GRAIN-DUST EXPLOSIONS. 47 
is similar to those used in experiments in the special mill built at The Pennsylvania 
State College for the purpose of this investigation. 
This company reports that no damage has been done, and, so far as they know 
explosions have not been transmitted beyond the safety pipe and damper. 
The double damper as installed in the present series of tests did 
not give the desired effect, probably because the two sets of revolv- 
ing paddles were too close together. If the flame can pass one set 
of blades, it is likely that it can pass an adjacent set of blades 
before it has been extinguished. It seems reasonable to suppose, 
however, that a sufficient volume or extent of passage between the 
two revolving dampers may cause the flame to expend its energy to 
such an extent that the second damper would be able to check it 
absolutely. 
The following conclusions can be drawn: 
1. A single revolving damper is a fairly effective device for pre- 
venting the violent propagation of flames, but is by no* means an 
absolute preventive. 
2. A double revolving damper does not appear to serve the pur- 
pose of a fire trap any better than a single damper. 
RELIEF VALVE AND PIPE. 
Although the three tests with the relief valve and pipe leading 
from the hopper of the attrition mill to the outside of the building 
showed that this device was not effective in preventing the propa- 
gation of flames, even past the double revolving damper, it is quite 
possible that a larger size of valve and pipe may aid materially in 
accomplishing the desired result. The principle of permitting the 
immediate escape of confined gases that have just been ignited 
seems to be a proper one on which to base a preventive device. 
Conclusions : 
1. The relief valve and pipe as installed alone are not successful 
in preventing flame propagation. 
2. In view of the correct principle involved, it may be worth while 
to try changes in the design and location of the relief valve and pipe. 
A study of the foregoing discussion of the tests performed at the 
experimental plant of the college may lead to the objection that not 
enough attention was paid to the quantitative side of the problem, 
involving such factors as the exact amount of grain fed into the 
mill, the rate of feeding, the physical and chemical analyses of the 
material, the minimum temperature and minimum amount of heat 
to produce ignition, the velocity and analysis of the dust-laden gases 
at various points between the attrition mill and the end of the con- 
veyor, the pressure developed by the explosion, the minimum 
humidity- and the minimum amount of inert gases to prevent igni- 
tion of inflammable dusts, and the limits of inflammability for vari- 
ous dusts. To such a criticism the authors would reply that the 
