American Society of Civil Engineers.

Transactions of the American Society of Civil Engineers (Volume 81) online

. (page 12 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 12 of 167)
Font size
QR-code for this ebook

In British Honduras "Moho"

• ■ • In Guatemala ''Lanillo" or "Moho"

■''"'"" ' In Spanish Honduras "Guano"

In Costa Rica ;' . T'? .W^R 7J \ "Balsa"

In Bocas del Tore "Moho"

Balsa wood, because of its lightness, has been known for a long
time as a desirable material for floats, for supporting other material,
and for rafts. Professor Gifford states that in the West Indies the
natives use it for poles* "somewhat as the Chinese use the bamboo
for shoulder poles, for all uses where a light, rather strong pole is
needed." So far as the writer could learn, balsa wood has no com-
mercial value with people in the tropics who know about it, the gen-
eral idea being that its lasting power is very slight, and that it warps
and checks when exposed to the weather so as to be of little or no
practical use.

Mr. W. F. Morgan, who recently made an investigation regarding
the growth of balsa wood in Costa Rica, states that the natives are in
the habit of cutting large balsa trees simply for the seed pods, which
grow a woolly fiber suitable for pillows and mattresses. Fig. 16 shows a
log of balsa wood, about 2 ft. in diameter, the tree having been cut down
merely for the seed pods. Mr. Morgan also states that it has been the
custom to use balsa wood in rafts of heavy timber for the purpose of
sec\u"ing buoyancy. At the end of a voyage, this wood is throvni away
as having no commercial value.

It is thought that the first person to make any extended commercial
use of balsa wood was Capt. A. P. Limdin, President of the Welin
Marine Equipment Company, and formerly connected with the Pacific
Mail Steamship Company. From his travels in tropical countries,
Capt. Lundin knew of the extreme lightness of this wood, and its
value as a buoyancy material in life preservers and life-boats was sug-

• von Schrenk.

Fig. 15. — The Leaf of the Balsa Tree.

Fig. 16. — Balsa Wood Log. Trees Cut for Seed Pods.


gested by its properties. When he undertook to apply the wood prac-
tically, however, he found that it was of little value because it absorbed
moisture in great quantities, and also because it soon rotted, and also
warped and checked when worked. He then undertook the discovery
of some means of treating the wood which would render it water-proof
and also prevent it from changing its shape. After testing nearly
every method that had been suggested, Col. Marr's method of treating
woods, which had been recently patented, was finally successful. In
this method the wood is treated in a bath, of which the principal
ingredient is paraffin, by a process which coats the interior cells with-
out entirely clogging up the porous system. The paraffin remains as
a coating or varnish over the interior cell walls, preventing the absorp-
tion of moisture and the ill effects as to change of volume and decay
which would otherwise take place; it also prevents the bad effects of
dry rot, which follows the use of any surface treatment for preserving
wood of the balsa type.

The Marr process tends to drive out all water and make the wood
water-proof ; it improves the quality of being readily worked with tools,
without material increase of weight. The treated balsa wood has
been used extensively by the Welin Marine Equipment Company in
the manufacture of life preservers, fenders for life-boats, and for struc-
tures requiring insulation from heat, as in the refrigerating compart-
ments of vessels, and in ice boxes.

Transmission of Heat.

Specific Conductivity. — The Bureau of Standards, Washington,
D. C, has determined the "specific conductivity", e, of balsa wood
with the following results, expressed in British thermal units per hour,
per square foot, per inch of thickness, per degree Fahrenheit of dif-
ference in temperature between the surfaces :

e e

for untreated balsa wood. for treated balsa wood.

0.394 B. t. u. 0.422 B. t. u.

0.352 " 0.350 "

0.403 " 0.424 "

0.383 "

The lowest results obtained with both the treated and the un-
treated wood indicate a "specific conductivity" of 0.350. The higher
results in other cases are to be attributed to imperfect specimens,


or to imperfect contact of the heat measuring devices. The
Bureau of Standards uses electrical methods in measuring the heat
supply and the temperature of the surfaces of the material, in order
to eliminate all surface losses. The "specific conductivity" corre-
sponds to e in the equation,

ff = -(e-e'),


where x = the thickness, 6 = the temperature of the entering sur-
face, ff = the temperature of the discharge surface, and B. = the heat

