John Ripley Freeman.

On the safeguarding of life in theaters; being a study from the standpoint of an engineer online

. (page 6 of 8)
Online LibraryJohn Ripley FreemanOn the safeguarding of life in theaters; being a study from the standpoint of an engineer → online text (page 6 of 8)
Font size
QR-code for this ebook

practical use and with solutions of the strength actually applied.
What little I have seen of men at work on fireproofing scenery
leads me to fear that in order to avoid discoloration and efflores-
cence, the solutions will be put on too weak for the best flame-
proofing, and that after the Iroquois is a little further in the past,
most of the scenery will no longer be treated for flameproofing.

This effect of aiitipyrine chemicals upon colors is a question
for the chemist rather than the engineer, and it is quite possible
that a full range of the necessary colors could be worked out
from pigments that would not be changed by the flame-proofing
liquids, particularly if the antipyrine chemical be applied to the
new canvas before sizing.

The Whipple and Fay Investigations on ' c Fireproofing ' ' Scenery.

Mr. Whipple and Dr. Fay gave much time to testing the rela-
tive efficiency of various solutions and to developing standard
methods of test, by which the relative efficiency of one fire retar-
dant solution could be compared with another, and their work
is so complete and instructive that I regret I can present here only
a summary of it.

The following brief outline will show its general scope.

The results may be summed up as follows:


(1) Phosphate of ammonia was found the most efficient

(2) Tungstate of soda, so often found recommended in
the text-books, was found to possess almost no value.

(3) The various proprietary solutions when analyzed were
found to be all based on one or another of the ammonium
salts, commonly the phosphate, but frequently -the cheaper
sulphate substituted in whole or in part.

(4) Linen canvas or cotton cloth, fireproof ed in the best
manner possible by any of these solutions, could be quickly
burned to total destruction if a sheet were rolled in a loose
coil with the axis vertical and a space of perhaps one-half an
inch between the folds, and a fire then lighted with a small
wad of excelsior at the bottom of the roll; this method of
test serving to confine the radiant heat and the gases dis-
tilled from the fiber. This was of special interest since strips
of the same cloth tested in the manner that has satisfied pre-
vious experimenters by holding the strips of treated cloth
vertically over an ordinary Bun sen flame could not be ig-
nited and appeared almost perfectly flameproof.

(5) The most efficient part in the fireproofing of fibers was
found performed by the covering of the fiber with a non-
volatile liquid that excluded the oxygen. Phosphoric acid
proved better for this purpose than any other substance
tested, but obviously could not be applied alone because of
its corrosive action on fibers and colors.

(6) The ammonium in combination with it in phosphate
of ammonia was found of value chiefly in locking up the cor-
rosive qualities of the phosphoric acid until released by the
heat of the fire, and thus giving a comparatively harmless
compound for application to color and fabric.

(7) The method of application of the fireproofing solution
to the canvas was found to have great influence on the de-
gree of fire protection secured. One of the best solutions,
when brushed cold over the back of old scenery, penetrated
the fiber so little as to be of no value, but when applied hot
was efficient. Under some conditions the linen canvas is
repellent of water, as one finds on trying to dry the hands
on a new crash towel. When the liquid is applied rapidly



to a vertical surface with a brush, linen cloth does not ab-
sorb it readily. Hot application or the solution adds much
to its efficient penetration of the fiber. For new scenery,
probably the best method is to saturate the canvas between
rollers in a bath. The next best method is to mix the
chemicals with the water of the glue size that the scene
painter puts on before painting his picture.

(8) Tests of the tendency of the various chemicals to in-
duce decay were made by sowing some of the treated sam-
ples with mold spores. Other tests were made by adding
various per cents, of phosphate of ammonia to nutrient gela-
tine and to mixtures of the glue size, and incubating these
for tests of bacterial growth.

(9) The effect of the solutions on the colors ordinarily used
by the scenic artist was not found bad, except in case of
some of the more delicate blues and greens, but a greater
length of time would be necessary before positive statements
about this can be made.

(10) When canvas that has been flameproof ed is actually
burned as it may be under practical conditions, it gives off
fumes that may be even more dense and suffocating than
those from the untreated canvas.

