . . . . . . . . . . . . . . . . . . 23

DATA COLLECTED IN THE LABORATORY ON HANDLING AND DRYING . . . . 24
DATA COLLECTION AT FURNITURE ROUGH MILLS. . . . . . . . . . . . 28
FIRST ROUGH MILL TEST. . . . . . . . . . . . . . . . . . . 3O
SECOND ROUGH MILL TEST . . . . . . . . . . . . . . . . . . 32
THIRD ROUGH MILL TEST. . . . . . . . . . . . . . . . . . . 33
DEMAND FOR DRAWER SIDES. . . . . . . . . . . . . . . . . . 34
RELATIVE COSTS . . . . . . . . . . . . . . . . . . . . . . 35

FUTURE PREDICTIONS O O I I O O 0 O O O O O O 0 O O O O O O O O O 37

TABLE OF CONTENTS CONTINUED

Page
CONCLUSIONS AND RECOWENDATIONS O O O O O O O I O O 0 O O O O O O O O 39
LITERATURE CITED 0 O O O O O O O O O O O O O O O O O O I O O O O O O 42

APPENDIX A. DRY KILN OPERATOR'S MANUAL KILN SCHEDULE FOR 4/4 RED
OAK T4-D2

APPENDIX B. DRY KILN OPERATOR'S MANUAL KILN SCHEDULE FOR 4/4 SOFT
MAPLE T8-D4

APPENDIX C. RESULTS OF GRADING INDIVIDUAL PALLET DECK BOARDS
APPENDIX D. TABLE VI. OUTPUTS PER MAN-MACHINE-HOUR (CUBIC FEET)
APPENDIX E. FIGURE 5. PROCESS FLOW CHART

APPENDIX F. FIRST FURNITURE COMPANY YIELD STUDY

APPENDIX G. SECOND FURNITURE COMPANY YIELD STUDY

APPENDIX H. THIRD FURNITURE COMPANY YIELD STUDY

APPENDIX I. DATA TABULATION FOR EIGHT PALLET MILLS

iv

FIGURE 1.

FIGURE 2.

FIGURE 3.

FIGURE 4.

FIGURE 5.

LIST OF FIGURES

DIFFERENT TYPES OF COMMON PALLETS

LABORATORY DRY KILN WITH ROUGH LOAD ON LEFT AND

PRESURFACED LOAD ON RIGHT.

LABORATORY DRY KILN WITH THIRD TEST CHARGE OF

3/4 SOFT MAPLE . .

BANDING METHOD WITH EXTRA TAILS FOR RESTRAINT
DURING DRYING AND EASE OF HANDLING

PROCESS FLOW CHART.

Page

0 O O O 29

Appendix E.

TABLE I .

TABLE II.

TABLE III.

TABLE IV.

TABLE V.

TABLE VI.

LIST OF TABLES

PALLET MILL COMPARISONS. . . . . . . . . . .

CANT UTILIZATION FOR INDIVIDUAL PALLET MILLS
BY SPECIES O O O O O O O O O O O O O O O O

CANT INPUT EXTRAPOLATION . . . . . . . . . .
DECK BOARD OUTPUT EXTRAPOLATION. . . . . . .
WASTE DATA FOR 1/2 INCH DECK BOARDS. . . . .

OUTPUTS PER MAN-MACHINE-HOUR (CUBIC FEET). .

vi

Page

. . . . 16

. . . . 18
. . . . 19
. . . . 21
. . . . 24

Appendix D.

INTRODUCTION

Wood is a renewable natural resource yet the need persists to
make efficient use of the existing and future raw material. Recognition
of the value of increased timber utilization led the United States
Department of Agriculture Forest Service to initiate the pallet project
at Michigan State University to determine the feasibility of redirecting
the higher quality wood from pallet mills to higher value uses.
Information for this thesis was obtained by the author from data
collected on the pallet project.

Previous studies have indicated the possibility of obtaining
furniture quality dimension from log bolts or from cants sawn from

pallet mills' bolts.10’17’19’23:32.36

Tests conducted at pallet mills
showed the most reliable approach to obtain lumber suitable for furniture
drawer sides was to grade the individual pallet deck boards. Previous
results of the Pallet Project19 present the feasibility of grading the
input to pallet mills, logs and cants (squared cross sections of logs
with little or no bark), in order to determine the potential for
furniture raw material, principally, 2 1/4 inch by 2 1/4 inch squares
for turnings. Among the conclusions arrived at were:

"Using ... oak in absolutely clear lengths is an economic

waste and unnecessary in even a quality pallet. With

increasingly higher prices offered for the upper grades

of these species (including oak), a definite effort

should be made to redirect that portion of wood which

qualifies for higher value use into furniture type

areas. Redirecting high value materials used by many

pallet manufactures will require effort on the part of
both the pallet manufacturer and the user of high value

1

wood. However, on the basis of actual experience to date,
under commercial conditions it has proven economically
advantageous in several trial cases to both the pallet

. l
bolter saw operator and the furniture dimension user." 7

Another furniture product besides turning squares which might be
made from pallet stock is drawer sides and backs. Drawer sides are
similiar in width and thickness and are shorter in length than the deck
boards cut for pallets. An area for investigation then is the economic
advantage, if any, of utilizing stock for drawer sides. This is to be
developed in the present study.

Two basic definitions should be stated early: pallets are
portable, materials-handling platforms consisting of one or two faces of
deck boards, separated or supported by structural members to provide a
method of carrying a load with a fork—lift truck, (see illustration on
next page). Dimension stock is short, graded pieces of lumber which
require little additional processing before fabrication into a furniture

component.

PALLET INDUSTRY

The wood pallet industry has expanded very rapidly to fulfill a
need in modern materials handling. William H. Sardo, Jr. wrote "WOoden
Pallets and Pallet Containers ... Present and Future Markets" which
contained the following pertinent passage:

"The pallet which had its birth in the mid-1930's achieved
major use when the military services began to use it to
speed the warehousing, handling, and delivery of supplies
during W’orld war II. The fork-lift truck and the wooden
pallet combination created the 'cubic concept' of modern
warehousing, which permits the use of air space as well

. 38
as floor space in storage areas and transportation vehicles."

 

Figure l.—-Expendable pallefs:

no-block Type;
l39 428

nine-posf Type.

A’ E)
M M |39 430

After World War II the use of pallets achieved rapid growth and popular-
ity. Today "wood pallets are used for material handling, shipping, and
storage by just about every conceivable manufacturing and distribution
industry, including government agencies. Company, industry-wide,
national, and international pallet pools currently exist and are
expanding."43
Despite the tremendous popularity, which made pallet manufacturing
the fastest growing wood products industry, consuming more hardwood
lumber than all other hardwood-using industries except furniture manu-
facturing, the pallet companies have experienced major obstacles. "The
main headache of the wooden pallet industry is that it probably has the
lowest profit margin of any fabricated wood product. The wooden pallet
industry probably has the highest financial mortality rate in the lumber
and wood products industries."38
Pallet companies being extremely competitive, market directly to
the user or others sell their products through a broker or manufacturer's
representative.43 Regardless which method of sales is used, the pallet
industry's "growing markets are aggressively threatened by competitive
materials and substitute systems."2 Acknowledging economic problems the
progressive mills are exploring alternative markets' potentials.
"Additional or better markets for the specific items that can be produced

"23 could include the sale of pallet deck

from the material available
board lumber for uses other than pallets.

At the last census in 1973 there were 258 pallet mills scattered
throughout Michigan's upper andlower peninsula with the heaviest con-

centration located in the northern half of the lower peninsula.28 Raw

materials are usually obtained in short logs (8 feet and less) or as

cants from independent loggers. The logs and cants are purchased either
as hardwoods or softwoods by a particular size corresponding to the
finished pallet but without regard to grade, other than soundness. The
product, pallets, require varied lengths (3 to 6 feet), widths (3 to 6
inches), and thicknesses (3/8 to 1 inch) of deck boards. Deck boards
nailed or stapled to runners according to the purchaser's specifications

form the finished pallet.

FURNITURE INDUSTRY

The furniture industry is an old, established business in contrast
to the pallet industry. While the size of individual companies varies
there are over 200 furniture manufacturers in Michigan. Years have been
spent by this hardwood-using industry to build a sound reputation for
quality and craftsmanship. While centered in Michigan primarily in
Grand Rapids and Holland areas, other firms needing hardwood lumber for
dimension do exist throughout the state.

Many factors mentioned in the description of the pallet industry
are the reverse for the furniture industry: the profit margin is much
higher, financial mortality low, and the changes in the economy are felt
with less severity. R. M. Hallet's article in the Forest Products
Journal, "A Method to Analyze Rough Mill Productivity" says the "efforts

"13 of

at cost reduction should be directed at yields-—not productivity.
course cost reduction ultimately leads to increased profit margins which
are the main objective in all furniture firms.

Furniture companies typically select their raw material for each

job based on the cost and yield of the graded lumber. "A manufacturing

procedure should involve scheduling raw material by grade and size for

cutting into a few well chosen grades and sizes of dimension to result
in the most economical utilization."45 Drawer sides are commonly manu-
factured from red oak or white oak, and maple and vary in length, width
and thickness depending on the particular furniture company requirements.
Most material for drawer sides comes from grade lumber either directly
or as the pieces left from other cuttings. There appears to be
considerable variation in the sizes needed and raw material sources for

drawer side dimension.

OBJECTIVES

The overall objective of this study was to explore the possibility

of redirecting the higher quality lumber from pallet mills into drawer

sides for furniture.

Some specific topics for investigation included:

1.

To determine the quantity of wood processed by Michigan
pallet mills in 1974.

To calculate the quantity of higher quality lumber
produced at the mills which would be suitable for more
efficient uses.

To propose a method for handling lumber in the pallet
mill, the furniture rough mill, between mills, and the
dry kiln.

To propose a drying method.

To estimate the marketability and demand for furniture

drawer side dimension stock.

PROCEDURE

PALLET MILL SELECTION

In the earlier studylg, of which the present study was a continua-

tion, a sample of ten pallet mills was randomly selected from the state-
wide population of 258 Michigan mills. The mills appeared to have
production operations similar to other pallet mills around Michigan. The
size of the study mills varied from 15 to 50 employees with corresponding
variations in equipment and output. The present study utilized the same
pallet mills for sampling and data collection with the exception of two
which had altered their production techniques making them incongruent

with the needs of the study.

GRADING SYSTEM FOR SELECTION OF HIGH QUALITY WOOD

To establish an appropriate grading system to be utilized in dis-
tinguishing between qualities in pallet mills' deck boards the end
product-drawer sides must be considered. Because of the non-homogeneous
nature of wood every possible combination of defects may be found in
lumber. "The size of knots permitted in various pallet parts is governed
by the proportion of their width to the width of the piece containing
them as well as by their location in the finished pallet part. Though a
wide range of knot sizes is acceptable in pallet parts, the referenced

pallet specifications should be consulted as they outline the limitations

controlling the location and size of permissible knots."43 Conversely,
drawer sides and backs must be free of all but minor defects.

Another consideration in designing the grading system was ease of
application. The worker at the pallet mill must be able to rapidly
observe a pallet deck board either as it leaves the gang rip saw or as
it lays in an output pile and make a judgement whether that piece of
lumber should become part of a pallet or part of a furniture drawer.

To meet the above requirements the following grading criteria
were established for selecting drawer side stock from cants and lumber:

"Clear" includes: Sound knots less than 1/2 inch in diameter

Maximum of 3 knots per cant or board
wane less than 2 inch width on any cant
wane less than 1 inch width on any board
Stain, mineral streak, fleck

Any defect within 2 inches of an end
No unsound wood, large checks, or ringshake

DATA COLLECTED AT PALLET MILLS

The data collected at a pallet mill encompassed inputs, measure—
ment of end product yields, classifying or grading the input, denoting
the species processed, recording of production time, saw characteristics,
and general observations.

Input volumes of cants were determined either from the exterior
dimensions of a load of cants or by examining each cant individually. In
some mills it was more convenient to measure the exterior dimensions of
a load of tightly piled, four-sided, random width cants of equal length
and thickness brought to the saw for ripping into lumber. In the case
of an in-line system the individual cant dimensions were recorded to the

nearest inch before sawing. Before the cants were sawn each was graded

10

by the previously given criteria and the species noted. The volume of
clear cants was calculated and divided by the total input of cants to
obtain a percentage of clear cant material for that run.