Heai Transmitted hy Melting Ice. — Table 3 gives the results of the
writer's investigation of heat transmission by determining the quantity
of ice which melted in boxes made of balsa wood and other materials
under known differences of temperature measured inside and outside.
The results were reduced to British thermal units transmitted per
square foot of mean surface between the outside and inside surfaces
of each box, per degree of difference of temperature of the air inside
and outside, per hour of time, which correspond to the "coefficient" of
heat transmission, h, in the equation,

H = Jc (T — T),

in which T = the temperature of the air on the entering side, and
T' = the temperature of the air on the discharge side, the other
symbols being as before stated. The coefficient, k, differs from the
"specific conductivity", e, of the previous equation as it is dependent on
the surface as well as the conduction capacity for heat transmission.
Relation Between "Specific Conductivity" and "Coejficient of Heat
Transmission." — It is evident that the quantity of heat transmitted
through any body is equal to that passing in succession each heat-
resisting part. For example, if heat passes through a simple homo-
geneous body from a higher to a lower temperature, and from the air
on one side to that on the other side, it must overcome: (1) the re-
sistance of the entering surface, (2) the resistance of the material com-
posing the body, and (3) the resistance of the surface from which it
emerges. The surface resistances, (1) and (3), are overcome by
radiation and convection, the interior resistance, (2), by conduction.
The surface capacity for transmitting heat may be considered as equal



to the convection capacity plus the radiation capacity. The coefficients
of radiation for most materials are known accurately, and may be
calculated for different temperature conditions by the application of
Stefan's or DuLong and Pettit's Law. The coefficients of convection
are not known so accurately, but the values as stated by German en-
gineers appear to give reliable results.

TABLE 3. — Heat Transmission Experiments in Sibley College.
Tests Made by Melting Ice.

British Thermal Units

Per Square Foot Per

of wood,

Degree op Difference

No. of

OF Temperature of Air.

Kind of material.

in inches.

Per hour i Per 34 hours

k. ! k\



0.131 ! 2.90

Balsa wood, treated.



0.120 2.89

Balsa wood, treated.



0.122 2.93

Balsa wood, untreated.



0.192 4.61

White pine.



0.103 3.45

Nonpareil cork, extra.



0.132 2.93

Balsa, single boards, treated.



0.121 2.90

Balsa, double boards, at right angles.



0.1194 3.67

Armstrong, XX-cork blocks.



0.191 2.18

Balsa wood, treated.



0.1194 2.87

Balsa wood, treated.



0.199 1 4.78

Balsa wood, treated.


0.660 ! 15.83

Bare zinc. Vie in. thick.

k = coeflficient of heat transmission per degree Fahrenheit of difference of temperature
of air near the sides, per square foot, per inch of thickness, per hour ;
fc' = 34 fc.

The Bureau of Standards has given the name "specific conduc-
tivity" to the quantity of heat conducted, in British thermal imits per
square foot per hour, per inch of thickness, per degree of difference of
temperature of the walls; and this term is used in this paper. This
method does not consider surface losses, which vary with conditions.
Engineering computations of heat transfer must usually be made by
considering the temperatures of the air on the two sides, and require
a knowledge of a coefficient of heat transfer per unit of area, per inch
in thickness, per degree of difference of temperature of the air on each
side, represented herein by k.

The following equations give the relation between "specific con-
duction", e, and "coefficient of heat transmission", k. They are rational.


and are recognized by French and German engineers as accurate. In
the equations which follow,

H = heat transmitted per unit of surface per unit of

e = "specific conductivity", or heat transmitted per de-
gree of difference of temperature of the sides of
the material, per unit of time, per inch of
thickness ;
X = thickness;

a^ = coefficient of surface flow entering the body ;
a^= " " " " leaving the body;

h = coefficient of total heat transmission per degree of
difference of air temperature, per inch of thickness ;
T — T' = difference of temperature of air on two sides ;
6 — ff = " " " of material on two sides ;

t ^T — ^ = drop of temperature entering the material ;
f ■=ff — T' := drop of temperature leaving the material.

