January 14, 1916.

THE COLLIERY GUARDIAN.

67

break device (such as the trembler of a bell) be employed,
it is found that the ability of a given spark to ignite
a gas mixture depends upon the duration of the sparking
as well as upon the circuit conditions recorded in the
investigations above described.

Trembler Sparks.

Fig. 1 shows a typical result following the use of a
trembler spark in an explosive atmosphere consisting
of a 10 per cent, mixture of coal gas and air. The
current was approximately 0*4 ampere throughout the
range of the experiment. At 9 volts ignition was
obtained instantly. On reducing the voltage to 7 a
single break spark would not ignite the gas, even when
repeated as rapidly as hand manipulation would permit;
but when the trembler was allowed to vibrate normally,
ignition due to the trembler spark occurred after one
second. At 5 volts ignition occurred after 10 seconds.
This fact appears to be of practical importance in con-
nection with that system of bell signalling commonly
used in mines, in which the circuit is closed by the
application of a piece of iron to a pair of bare wires.
Sometimes an old file or a knife is used, and whatever
be the implement employed, the surface is usually rough.
In drawing this implement across the wires there is not
obtained the single spark of carefully maintained
laboratory apparatus, but a rapid succession of sparks
•approximating to that of the trembler.

Reverting to the question as to whether the measure
of the circuit volts and amperes, or amperes and induct-
ance, can be regarded as a measure of the incendivity
of the spark, Prof. Thornton has stated that the energy
of a break flash (referred to sometimes in this paper
as a single spark) is proportional to the power of the
circuit, and is equal to 4 Li2, where i is the circuit
current.* In a recent leading article, the Electrical
i Review} has gone a step further. Arguing from Prof.
Thornton’s work and basing its calculations specifically
on certain curves by Dr. Wheeler (of which fig. 5 is
one), it is stated that the igniting power of a break flash
depends on the Li1,51, or approximately the Li3/2, of
the circuit. Finally, it is stated that for every gaseous

3-o

2-5

o-5

i-o

i-5

•o

4

~ mg Mt Jig Mt

12

> Crd
minify-

8
voice

^Ordinary Deli
with short-circuited
magnet coils-

--------1-----------------1----------------------

Mg ~ Maximum gap at trembler

mg= Minimum ,	,	,

Ml = Maximum tension on trembler spring

-mt = Minimum »	>•	»------------—

X = Ignition of coalgas at trembler spark

O =	"	»	'•	" " breu.fr in externa/ circuit

@ = Extra tension on trembler spring,	.s'

-------m opempon of trembler below 9 vo/ts.—s^—

-Mg mt—mg mt^—

Fig. 2.

the incendivity of a spark depends to some extent upon
' its duration. ’ ■ ". " . ' '

The general conclusion to which the .author has been
led by a variety of experiments on the electrical ignition
of gases is that it is necessary to ‘distinguish between
the energy which produces a spark and that quality of
the spark termed by him “ incendivity ” which enables
the spark to cause ignition, and that the magnitude of
the one is not a measure of the other, although there may
be a more or less regular relation between them when
crtain physical conditions are kept constant. Ignition
seems to depend on the ionisation caused by the spark.
During the interval of sparking the ionisation may be
rapidly dissipated or neutralised. If the neutralising
action predominates there is no ignition of a gas
mixture. If there is little or no neutralising action,
ignition occurs immediately. Between these two limits
‘ there are a variety of intermediate conditions which
apparently account for the delay of ignition indicated
by fig-. 1, and much of the great irregularity that is often
experienced in experimental work on this subject.

The foregoing remarks describe the main features of
the theory of ignition by sparks so far as it relates to
electrical apparatus for mining and other operations in
which an explosive atmosphere is liable to occur. As
already stated, the outstanding fact of practical import-
ance is that circuits carrying relatively small amounts
of electrical energy are capable of producing dangerous
sparks. No fact is more familiar, and. at this date no
justification could be found for addressing to electrical
engineers a paper on the subject of gas ignition by sparks
were it not for the existence in many' mines of a crude
but indispensable piece of apparatus—the electric bell
signalling circuit—which in consequence of at least one
serious disaster has excited the suspicion and attention of
those responsible for the safety of mines, and caused a
revival of enquiry into the facts associated with spark
ignition.

