You are on page 1of 6

AD-A268 652

OU I93-19464

Acceleration Measurements on Naval Vessels Undergoing


Explosive Shock Testing

G. YIANNAKOPOULOS, BE(Mech), BSc


Scientific Services Division, Materials Research Laboratory, DSTO
SUMMARY
When shock testing naval vessels, the structural response to a dynamic load is usually measured by accelerometers
positioned at selected locations throughout the vessel. The complexity of the structure under test presents a
formidable task in identifying key measurement locations. The aim of this paper is to present the issues raised in
choosing such locations.

1. INTRODUCTION 2. BACKGROUND
In shock testing a naval vessel many accelerometers Shock testing the structure and equipment of a naval
may be located on board the vessel to measure its vessel is part of the shock qualifying process that
response from the underwater blast wave. ensuresthatthevessel hasmetitsdesign specification
Accelerometers are positioned on the structure at and identifies deficiencies in the design or construction
locations such as the outer hull plating, bulkheads and of the vessel. The performances under shock loading
superstructure. Other locations include ship's equipment of equipment, such as the propulsion unit, the navigation
such as the power plant, control systems and electronic system and combat systems are evaluated. The shock i
instrumentation. It is a difficult task to identify suitable test parameters are defined by the shock standard to
locations for measurements because of the large size which the vessel was designed.
of the naval structure and the complex behaviour of the The test, as performed by MRL in conjunction with the
structure arising from the interaction of the blast wave RAN, consists of a series of experiments conducted at
sea using explosives to generate underwater
shock
The aim of this report is to present some of the work waves. A typical experimental configuration consists
undertaken at DSTO-Materials Research Laboratory of mooring the vessel at a defined stand-off distance
(MRL) on acceleration measurements, and to discuss froio- the explosive charge of known yield located on
the issues raised in selecting accelerometer locations, the sea bed (Figure I).

The selection of accelerometer locations begins with


recommendations by naval architects, but the decision
as to which locations are most suitable involves further Nava, Vessel
refinements that depend on considering the
instrumentation requirements. This report is confined Air
to the instrumentation considerations; it discusses the
selection of locations by considering the limitations of
the accelerometers used, the manner in which they are Wae _-5
mounted, and the nature of the signal recorded. -

Acceleration measurements are only a small part of Stand-oF


our work on shock testing. Other current activities
include measurements of underwater blast waves,
vibration analysis, and the shock loading of materials.
n
Explosive Charge

//
D,< tance

77//
However accelerometer measurements are important See Bed
because they provide valuable information on the
vessel's response in the most extreme operational Figure 1: A typical experimental configurationfor shock
environment. This response is difficult to predict in testing a naval vessel. The vessel is mooredat a defined
detail owing to the structural complexity of such a stand-off distance from the explosive charge of a known
vessel. yield.