As the flow of heat is continuous, we have the following expres-
sions, all equal to each other, as representing the flow through a single
wall of homogeneous material without air space, per unit of surface, per
unit of time:

Flow by conduction through interior, .fl"^ {6 — $').... (l)

Surface flow entering H = a^(T — e) (2)

Surface flow leaving H = a^ ($' — T') . ..(3)

Total heat transmission H = k (T—T)...(4:)

From these we can readily deduce

l = - + -+- .-■• (5)

k aj «Q e

from which the relations between k and e can be computed, provided
the values of a^ and a^ are known.

The foregoing equations apply to a single thickness without air
space; but it is easy to calculate, by the same process of reasoning, the
heat transmission through walls made up of various materials with


or without air spaces. In the case of a wall made up of two different
materials with an air space between, we should find by calculation

l^L + ^ + l + L, + % + ±, (e,

The temperature of the surface can be readily deduced by trans-
position in the equations given, thus, from Equation (2),

e = T—— (7)


The surface transmission coefficients have been worked out care-
fully for numerous cases by Rietschel and other German engineers.
These are given by Kinealy,* in English units, as follows :

a = c + d + ^^^\^QQ "^^ ^ = c (1 + 0.004 + d (1 + 0.003

c = 0.82 for still air ;

c = 1.03 for air with moderate velocity;

c = 1.23 for air with high velocity.

Values of d.f

Brickwork 0.74 Iron, rusted 0.69

Mortar 0.74 Cast iron, new 0.65

Plaster of Paris 0.74 Sheet iron, polished . 0.009

Stone masonry 0.74 Brass, polished 0.053

Wood 0.74 Copper " 0.033

Paper 0.78 Tin " 0.045

Glass, dry 0.60 Zinc " 0.049

Glass, wet surface. . .1.09 Zinc, dull 0.10

For thick walls and poor conductors, t is so small that it can be
neglected. Rietschel gives values of i, for windows, as 36 ; for brick-
work 5 in. thick, as 14; for brickwork 30 in. thick, as 5, for wooden
doors, as 2.

Comparison of "Specific Conductivity", e, with "Coefficient of Heat
Transmission", k. — The following computation is based on an assumed
value of the "specific conductivity" (e = 0.35), as determined by the

• "Formulas and Tables for Heating", New York, David Williams Company.
t This table is the same as the Peclet table (see E. Peclet, "Traite de la Chaleur")
for coefficients of radiation reduced to English units.


Bureau of Standards, a^ and o^ as computed from the preceding equa-
tions, and coefficients, which are as follows:

First, For a Single Wall Without Air Space. —

a^ = a^ = 0.82 + 0.74 + 0.01 = 1.57.
These values substituted in the general equation, give



1 . 57 1 . 57 e

This, solved for e =

0.35 and for values of a; = 1, 2, and

the following results:


Thickness, in ' " Coeflficient of
inches, heat transmission.''

e = 0.35

X = 1 h = 0.242

e = 0.35

X = 2 k = 0.143

e = 0.35

a; = 3 Jc = 0.1015

Second, Single Balsa Wall, Ice in Zinc Box; Representing Test Re-
sults. — A box of balsa in still air, having thicknesses of 1, 2, and 3 in.,
respectively; melting ice confined in a dull zinc box, ^^ in. thick, the
inside of which is in contact with the ice but separated from the balsa
wood by an air space. These conditions correspond to those of the
ice-melting test of which the results have been stated, and to which

Equation (6) applies. For these conditions, a^ = 0, and — is so small

that it is negligible.

Substituting numerical values, we have

a^ = c -\- d = 0.82 + 0.10 + 0.92

2^ _ 1 1 j^ _1_

k ~ TT57 1757 ^ 0T35 '^ 0.92'

This, solved for different values, gives the following results :

"Specific Thickness in "Coefflcienl of

conductivity." inches. heat transmission.'"

e = 0.35 X = \ Jc = 0.192

e = 0.35 X = 2 Tc = 0.124

e = 0.35 X = Z Jc = 0.091

Table 4 gives a comparison of the determinations of Jc by compu-
tation (assuming e = 0.35) and test.



TABLE 4. — Eesults with Balsa Wood Box.

"Coefficient of Heat Transmission," h.
"Specific Conduction " e.

By Test, Melting Ice, in a Zinc Box,
Inside Balsa.
By computation.

in inches.

assumiDg e = u.So

(for 1 hour). | ,

for 1 hour.

for 34 hours.