Two basic types of cants were measured in the pallet mills--
three-sided and four-sided. While each cant type has four sides because
of its square or rectangular profile, the three-sided cants have wane
(bark) on one side. Since only a few mills sawed three-sided cants it
became desirable to convert their volume to an equivalent volume of four-
sided cants. Laboratory tests at Michigan State university which were
later confirmed by industry buying practices showed reduction of the
three-sided cant volume by 25 percent closely approximated a non-waney,
four-side cant. The 25 percent reduction factor was applied to all
three-sided cant volumes.

Two methods of measuring the output,which proved to be equally
valid, were similar to the measurement of the input. The first method
was to measure the tightly-stacked output pile's dimension and compute
a cubic foot volume. Another method which gave insignificantly
different results was to obtain a piece count of boards in the output
pile. The number of boards per pile multiplied by the volume per board
also gave the cubic feet of output. When divided by the input and

multiplied by 100 the yield in percent may be found:

board output in cubic feet
cant input in cubic feet

 

x 100 = Yield in Percentage

The time factor was subdivided into lapse time and production
time (P/T). Lapse time was recorded from the time an observation began

at a pallet mill until production and/or the observation terminated for

that day. Over 25 hours of total lapse time distributed over a seven

11

month period was the designated amount of time to collect data at each
mill. The down time (the periods of time the saw did not produce
lumber) was subtracted from the lapse time to give the actual total

production time.

DATA COLLECTED IN THE LABORATORY ON DRYING LUMBER

Using the Michigan State University Department of Porestry's dry
kiln, data was collected on seasoning of the better quality, "clear",
pallet lumber. Three different charges were run to examine the effects
of thickness and species of boards, kiln schedules, and loading methods.
The first charge of oak was subdivided into two 200 board groups, one
hit-or-miss presurfaced to 5/8 inch and the other a rough 3/4 inch
thickness. Both groups were strapped with steel banding in self-stickered
loads capable of being transported by fork-lift truck. A second charge
of oak, 3/4 inch thick, unsurfaced, deck boards was dried using typical
3/4 inch stickers spaced about 18 inches apart between layers of deck
boards. No steel banding was applied in this latter case. The first
and second tests followed the same kiln schedule, T4-D2 from the Dry
Kiln Operator's Manual30; a third charge followed a slightly different
schedule T8-D4 because of a species change to soft maple (kiln schedules
appear in appendix). The soft maple charge was self-stickered and

banded in two loads suitable for handling by a fork-lift truck.

DATA COLLECTED AT FURNITURE ROUGH MILLS

Three furniture companies' rough mills were another source of

data. Three demonstrations were conducted, one per company, to collect

12

data on the feasibility of sawing drawer side dimension from the pallet
deck boards which had been graded, kiln dried, and inspected for drying
defects. Similar to the methods in the pallet mills the input and output
volumes were computed and the yield of dimension stock calculated. Costs
of the standard material utilized by the company for drawer sides versus
the cost of pallet mills' stock were compared. Figures were obtained on
annual volume of drawer sides produced and the amount of lumber purchased

to meet that production requirement where available.

RESULTS AND ANALYSIS

DATA TABULATIONS FROM PALLET MILLS

 

Most of the data collected came from pallet mills with similar
types of sawing equipment. Variable factors such as cant input, time,
kerf, and end products were equalized for purposes of making comparisons
as explained below. Information concerning the amount of clear input,
output, output potential, and waste percentages are tabulated and

discussed.

EQUIPMENT

At the beginning of the study the majority of the pallet mills
had circular-bladed gang rip saws, but by the end of the study all the
mills employed the gang saw type of production. A gang saw is a multi-
bladed piece of equipment capable of ripping (sawing parallel to the
grain direction) cants into lumber. Because it would not be possible to
determine the precise percentage of each equipment type in the state,
the sample selected and used in this study is assumed representative of
the population.

Two types of saw blades were used in the eight pallet mills--
insert tooth type and carbide tip. While each type possessed certain
advantages and disadvantages, the decrease in kerf was the most signifi—
cant advantage with carbide—tipped blades. The insert teeth, held in

the saw blade by springs, ranged in width from 3/16 to 9/64 inch. A

13

l4

cemented carbide tip on a circular saw is welded to the blade and varies
in width from 1/8 to 3/16 inch. For the purpose of comparison of
different pallet mills' data the saw tooth type variable was equivocated
by assuming that like the sample, the population of Michigan pallet

mills was equally divided between carbide and insert-tooth saws.

TIME

Delays in production due to machinery failures, input or output
delays, rest steps, and other contingencies created the need for estab-
lishing two measures of time--production time (P/T) and lapse time as
defined earlier. The following formula was used to calculate down time
for the sample mills:

_ _production time

(1 lapse time

) 100 = Z down time

Down time expressed as the percent of time the saw is not producing
averaged 54.0 percent. By employing more efficient production techni-
ques a portion of that precentage of time could be used to increase
output. The average was calculated from the eight mills which ranged
from a low of 37.7 percent to a high of 76.3 percent with a standard
deviation of 12.9 percent (see Table I on page 16). The input and
output quantities were necessarily figured using production time and
then corrected by the 54.0 percent average down time to allow comparison

between firms in Table I.

PRODUCTS
From specific cant thicknesses came corresponding pallet deck
board widths. Common deck board widths ranged from 3.5 to 5.5 inches

with thicknesses for deck boards beginning at 3/8 inch and merging with

15

runner-sized stock at l to 1 3/4 inches. The most observed deck board
size, 1/2 inch, was used in making comparisons in Table I. Pallets
encompass a wide variation in sizes according to the purchaser's speci-
fications. All board outputs from the gang saws are stacked into lumber
piles before nailing operations. For separation of quality lumber for
drawer side dimension stock an intermediate step could be imposed to

grade and stack the best appearing lumber.

SPECIES
Pallet manufacturers use conventional lumber terms to classify
wood as "hardwood" or "softwood"; the exception to the pallet industry is

aspen which with its relatively low density is treated as a softwood.

 

Aspen, commonly called popple in the trade, (Populus grandidentata and P,

tremuloides), forms a dominant species in many stands located in Michigan's

 

northern lower and upper peninsulas. Accordingly, for pallet sawmills
established within these areas aspen usually constitutes a principal
species utilized in pallet production. The other most prevalent species
labelled "oak" in the trade is officially several species of red and
white oaks (QuerCus, spp.). When a purchaser failed to specify the
particular species or classification of lumber, the pallets were made
with mixed, locally available species including oak, aspen, maple (5223,

spp.), birch (Betula, spp.), basswood (Tilia americana), elm (Ulmus,

 

spp.) beech (Fagus grandifolia), ash (Fraxinus, spp.), pine (Pinus, spp.),

hemlock (Tsuga canadensis), etc.

 

0f the mills observed three processed predominently oak and three
processed predominantly aSpen. Every mill was capable and actually

obtained either oak or aspen as dictated by their purchase orders. While

l6

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

em.o Nm.e «mama 20HH<H>NQ
o.NN oo.oN as.NN mm.e NN.NN N\m nmaazaem
«N.NN mo.Nm mason ANNaa page
o.qm No.mo «N.om NN.eN No.mq H\m Nuamm><
Hmo.omm momma
mma.NmN.N we.Nmo.m Ne.mqe.qa e\m gases
ano.eo om.oN so.NN Nm.mm magma
N m m.mm ONm.omN N.NN Nm.eN 00.0NN.N «.0N mm.mm mm.mNm.N om.mN a\m w
NNN.oe Ne.oN mm.mm mm.NN Omama
m m N.Nm «NN.NON N.mo 0N.oN mm.maN m.aN NN.am qm.qu.N Nm.mN a\m N
mmN.Nw mo.eN om.om 0N.om «mama
N m m.Nq amm.ooN «.mN 0N.NN ew.mNm.N q.NN Hm.mm ow.ooN.N Nm.mN H\N o
er.om Ns.m Nm.oe wq.em magma
N N N.oq som.mON N.Nm No.NN mo.mNm o.NN mm.om No.NmN.N Nm.NN ia\m, m
mNN.mm Nm.mN mN.mN so.NN amass
N N m.oN mos.mofl o.Nm No.0N am.qu o.mN wo.Nm me.Nam em.o H\m e
mam.Nm Nm.eN Nm.em Nm.Nq mmamn
N m e.No NNo.sw «.mm NN.¢¢ am.moN o.w mw.No mm.Noo.N Nm.NN e\m m
NoN.wm SN.oN me.Nm mo.NN mmdma
N m s.eo emm.oqN m.mm oo.mm oo.ooN.N N.NN mm.m¢ Nm.mNN.N Ne.m N\N N
oNN.NmN as.oN NN.NN N.NN magma
m N o.Ne qwo.meN m.oN 0N.os NN.NNN.N N.mN we.eo Ns.mNe.N N.mN e\m N
N O S
0 T. “I .d a I
P... . a 93 N umo Sam. 2 NWN 2....
s. m. we we... aim. “mm mmn mm mmm mmm n
m... was 1% “an 91.3 G d I. u 80 .... a I nW
/ 38 0 0.1.3 ual 33m 300 1. n1 DOT. I. mt.
22 ex 31 o/n 3.19 an... 0.9 I )Ju a. m QT.
3 IA ul. BA. Poo M../ is... US 3 d 383 8 8.1.
1.1 8 3m 18 T. a N 33». dT. nTn 88W 1
cu 3 3 P85 / and? 3M na ./3 am
no I. 313 Z Plu 3 an .49 Z/ 3 3
SD. 0 . u 8.. 80. 1 in (a
u . P. 3 P
s “x
.maomaumaaoo Haas umNNmm .H mnm<a

 

17

some Michigan furniture manufacturers use aspen, pine, and other
softwoods for drawer sides,the majority are oak. Based on past pallet
manufacturers' response to buyers' species preferences the pallet indus-
try throughout Michigan would be able to produce high quality stock of
the species desired by the furniture manufactures.

Details of the species used at individual pallet mills included

in the sample appear in Table II on page 18.

INPUT

The cant input for the eight sample mills totalled 14,449.5 cubic
feet (173 MBF or 173 thousand board feet). When divided by the produc-
tion time for deck boards the result is the quantity of wood used per
actual hour of saw operation (input per machine-hour). While the input
per machine-hour varies according to the operator speed and infeed rate
the range was from 34.4 to 308.5 cubic feet with an average of 139.1
cubic feet per machine-hour and a standard deviation of 57.6 cubic feet
per machine-hour (see Table III extrapolation on page 19).

Using the population of 258 pallet mills with 1.125 machines per
mill a total of about 80.8 million cubic feet of input per year could be
expected in Michigan if there was no down time.

Down time for unavoidable contingencies such as infeed and outfeed
delays, rest periods, fatigue allowance, equipment failures, machine
maintenance, and so on must be included to provide a realistic estimate
of the quantity of wood processed in Michigan pallet mills in 1974. When
all the above factors which decrease production were included, the eight
sample mills averaged 54.0 percent down time. While this figure may seem

high it must be realized that only a few mills operated their gang saw(s)

18

 

 

 

 

 

 

 

 

 

TABLE II. Cant Utilization for Individual Pallet Mills by Species.
Mill Oak Aspen Mixed Species Lapse
Number Volume Percent* Volume Percent Volume Percent Tame In

Cu. Ft. Clear Cu. Ft. Clear Cu. Ft. Clear ours

1 1265.6 11.7 807.8 10.6 405.0 21.1 22.72

2 1109.0 23.0 263.7 8.3 342.6 5.3 27.05

3 ------ ---- ----- ---- 1662.9 8.6 47.51

4 ------ ---- 991.5 23.0 ----- ---- 27.66

5 1752.0 11.6 --—- ——-- ----- —-—- 34.48

6 ----- "'_-- 1790.8 1104 ----- --_- 30016

7 ------ ---- ----- ---- 1134.9 19.3 22.33

8 ------ ---- 1120.7 10.3 1802.9 10.5 53.31
TOTAL 4126.6 --- 4974.5 ---- 5348.3 --- 265.22

 

*
Based on volume of

clear cants divided by volume of cants processed.

 

19

continuously for eight hours; rather the machine would be shut down when
the crew switched and performed some other operation or ran another
piece of equipment.