I This document bcE '-"n zpp.,cved


'for .l.. Pu -. z...
-~~ di~....Lirnte&
-C
CAMER NTICE
18~l

THIS DOCUMENT IS BEST


QUALITY AVAILABLE. THE COPY
FURNISHED TO DTIC CONTAINED
A SIGNIFICANT NUMBER OF
PAGES WHICH DO NOT
REPRODUCE LEGIBLY.
on a small cantilever [2]. The cantilever has two slits ?-
I olt
d one end and is rigidly attached to the accelerom,_aer
e' e rI-P 2 t_ base at the other (Figure 3). The strain gauges are
/ o-located across the slits; with one strain gauge in
tension and the other in compression. The strain
Referen.ce Surface Cut-oFF gauges are part of a wheatstone bridge and form a half
Time Zero / / _ bridge.
d :-The two major reasons for using piezoresistive
__ _ accelerometers are that they have DC response and a
stable zero baseline. We evaluated piezoelectric
Figure 2: A typical shock profile from an underwateracerotrsndfudhtinhsevrnmtte
explosion. Irrportantparametrsof the profile are the peak accelerometers and found that in this environment the
pressure, P.. the rise time of the signal, the duration,t, the baseline drifts. This drift arises from heat exposure
surface cut-off time, t, the impulse, i, and the energy flux
density, e. Note p. and C.denote the density and acoustic Siltcon Stran Gauge
velocity of the water, respectively. (OnLy one etement sho.n)
The dynamic load on the vessel caused by the shock Z
profile can be quite complex. Its shape depends on the .E- Ca•oever
explosive geometry and the propagation of the shock
wave. Interface boundaries where there are differences
in density, such as the surface of the water, give rise < Accelerometer base
to reflection and refraction of the shock wave. Typical
features of the shock profile include a very rapid
pressure rise to a peak value, followed by an exponential
decay until it drops sharply depending on the distance Figure 3: Schematic of an Endevco piezoresistive
from the free surface of the water (Figure 2). The peak accelerometer.
pressure, decaytime (time constant), duration, impulse, but, more importantly, from the amplitude linearity
energy and time to surface cut-off are parameters that characteristics ofthe piezoelectric crystal [31. Increasing
characterise the pressure profile. Thereforethe pressure the acceleration level increases the sensitivity of the
profile is measured by underwaterpressuretransducers accelerometer. The relationship is linear depending
placed in the vicinity of the vessel in order to determine on the accelerometer design and crystal used. A
these characteristics. spurious DC signal is produced in the acceleration
The dynamic load, defined by the parameters of the record. This is highly undesirable because the results
pressure profile, providesthecontrolledconditionofthe are integrated numerically to obtain velocity and
experiment. Similitude equations which predict the displacement. The baseline drift introduces an offset
shock wave parameters [1] can be used to calculate that appears in the integration as a ramp and parabolic
the amount of explosive and stand-off distance of the error, respectively.
vessel required for the desired dynamic load. The To ensure that the accelerometer faithfully measures
load, defined in this manner, complies with the shock the dynamic load it must be mounted rigidly on the
standard. surface [4J. Although the accelerometer is small and
light, minimising material mismatching, precautions
3. ACCELEROMETER are needed in mounting the accelerometer. These
CHARACTERLSTLCS AND MOUNTING include applying thecorrecttorque on the accelerometer
mounting screws, using mechanical filters, and avoiding
Accelerometers located on board measurethe response locations that have high local resonance. Applying the
of the vessel to underwater shock. The choice of the correct torque ensures that the accelerometer and the
type of accelerometer and its location on the naval component under test are an integral part so that the
vessel is a vital decision determining the validity of the entire system responds as one. Mechanical filters are
results. The accelerometer must have suitable used to isolate the accelerometer from unwanted
characteristics to faithfully record the signal. The short frequencies. For example, measurements on a steel
duration and intensity of the signal place demands on panel contain a high frequency resonance which is not
the accelerometer'scharacteristics such asthe sensitivity, representative of the structural response of the vessel
linearity and response. Furthermore, the accelerometer and should be eliminated.
must be mountcd at a locatioll representative of the Isolating the accelerometer from vibration requires an
manner in which the vessel responds to the load. For understanding ofthe modesandfrequenciesof vibration
example, locatillg the accelerometer on components ofnthe ship. As an example, a merchant ship undergoes
not rigidly mounted would yield meaningless rcsults on several vibration modes in its normal operational
the vessel's structural response. environment (5]. These include hull vibration (1-6 Hz),
We have found that piezoresistive accelerometers are vibrations at interfaces between the hull, bulkhead
most suitable for these measurements. The and/or ship's equipment (3-12 Hz), superstructure
accelerometerconsistsofsilicon strain gauges mounted vibration (7-14 Hz) and local vibration of plates and
REPORT DOCUMENTATION PAGE

3 TYPEIDATES COVERED

4TITLE Amn m ACCELERATION MEASUREMENTS ON


NAVAL VESSELS UNDERGOING EXPLOSIVE SHOCK
TESTING.
6AUTH G YIANNAKOPOULOS

7 w•
IOORON mESmms DEFENCE SCIENCE AND
TECHNOLOGY ORGANISATION, MATERIALS
RESEARCH LABORATORY, P.O. BOX 50, ASCOT
VALE, VICTORIA, 3032 AUSTRALIA

09 - AGENCY NAMES AND ADRESSES

I ISUPPLFAENTARY NUM

12 DISTIUflQ4IAVALAZLI"T STATEMEr .... COWM


128 DISTILJBUTION
UNCLSUE

13 A•SrAC(MAx 20A WORDS) WHEN SWXoL T1M540 NAVAL VESSELS. 1IM STRUCTURAL RESPONSETO A I)YNAM LOAD IS USUAILY WMA3URED
BY ACCEjEMSIERS POSITIOR•ED AT SELECTED LOCATIONS THROUGHOU THI VESSEL.. THE C EXfMlMY OP THE STRUCTURI U1NDERTEST
PRESENIS A PORIdDABLE TASK IN IDENTIFYINO KEY MASUREMENT LOCATIONS. THE AEMOP THIS PAPER IS TO PREN THE ISSUESRAISED IN
CHOO SUcM.LOCATIONS.