The prevention of ignition by sparks in ;he power and
lighting circuits of mines has been, thoroughly investi-
gated, and, excepting accidental’breakage of cables, no

(1) to avoid intercommunicating chambers or chambers
with partitions so arranged that it is possible for ah
explosion in one chamber to compress the gas in the
other.before ignhing it, and (2) to construct the case so
that it will withstand the increased explosion pressure
due to turbulence. Theoretical investigations seem to
show that the maximum pressure attained is but little
affected by turbulence, although .the speed with which
the explosion attains its maximum pressure is greatly
accelerated. Mr. Simon mentions, however, that a
motor which successfully resisted an internal gas explo-
sion when at rest had its cover blown off by an explosion
when the motor was in motion. At first sight this does
not appear to agree .with what is known of turbulence
effects; but the explanation is probably that when an
explosion occurs in a chamber in which the cooling
surface is relatively large, turbulence may produce no.
only an increased rate of explosion, but also an increased
maximum pressure. A further fact which might have
a practical bearing upon the design of covers is the
character of the internal surface, if the configuration
involves restricted passages. A rough surface has an
accelerating effect both in dust and gas explosions,*
probably due to turbulence. Whether this effect is
appreciable in closed chambers does not appear to be
definitely known, but in tubes the effect is very marked.

Mine Bell Circuits.

As previously mentioned, the signal bell circuits
employed in mines commonly comprise a pair of bare
wires connected to a irembler bell and battery. A signal
is given by connecting the wires with the aid of a metal
instrument, such as a file, a knife, or an iron bar. The
bare wire system has much to commend it on the score
of simplicity and convenience, but latterly it has become
the subject of much suspicion, especially since the
Senghenydd disaster in 1913. The danger associated
with bare signal wires has, in the author’s opinion, been
somewhat exaggerated; nevertheless some danger does
exist, and it is well that the risk should be eliminated.
Suggestions have been made to abolish bare wire circuits
and employ enclosed switches arranged at suitable
intervals along insulated circuits, but the inconvenience

of this, is sufficiently evident to make it clear that miners
will not abandon the bare wire system unless it is shown
to be irremediably dangerous. The facts appear to be
that the system is not seriously risky and that what risk
does exist can easily be removed. A very careful
investigation of the subject has already been made by
Dr. Wheeler,f and . his report contains valuable
suggestions of a simple and practical nature.

The spark at the trembler of a bell is rightly regarded

	1	XW	; '  i	W	1	
			 ■	mgMt^	.j.	—	
b		£—r	q	 ;
	1 |

J_______________I___,__________1_____:_,_______1___________:--

—	g	g	I0

Volts

><3,t trembler coi/s interwound with tui-fo//

Fig. 4.

-0-046 sec

0-6 —

Amperes

Fig. 3.

mixture there is a constant value of the product L i3’2
beyond which the break flash will be capable of igniting
the mixture. This is an interesting deduction, but the
truth does not seem to be expressible in such a simple
form. Igniting sparks can be produced in practically
inductionless circuits carrying but a few amperes.
These sparks, which may be termed “ hot point ” sparks,
are not included in an expression based on inductance.

When a pair of contact points in a non-inductive
circuit are separated so that an arc is, maimained
between them it is true that-the product v i is a measure
of the power of the arc. The value of v i is not
necessarily the same during arcing as when the con acts
are together, and there is no reason for assuming them
to be the same when the arc is only of momentary
duration. Therefore v i prior to sparking . is not a
measure of the power of a hot point spark.. Further, in
non-inductive circuits carrying the same power both
igniting and non-igniting sparks can be produced by
simply altering the shape or material of the contacts.
Again, when the circuit is inductive and the above
effect does not enter or is negligible, then the energy.,
which produces the spark on separating the contacts is
expressed by 4 L i2. With the same energy either
igniting or non-igniting sparks may be produced accord-
ing to the shape or material of the sparking points.

As already indicated, curves connecting volts ant
amperes, or amperes and inductance, in the.circuit are
useful as giving a general idea of the magnitudes
involved in the production of igniting sparks, . but
caution must be exercised in using them, since (1) a
variation of the size-or material of the sparking points is
attended by an alteration of the value of the circuit
current required to produce an igniting spark; and (2)

* W. M. Thornton. “The Least Energv Required to
Start a Gaseous Explosion.’’ Philosophical Magazine, vol,
28, p. 734, 1914.

f Electrical Review, vol, 77, p. 65, 1915,

danger need be apprehended if the usual protective
devices are maintained in proper condition. The open
sparking on the bare wires of the bell systems has,
however, been neglected, and an urgent necessity has
arisen to render it harmless. Before dealing with this
subject, >a brief digression may advantageously be made
on the protection of enclosed apparatus.