When the previously calculated average input based on production
time, 139.1 cubic feet per machine-hour is reduced by the 54.0 percent
down time the result is 64.0 cubic feet per machine-hour of cant input
on the average based on lapse time. The extrapolation to an annual input

figure is presented in the following table:

TABLE III. Cant Input Extrapolation

Cubic Foot Cant

 

 

 

Input for Production Time Lapse Time
Machine-hour 139.1 ft.3 64.0 ft.3
8 Hour Day 1112.8 511.9
40 Hour week 5564.0 2559.4
50 week Year 278,200.0 127,972.0
258 Pallet Mills 6 6
at 1.125 machines 80.8 x 10 37.1 x 10
per mill

CLEAR CANT INPUT

As previously explained in the procedure the cant input to a
circular gang saw was graded by the simplified grading method. The
overall average clear material for the study was 12.9 percent, calculated
as follows:

Cubic feet of clear cants 1,859.13
Total cubic feet of graded input 14,449.47

 

x 100 = 12.9%

The wide range in percentage volume of clear cants, 34.83, corres-

ponds to the nature of material used in pallet production. The highest

20

value observed, 37.52 percent came from cants sawn from old growth,
northern hardwoods. Conversely the low value of 2.69 percent was derived
from grading young, pulpwood-sized cants of aspen, many which had traces
of wane on each corner. Overall the clear cant input for 258 pallet
mills in Michigan for 1974 is estimated to be 12.9 percent times 37.1
million cubic feet of total yearly input yielding 4.8 million cubic
feet.

It would be difficult to improve the grade of output from the
cants processed presently in Michigan's typical pallet mills. If a
system of using deck boards for drawer sides which are primarily oak was
adopted, the pallet mills could purchase more hardwood logs to meet the
demands. Since the overall percentage of clear material available when
the softwoods including aspen were removed was 12.03 percent the amount
of clear material would be expected to remain unchanged. Adoption of a
grading and selling system for high quality pallet deck boards could
provide the economic incentives back through the pallet production system
and logging operations to direct efforts into obtaining quality inputs.
"Without a strong economic incentive to both buyer and seller it is
doubtful that the concept of 'highest value' usage would be accepted."17

Grading of the input, which was usually small cants was intended
to indicate the amount of clear boards or clear wood which was available.
Unfortunately almost all the cants observed were obtained from poor
growing, unmanaged Michigan hardwood forests which contain predominantly
small trees. Cants sawn from the small trees have "boxed-hearts" (con-
tain the heartwood, pith, and juvenile wood) and therefore do not always
yield the same grade deck boards as indicated by the exterior grading

prediction, particularily in the case of 'clear" cants. The interpolation

21

of the quantity of clear pallet lumber available in Michigan would be

valid only in the case of non-boxed-heart cants.

OUTPUT

The lumber output including runners and deck boards for the eight
pallet mills observed in the study totalled 10,586.9 cubic feet (127 MBF).
Like the input, the division of this total by the production time gives
the quantity of wood produced perzn1actual hour of saw operation. Again
the output per machine-hour varies according to operator speed, infeed-
outfeed rate, and the tailers' ability. The average output per machine-
hour averaged 92.8 with a 44.4 cubic foot standard deviation. The range
245.4 cubic feet was calculated from a high of 266.0 cubic feet and a
low value of 20.6 cubic feet. Expanding the output for production time

and lapse time by different units of time is shown in the following table:

TABLE IV. Deck Board Output Extrapolation

Board Output in

 

 

 

cubic Feet for Production Time Lapse Time (54.02 down)
A Machine-hour 92.8 ft.3 42.7 ft.3

8 Hour Day 742.1 341.4

40 Hour Week 3710.4 1706.8

50 Week Year l85,520.0 85,339.2

258 Mills with 6 6

1.125 saws per 53.9 x 10 24.8 x 10

Midhigan Mill

To calculate the output per man-machine-hour the output per
machine-hour is divided by the number of persons operation and tailing

the saw. The Table Number VI, presented in the appendix, indicates the

22

statistics for the output of lumber in cubic feet per man-machine-hour
for the different conditions involved when two, three, or four men were
working the gang rip saw.

The output potential for the Michigan pallet mills is somewhere
between the extrapolated production time of estimated output of roughly
53.9 million cubic feet and the lapse time calculated value of 24.8
million cubic feet for 1974. While many factors influence a pallet mill's
output including economic conditions, worker motivation, quantity and
quality of logs or cants purchased, demand for pallets, working conditions,
equipment condition, and so on, the present systems could definitely be
improved. If the above conditions were favorable or upgraded the pallet
mills could increase production.

Assuming that 28.08 percent of all lumber is clear the total clear
quantity of deck boards which could have been redirected into higher
quality uses was about 7.0 million cubic feet or 83,464 thousand board
feet. Of course almost all of the clear wood went into pallet production
in 1974. If as previously discussed the lumber production was increased
in the future either by employing more favorable and more efficient
conditions the quantity of clear output could be expected to increase
proportionately.

While this study began by grading cants for furniture dimension
some additional data were gathered by grading stacks of pallet deck
boards. The same grading system adapted well to grading deck boards.
Obviously some clear deck boards originated from clear cants but other
clear deck boards could be sawn from non—clear cants depending on the
nature of the cants' defect(s). The results of grading stacks of pallet

lumber showed a wide range, 25.3 percent of clear boards, with a high of

23

37.72 percent and a low of 12.39 percent. The average percentage of
clear boards in a stack of lumber, 28.08, with a standard deviation
of 9.24 percent shows a higher value than projected by grading the
incoming cants, the quantity of clear wood will be best determined
from the output figures calculated and discussed in the next section.

(Data sheet presented in appendix).

WASTE

The amount of waste encountered in ripping cants into lumber
depends upon kerf, lumber size, quality of cants, and the pallet's
specifications. While pallet specifications and the cant input quality
remained stable from mill to mill the saws tended to be either carbide
(about 3/16 inch kerf) or insert-type tooth (roughly 1/4 inch kerf).
For the purposes of comparison the most common thickness of lumber
produced, l/2 inch, was used. Investigation into the amount of waste
(sawdust, defective wood, etc.) is important because it is a direct
reflection of the yield of lumber. Data on waste generated from end
trimmings was not measured and is therefore not included.

To calculate the data for Table V a given input quantity of
wood from a day's production of 1/2 inch deck boards was divided into
the actual quantity of 1/2 inch boards produced. The result is the
yield in percent which when subtracted from 100 percent gives the

percent of waste. The composite data appears on the next page:

24

TABLE V. Waste Data For 1/2 Inch Deck Boards

 

Saw Tooth 'High Low Range Average Standard Number of
Type Propor- PrOPOI’ Deviation Observations
tion (Z) tion (Z) (perceng) (percent) (percent)
Carbide 29.51 21.57 7.94 ' 26.29 4.18 9
Insert 47.80 28.86 18.94 35.52 7.63 5

These comparisons were difficult to make due to limited observations of
1/2 inch lumber. The above data used three-sided cants with volumes
reduced by 25 percent to equate to four-sided cants to calculate the
percent waste from inserted-tooth saws. The inserted tooth data also
came from five observations at one mill's machine. Others showed waste
factors but not enough observations were made at each mill to calculate

significant statistics.

DATA COLLECTED IN THE LABORATORY 0N HANDLING AND DRYING

 

Experimentation with various drying techniques at the Michigan
State university dry kiln resulted in information relative to both
drying and handling. "One of the major methods of reducing costs in the
sawmill (and rough mill) today is in developing good and efficient
lumber-handling techniques."12

In the first kiln charge (see photo, page 25) of unsurfaced
(rough) and presurfaced 3/4 and 5/8 inch respectively red oak pallet
deck boards the presurfacing had significant advantages. The thinner
lumber dried faster and more evenly to the desired final moisture content
of seven percent from an original 33.7 percent than the rough lumber; the
kiln conditions followed schedule T4-D2 from the Dry Kiln Operator's

Manual.30 A major advantage of presurfacing was that the thinner boards

25

J52 no 33

vooouusuoum van uuog no

.33 8:2 fin. ads .05 E3333

.N

ouswam

 

26

consumed 11.2 percent less kiln space. Although the smoother, lighter
presurfaced boards were easier to handle, the true test-conversion to
drawer side stock at the furniture company rough mill showed handling
and processing of either rough or presurfaced wood to be equal. No
difference in surface checking was observed and stress tests showed all
stresses to be relieved after conditioning.

The second drying experiment employed a similar material and kiln
schedule--3/4, rough sawn, red oak, pallet deck boards, and T4-D2. The
method of loading the kiln proved to be the significant outcome of this
second charge of oak. The deck boards were stacked tightly side-by-side
and stickered with dry, 3/4 by 1 1/2 inch kiln stickers. Because no
steel strapping was used to keep the stacked lumber tight during drying,
the boards warped significantly more than in the previous charge of
banded deck boards. The effect of increased warp, crook, bow, and cup,
was partially nullified at the furniture company's rough mill due to
short, thin size of the cuttings produced for their particular drawer
sides.

The third test charge of 4/4, soft maple, pallet deck boards was
again banded which restrained warp (see photograph, page 27). A disad-
vantage of banding was the discoloration to the six self-stickered
stickers on the outside of the pile. No loss was incurred as the stain
was expected; therefore low grade deck boards of lesser value were used
in these six positions. The anticipated result of quick drying from
green to seven percent occurred in approximately 100 hours without drying

30

defects. The Dry Kiln Operator's Manual kiln schedule, T8-D4, may be

 

found in the appendix.

27

.—
..
L. .. .6.

$4..._.... WV... 1. .. A

Juvrokfl gm]. ... fat 3 W?» ,.
.39.: “com in No 83.6 38. BEN. fit.

oHHM mun huououonou.

     

.n enough

28

The method of restraining the deck boards from warping during
drying requires further elaboration. Three rows of steel strapping
encircled each pile which were fastened at a convenient location in the
normal manner using strapping clips. A major difference was that two
"tails" were fastened to each band with the same slips allowing reten-
sioning of each strap as necessitated during drying, (see the photo on
page 29). Old clips were cut off with tin snips and a new tail pulled
taunt before another clip was applied. The tensioning procedure was
repeated once or twice for each package in the test charges.

The bundling of the lumber for drying also facilitated its hand-
ling by fork-lift trucks which are common pieces of equipment in both
pallet mills and furniture companies. The self—stickered packages were
designed to be easily stacked and strapped on a pallet at the lumber
producers. Jigs for uniform, square-sided stacks with sticker guides
for correct alignment would ease stacking at the pallet mill. The
package could be fork-lifted to a truck, transported to a furniture
company, fork-lifted directly into the dry kiln, and finally fork-lifted
or carted into the rough mill. (See Process Flow Chart in the appendix).

Overall the kiln schedules applied to the oak and maple lumber
packages dried the deck boards to the desired moisture content, with a
minimal amount of defects, in a minimum amount of kiln time. The
packages will facilitate handling between mills as well as restraining

the lumber during drying.

DATA COLLECTION AT FURNITURE ROUGH MILLS

Samples of pallet deck boards processed at three cooperating

furniture rough mills became high quality furniture drawer side stock.

29

 

....

‘4

.9":ch mo Oman new $5.95 wows—6 unassumom wow «:3. ouuxm new: venom: medium ..N 0.3m:

30

Attention was placed on material costs and yields as information on the

handling, drying, gluing, and machining costs was not available.

FIRST ROUGH MILL TEST

The production sequence at the first rough mill involved cutting
a deck board to the desired drawer side length, surfacing to the rough
thickness, and ripping to width which would be glued together to the
rough dimension of a drawer side; other mills used a different produc~
tion sequence. Although the complete data appears in the appendix some
pertinent results will be presented here.

The first test utilized some of the 3/4 by 3 1/4 by 47 inch, red
oak, seven percent moisture content, pallet deck boards dried at Michigan
State University. After all inputs and outputs were measured the yield
was computed by surface measure (disregard thickness) to be 60.6 percent.
It should be explained that only two cutting lengths were sawn, 16 5/8
inches and some 46 7/8, which is contrary to normal production procedures.
Usually three or more cutting lengths provided to the cut off saw make
the most efficient use of the incoming stock's lengths. With a variety
of possible cutting lengths, clear cuttings may be obtained that best
fit the rough lumber lengths (especially with short lengths), avoid
defects within the piece, and produce the least waste. In Aubrey E.
Wylie's article, "Lumber Grades and Sizes in Relation to Product Yields"
he comments on the necessity of increasing the number of cutting
lengths:

"There can be no doubt that one cutting length at a time

would be ruinous. Any improvement beyond three or four

sizes would depend upon a reduction of the differences
between sizes chosen. As the number of cutting lengths

in the group increases the maximum loss becomes a
function of the spacing of lengths chosen."45

31

Another peculiarity of this case was that the widths to rip saw, 2 5/32
inches, were incongruent with the deck board width of 3 1/4 inches. The
excessive waste at the rip saw only compounded the cut-off saw waste to
give that final low yield figure, 60.6 percent. Many furniture companies
avoid the extra waste occurring from ripping by edge gluing boards to form
a long continuous panel moving through an electronic bonding machine such
as the Raybond model. The panel's width corresponds to desired drawer
side length; the rough width of each drawer side is Obtained by ripping
the long, glued panel to the designated size. The advantage of this
system, used in the other two sample rough mills, is a reduction in
waste and increase in yield.

