A v a l a n c h e

D y n a m i c s

a n d

D e f e n ce

SNOW FLOW EXPERIMENTS

In these experiments,weights of natural snow up to300
kg were released at the top of the approach.A container
set on top of the approachwas filled withthe snow chunks,
density and temperatureof which were 290 kg/m³and -10
ºC respectively, and all the snow was released by opening
the sliding gate. Upon release, the snow accelerated down
in a rectangular cross-section track, which is 0.8 m wide,
0.3 m deep and 70 m long.The track was covered with a
polyethylenesheet toreduce the frictionand toachieve
maximumflowvelocities.Thefloweventuallyreached
more than 10 m/s and then with approaching the Kantein
Figure 1 decelerated. Above flow behaviors including the
leading edge position as a function of time were recorded
with several sets of video cameras. Duringthe experiments
it was overcast and the ambient temperature was -2 ºC.
Figure 2 shows the measuring apparatus set up on the
snow flow track. The observation point was55m away from
the gate. Most equipment was installed in two sets of steel
towerswhich consistedofcylinders 0.05mindiameter
and 1.2 m in height. The distance betweentwo towers was
0.87m.Impactforceswere measured withstrain-gauge-
typeloadcells,KYOWALUB20KB,attachedto10cm
diameter pressureplate. Three loadcellswerecitedon each

tower,at height of0.14,0.24 and 0.34 m above the track
surface.Theairvelocityinduced bythesnowflowwas
recordedwith anultra-sonic anemometer, Kaijodenki WA-
200, which hasbeeninstalledin Kurobe canyonto observe
the natural snow avalanche and revealed the snow cloud
structure of one powder snow avalanche once( Nishimura
et al. 1995).However,there are a couple disadvantages in
thismethod.Firstlyitcostsfairlyhighalthoughlarge
avalanches sometimestake awaythe sensor.Secondly it
oftengivestheabnormalsignalwhen the densityofthe
snowcloudisrelativelyhigh.Inthisstudy,thus,we
measured the static pressure depression in the flow at the
same time and tried to transfer the obtained value into the
airvelocity.In general, we can expect the flow velocity u
is related to the static pressure depression[!]Pas follows,

= --¹ru²
whereris the density^[!]Pof
2^{the air. However, since the diam-}
eterand the length of tube give substantial effects on the
above relation,we obtained the correction factor with the
wind-tunnelin advance.Inthe experimentstwosetsof
tube,inner diameterof which is 0.01 m,were set as each
cut ends looked downward.Besides a drag meternewly
designed wasset at the end of the track. It consisted essen-
tially of a plate,which was 0.3 m x 0.2 m in size and was
coveredwithsnow,andastrain-gauge-typeloadcell,
KYOWA LSM-20KBS,which could sense the three com-
ponents of stresses. This equipment was utilized to inves-
tigate both theshear andnormal stresses actingonthe snow
surface during the snow flow passing by.

Ping-pong ball experiments

Snow avalanchesare made upof granular materials. Upon
breaking out the dry snow avalanche, the snow blocks are
brokenintosmallerlumpsorevenice particles.Onthe
other hand, after stopping the wet snow avalanche we can
find a numberof snow balls in the debris. Hence, some of
the results studied in the granular flow can be applicable
in the snow avalanche modeling (e.g.,Savage, 1983),but
unfortunately most of theories and numerical simulations
developedsofarlooktoosimplifiedandnotenough to
formulate the snow avalanche motion at this stage. Before
muchprogresscan bemade inthisarea, moredata sets
should be compiled in order to check the models.
Nishimura et al.(1991) carried out inclined chute ex-
perimentswithice spheresina coldlaboratoryand ob-
tained the profiles of density and velocity as functions of
inclinationandtemperature.However,thequestion
whether the flow reachedthe steady-state in the5.4 m long
chute was remained.
In this study,we have used a ping-pong ball with 37.7
mmin diameterand 2.48 g in weight.Since the effect of
the air dragon the ping-pong ball was fairly large, the flow
velocitiesareexpected toarrive atsteadystate withina
short distance. In fact, Nohguchiet al. (1996) found in their
22 m long chuteexperiments that thefront velocity of ping-
pong ball flow became nearly constant at 10m downstream
ofthestartingpoint.Furthermore,Nohguchi.(1996)
concluded with his similarity analysis that the ping-pong
flow on the 100 m long slope corresponded to the natural

Figure 2. The measuring apparatus set in thesnow flow path.a:
impact pressure sensors, b: ultra-sonic anemometer, c: tubes for
measuring thestatic pressure depressions, d: video camera.

204