Proof that the Kind of Paint Used on Scenery Makes it Less
Readily Combustible.

At the beginning, Messrs. Whipple and Fay made tests of the
comparative combustibility of old painted scenery canvas with
new unpainted canvas by taking strips all of the same size, thirty
inches high by three inches wide, and burning them while hang-
ing vertically from a nail, in a box that shielded them from cross
drafts of air. The specimens were all lighted at the bottom and
all under similar conditions.


Specimen No.

Original Weight of
Sample Grains.

Time of Burning,

Weight of Pigment
Compared with
Unpainted Canvas.



40 %




48 %




61 %




121 %


In ISTo. 3 the paint was heavier than the canvas.

It will be noted the retardation of the flame was proportional
to the amount of paint. The flames reached a maximum height
of one and five- tenths feet.

Order of Experimenting on Effect of Various Antipyrines.

The general course of the subsequent experimenting followed
by "Whipple and Fay, stated briefly, ran as follows:

As a preliminary experiment, strips of cotton cheese cloth were
dipped in various saturated solutions, and after drying were
held in a Bunsen flame. The substances were thus quickly proved
as to relative efficiency as follows: the ammonium phosphate in
saturated solution proving most efficient of any.

Full notes, which I will not take space to reproduce here, were
made of the behavior of each sample as a guide to further tests.

Relatively Poor Besults. Relatively Fair Results.

Common salt. Ammonium chloride.

Boric acid. Martin & Tessier's formula.


Borax ^j and sodium sulphate, ^3.

Sodium phosphate.

Sodium sulphate.

Sodium tungstate, 15% sol.

Bicarbonate of soda. Relatively Good Results.

Ammonium sulphate, half sat. sol. Ammonium phosphate.

Ammonium phosphate ^, and sodium sulphate ^. Phosphoric acid.

Aluminum sulphate. Borax and ammonium sulphate \^.

Potash alum of various strength of sol. Ammonium sulphate.

" Paris Theater solution." Calcium chloride, 25% sol.

" Subrath's Formula." " New Paris solution."

Next, explanation was sought of the reason for the behavior
of the various li reproofing compounds.

Joints in Theory of Flameproojing Established by the Whipple

and Fay Tests.

(1) The influence of the water of crystallization in retarding
ignition was studied. It was found that although different
samples of cloth treated respectively with alum, borax and
sodium tungstate and each loaded with all it could carry, the
large amount of water of crystallization in these salts did not
make them efficient fire retardants. The subject was studied
farther by selecting two salts, both compounds of the same
phosphoric acid, but one possessing twelve molecules of crystal


water, or over sixty per cent., while the other possessed none;
sodium phosphate and ammonium phosphate being chosen. The
chemical reactions were studied through the successive stages,
and the relative effect judged by weighing the amount of char
left after ignition of the treated cloth. It became plain that
water of crystallization played a much less important part than
the fluid, varnish-like residuum.

(2) Tests were then made for learning of the influence of the
ammonia given off from the phosphate of ammonia when heated
by comparing the effects of potash alum and ammonia alum. The
ammonia alum proved somewhat the better, indicating that the
evolution of ammonia had some small value.

(3) Tests were made to learn of the efficiency of the phos-
phoric acid left from heating the phosphate of ammonia by start-
ing with canvas treated with phosphoric acid. The phosphoric
acid proved nearly as efficient a fire retardant as the phosphate
of ammonia. The chief value of the ammonia in the phosphate
of ammonia appeared to be the rendering of the phosphoric acid
less harmful to canvas and colors.

(4) A study was next made of the absorption of heat by the
volatilization and decomposition of the fireproofing salts, and it
was, for example, made apparent that the number of thermal
units absorbed in driving out the combined water from a given
weight of ammonium chloride was nearly four times as great as
for an equal weight of sodium phosphate, and this helps make
clear why ammonium chloride has flameproofing qualities of
some value, while the sodium phosphate is comparatively worth-
less for this purpose.