The actual cost of the drawer sides is obtained by dividing the

original cost of lumber by the final yield:

Cost of Lumber $/MBF
Yield Z

 

= actual cost $/MBF

The calculating of actual yields and costs allows comparison of the use
of pallet deck board lumber to normally-used grade lumber. In this
first furniture rough mill the estimated cost of "clear" pallet lumber
was $200.00 per thousand board feet delivered to the plant green.
Dividing by the 60.6 percent yield gives a true cost of lumber to be
$330.03. Normally the furniture company experiences a 60.0 percent
yield from the 3/4, #1 and #2 common oak, including delivery, which
costs $171.95 per thousand board feet green. Division of the cost by
the yield shows the actual cost of drawer sides and backs to be $286.58
when using graded lumber. The cost of kiln drying will increase the

costs of the drawer sides.

32

Obviously, the cost of using pallet lumber would mean an increase
in raw material price if yield were always this low. Unfortunately the
uncustomary production operations employed during an economic and pro-

duction decline distorted the value of the above data.

SECOND ROUGH MILL TEST

The pallet stock cut at the second furniture company's rough mill
was the kiln dried, 3/4 soft maple. In comparison with the former test
this mill had four cutting lengths, 17 5/8, 20, 24, and 27 inches, which
made more efficient use of the 40 and 48 inch pallet deck boards. After
cross—cutting to length and ripping to width for edge gluing into drawer
sides the output divided by the input gave an average yield, (two loads
were run), of 87.2 percent. This high yield partially resulted from the
fact that the drawer sides were an exact multiple of the deck board
dimensions. With 68.0 percent being the normal yield in this company's
rough mill production of furniture drawer sides, the 87.2 percent yield
compares favorably.

To further emphasize the difference between the use of pallet
stock over the normal use of grade lumber the true cost of one-thousand
board feet used to produce drawer components may be calculated. The
purchase cost of the delivered, green lumber divided by the yield gives

the real cost of the processed drawer sides (see below).

 

 

 

 

Pallet Stock Cost _ $200.00/MBF _
Yield ‘ 88.4% ’ $227'27/MBF
Average Normaligizde Lumber Cost = $202é6géMBF = $300.98/MBF

Excluding kiln drying costs, which should be lower for the pallet

stock, a drawer side producer could expect to save $73.71 per thousand

33

board feet of actual dimension produced. Compounded over a month's or
year's time the cost reduction could be substantial; unfortunately the
company involved in this test could not estimate their yearly production

quantity of drawer sides.

THIRD ROUGH MILL TEST

The third kiln dried charge of 3/4 rough and 5/8 inch presurfaced
red oak contained approximately the same amount of lumber as the two
other trials, 371 board feet. The expected advantage of no rough
surfacing needed at the furniture company was negated when all of the
deck boards were planed to 9/16 inch. Correlation between the three
cutting lengths, 15 1/8, 18, and 29 inches, and the 49 inch lumber was
high but failure to remove previously set aside boards marked for defects
led to a lower than expected yield of 77.4 percent based on surface
measure (see complete data in appendix).

Calculation of the actual cost of one-thousand board feet of
lumber used to produce drawer sides allows comparison of normal input of

grade lumber and the alternative raw material of pallet stock:

 

 

 

 

Pallet Stock Cost _ $200.00/MBF _
YiEld - 77.47. "" $258.40/1’EBF
Average Normal Grade Lumber Cost _ $244.00/MBF _
Normal Yiéld - 55.0% - $443064/MBF

The difference in actual cost per thousand board foot of drawer
sides, $185.24, was larger than savings calculated for the two previous
trials. Part of the reason for the higher amount of savings became
apparent when this third firm disclosed that their drawer side stock was
purchased along with lumber for other parts. There is no distinction

made as to lumber for exclusive use as drawer sides, but rather the

34

lumber for drawer sides allegedly comes from other components' lumber
cut backs. Obviously the drawer side raw material can not be termed

"scrap"--the furniture manufacturer pays for the lumber.

DEMAND FOR DRAWER SIDES

Sampling in three furniture companies showed that drawer side
dimensions vary widely; it would be impossible to choose a common size.
Similarily the demand for drawer sides fluctuates within each firm as
well as between different firms. One company was able to offer an
estimate of their yearly annual consumption of drawer sides-300 thousand
board feet. Another company calculated that "a portion" of their purchase
of 25 thousand board feet of soft maple went for drawer side production.
Still another major Grand Rapids furniture manufacturer stated that the
25,000 pieces of furniture produced annually required about one million
board feet of drawer side dimension stock. The president of a furniture
company making lower-cost furniture stated that scrap from four by eight
foot sheets of particle board was used for their drawer parts.

The preceding figures indicate the inconsistent quantities of
drawer sides and backs that the different furniture companies consume.
Factors which influence the board footage of drawer sides produced
include style of furniture, price of furniture, and quantity of parti-
cular pieces produced. Because many manufacturers did not know the
amount or would not give the amount of drawer sides required, no accurate

figures could be estimated for drawer side demand.

35

RELATIVE COSTS

To offer some tangible means for evaluating the advantage of
using graded pallet lumber as the raw material for the manufacture of
furniture drawer sides the actual monetary values must be attached to
the respective materials and processes.

Discussions with pallet lumber producers gave insight to a few
costs. The costs of lumber production in pallet mills as described by
Yaptenco, Harold, and Wylie in their publication on pallet manufacturing
costs, include handling, trimming, slatting, canting, material, and
overhead (composed of "administrative overhead, taxes, insurance,
depreciation, supplies, and maintenance.")46 Based on general economic
conditions, which were relatively poor at the end of 1974 and the
beginning of 1975, pallet manufacturers quoted a price of about $200.00
per thousand board feet of green, 4/4, oak deck boards including shipping
or delivery to the furniture plant in stacked, steel—strapped, palletized
packages.

Before conversion to dimension stock at furniture rough mills
additional handling to and from the dry kiln and to the rough mill would
be incurred. A drying cost and possible storage cost might be included.
Self-stickering coupled with unitized packages should reduce appreciably
the normal price of kiln drying lumber. Edward M. Conway in a Forest
Products Journal article titled "Factors That Affect the Cost of Hardwood
Kiln Drying"8 stated:

Costs depend "upon its beginning and final moisture content,

the uniformity of moisture content required, the quality of

the drying, the efficiency of the kiln plant, the know-how

of the operator, the local labor costs, the proper quantity

for the charge, the mechanization of handling, the assigned

proportion of company burden, and many other factors. These

considerations are in addition to the obvious variables of
power rates, fuel costs and heat transfer efficiency,

36

maintenance, fire insurance, sticker breakage, and depre—

ciation on kilns and equipment. In America today, few

industrial processes in any field have as widely varied

costs as the kiln drying of a unit of hardwood lumber."

Typical prices paid by furniture manufacturers for lumber used in
drawer side production vary between plants and their lumber sources. A
low figure of $151.00 per thousand board feet of number 1 and number 2
common grade was found at one furniture company. Another firm used
better quality lumber for drawer stock and paid $244.04 per thousand
board feet. This latter firm purchased drawer side stock by obtaining
excess stock for other furniture parts and using what they termed "scrap"
for drawer sides.

As reported earlier the actual cost of the drawer sides was deter-
mined by dividing the purchase price of the lumber by the yield at the
end of the rough mill. Actual cost of grade lumber was $286.58 to
$443.64 per thousand board feet based on yield; but the actual cost of
the selected deck boards was $227.27 to $330.03 per thousand board feet,
including delivery charges.

The above figures were collected at furniture company rough mills
but would be applicable to dimension stock producers who could also
manufacture drawer sides. From "Cost Analysis for WOod Products Manufac-
turing" by Henry A. Huber and George Vasiliou comes the following quote:

"For example, 7/16-inch drawer sides and backs can be

purchased, FOB mill, for about 82 percent of the cost of
making them. The company should investigate the possi-
bility of a standardization program to reduce the number
of cross-sections of drawer stock required. It then
might be possible to buy such stock processed through
the moulder (but not to finish length), in a few single
or multiple lengths. In this way, the company could
maintain control over the critical areas of shoulder

lengths and joint details, but still take advantage of
volume purchasing."15

37

While buying drawer stock dimension might have advantages; each company
would have to accurately assess their own costs. The dimension producer

certainly is a key market for the pallet mill lumber companies.

FUTURE PREDICTIONS

 

In these unstable economic times it becomes difficult to evaluate
and predict future trends. Some values will be pertinent regardless of
general economic conditions. Those values are the relatively high
overall savings incurred when sawing drawer sides from quality pallet
stock rather than grade hardwood lumber.

To the pallet industry whose business fluctuates much more than
the normal economy, alternative product markets are needed to smoothen
Operations. Fluctuations occur because "pallets are required before any
manufactured parts or finished goods can be delivered. As the manufac-
turing tempo increases, more pallets are required and when orders for
manufactured goods decline so does the need for pallets."16 One strong
market for products which could be produced by Michigan pallet mills
without altering basic present production methods would be to produce
drawer side quality material for furniture companies.

Regardless which way the United States' economy moves the Michigan
pallet mills will have to broaden their products and/or increase pallet
sales to operate more profitably. "More time should be spent trying to
find additional or better markets for the specific items that can be

"23 Obviously the sale of deck

produced from the material available.
boards for dimension stock to furniture manufacturers or dimension

producers would be one feasible market.

38

The furniture producer cannot escape the unfavorable economic
conditions surrounding the United States, including the wood products
industry. Even in more prosperous times the furniture company would
aspire to increase their profit margin by lowering production costs or
material costs. In R. M. Hallet's method of analyzing rough mill pro-
ductivity he stated that "efforts at cost reduction should be directed
at yields-~not productivity."13 From previously presented data the
results indicated that yields at furniture rough mills are very signi-
ficantly higher when employing the graded pallet stock as the raw
material.

Although economists disagree as to the immediate economic future
of the nation, the preceding leads to the conclusion that the producti-
vity and efficiency of the furniture and pallet firms could be improved.
The more progressive firms will be first to analyze and implement the
application of the new product market or raw material source in relation
to their business. Assuming that the data previously presented is
correct then adoption of the proposed system of redirecting pallet mill
produced deck boards into furniture drawer sides should prove profitable,

as well as conserving natural resources in the long run.

CONCLUSIONS AND RECOMMENDATIONS

Based on the data collected and analyzed from Michigan pallet and

furniture manufacturing companies the following conclusions on redirect-

ing pallet deck boards into furniture drawer sides may be drawn:

1.

The sale of pallet deck boards possessing the best appear-
ences would be a profitable, alternative market for pallet
sawmills.

Utilizing pallet lumber for dimension stock at a furniture
company or dimension producer for drawer side production
would result in cost reduction over the use of grade lumber.
It is estimated that 24.8 million cubic feet of mixed
species lumber, of which 7.0 million cubic feet was clear,
was produced by Michigan pallet mills in 1974.

Since graded pallet lumber would have few defects the high
yields resulting from integration of this material into
furniture drawer side production would be a function of
the correlation between cutting lengths and the original
lumber length.

The grade of a given cant is not always indicative of the
quality of each piece of lumber sawn from said cant,
especially in the cases of boxed-heart cants.

With only minor modifications to production operations the
pallet mills could meet the demand for high quality wooden

drawer side dimension stock.

39

40

7. The inclusion of kiln dried pallet stock for remanufac-
ture into drawer sides would have a negligible affect
on the normal production operations throughout a furni-
ture plant beyond the rough mill.

8. The two class, clear or non-clear, grading system was
applicable to both cants and deck boards; grading was
quick, accurate, and simple.