Enclosed Apparatus.

In a paper by Air. S. A. Simon* an interesting account
is given of an investigation carried out by the Committee
of the Westphalian Alining Fund and six leading
electrical manufacturers in Germany, to determine the
value of totally enclosing apparatus as a protection against
explosion. Among other things mentioned is the effect,
upon the pressure attained in an explosion within an
enclosing.case, of the configuration of the case and the
rotation (when this occurs) of the parts enclosed.

Various experiments have been made by different
investigators into- the effect on an explosion of the shape
of the gas containing chamber, and it has generally been
found that whilst with cylindrical and spherical
chambers the shape does have a modifying effect, it is
not sufficient to be of any practical importance in the
design of a case. The German inves igation rightly
emphasises, however, the important fact that if the
shape is such that the case virtually comprises two or
more communicating compartments, then an explosion
initiated in one of them might spread to and produce
destructive effects in the other. Another fact is that a
given gas mixture which explodes harmlessly in a motor
case when the armature is at rest might explode wi;h
sufficient violence to wreck the case when the armature
is in motion.

These phenomena are quite well known in connection
with the study of gas explosions. The increased pres-
sure in the first is due to compression prior to ignition,
and in the second is due to turbulence. Both are
operative in the working of an internal combustion
engine, and although the effect of the first has been
familiar for many years, it was not until recently that
the importance of the second was discovered.! It is
now known that the gas. or petrol engine is largely
dependent for its action upon the state of commotion of
the gas immediately prior to ignition. In the design of
protective cases for electrical apparatus it is necessary

* S. A. Simon. “ Notes on Safety of Working Electrical
Plants in Coal Mines.” Journal of the Institution of Elec-
trical Engineers, vol. 43, p. 10.7, 1909.	•’

f Dugald Clerk, Gustave Canet Lecture, 1913,

in a general way as more dangerous than the single break
spark which occurs on the bare circuit when a. signal
is given. To avoid the use of a trembler it has been
proposed to use single stroke bells; but the trembler bell
is apparenily preferred by miners, as the liability to
mistake in the interpretation of a signal is less with the
trembler than the single stroke bell.

Danger of gas ignition at the trembler has been
minimised by the use of an enclosing case, and the risk
due to the spark in the external circuit has been tolerated
partly for the reasons that such spark is supposed to
be comparatively harmless and that the ventilation of
the mine is sufficient to prevent the accumuulation.near
the wires of an explosive mixture. Laboratory investi-
gations show that using a bell of the ordinary type the
break spark can be not less dangerous, than the trembler
spark. If it is deemed necessary to enclose the trembler
there is no justification therefore for disregarding the
break spark. The first necessity is to abolish the-break
spark, or reduce it to a harmless, dimension, and the
secpnd' is to render impossible such a variation of the
bell adjustment as would enable a dangerous break spark
to be produced.	■

The results of tests by the author on a mining bell in
an atmosphere of coal gas are shown in fig. 2. One
object was to find the variation of current in the bell
circuit with variation of voltage, and the manner in
which these variations are affected by altera, ions of the
bell adjustments. The otlffr object was to find with
what voltage and current ignition could be obtained
both at the trembler and the break, or contact in
the external circuit. By adjusting the movable con-
tact screw at the trembler so as to produce the
minimum spark gap when the armature was vibrated
by the magnet, and reducing the resistance of the
armature controlling spring to the minimum, the relation
of volts and amperes indicated by mg.mt. was found.
At 4 volts the current was 0*1 ampere, and at 12 volts
the current was 0’2 ampere. Retaining the same tension
at the spring and increasing the trembler gap to the
maximum, the curve AIg.mt. was obtained. With this
adjustment the current is 0’05 ampere a: 4 volts and
•0'2 ampere at 12 volts. Keeping the gap at the

* (1) Home Office Reports of Explosions in Mines Com-
mittee, No. 4.	(2) J. D. Morgan. “Coal Dust Explo-

sions.” Transactions of the. Institution of. Mining
Engineers, vol. 49, p. 220, 1915. (See Colliery Guardian,
March 19, 1915, p. 567).

! R. V. Wheeler. Home Office Report on Battery Bell
Signalling Systems. 1915, '