'4 SU11, CT TIMS I IU m OFsPAGES 4


16 PRIM CODES

17 SECUrrY CASS.REPST IS SECCLASS PAGE 19 SEC CLASS AmSr. 0 LUMtAltI UP AESTRACT

UNCLASSIFIED UL

I
frames (10-40 Hz). Vibration is caused by propulsion structure and the location of the accelerometer on
induced forces and the surrounding water. The board the vessel. By these comparisons we estimate
frequencies of interest during shock testing are in the the levels and select the appropriate accelerometer
range of 0-300 Hz [6]. During the shock test ship and mounting configuration. At present this is the only
induced vibrational effects are of secondary interest, quantitative means used to estimate acceleration levels
The accelerometers can be mounted in three orthogonal and to determine the suitability of a particular location.
axes but are commonly mounted either in the vertical Guidance on suitable accelerometer locations is also
direction or horizontal plane (Figure 4). We have used provided by qualitative considerations of materials
the vertical direction to study the structural response of science and structural engineering. For example, the
the vessel. The horizontal mounting technique is used shock wave causes local motion, which occurs as
on ship's equipment such as electronic systems and deformation of hull plating, and overall structural motion,
the propulsion system. However, both usually are which results from inertial effects. The accelerometer
required at the same location, must faithfully record both types of motion. Locations
of the accelerometers are determined by considering
their suitability for providing information on these
motions.
Consideration is also given to the nature of the signal.
ve•-t•caL A typical result is shown in Figure 5. The general
AcceterK!:e- •feature of an acceleration signal is its oscillatory and
dissipative nature; that is, it resonates at a particular
Longitudinat frequency and decays quickly with time (-=80 ms). The
signal is complex and difficult to analyse and exhibits
T-ansve-se recurrence (ie. there is a reappearance of the signal
after some time has elapsed). By analysing the signal,
thevalidity of the results is determined and the suitability
:,cce~e'-o-•e~e'] NavaL Vessel of the accelerometer location and mounting are
mo•rvtin P.-e evaluated.

Figure 4: Possible mounting orientationsof an I


accelerometeron boarda naval vessel.

4. SELECTION OF ACCELEROMETER
LOCATIONS Figure5: A typical accelerationsignal from a dynamic
Shock testing of naval vessels is a complicated field load due to an underwaterexplosion.
and requires the understanding of many aspects of
shock waves and dynamic loading. For example, in 5. FUTURE DIRECTIONS
designing the shock test, hydrodynamics, shock wave
physics and explosives' science are used to predict Another aspect to be considered is the need to develop
the load on the vessel and understand the propagation tools to identify measurement locations so that the
and interactions of the shock wave. Studying the selection procedure is quantitative and consistent,
response of the vessel requires knowledge of materials and acceleration levels for a given shock configuration
science, and structural and vibration engineering, are predictable. A quantitative selection procedure
Because measurements involve the use of electronic would identify measurement locations and would provide
instrumentation, metrology, computer science and a systematic selection procedure. The prediction of
numerical analysis, an understanding of these branches acceleration levels would ensure that the locations
is needed when selecting transducer locations, chosen would represent the severest loading conditions
Our preliminary approach in the selection of suitable on the vessel.
locations has been to review all of the results from Thedevelopmentofanaccelerationpredictivecapability
previous shock experiments. These are categorised is made possible bythe availability of relatively powerful
according to the magnitude of the dynamic load and and cheap computers. To predict the acceleration
correlated to the accelerometer locations. Comparisons levels on the vessel, finite element and boundary
between results show that the acceleration levels and element codes can be used [71 which are now relatively
frequencies are dependent on the geometry of the accessible. Major disadvantages are that tiiese cooes