9. Self-stickered, steel-strapped, palletized packages of
deck board lumber stacked at the pallet mill may be
transported by fork-lift truck from the pallet mill to
the dry kiln, dried, and fork—lifted to the rough mill.

10. Close attention to the moisture content—based steps of
the Dry Kiln_Qperator's Manual kiln schedule success-
fully dried the 3/4 inch lumber quickly and evenly to
the desired seven percent moisture content; with
experience time-based steps for the same kiln schedule
will allow increased ease of operation.

11. It was impossible to accurately estimate the demand for
drawer sides by Michigan furniture manufacturers due to
complicated variables in each company's drawer side

production.

The recommendations presented here stem from the research done on
increasing utilization of hardwood lumber by redirecting high quality
pallet deck boards into dimension stock:

1. Several furniture companies should attempt a series of

trials to evaluate the potential cost reductions for

41

their particular firm; publication of their results would
provide valuable decision making tools for other companies.
The pallet deck board lumber would have other uses aside
from drawer sides such as structural components for

furniture which should also be investigated.

LITERATURE CITED

10.

11.

12.

13.

14.

LITERATURE CITED

Anonomous. 1970. Container maker find gold in furniture parts.
Wood and Wood Products 75(6):42, 46.

 

Baldwin, W.C. 1968. Pallets from low grade hardwoods. Forest
Products Journal 18(3):ll-13.

 

Blyth, James E., Boelter, Allan H., and Carl W. Danielson. 1975.
Primary forest products industry and timber use, Michigan, 1972.
USDAFS Resource Bulletin NC-24. 45 pp.

Bruce, Richard W. 1961. Factors affecting sawmill output. Forest
Products Journal 11(10):471-476.

 

Catteric, James W. 1973. Seasoning short boards or cuttings.
Southern Lumberman 226(2811)=13-20-

 

Compton, Kenneth C. 1956. Increasing sawmill efficiency. Forest
Products Journal 6(1)319‘27.

Conner, Richard M. 1952. Packaging and handling of lumber, dimension,
and hardwood flooring at a hardwood sawmill, furniture factory, and

flooring plant. Journal of the Forest Products Research Society
2(5):28-3l.

 

Conway, Edward M. 1958. Factors that affect the cost of hardwood
kiln drying. Forest Products Journal 8(1):21-A-24-A.

 

Creighton, J.W. 1956. Choosing a hardwood lumber grade for furniture
manufacture. Forest Products Journal 6(1):ll-15.

 

Daignault, L.G. 1969. Economics of manufacture of wood components.
Forest Products Journal l9(l):13-15.

 

Fowler, J.C. 1958. Packaging lumber saves time and money. Forest
Products Journal 8(10):80—A-82-A.

 

Glesne, Nels. 1959. Small mill lumber handling. Forest Products
Journal 9(7):9-A-lO-A.

 

Hallett, R.M. 1972. A method to analyze rough mill productivity.
Forest Products Journal 22(11):22-27.

 

Holemo, Fredrick J. and Peter J. Dyson. 1972. Ratio-delay-a method
for analyzing down time in sawmills. Forest Products Journal 22
(8):56-60.

42

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

43

Huber, Henry A. and George Vasiliou. 1968. Cost analysis in the
wood products industry. MSU Cooperative Extension Service, East

Lansing, 54 pp.

Huber, Henry A. 1975. The use of financial indicators for wood
products firms. MSU Cooperative Extension Bulletin E-770, 7 pp.

Huber, Henry A., Sliker, Alan W., and Michael J. Kroon. 1975.
Production of high quality wood from hard maple bolts. MSU
Cooperative Extension Bulletin E-805, 7 pp.

Keppler, W.E., Huxter, W.T., Jr., and P.J. Dyson. 1962. WOrk
measurements. North Carolina Agricultural Extension Service,
36 pp.

Kroon, Michael J. 1974. Sawing furniture quality dimension from
short log bolts. Master's Thesis, MSU, East Lansing, 67 pp.

Leney, Lawrence. 1964. Checking of planed and rough red oak during
kiln drying. Forest Products Journal 14(3):103-105.

 

Lockhard, C.R., Putnam, J.A., and R.D. Carpenter. 1963. Grade
defects in hardwood timber and logs. USDA Handbook No. 244, 39 pp.

Lunstrum, Stanford J. 1972. Circular sawmills and their efficient
operation. USDAFS, Southeastern Area, Atlanta, Georgia, 86 pp.

Malcolm, Fredrick B. 1973. WOod products manufacture at small
sawmills and woodworking plants. USDAFS Research Note FPL-087,

16 pp.

Manthy, Robert S., James, Lee M., and Henry A. Huber. 1973. Michigan
timber production - now and 1985. MSU Agriculture Experiment
Station and Cooperative Extension Service, Research Report 192,

23 pp.

Masengill, B.W. 1952. Rough mill efficiency. Journal of the Forest
Products Research Society 2(4):33-34, 68.

McLauchlan, T.A. 1972. Recent developments in circular rip sawing.
Forest Products Journal 22(6):42-48.

McMillen, John M. 1969. Accelerated kiln drying of pre-surfaced
one-inch northern red oak. USDAFS Research Paper, FPL-122.

Michigan Department of Natural Resources Forestry Division. 1974.
Directory of primary wood using plants in Michigan, 1974.

MSU Forestry Department. 1973. Directory of Michigan pallet
manufacturers.

Rasmussen, Edmund F. 1961 Dry Kiln Operator's Manual. USDAFS
Agricultural Handbook No. 188, 197 pp.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

44

Reynolds, Hugh W. 1968. Grading and unitized seasoning of oak
cuttings. WOod and WOod Products, Reprint by the USDAFS.

Reynolds, Hugh W. 1969. The sholo (short log) mill. Southern
Lumberman 219(2728):182-l85.

Rice, William W. 1971. Field test of a schedule for accelerated
kiln drying of presurfaced one-inch northern red oak. Research
Bulletin 595, Agriculture Experiment Station, University of
Massachusetts, Amherst.

Rietz, Raymond C. and Joseph A. Jenson. 1966. Presurfacing green
oak lumber to reduce surface checking. USDAFS Research Note

Rietz, Raymond C. and R.H. Page. 1971. Air drying of lumber: a
grade to industry practices. USDAFS Agriculture Handbook No.
402, 110 pp.

Sage, Roy L. and John Olsen. 1975. Circular gang rip-saws for
short hardwood pallet cants. Southern Lumberman 230(2856):13-14.

Sardo, William H., Jr. 1958. Pallets play an important part in
materials handling. Forest Products Journal 8(7):22—A-25~A.

Sardo, William H., Jr. 1961. Whoden pallets and pallet containers
... present and future markets. Forest Products Journal 11(12):
576-578.

Schumann, David R. and George H. Englerth. 1967. Yields of randomr
width dimension from 4/4 hard maple lumber. USDAFS Research Paper
FPL 81, 13 pp.

Simson, William T. and R.C. Bates. 1972. Accelerating oak air
drying by presurfacing. USDAFS Research Note FPL-0223.

Smith, walton R. 1967. Simplified guidelines to hardwood lumber
grading. USDAFS Agriculture Handbook No. 402, 110 pp.

Thomas, Richard J. 1965. Analysis of yield of dimension stock from
standard lumber grades. Forest Products Journal 15(7):285-288.

U. S. Forest Products Laboratory. 1971. WOod pallet manufacturing.
USDAFS Research Note FPL-0213, 35 pp.

U. S. Forest Products Laboratory. 1974. WOod Handbook: Whod as an
Engineering_Material. USDAFS Agriculture Handbook No. 72.

Hylie, Arbrey E. 1954. Lumber grades and sizes in relation to
product yields. lggurnal of the Forest Products Research Sociegy
4(4):l76-180.

Yaptenco, Rodolfo C., Harold, Monte R., and Aubrey E. Wylie. 1968.
Pallet manufacturing costs: a case study. Quarterly Bulletin of
Michigan Agricultural Experiment Station, MSU, 50(3):365-383.

APPENDICES

 

DRY KILN OPERATOR'S MANUAL KILN SCHEDULE FOR 4/4 RED OAK T4-D2

APPENDIX A

 

 

 

. 5'81 3‘ n. k‘ E;
a a. E“ ° ° a
326: can: c>h. El 25 E; §§c>
=> LDC) :3 C> L)
5:5; #35; E4 5394 #35; :35. gig: g;
1% 55 Egg; :3}: 5503 53;: Pdhd flag;
a: m: man: mafii can: Pic: #1
o. 6. P454 cu :5t4 .4
8%: “a saw 22% Ha SE a as
Eden Egcn c>ea c154 §§:: E: 22:: crca
zz<z s12:
l 1 50+Z 110° 4° 106° 872 17.52
1 2 50 Z 110° 5° 105° 842 16.2%
1 3 40 Z 110° 8° 102° 75% 13.3Z
1 4 35 Z 110° 14° 96° 60% 9.92
2 5 30 Z 120° 30° 90° 31% 5.42
3 6 25 Z 130° 50° 80° lOZ 2.0Z
130° 23° 107° 47% 7.3%
4 7 20 Z 140° 50° 90° 142 2.61
130° 10° 120° 73% 12.12
5 8 15 Z 180° 50° 130° 262 3.32

 

 

 

 

 

 

 

 

 

*
The figures above steps 7 and 8 are suggested equalizing and
conditioning treatments by Dry Kiln Operator's Manual.

**
Figures for steps 7 and 8 are FPL 1951 schedule.

APPENDIX B

DRY KILN OPERATOR'S MANUAL KILN SCHEDULE FOR 4/4 SOFT MAPLE T8-D4

 

 

 

 

 

 

 

 

 

 

*‘Ei

a. .. .. E
e: E2"’ ° 3‘ ° E;
M land C>h Egcp
gm DE. 00 E g g HO
:5 5% ES? 5. 55 a; 55 a;

mm mm [as mm 0022 m E-IQ 0-3
55 “E3 8.. RE *8 PE 95 '33
Hm Em z< cu. Ea g9 88
50+2 130° 7° 123° 81% 14.01
50 2 130° 10° 120° 73: 12.1:
40 2 130° 15° 115° 62Z 9.72
35 2 130° 25° 105° 43% 6.82
30 2 140° 40° 100° 25: 4.1:
25 2 150° 50° 100° 8: 2.9:
20 2 160° 50° 110° 212 3.21
15 2 130° 50° 130° 26: 3.3:

170° 160°

 

 

APPENDIX C

RESULTS OF GRADING INDIVIDUAL PALLET DECK BOARDS

Data Collected at Mill #5 Species:

Oak Stock

Size

 

Trial #1 113 Boards (175.6 bd. ft.)
Trial #2 228 Boards (225.5 bd. ft.)
Trial #3 127 Boards (125.6 bd. ft.)

Trial #4 216 Boards (337.5 bd. ft.)

Data Collected at Mill #4 Species:

Trial #1 320 Boards (226.7 bd. ft.)

Trial #2 554 Boards (1,108.0 bd. ft.

Data Collected at Mill #5 Species:

12.39% Clear 7/8 x
37.72% Clear 7/8 x
37.00% Clear 7/8 x

28.70% Clear 7/8 x

Soft Maple Stock

34.38% Clear 3/4 x

) 21.48% Clear 1 1/2

Oak Stock

Trial #1 1,737 Boards (1,411.3 bd.ft.) 24.87% Clear 3/4

Data Collected at a Non-sagple Mill

Species: Oak Stock

Trial #1 1,092 Boards (1,251.3 bd.ft.) 6.14% C1ear* 3/4

Trial #2 327 Boards (374.7 bd. ft.

) 6.42% Clear* 3/4

5 1/2 x 46 1/2"
3 1/2 x 46 1/2"
3 1/2 x 46 1/2"

5 1/2 x 46 3/4"

Size
4 x 40"

x 4 x 48"

Size

x 3 1/4 x 48"

Size
x 5 1/2 x 40"

x 5 1/2 x 40"

*
High quality material had been previously graded and removed

APPENDIX D

Table VI. Outputs Per Man-Machine-Hour (Cubic Feet)
Number High Low Range Average Standard Number of
of Men Value Value Deviation Observations
2 68.33 21.85 46.48 39.17 15.11 23
3 88.66 12.02 76.64 42.67 22.75 20
4 38.06 25.52 12.54 33.47 5.81 4

Combined Average 40.17

Some interpretations concerning the statistics presented in the
table should be offered. The ranges for the outputs were wide due to
the variety of lumber sizes produced. Saw speeds kept approximately the
same revolutions per minute whether 1/2 or 1 1/2 inch lumber was cut.
But the infeed-outfeed rate for 1/2 inch boards was relatively slower
than for thicker lumber. This was especially true when the less-powerful
saws had to cut six, 1/2 inch deck boards from a five inch wide cant

versus three 1 1/2 inch wide runners from the same width cant for an

example.