(5) A study was then made of the combustible quality of the gases
distilled off when canvas that had been treated by various flame-
proofing compounds was ignited, in order to learn if inert gases
derived from the chemicals used for flameproofing diluted the
combustible gases from the cellulose, to the point where the
combined gases would not ignite. For this purpose little rolls
of linen untreated, and treated by various chemicals, were heated
to destruction separately in glass ignition tubes, five-eighth of an
inch diameter x 6 inches long, placed with the end in a muffle,
heating the muffle by gas to a temperature which, judging by
the color, was from one thousand degrees to twelve hundred


degrees Centigrade. This temperature, as shown by the color,
was maintained nearly constant all through these ignition tests.

These test rolls were two and one-half inches long and lay
only within the uniformly heated zone at the bottom of the test
tube. The distilled gas issuing from the end of the test tube was
ignited. The progress of the charring of the canvas could be
observed through the glass tube. The relative amounts of tarry
matters condensed at the cooler, outer portion of the tubes was
also compared.

It was found that canvas, flameproof ed so that a strip of this
canvas could not be made to ignite from a Bunsen flame, would,
when tested in the ignition tube, not give off ignitible gases from
the tube. The rapidity and simplicity of the ignition tube test
were found such as to commend it.

Therefore, series of tests with the whole line of known efficient
fire retardant compounds was made in this manner, and full notes
of their behavior kept.

As a result of the tests thus far, it was concluded in brief:

(a) That inert chemical substances can exert but very ^slight
fire-retarding action.

(b) The fire-retarding action of salts which depend for fire-
retardant quality]only upon their water of crystallization, like pot-
ash, alum, sodium phosphate and borax, is'slight and unimportant,
although somewhat "'superior to that of inert substances.

(c) Fire retard ants of the class which suffers chemical de-
composition under heating are decidedly more- efficient than thos
which depend on the driving off of water of crystallization, but
still far less efficient than the class that follows.

(d) The most efficient salts are those which on decomposing leave
behind a non-volatile residue which is fluid at the temperature of
the burning canvas, and covers the charring fabric with a thin glaze
which prevents further access of air, and of this type, phosphate
of ammonium was found to be the best.

Analyses of Sundry Proprietary Fireprooflng Solutions in
Use in 1904 to Meet the Recent Requirements of the New
York Building Law.

The following table gives the result of chemical analysis of the
most prominent fireproofing solutions found at that time on sale


in the New York market for the purpose of fireproofing
scenery :

Grams per 100 Cubic Centimeters.

1 Fireproofine ....

ii li i! it

i |f II a|

w | M < |

63 15

a ^ a 2

II gs & ii

gl. |& &% 'gl
^^ s. g

25 23*

2 H S Fireproofin ' Solution

16 5 12 6

3 Electric . .

24 5 17


4. H. W Johns' Compound

16 7 10 7

. . 5.3 ..

5 No flame

15 5 84 50

20 7 37 15 6

1 4

7 Salamanderine

17 4 81 58

3 6

8 Antipyros Klu p iana

23.9 18 5 ...

2.0 1.0 ..

.. 2.5

9 Van Ripper Solution


10 Lamb & Finlay's prepared canvas

nrpsent nresent.

* Combined as glyceride, 3.9 grams per 100 C.C.

Tests were made on scenery canvas that had been treated, the
fabric being thoroughly impregnated by soaking and wringing
out or by brushing on both sides of the heavy canvas, with each
of the foregoing, both in the Bunsen flame and in the glass tube
in the furnace ; all of them were found to be fairly efficient, JSTos.
9 and 10 being perhaps the least so. It should be noted that the
doth tested was more thoroughly impregnated than old scenery will
be when brushed over on the ~back at a single application. Those
solutions containing the larger amounts of ammonium phosphate
were found the most efficient. The only apparent advantage of
the chloride or sulphate of ammonium is the fact that it costs only
half as much as the phosphate ; it is less efficient.

Sodium sulphate, boric and boric acid are present in some of
the solutions. These were found to contribute relatively little
to the flame-resisting power, and the sodium tungstate came to
be regarded by these chemists as worthless for this purpose.