Lal,
are complex and need to be calibrated by using relatively powerful and cheap computers mak4- it
representative measurements from shock experiments, possible to use finite element and boundary element
A further development that can improve the selection codes to achieve this. These can be supplemented by
procedure is the availability of an extensive shock experimental data from laboratory experiments using
induced acceleration data base. Shock experiments shock machines and small scale shock tests. It may be
are expensive and a limited amount of data is currently possibletoobtain furtherdata from slamming experiments
available at MRL. The data base can be increased under conditions that simulate the structural response
without the need for expensive full scale shock of the vessel from shock tests. The feasibility of such
experiments. To study the local effects on components comparisons requires further investigation.
and ship's equipment, results can be obtained from
laboratory experiments using high impact shock machines REFERENCES
[8] and small scale shock experiments [9]. These
results would indicate acceleration levels for full scale 1. M. M. 'Explosion Effects and Properties:
shock tests. Part Swisdak,
II - Explosion Effects in Water", Naval Surface
Weapons Center, 22 February 1978, NSWC/WOL
TR
To obtain data on the structural effects, it may be 76-116.
possible to usevessl
slamming experiments
oadng ninrughsea
uderoesdynmic which the .
2. Whittier, .
R.M. and Wilson,.
J. 'Application of Silicon
vessl uderoesdynmic
oadng n rughsea Strain Gauge Technology to Aerospace Acceleration
conditions. Slamming experiments also give rise to an Gauge Technolog Pac eration
hull whipping [ 10]. In fact, studies have shown that the and Pressure Measurements", Technical PaperTP28o,
whipping component is of much lower amplitude and Endevco Corporation, Dynamic Instrument Division,
frequency than the dynamic load (in this case the San Juan Capistrano, California, 1982.
slamming force) and persists over a longer period of 3. Bouche, R.R. and Lovelace, D.E. "Acceleration
time. The load, on the other hand, has a much higher Characteristics Used in Transient Motion and Nuclear
frequency and amplitude but decays very rapidly. It is Applications", Technical Paper TP250, Endevco
recommended that investigations should be undertaken Corporation, Dynamic Instrument Division, San Juan
to determine the validity of comparisons between Capistrano, California.
results from slamming experiments to results from
shock tests. 4. Stathopoulos, G. "Effects of Mounting on
Accelerometer Response", Electronic Industries May
Data obtained by these techniques can be compiled 1962, Reprinted by Endevco, San Juan Capistrano,
together with the limited data available from shock California, June 1975.
tests. Models and/or guide lines can then be developed
to identify suitable locations for instr'Jmentation in 5. Jensen, J.J. 1983, "Beam Models For Ship Hull
future shock trials. Vibration Analysis", Shock and Vibration Digest, 15
(2), Febmary 1983, pp 23-27.
6. Macintosh, A.M., "Shock Design Methods - An
6. CONCLUSION Overview", Paper presented at 1988 ABCA-7
Quadripartite Conference, Bath.
on board naval
Selecting locations for accelerometers
vessels begins with recommendations by the naval 7. Haxton, R.S. "Predicting the Shock Response of
architects on possible locations of interest. The actual Submarine Platforms", J.N.S. 16(3).
locations chosen take into consideration the 8. Clements, E.W."Shipboard Shockand Navy Devices
accelerometer type, mounting and the nature of the for its Simulation", NRL Report 7396, 1972.
acceleration signal. In choosing the locations, data are
usedfrompreviousshockexperimentsandthesuitability 9. Bishop, J. "Small Scale Underwater Explosion
of the location is based on these measurements. Shock Tests",20th International AnnualConferenceof
Qualitative considerations are also used to identify ICT, 18th Annual Technical Meeting of GUS, June 27-
items on the vessel that are prone to damage. 30, 1989, Federal Republic of Germany.
Itis recommended that computational tools be developed 10. Bailey, F.C. Fritch, D.J. and Wise, N.S. "Acquisition
to quantify the selection procedure and enable the and Analysis of Acceleration Data", Ships Structure
prediction of acceleration levels. The availability of Committee rt SCC-159, February 16,1964.
A.:•L •. Or

&.3
MiC QUAIY Df-l-ED

Dist o - .

Dist b~•J
A-/ tL