The average output per mandmachine-hour for two men and three men
was only 3.5 cubic feet different. As in the input per man-machine-hour
this slight gain in output by using one more man was negated by increasing
the infeed material by the saw operator. While the output increases by
adding a second tailer to a saw, the output per man~machine-hour remains

approximately the same. One additional man tailing, four total men,

proves to be extra labor since the output per man-machine-hour
decreases. More idle time occurs when four men work one saw since the
operator cannot feed beyond a saw's maximum rate. The idle time did
not constitute wasted labor however if the extra time was used for

sorting, stacking, off—bearing loads of lumber, etc.

APPENDIX E

Figure 5. Process Flow Chart

 

LOG
CANT PALLET MILL
LUMBER
"Not Clear" "Clear"
Grade Grade
Pallets Self-Stickered

Strapped Package

 

 

Fork — Lift to truck

then Fork - Lift again to

 

\V

 

DRY KILN
Fork - Lift Truck
or Kiln Car

FURNITUM

COMPANY
ROUGH MILL

 

 

 

 

APPENDIX F

FIRST FURNITURE COMPANY YIELD STUDY
March 10, 1975
RUN #1 -- NO DRYING DEFECTS
INPUT
289 pieces of pallet deckboard, oak

289 (3/4" x 3 1/4" x 47") = 307 bd. ft. by Surface Measure

CUT OFF + SURFACER +RIP SAW

 

Because of a sales decline the operations were observed at limited pro-
duction. Only two sizes of drawer sides were sawn; much higher yields
would be possible with more cutting lengths and random widths.

OUTPUT
325 pieces (5/8" x 3 1/4" x 16 5/8") = 122 bd. ft.
91 pieces (5/8" x 2 5/32" x 46 7/8") = 64 bd. ft.

 

186 bd. ft. total
YIELD

 

186 bd. ft.
307 bd. ft.

************************************************************************

 

100 = 60.6%

RUN #2 -- CONTAINED ALL DRYING DEFECTS
my:
60 pieces (3/4" x 3 1/4" x 47") = 64 bd. ft.
CUT OFF + SURFACER‘+ RIP SAW

 

 

OUTPUT
89 pieces (5/8" 3 1/4" x 16 5/8") = 33 bd. ft.
YIELD
33 bd. ft. _ a
64 bd. ft. 100 ' 51'6‘

************************************************************************

COSTS
Normal material used for drawer sides: 3/4", #1 and #2 Common, Oak

Cost of Lumber
Yield

 

= actual cost of drawer sides produced in the rough mill.

$171.95 MBF not kiln dried / Yield 60.0% = $268.58

Pallet deckboards - an alternative drawer side material 3/4", green oak,
clear, delivered, not kiln dried.

Cost of Lumber $200.00 MBF
Yield 60.6%

 

$330.03 MBF actual cost plus drying costs.

APPENDIX G

SECOND FURNITURE COMPANY YIELD STUDY
March 14, 1975

 

 

11111112
166 pieces (3/4 x 3 3/4 x 48") = 207.5 bd. ft. by surface measure
maple
159 pieces (3/4 x 3 3/4 x 40") = 165.5 bd. ft. by surface measure
373.1 bd. ft. by surface measure
CUT OFF
332 pieces (3/4 x 3 3/4 x 24") = 207.5 bd. ft. by surface measure
214 pieces (3/4 x 3 3/4 x 20") = 111.5 bd. ft. by surface measure
319.0 bd. ft. by surface measure
52 pieces (344 x 3 3/4 x 40") = 54.1 bd. ft. were set aside for running

full length through the moulder for
conversion into 3/4 x 3 1/2 x 17 5/8

 

 

and 27".
Percent of input yielded = 319.0 bd. ft. _ .
319.0 bd. ft. ‘ 100.02
(319.0 is 373.1 less 54.1 removed)
RIP SAW
21 Layers 48.75 x 24" = 170.6 bd. ft. by surface measure
1 Layer 41.75 x 24" = 7.0 bd. ft. by surface measure
14 Layers 50.50 x 20" = 98.2 bd. ft. by surface measure
1 Layer 17.50 x 20" = 2.4 bd. ft. by surface measure
278.2 bd. ft. by surface measure

OUTPUT

_ 278.2 bd. ft. _ .
Percent of input yielded - 319.0 bd. ft. - 87.2%

 

**************************************************id:********************

 

 

 

COSTS
Normal drawer side material cost _ $204.67/MBF =

Normal Yield - 68.0% $300°98/MBF
Pallet Lumber Cost _ 4§200.00/MBF = $229.33/MBF

 

 

Yield " 87.2%

APPENDIX H

THIRD FURNITURE COMPANY YIELD STUDY
April 16, 1975
INPUT
342 pieces of pallet deckboard, oak
342 (3/4" x 4 1/4" x 49") = 495 bd. ft. by surface measure

SURFACE to 9/16"

 

 

 

 

CUT OFF
Length P332991 MEI; Bd. Ft .
29 326 4 1/4" 279
18 299 4 1/4" 159
15 1/8 64 4 1/4" 29
466 Bd. Ft.
11m
Percent of input yielded = 94.2% = 2%? :3: it: x 100
‘RIE
Length Width of Panel Pieces (panels) Bd. Ft.
29 10 1/4 96 198
18 10 1/4 112 144
15 1/8 10 1/4 19 21
363 Bd.Ft.
315.119.
Percent of input yielded = 363 bd' ft' x 100 = 73.36%

 

495 bd. ft.

CUT BACK SALVAGE;

 

 

Length Width Pieces (panels)
15 10 1/4 19
OUTPUT

TOTAL YIELD

 

383 Bd. Ft.
495 Bd. Ft.

 

100 = 77.4%

Bd. Ft.

20

 

383 Bd.Ft. Total

**************************************************************************

COSTS

Normal Drawer Side Material 4/4, #1 Common and better, green oak

$244 MBF

Cost calculated as follows: #1 Common and better includes:

50% FAS @ $274/MBF
20% SEL @ $264/MBF
25% #1C @ $189/MBF

5% #2C @ $130/MBF

$244
55%

 

Standard yield at rough mill = 55% = $443.64 MBF

$137
$53
$ 47
$ 7

$244

actual cost without
kiln drying.

Pallet Deckboards - an alternative drawer side material 3/4, green oak,

$200
77.4%

 

clear $200 MBF delivered

cost without kiln drying.

$258.40 MBF actual

>mmmchx H.

O>H> H>mCfi>HHOZm flow mHnIH w>brma KHrrm.
ZHfir zczumx H

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

654m z>o=Hzmxa ”axe nsza «accuse gazes memw “mmmwm “MMCMM mwwnm mmwmw\ hummwv nmm>w wwwwmw mmmmmm hwmmmmu ommwme N
szca as we mocx nc -ma. ac 14 :somumw. :sz-mw. o>za as as mos» n: ma. N om zsmam
. . ca. 44. . . . . oc. ma. oc. ma. Hzmca . . as. we. . . szca
HH\HH\us nonawwp “\Hm sumaama H\~= w\a_ H.m~ map.sw wom.sm Ho~.s~ as.mu sw.6~ Hm.wo Hs.mm wso.oo www.8o s~.sm mo.us us.~m
our r66. u.~o Hwo.om no.66 Ho.um u.ss os.m~ NN.~6
m\~w\us nonsopp u\pa s-6»a6a H\~= w\a p.66 spm.ss -u.up ms.su wo.us No.pa 6.0» 6.5» ~s~.oo puo.- up.uu No.Nm No.-
>666: emu. ~.sm HuH.ms us.mo ps.up w.um Hos.mp ~o.ou .
m\~s\ss nonsmHH u\Hs 6-6Ha6a H\~= w\a. H.6m uos.oV HsV.uH sm.ss ss.Ho NU.NS pp.os Hp.mp som.mu Ham.ou um.OH us.oo NH.oo
>6666 was ~.so Hus.mu su.su Hm.ou s.oo Hoo.uu uu.~o
o\8\us nouawpw “\Hs b-6Hama ~\~= m\a _p.o~ VH.Hs 66.60 ~u.~m HS.H~ Hs.mu u.o~ Ns.Hs sw.~s so.os ~o.ow ms.os Hm. s
,sz6a was _H.HV oo.m~ ~o.~s Hs.ss s.mo u~.us H~.su
o\u\ss 06666HH S\Ho sumaaaa H\~= w\a N.wp www.ms ku.oo ss.us so.~m Ns.oo m.oo Hm.ou ”so.sm om.- w~.ss su.mm ~s.p~
szma ~66 _s.mo us.~m Ns.su Hm.mm a.~o ms.m~ Hs.~o
aoa>r w\a _o.~m H.qu.~s H.~uo.sw so.so No.up
r66 .Hs.s~
><o. was.so os.sm Hm.so Huo.um so.po
mao.sm<. _ ms.ss ms.mm u.ss 60.06 ~w.ms
HOHZ. EH. _ 5.0m Quota
r66 .-.u~ s~.m
pH\HH-H~ nonsopw w\po suwsama “\s= w\a H.so ”Hm.~m HMM.¢H mH.os Hm.o~ p~.ms s.No m.Ns Huo.u~ Hos.u~ uu.uo om.mu wH.Hu
oar r66 _~.uo os.oo up.sw s.mw ~.6H ow.us NH.so
W\HS\um nonamyw w\ps sumaama u\m= m\a _~.wu smm.so Nom.o~ sm.ss um.au Ns.ss m.~u H~.o~ mNm.Ho www.6o sm.mu ou.pp u».mo
omam was away. our 666 PFoo HNH.sm so.ss Hs.as s.ms mp.mu ~s.pm

 

 

>wwm25Hx H.

U>H> H>mCfi>HHOZm mow mHomH w>firma zHrbm.
KHFF ZCZwmw N

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

c>am z>omHzmxe xmwm osza mwoecoa aHzmm mmwww “mmmwm mewmm ”wwwm mmwmm\ ”wwww\ nrm>w magma» mmmmmm ,mmmmmm oMMmma N
Hzmqa an 44 20cm 0: ma. me we zsomuzx. g>znmw. o>za a: ma moan ac -ma. N ow zsmam
. . oe. ma. . . . . ac. ma. me. 44. szce . . ac. ma. . . szca
.411
H\Hs\um :HHH .Hsm sumsama u\s: w\a _o.sm ~Hs.ss NNN.Hm Vs.os mp.uo mu.u~ ~u.~o wu.m~ ps6.om Ho~.sw um.m~ so.o~ wo.u
m\w~ our emnwsam. r66 m.oo ~6.- s.o~ Ho.~p u.so No.up o.ms
HH\os\ss s16Hama w\a .o.o~ Nsw.oo Nmm.sm ou.ms ~H.ss ~w.¢¢ ~.mo m.~s Hou.us NHo.H~ so.os uu.uu ~s.s~
>666: r66 _~.H~ H-.m~ so.mH Ho.0u s.ws mo.HH No.uo
HN\Hm-No sum»ama m\a H.mm ~m6.o~ Hs~.us mu.sm v.66 w.um H.~s N.us Hsm.uo Hos.so us.mo os.m~ uu.pu
zwxma r66 .u.so oo.om ~w.0~ H.o~ .ws 7 ss.mp Hs.ss
snmwama w\a . .so mw.sa Hso.p~ uo.mu Hs.su Ns.om s.oH Hs.ss mp.~u Ho~.om ~m.m~ ow.ms uo.ps
:Fxma 666 H.»» so.mm Hu.- Ho.o~ ~.uo s~.uu Ho.ou
sumsama w\a _u.VH map.ow Hmp.so mo.so s~.~o -.V~ s.~s m.so us~.mu ow.ou up.- sp.mm um.pm
omx r66 .m.oo so.~s Nu.sH v.00 H.6u su.ss Hs.sm
aos>rm w\a u.uo H.umu.om ou~.mm o~.- u~.~2
Haw .meO
2m>z _ www.ms s~.om Hs.oo Hum.oo su.uo
was.sm<. oo.~o Nu.o~ up.o~ Hm.~s
. w}. .996. cot:
soa>r r66 _-.ou as.sN
m\Na-~o xHHH .Hsm sumnama u\m: 6\a pso.um www.ms Hos.mm Hs.mu uu.ss H~.uo Hp.om HHH.wa ~s~.so mN.uo so.HH ~o.ms
gum omr Umnwvmm. run _ mokG No.3 0.2. aid. aura» BIB.
s\S\o snmnama H\~: m\a _ “mm.us Hwo.os sm.mm Nu.uo Ho.mu o.o~ Hs.u~ Huu.mu HHs.HV um.oo mu.ms Hs.so
our Ap6\wwv r66 66.80 up.su Hu.so s.mo um.au No.~s
P

 

 

 

 

>wwmzuHN H.