U. S. Patents on Fireproofir.g Solutions.

Previous to the investigations made for me by Messrs. Whip-
pie and Fay I had procured from the U. S. Patent Office a
complete file of the patents issued during a period of about 30
years, for the purpose of studying them for suggestions as to
chemicals or processes to be used. I found in them nothing of
particular interest. The compounds in most cases were made up
by mixing one and another of the salts, alum, phosphate of am-
monia, borax, sulphate of ammonia, etc., that have been in
common use and recommended over and over again for 50 years,


the novelty consisting in the precise formula for proportioning
the mixture and in the selection of ingredients. Alwin Kieske,
of Dresden Germany, however, went somewhat outside the
beaten path in patenting in 1901 molybdate of sodium in a
10 per cent, solution for application to fabrics for fireproofing
and preserving them. In general, fabrics other than theatrical
scenery appear to have been in the mind of these patentees, and
the number of patents for fireproofing textile fabrics is not
nearly so numerous as for the fireproofing of wood.

Difficulty of Proper Application of Fireproofing Solutions.

In the foregoing tests the effort had been to test the efficiency
of the solutions, it being assumed they would all be most thor-
oughly applied.

The method of application of fire-proofing solutions was next
made an object of study by Whipple and Fay. Samples of old
scenery were subjected to treatment by the various more efficient
solutions in different ways, and finally in order to produce uni-
form results and ensure the uniform distribution, application was
made by immersion in a bath containing submerged rollers, while
dipping, followed by a wringer with rubber rollers was used as
an alternative method.

It became plain that the method of application and the thor-
oughness with which the solution was absorbed had much to do
with efficiency. In order to completely saturate the fibers, many
dips and wringings were found necessary. The glue size of the
unpainted back of scenery canvas prevents to a considerable de-
gree the rapid penetration and absorption of the liquids applied.
Under rapid application with a brush, the best of the solutions
may fail to render the canvas flameproof, particularly if applied

No inspector can tell from the appearance of one of these large
sheets of canvas whether the solution has been properly applied
all over its surface, and probably all that inspection will ordi-
narily amount to in practice will be equivalent to what would
be shown by the touching of a lighted match to the edge of the
canvas sheet.

As to the permanence of the residue left in the canvas, it was
noted that when cloth that had been treated by one of the best



f tbe fireproofing solutions was shaken and brushed, the white
po \vder could be shaken off in the form of dust, and that more
was removed by brushing. This indicates that although a freshly
treated canvas may be well flameproofed, it may lose this quality
to a noteworthy extent after the rough usage which scenery re-
ceived on the stage and on the road. It would be interesting to
follow this matter further by tests of pieces taken from the mar-
gins of scenery that had been treated, and then had one or two
years of travel and use.


It has been claimed that the application of fireproofing solu-
tions weakens the canvas. If true, it is important to know
whether this comes from slow chemical action or from the rotting
of fiber due to bacterial action or mold, since in the latter case
a germicide could perhaps be incorporated in the solution.

Tests of effect of certain fireproofing compounds in promoting
mildew and mold were made; first, by adding varying percentages
of ammonium phosphate to nutrient gelatine which was then ex-
posed and incubated by methods common in bacteriological work,
and, secondly, by seeding the worst treated canvas with mold
spores. Time w r as lacking to carry these tests to the desired
length, but so far as they went it was found that the glue used
to size the canvas is probably a more potent promoter of mil-
dew than the salts employed for fireproofing. Concentrated
applications of the fireproofing salts will doubtless retard these
organic growths, while dilute applications of some of the salts,
phosphate of ammonia, for example, will very likely stimulate
mold and bacterial decomposition, particularly if hygroscopic.
Time did not permit the working out of experiments to find a
suitable germicide for addition to the solutions.

The Fireproofing of Wood.

Since the pine frame work of the set pieces and wings present
a greater quantity of fuel than the canvas itself, it would be de-
sirable to flameproof this wood. A simple brushing over with
phosphate of ammonia or other chemical solutions is found ineffi-

Yarious processes for making wood fireproof have long been
known and have been used on wood for interior finish and trim


of fireproof buildings, more here in ^"ew York City than any-
where else, because of certain favoring clauses in its building

The various tests made by Professor Norton of the Massachu-
setts Institute of Technology* and others have shown that, al-
though the wood, after treatment, is much less readily ignited
from a small blaze, as from a match or an electric spark, no real
fireproofing results. Previous tests have covered this matter so
thoroughly, and have shown the loss of strength and tendency
to gather moisture and other objectionable qualities that follow
treatment, that I gave little attention to testing this matter
further, but rested mainly on the tests of previous experimenters.
I obtained sundry specimens of wood that had been lireproofed
in the commercial way from two prominent shipyards that had
war vessels under construction and I made a few simple tests.