0.192 0.199
0.124 0.121
0.091 0.091


Engineers using insulating material for cold storage generally
express the heat transmission coefficient on the basis of the heat, in
British thermal units transmitted per degree Fahrenheit of difference
in temperature of the air on the two sides, through material 1 in. in
thickness, and for a period of 24 hours, which corresponds to h' in
Table 4. The coefficient for 1-in. material is generally assumed as equal
to twice the result obtained in a test of 2-in. material. In practice,
only materials 2 and 3 in. in thickness are used, so that the coefficient
obtained on such a basis, though not scientific, gives results which
are fairly close. On such a basis, the "coefficient of heat transmitted"
through balsa wood, per inch of thickness, computed from the Bureau
of Standards tests, is 5.98; and, as determined by ice-melting experi-
ments in Sibley College, is 5.80.


I) I s c u s s I o isr

Mr. A. P. LuNDiN,* EsQ. — The speaker's attention was drawn to this

wood many years ago, during voyages to tropical countries, viz..
Central and South America, Central Africa, and also India, in all
of which the same species can be found; and he first remarked it
when a number of natives came floating down a river on a raft
made up of balsa logs. The logs were covered more or less with bark,
and where this was chipped away there still remained the hard surface
which is found on the outside of balsa logs, under the bark; the
ends were covered with tar, or some waxy substance.

The natives, particularly in Central and South America, use such
rafts to float their products to the sea coast, and seldom use them
more than once, for one reason, because it would be difficult to bring
the rafts up against the stream, and because the wood absorbs water
very readily and the raft is more or less water-logged after arriving.
Of course, in solid logs, with the ends closed up, the absorption is
not so rapid as when balsa is freed from bark and cut up in planks.

Later, when engaged in the life-saving equipment business, it was
brought to the speaker's attention that some crude attempts had
been made to use balsa in life preservers. On taking over a boat
shop in Long Island City, a quantity of balsa was found there, and,
on inquiry as to its purpose, it was learned that experiments had
been made with it in life-belt manufacture, but that the wood absorbed
water so rapidly that the belts had to be made two or three times as
large as the ordinary cork life-belt to assure the required buoyancy.
Subsequently, chemical experts were put to work to devise methods
for making the wood non-absorbent.

First, painting was tried, but, owing to the peculiar nature of the
material, the paint was rapidly absorbed, and coating it over and
over meant having just as much paint as wood, which added greatly
to the weight. Then varnishing was tried, but, owing to the moisture
left inside, the varnish cracked and blistered off.

Next, several mixtures of paraffin, asphaltum, gilsonite, etc., were
tried, which gave a fairly good outside coating, penetrating about
i in. on the ends and about xs ^^- o^i ^^^ sides, and the problem
was apparently solved. However, before long it became evident that,
owing to the cellular structure of balsa, which is mostly pith, and
the great quantity of moisture sealed up in the wood by the impervious
surface treatment, dry-rot developed even sooner than in the untreated

• New York City.


Just at the time these difficulties became apparent, Col. Marr's Mr.
water-proofing process was brought to the speaker's attention, and ""'''°-
after numerous experiments with the new method, it was successfully
and practically applied.

The United States Government tried out balsa life-preservers,
life-buoys, etc., as compared with the cork articles, for a period of
49 days (24 hours per day), at the end of which period the cork
preserver had lost all its buoyancy and the balsa preserver still retained
the buoyancy stipidated in the Government requirements.

A few years ago, while working on the buoyant-material proposi-
tion, it was considered that, as balsa, owing to its peculiar structure,
was so advantageous for use in buoyancy products, it might also be
adaptable for insulation purposes, and accordingly experiments in that
direction were begtm. The first ice-box made of the new material
was on the speaker's motor boat, and the results were surprising.
All during the hot summer weather, ice was put in the box on Friday
or Saturday, and on the following Friday or Saturday the temperature
in the box would still be quite low and some ice was still left in the box.

Naturally, all first work in the line of balsa insulation was more
or less crude, and the importance of scientific investigation was soon

It was particularly fortunate that Professor Carpenter became
interested in this material. The speaker well remembers that when
he first spoke to him about this wood, and stated that it was all pith
and no fiber, he and the gentlemen in his company looked very
skeptical. However, he was sufficiently interested to visit the Welin
plant, intending to remain there half an hour, but he spent prac-
tically a whole day, and when last seen on that occasion, he had all
his pockets full of balsa, and he has been steadily devoted to the
investigation ever since.