U>H> H>wcfi>HHOZm mow mHnme w>hfimfi zHrrm.
zHrh zcxbmw u

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.... ...... in 25 a... fit ”is ...”... ...... 156-..... .fi. .
95m §8Hzm§ g 38:8. ass... 35.. . E215. 5.5 0565 . E219.
Hzmca me we was» as we as we :>om.=~. z>2-mw. 0524 a: we seem as we a on =>mam
. . as. ma. . . . . ac. ma. as. we. Hzmca . . on. ma. . . Hzmca
6\Ho-~m mmwz>znm HS\ss w-6»aaa U\s= m\e _ s.pu moo.su www.mo o6.os u~.sp.» q.ss w.mo HH.VH ~w~.mo um.uo ~a.as so.op mo.oo_
swung 666x666. r66 _ m.oo so.mw uu.s~ s.ou ~.om ~m.om as.so
6\Houwn m\a . H.Hs ~H6.so www.6s ow.os ~.oo.. ~.~s H.HN ~.ou Hop.ou mq.su su.m~ s~.ou u~.u~
was u.~m mu.ss u~.~u .mo .so up.pm pu.uo
m\~wnno AH\Hsv «\e _ H.HN ~ow.~o Hu~.os ma.o~ m.HN 6.6m u.so s.ow H~s.s~ Hoa.sp mu.sp m~.oo ~m.op
r66 _ ~.oo Hoo.ss no.8» s.os ~.o~ o~.sm uw.~s
H~\Ho-no w\a o
ruv _ a. co
Hoa>h w\a _ a. mo omH.ub bum.b~ &O.NH mw.~o
r66 _~ NH. Na
:msz _ Hsu.6N «H.6m o.p~ mu.ws sw.ss
mad.cm<. _ Nu.o» H~.m~ ~m.mm H~.oa
_ :93. 2.3.
s\po-p~ ”\m= w\a . u. m Hao.su sw.mm Hm.us H~.m~ u.sm H.6s s.so u~.- No.um m.~u as.ms mu.pw
r66 . s.ss uu.so Hw.mm ~.6H H.0m Hm.~s s.po
_ :81 5:3.
85:. in. .8. .3ch
run .Ns. oo sp.un
s\Ho-Hm scram» H\s= Susanna u\s= m\a _ H.wm s~.u~ ss.wu sm.uu ~.mw H.ma H.ms s.o~ ~u.~H Hm.su Hm.ou so.~o 66.“;
“H6\ zwxaa courage. r66 . ~.Ho No.mo No.mo H.~o H.~o H~.oo H~.oo
m\~p-~u 66V AHH\HsV w\a . H.ma Hpo.ou mo.ws us.- s.~o u.mm u.mw s.up su.s~ ~s.so Ns.so sp.~u um.uq
r66 w.~u No.6H No.6“ H.6N H.6N H~.H~ H~.H~ .
H~\Ho-~o w\e . 0
run _ m.oo
p

 

 

 

» u <mwcmm Hmmm NwN

4» I sen opp Haven «manna

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

smcmchx H. chc zczucx u Aaoua.a.v
a
at: can? “flaw Hzccd arm; Nazca”? wwwscv an“; coat. “Ham ocacca ”Name a
ct...” EacHzcwa scum 2683 can... HzccH. . 35-..». Hzcca 848a 521m?
szca ac we coca ac ca ac we z>acucw. :szucw. asza ac we mac» ac me a om c>m
. . ac. we. . . . . ac. ma. ac. ca. Hzcca . . ac. ca. . . Hzcca
[1 d
aoesc c\a_ u.- ch.- so.cu
ran Hu.mu
mm>z . u~.aa u~.~a u.~a NH.ua HH.ua so.ao ma.HH
mac.ca<. . o.os o.o~ a.ou a.cm
ac.ua. ao.ua.
_
c\HauH~ uuuHana ~\Ha c\a_ c.~o ma.aa au.ao ca.aa Hc.ac Hu.aa Hu.aw Ha.aH wH.wa aa.- aa.- ma.aa aa.u
:ona cps H.~a aa.ua aa.ua c.ua a.u~ Ha.ca ~a.ac
_ a~.au. a~.au.
a\HanHa cocacw H\a= auuHaoa H o\Ha= c\a_ H.~H Huu.~a 6H.Ho 6H.Ho o.aw m.au u.au a.Ha us.uo uH.aH aH.aH ua.aH au.o
aHa\ :Huoa weapons can. ~.am um.~o ua.sa u.aa a.aa ~o.ua ao.ua
sac
a\~6-~o H H\~: c\a. H.HH aa.ca u~.~s “6.6a c.0a s.~a V.Nu H~.mu w~.~H ao.o~ No.o~ uo.~u aa.um
cue. H.cH au.aH uu.aH a.aa a.aa Hu.aa Ha.co
aoa>c c\a. ~.c~ ~H6.cu aa.aa
hmv_ b.ao
zmsz _ Va.a~ as.a~ a.ca uo.a~ ac.a~ ao.uo ac.aH
.mac.ca<. ~a.ua ~a.ua H.ca H.6c
_ us.a~..» ma.a~.».
a\HouHN H ~\a= c\a. H.m~ ac.ac sa.ca aa.ac H.am H.Ha H.Ha N.ac mo.Hu uH.oa wH.cu s~.- ~c.~u
”cabana cps H.6a um.uH um.MH .ou .ou ~u.aa ~u.ca
. u~.cH» a~.oH.
goat...“ 3a. cLH c3:
vac. NH.uH aH.au
all

 

 

 

>wwm2UHx H.

U>H> H>wCH>HHOZm mow mHO:H w>rrma zHrrm.
ZHrF ZCmew b

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hzccc\ acm>w acm>w a ocawca ocacca
c>aa z>acHzmxc cmwc a>za cmoccag aHzcm mmwww :>ac.u memmm ”wwwm Hzcca\ Hcha\ acm>x mwmwmm :>ac.u "mmwwqa >m > a
Hzcca ac ma cocw ac ca. ac «a z>acucw. :szumx. a>za ac ca coca ac cc. a as cwma
. . ac. ma. . . . . ac. ma. ac. ca. Hzcca . . ac. ca. . . szca
H
HN\Hau~o cocamm H\a a16Ha6a c\a o
+ who: >mwmz rmv _ m.oo
a\~Huaa cam»: H\~= c\a . W.Ww aHa.aa HaH.aa Nu.aa am.HN Ha.as 5.0a Ha.aa as.Ha sa.HH HH.ua au.~a ma.ua
cmnwccm. cmc _ . a su.aa Ha.au HN.ao a.Ho am.aa a.aa
a\--~c c\a c
6\Ho-H~ MMH _ N mm
6 O
a\Ha\Vs H\~= c\a . ~.Ha “NH.sa Hsa.aa ca.uo aa.aa» aa.aa 6.66 No.0a Hcc.ca c~.Ho No.u~ uu.au sa.o~
cps ~.mo Haa.ma aN.Hu as.aa a.aa cH.c~ Hc.cm
Ada; M\H _NM.WW mbo.wm Num.oN
mv _ .
zm>z _ Haa.aa aH.ca -.aa aa.HH Ha.oa uH.ao aa.ao
aac.cac. SH.VV a.am aw.aa a.aa
_
c\Ha\us scream H\s c-6Ha6a H H\~= c\a . ~.- auo.as aca.Ha ma.sa Hca.aa so.0m HN.Na as.Hm aau.a~ HHa.sm as.ao ma.aa aH.us
+ 666: cinmc wannanm cmc a.ac Ham.a~ ca.cm a~.aa a.Hs so.Ha Ho.uc
aoa>c chm»: c\a . a.ua cocz
cmc .Na.aa Na.aH
b

 

r
zen mHH Haven mummma

 

>WMMZUHX H.

U>H> H>wcfi>HHOZm mow mHoza w>firme ZHrrm.
KHfir zczwww u

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9.24 .843. :Hmwcd Hzcci are; WWW”? ahwscv new; 8.2:. mmmmcm ocawca oMMcMH N
U>Hm z>o=Hzmxm warm wwoocna HHXMM szCH . z>21=x. szca ocawCH . zbzImw.
szca n: we soc» me me n: we z>omo:w. z>z|=x. n>zH 0: ma mocw 0: we N om z>mem
. . oc. ma. . . . . 0c. ma. 0:. ma. szce . . 0:. we. . . Hzmca
HH\Hm\ua :HHH aw «\un ulmnama wa: m\4Wfi1H.~o HmH.»H HmH.Hm mo.wo a.o¢ ».uH H.wu ~.uu au.~u ua.oo H~.oN Nu.mm ua.Hm
non<mnu raw . H.uo HNo.es bo.uH - u.uH H.Ho ~m.mm o.m~
mac: no .
almnama
our .
HN\Ho\~m mHHH «H .NHo omx w\s: w\a . u.wu aa~.uH HuH.Hu mm.u~ ab.om Hw.wb o.mu Ho.Hu ~wm.om uo.mm um.No uu.mu ao.wu
“voow but o.~m ow.H~ ou.mo e.>u a.~b mo.H~ Nu.om
mHHH a» .NHo our u\b: w\a . H.au Hum.uo ob.Hb uH.wm Nn.uo Hm.mo m.Hu Hw.uo mm.mm aH.H~ no.wu oa.wu uu.o~
uvoom run . ~.oo mo.~o Nu.o~ HH.wm u.uo s».o~ Ha.ou
H\Hm\uw mHHH HH our H w\b: m\H H.mm Hou.uo Hoa.so ua.mu HN.N¢ m.oo N.uo u.o~ www.mb om.um -.mm mu.o~ u§.um
has . N.Nm ms.~m ~m.~m u.oo H.mu uu.ou Hm.ua
Hoa>r u\>: wxa: m\a _ a.ma mmo.uo unu.wu um.ou su.uw
who _ m.uu
><o. u\b: u\a: w\a HHN.oa om.bm Ha.oo au.mu nu.mu
mec.cm<. u\a: u\a: w\a _ No.Ho ~b.Hm o.uo ¢.ao Ho.um
mHHH aw w\a _ o
rmv . ~.oo
H\~m\um mHHH HH m\a _ o
bun H.oo
HH\Hm\~a =HHH a» w\a _ 0
row a.oo
~\Ho\uu :HHH aw ~\m= w\a _ u.oH boo.~o HH~.mH u~.u> HH~.~o u~.au Ho.m~ Nm.mu ~w».uu oa.o~ NH.ma mu.uo u~.uo
our our ran _ s.uu mu.u~ nm.uH ~o.ou m.N~ so.ua Ho.bu
82cm Ea _ Hobs
raw _uH.ow
H\~5\cm :HHH an an “\ww H H\m: w\a . H.»b ~o~.oo Hbo.oo oo.oo Hu.oo H~.- o.ou m.oo H~m.OV mm.op Nw.mu mu.~H uo.um
mHama row . ~.oo How.uo ww.uu m.mo ~.ow oa.oa NH.wu
our _
HII

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>0wmchx H. zHcc zcz000 0 Aaosa.a.0
a>za aoa>c “Hmmam H20c40 acm>0 hwmwwv “mmwwv acm>0 aoa>r mmmmcm ocawqa ommwwa a
0000 0.5202000 080.0 00080.0. 0500 Hzcca . 0.52-00. 0205 8.005. . 202-00.
szca ac 00 sec” ac 00 ac 00 z>ac1cw. 002-00. a>za ac we 00:» ac 00 u 00 2000
. . ac. 00. . . . . ac. 00. ac. 00. Hz0c0 . . ac. 00. . . szca
HH000000 cHHH 00 our H 000: 000;fi|..00 H00.0H 000.00 00.00 0.00 0.00 0.00 0.00 HHO.H0 H00.00 00.00 00.00 00.00
000 _ H.00 H00.00 00.H0 0.0H H.00 00.H0 00.00
00000 000 _H0.00 00::
000 .00.00 00.00
HH0H0000 >Ham: 0000 0-00000 _
000 H 000= 00a _ 0.00 000.00 H00.00 00.00 0H.00 H0.00 0.00 H0.00 000.00 H00.00 00.00 00.00 00.0H
320w» _
000 000 0.00 00.00 00.00 HH.00 0.00 0H.00 00.00
our _
“(H . o
000 0.00
H000000 000 _ 0
r20. ~.oo
00H0000 000 0
ram. m.oo
00:.“ “Too coca
0000 H0.00 00.00

 

 

 

>wwmchx H.