Fireproof wood was at one time much used on the war vessels
of the Navy, but has been almost wholly abandoned by reason of
its gathering moisture badly and the lessening of strength and
the increased difficulty of working it.

^The frames of scenery must be particularly light and strong,
and the wood must possess its maximum strength, and should
not be liable to warp. I do not find that " fireproof " wood
has ever been used practically for this purpose at any theater,
in this country or abroad, notwithstanding the activity of its
promoters. I soon concluded that in the present state of the
art it was too much to expect that the wood flameproofed by
any of the ordinary commercial processes could come into gen-
eral use for battens, frames, profiles, etc., of stage scenery.

* See "Report on Fireproof Wood," so-called, by Prof. C. L. Norton, August,

Professor Norton summed up the results of his tests on samples of wood " fire-
proofed " by three of the more prominent commercial processes as follows :

" Fireproof ed wood is almost identical with untreated wood in the following

' It smokes at about the same temperature.
' It can be ignited at about the same temperature.
' It will continue to burn in many cases.
' It is a good fuel.
' It makes a very hot fire.''

The ordinary method of test of little samples in the flame of a laboratory lamp
tends to greatly exaggerate the extent of protection against fire gained. A better
test is to make a small long vertical box of the wood, open at top and bottom, and
et this serve as a chimney for a small fire kindled inside at the bottom.


( ( I 1 ireproof ' ' Paints.

" Fireproof paints" are sometimes required by law to be applied
to wood- work about the stage. The underwriters' laboratory at
Chicago had a short time previously made an extensive series of
tests of all of the prominent ones in the market. The unpub-
lished records were placed at my service. These tests had shown
that none of these paints had any noteworthy value in flame-
proofing wood, but for confirmation I requested Messrs. Whipple
and Fay to make tests of a few of those most prominent in
the market. They purchased commercial samples and made
chemical analyses of several; each was found to be mainly a sort
of whitewash consisting of slaked lime, finely pulverized asbes-
tos, with also a little alum, gypsum and glue. The paint adhered
well when applied to canvas, but was quickly proved by test to
have almost no flame-proofing quality whatever.

It is difficult or impossible, on precise scientific grounds, to see
how these paints can have any noteworthy value against any-
thing but a very small momentary blaze, like that of a match or

None of these paints were found to penetrate below the sur-
face of the wood as phosphate of ammonia, for example, pene-
trates into the fiber of cotton or linen cloth.

Obviously, so thin a film can have only exceedingly small effect
as a non-conductor of heat. Radiation or contact must char the
wood beneath almost as quickly as if the paint were not there.
The destructive distillation will give off gas which will push
out, blister, and peel off the paint, and this gas will burn.

In the " asbestos paints," the pulverized asbestos, glued into a
thin crust less than jfo mcn thick, can obviously be of no more
fire retardant value than so much carbonate of lime or clay.
The special value of the asbestos in paints is chiefly as a name
to conjure with in attracting purchasers.

From all these tests a common lime whitewash appears to be
as efficient a fireproof paint as anything yet found in the market.



Finally, much attention was given to devising a standard
method for testing the relative efficiency of various chemicals



used for the flameproofing of scenery canvas. It has already
been explained that no fireproofing of cloth is effective against
severe heat, but it was plain, from the preliminary trials, that
some of the solutions were much better than others in protection
against a little blaze like that from a match, a cigarette or an
electric spark.

Since all samples, however flameproofed, were destroyed by a
severe test, all of these tests, of necessity, had to be merely com-
parative, and canvas treated with a saturated solution of phos-
phate of ammonia thoroughly worked into the fiber was adopted

1 2 3 4 6 8

Online LibraryJohn Ripley FreemanOn the safeguarding of life in theaters; being a study from the standpoint of an engineer → online text (page 6 of 8)