The speaker does not pretend to be an expert on insulation or non-
conductivity, but looks at this material from a practical rather than
a scientific viewpoint.

The principal feature in insulation material is, of course, that it
must be a good non-coijductor, but no doubt in the future engineers
will also consider structural strength, and the possibility of making
up complete homogeneous units will also be considered in judging
the efficiency and value of insulating material, particularly where it
is to be used in making ice-boxes and as insulation for buildings, in
ships, and in railroad cars.

The principal consideration is a commercial one; in other words,
good engineers always try to obtain the highest total efficiency, and
commercial men want it and are willing to pay for it.

In shipping, for instance, it is known that almost any ship can
be insured if built to certain requirements, such as Lloyds', and it


Mr. can be insured at the lowest rates (except, of course, in time of war),
un in. ^Yi cargo on such ships can also be insured at reasonable rates, except
one class, and that is perishable food stuffs carried in refrigerator
compartments, which are always carried at the shipper's or consignee's
risk. Neither the steamship companies nor the insurance organiza-
tions will place insurance on such goods, except for the event of
total disaster to the ship. Why ? Not because the science of refrigera-
tion has not kept pace with the science of naval architecture, but
because of the fact that if a break-down occurs in the refrigerating
machinery, the cargo will spoil, as the insulation is not reliable
enough to keep the temperature sufficiently low until the machinery
can be repaired and the system put to work again.

The speaker had an experience of this kind, many years ago,
when on board a meat ship running from Australia to London. The
machinery broke down while running through the Red Sea, and
more than half the cargo spoiled in less than 24 hours.

A few years ago, in a meat market in Seattle, a new refrigerating
plant, which had what might be called ordinary commercial insulating
material, broke down, and before new parts could be obtained to
replace the broken ones, and the machinery put in working order
again, the damage to the meat in the market amounted to more than
the cost of the refrigerating plant.

Therefore, when it comes to efficiency in insulation, there is still
room for improvement. Perfection cannot be attained, but balsa
wood will surely help to make insulation perfect, as it is not only a
very efficient non-conductor, but has sufficient structural strength,
and, through its physical properties, permits of constructing units
with practically unbroken insulated surface walls.

A small ice-box, or pony refrigerator, made by the Welin Com-
pany, is of balsa wood, 2 in. thick, about 36 in. long, 21 in. wide,
and 22 in. deep, and weighs about 30 lb. Such a box could not be
made up of any other known insulating material. It is strong enough
to stand severe jars, and a man could jump on it without straining
it unduly. Of course, other woods will stand more rough usage than
balsa, but, to meet this factor of additional strength, particularly
on the surface, paneling made from the bark and waste of the wood
is applied to the outside.

A small container, on the order of a thermos bottle, but in the
form of a box made of 1-in. material, has a capacity of about 1 cu.
ft. and weighs 6 lb. When going on an automobile trip, or something
of that sort, the "lunch" can be placed in such a container — whether
it is to be kept hot or cold. If it is wanted cold, a little ice should
be put in to keep it so.

Butter has been sent all the way from Virginia to Southern Cali-
fornia in such boxes, at an average outside temperature of 82°,


and the trip took 8 days, by the slowest route. Yet, when the boxes Mr.
arrived at Los Angeles, the butter was still hard and frozen.

Even if balsa had 30% less efficiency as a non-conductor, the
speaker believes that it would meet a very common requirement for
insulation, not satisfied by another material lacking in structural
strength, which makes it possible to eliminate all leakage of heat
through imperfect joints, or by use of cement or nails which may be
classified as good conductors. Other insulating materials are almost
entirely limited to use as a lining for a structure built of good con-
ducting materials.

The speaker has a dream of perfect refrigerated transportation
and conservation which may be of interest : It is growing more and
more expensive to live in America and in many other countries.
Food seems to be getting more scarce and more expensive all the
time; heat in the winter and cold in the summer are rising in price,
in short, the cost of every daily need is "going up."

Much of the increase in prices is due to waste, and, if this waste
can be eliminated, it stands to reason that prices will go down. If
foodstuffs could be transported, stored, and kept in first-class condi-
tion, not only the tremendous waste which exists to-day would be
done away with, but food would be purer and would retain its full

Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 12 of 167)