U>H> H>wCH>HHOZm mow MHOSH w>rfima KHrrm.
KHrb zczumw o

 

 

 

0000 00000 “mmm0m 020000 00000 mmwmwv “wwwwv acm00 00000 mwmmcm 000000 000 00 a
0000 000002000 0000 0000ca0 00000 02000 . 002-00. 02000 000000 . 002-00.
0205 ac 00 0000 ac 00 ac 00 002-02. 002-00. 0000 ac 0.0 0000 ac 00 a 00 000000
. . ac. 00. . . . . ac. 00. ac. 00. 00000 . . ac. 00. . . 02000
H000000 00000000 .H00 0-0H000 H00: 000_-0.00 000.00 00.00 00.00 00.00 0.00 0.00 H0.00 0H0.00 00.00 00.00 0H.00 00.00
00000 000H00 000 _ 0.00 0H.00 H0.00 0.00 H.00 00.00 H0.00
f00-0000 000 . 0.00 000.00 00.00 00.00 0H.00 H0.00 0.00 H0.00 00H.00 00.00 00.00 00.00 00.00
000 0.0H 0H.00 00.00 H0.0H 0.00 00.00 00.H0
00000 000 . 0.00 000.00 000.0H 00.00 0H.00
rbv .H~.0o
00.0H 00.0H H0.00 00.00 0H.00
00.000. _ 00.H0 H0.00 H0.H0 0.00
00-0000 0-00000 H00: 000 _ 0.H0 H00.HH 00.H0 00.00 H0.00 0.00 0.00 H0.00 H00.00 00.00 00.H0 00.00 H0.00_
00009 ruv _ 0.ou uo.0H H0.0H 0.00 H.00 uo.0m Hu.wo
00000 000 _H0.00 cat:
can 00.00 00.00
00H-000 00000000 .H00 0-00000 000: 000 _ 0.0H 00.00 HOH.00 00.0H 0.00 H0.00 0.00 0.00 00.00 H00.00 00.0H 00.00 0H.00
0000: AHHOH00 000 . 0.00 H00.0H 00.0H 0.0H 0.H0 00.00 0H.00
0000-000 H 00H0= 000 _ 0.H0 000.00 H00.00 00.H0 00.00 00.HH H0.00 H0.00 000.00 HH0.00 00.00 00.00 00.00
0062000. 000. 0.00 H00.0H 0H.00 00.00 H0.H0 00.00 00.00
0000000 cHHH .H00 0-0H000 0: 0H0000 000. H.00 000.0H 000.00 HH0.00 00.00 00.00 H0.H0 0.00 00H.00 H00.0H 00.00 00.00 00.00
>mvmw won «on»: run m.00 0u.eo -.oo u.mH H.0H Nu.uN H~.ma
_ H00.0 00.00
0000-000 .H00 0-00000 0= 0H0000 000. .00 H00.00 H00.00 00.00 H0.00 H0.00 0.00 0.00 H00.00 H00.00 00.00 00.00 00.00
00000 000. .00 HOH.00 00.00 H0.00 0.00 HH0.00 00.00
00000 000. 0.00 000.00 000.HH 00.00 00.00
000 0.00
000. _ H00.0H 00.00 0.H0 H00.00 00.00
_ H0.00 0.00 H.00 .00
0.0000 000 0.00 023
ram_ o.¢m 0m.HmN

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

»
”macnma no 0100000

>wwmchx H.

U>H> H>mcfi>HH02m wow mwoma w>hfime ZHfirm.
ZHFF zczwmx 0

 

 

 

0020 . 00000 mummfi 020000 00000 0a2000%000 002%“,000 00“.; 00000 “Mmmcm 0c00c0 0NM0M0 0
0000 200002000 0000 02000 00000a0 00200 amzwmw 0000 mwz-mw. am20mw 20a0-00. 202-00. a020 0mw0mmw 0000 mwz-mm. N 00 0000L
. . ac. 00. . . . . ac. 00. ac. 00. 02000 . . ac. 00. . . 02000
0000-H0 000000 H00: 0-00002 H00: 0H000~ 000 0 0.00 000.00 HH0.00 00.00 00.00 H0.00 0.00 H0.00 000.00 0H.H0 00.00 00.00 00.00
20000 00- 20000 on 00000 000 . 0.00 00.0H 0H.00 H0.00 0.0H 00.00 00.H0
0000-HN 0000 000" 000 . 0.00 000.00 H0~.H0 0H.00 00.00 H0.00 0.0H H0.00 000.00 00.H0 00.00 0H.00 00.00
000 _ 0.0H 00.00 00.00 0.00 0.00 00.00 H0.H0
0000-00 000 _ 0.00 00.00 00.00 00.00 00.0H 00.00 H0.00 00.00 00.00 00.00 00.00 00.00 00.00
000 . H.00 00.0H 00.00 H0.00 0.00 00.00 H0.0H
0000-00 000 _ 0.00 000.00 00.00 00.00 00.00 00.00 H0.00 0H.00 H00.00 00.00 00.00 0H.00 00.00
000 _ 0.00 00.00 00.00 H0.H0 0.00 00.HH H0.00
H0000-00 000 _ H.00 000.00 H00.00 00.00 00.00 00.00 HH.00 H0.H0 H00.HH 00.HH 00.00 00.H0 00.00
000 _ 0.00 00.00 00.0H HH.00 0.00 00.00 H0.00
_ 00.00 00.H0
00000 0oz: 000 .H0.0~ H.000.~0 000.00
0 00.00 000 .N~.00
.2002 ” H00.00 00.0H H0.0H 00.00 00.H0
~000.00<. --_ H0.00 0.00 0.00 0.00

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>wwmzqu H.

U>H> H>uCH>HHozm wow mHnmd w>rrma zHfirm.
KHFfi zcxumw m

 

 

 

at: 8.5. “Manna“ szS< 9.5 mg.“ anwmfimw or“; .83. flaw 8.3.5. ommma N
o>am :panzmxw nmwm quca wwoocna aHzmm szca mos» z>z-=w. szce z>n=u=w z>zumw n>za ocawca mocw z>zumw. a om zpme
ac. we. ac. ma. 0:. «a. . . ac. me. me. we.
ac. ma. ac. ma. ac. ma. szca ac. ma. szca
ax]
\~V-o\» app“ .N m\a o
rmv _ p.00
\wu0\qo supp *~ H\a #:apaoa U\m= w\a . ~.oo uu~.- Hum.oo m~.eu Hu.~m m.~u ~.~m m.~o -u.~o Hoa.~w uu.mo ou.~o u~.mo
>uuoa umowuam. was _ u.uu oo.~u u.pe H.gw au.o~ ~N.u»
o\p~\ua mppp ‘N “\5 wuuuaoa H M\Ho= m\a _ ~.oH www.mm ap~.uu Huo.w~ gm.uu ~a.mu m.wm m.ma mu».au Nou.ou mm.oo mu.~u ua.-
zpuon uonwunu. run u.~m ~um.- mo.o» Hu.pc u.oo Ho».uo u».mu
_ upw.ow* Hm».ua.
_ zen Han. zen Han.
o\H~\~b mwpp “N H\» uuuwnoa u H\~= m\a . .Hu H¢~.~o omw.uu u-.~p ~¢.~b Hum.~u mo.»~ Hm.Hu mp.uo u»~.~u Hmo.~a uo.~a ao.ub
zuxoa raw .un Nmu.om o».uo up.»~ H~.H» Hum.»~ m~.H»
supp *p zwuma w\a _ ~.m~ uqu.wu ~Nu.~m www.mm -~.mu ma.w~ b».p¢ um.- op.~u
run _ m.oo “H.u» uu.m~ Nu.m~ Hw.cp
Hwo.mp» mu.»~»
mgpp a» guupaoa u H\~: w\a _ H.Hu ~a~.mo Nou.mp ao.~o Hm.w~ Hu.~o m.~w ~.mo www.mu Hpa.»o wm.pu um.p~ aa.mo
:Hxna raw _ H\um Huo.oo mo.cu Hu.oH p.ua wo.wu uu.ou
aoa>r w\a _ u.ma oou.oo >u~.-
raw o.mu pu.»p up.uo
zm>z _ Ham.po* “V.wo» H~.mu Hop.ua ap.p»
mau.om<. _ Nm.oc* ug.»u» Hm.au a.~¢
pp\o\go mwwp .H H\» a-u»aoa u\»= m\a _ H.HV om.~u m~.oo ap.um o.eo m.§o ».~u Ho.~u a~.~» mu.oo Nm.om go.o~ no.0m
>mvmn canrvam. raw _ H.~m uu.ao wm.~o u.o~ u.oa mb.Ho Nu.oo
1\~quo\b w\a _ o
raw p.00
m\w-o\~b “\N: «\a . ~.Ho Hbu.m» ou.om au.~m HH¢.H» mu.o» uu.wm mo.u» Ho.~o
cmnxaam. raw ” ~.mm uu.~w w~.HN au.o~ wo.op

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

»
>mucmnma no blmwmma

 

gwMZUHx H.

zHrr zczwmw m Anoan.a.v

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0% 32. “fig” QE 0;; ”2“.”wa Hfifi 9m; .82. “mum 8.2.5 0mm”: N
u>am z>o=Hzmw« xmxm mwoucna aHsz szqa . z>z-=w. szca ocemca . z>z-mw.
Hcha a: we mac» n: ma ac ma z>n:-:x. z>z-=w. n>za ac ma menu a: ma N om zymam
. . nc. ma. . . . . nu. ma. ac. ma. Hzmca . . nu. ma. . . szca
_
aoa>r w\a Hw.um not:
run _No.ow mw.uu
_
p~\pm\ub err §u ”\wN aum»ama H.“ w\a H.wo HuH.- Vo.om uo.u~ u.uo ».H~ ~.om ~.um om.~o Vu.o¢ uo.mw q~.om -.o~
>mum= ran _ ~.o~ o».o~ u~.oo ~.ou H.oa u¢.mm Hm.~m
_
u.m w\a H.o¢ www.mu Ho.om wo.u~ bo.up Nm.pa Hb.oV ~o.o~ HuH.- No.0m wo.m~ Nu.oo ~¢.op
vmu _ ~.om mu.um au.~o -.No Hp.uo a».o~ u~.oo
_
pw\b\ub xHHH g“ m\u~ a-mfiama H\~= w\a .mu §m.mu mo.bu bo.~b H.0e H.0H .oo ~.~m ~u.m¢ a».mp -.aH um.a~ »a.uw
and: >mvm= cmoxcmm. rmv _ H.00 bm.mu Nu.ob H.00 .um ~u.mn H~.uu
cwam _
p~\pm\ub w\a H.NH Haw.»o www.mo wo.w~ N~.Hu Hm.up ¢.Ho Hm.»b HHH.Ho op.m~ uo.a~ Nu.»o -.uw
rug _ H.mu um.mo No.Hw H~.Ho ».ow oo.~b no.um
_
aoa>r w\a _ H.~m Hmo.ww Hu¢.~o
rmv ~.mu
2m>z _ oo.m~ uo.mm 0.H0 om.w» ~¢.m~ -.~o -.mo
mau.um<. _ No.0» .up wu.~m u.m~
HH\»\Vb “\b: m\e _ .m» »~.mb mm.mo ab.~o b.ma m.oo 9.5m Ho.H~ uu.uo ¢~.Hu wH.om Vo.pm ~o.m£
oonxcam. raw . ~.oo »~.mb ~u.w~ b.m» ~.b~ ww.mo Ho.um
acaph w\a _Hw.um cots
rmv\_Nb.~m up.mN

 

 

 

'11111mu1|11111111111111 '

 

31293108012141