This action might not be possible to undo. Are you sure you want to continue?

BooksAudiobooksComicsSheet Music### Categories

### Categories

### Categories

Editors' Picks Books

Hand-picked favorites from

our editors

our editors

Editors' Picks Audiobooks

Hand-picked favorites from

our editors

our editors

Editors' Picks Comics

Hand-picked favorites from

our editors

our editors

Editors' Picks Sheet Music

Hand-picked favorites from

our editors

our editors

Top Books

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Audiobooks

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Comics

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Sheet Music

What's trending, bestsellers,

award-winners & more

award-winners & more

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

AD 401 818

DEFENSE DOCUMENTATION CENTER

FOR

**SCIENTIFIC AND TECHNICAL INFORMATION
**

CAMERON SIATION, ALEXANDRIA, VIRGINIA

UNCLASSIFIED

NOTICE: Nhen government or other drawings, specifications or other data are used for any purpose other than in ccnnection with a definitely related government procurement operation, the U. S. Government thereby incurs no responsibility, nor any obligation whatsoever; and the fact that the Government may have formulated, furnished, or in any way supplied the sali drawings, specifications, or other data is not to be regarded by implication or otherwise as in any manner licensing the holder or any other person or corporation, or conveying any rights or permission to manufacture, use or sell any patented invention that may in any way be related thereto.

Wyauin .EROMETER by ()Richard H.USAF Plan B Paper Submitted to the Department of Mechanical Engineering and the Graduate School of the University of Wyoming in Partial Fulfillment of Requirements for the Degree of Master of Science University of Laramie. Jessen Captain./{)A METHOD of NUMERICALLY EVALUATING the ERRORS RESULTING from MEASUREYENT of SPECIFIC SHOCK IFULES by ACCELEOMETERS due to the FRENUE•.1C.' RESFCME CHARATERI6TICS of the ACCi.

for his guidance and assistance in the preparation of . this paper. Department of Mechanical Engineering. University of Wyoming. J.ACKNIs4tIEDGEMENT The author wishes to express his sincere appreciation to Dr. E. Foster.

... HALF SNE WAVE SHOCK PULSE .. .... . 1 15 16 IV APPENDIX CONCLUSIONS . .... ... . 0 o a .. .... . 11 . . SQUARE 1AVE SHOCK PULS .. ..... .. . ..... 0 0 ..iii TABLE OF CONTENTS CHAPTER I II INTRODUCTION .. . . . .. .. . 4 o 5 . .... PAGE 1 SHOCK INPUTS TUANSFORMED TO THE FREQUENCY DCMAIN ..... III INDUCED ERROR DU. ERROR CALCULATION RESULTS ..... . .... ... 6 9 o .. . . .TO THE FREQUENCY RESPQNSE LESS THAN UNITY ... TRIANGULAR WAVE SHOCK PULSE . ..... . ........... ... ... o * o .. ... .

..... Curvw for •nde'-co Model Accelerometer (Capacitance 50rnmfd and Impedance 100megohms) ... Low Frequency Range . Square Wave Shock Pulse .. . .... . ...0.. "Crystal Accelerometer . 6o . 12 Frequency Responm. Frequency Response Curve of a Conventional .... 8.. ... 11 7.. Inout and Outnut Vave Forms for Accelerometer with Frequency Response Less than Unity in the .. Frequency Curve Divided into Small Incremental Areas ............ Triangular Shock Pulse .............. 13 ..... ............... 3 5 6 9 3. Half Sine Shock Pulse ... 5............ .. 4. . . PAGE 2 2........iv LIST OF FIGURES FIGURE 1.... .. ..... Product (rurve C) of the T'requency Spectrum (Curve •) for a 10 Millisecond Square Pulse and Frequency Response (Curve B) of the Pc(elerometer ..... ..........

Barium Titanate.CHAPTER I INTRODUCTION Vibration and shock have become important fields of mechanical engineering with the advent of high speed aircraft and missiles. Many times knowledge of the destructive force the indicated acceleration cannot be or no value. acoustical is a desirable quantity and if accurately determined the calculated force will be of little Factors such as mounting techniques. line materials are: Some of the common crystal- Rochelle Salt. temperature effects. while shock is Vibration velocity. calibration errors. effects. The type of transducer measurement is most frequently used for vibration and shock The ability of a crystal or piezoelectric accelerometer. can be defined in terms of displacement. and acceleration generally defined as a change in velocity which gives rise Acceleration is the value most often desired to high accelerations. because the destructive force is proportional to the mass times the indicated acceleration. Quartz. certain crystalline material to generate a charge when subjected to a physical stress has been known for many years. Lead . and frequency response all affect the The effect of the frequency accuracy of the recorded environmental data. response upon the accuracy of the acceleration data will be the main factor discussed in this paper. The equipment or "black boxes" installed in today's air vehicles are subjected to rigorous vibration inputs and sudden shock forces and must be designed to operate or at least withstand these conditions.

2 Zirconate-Lead Titanate and Lead Metaniobate. insofar as environmental testing. The shock inputs most frequently associated with the airframe and missile field. are a 10 millisecond square wave shock pulse. stress in the crystal causes very small displacements of the crystal domain which generates a charge on opposite surfaces. 1 Frequency Response Curve of a . and the seven millisecond triangular shock pulse. The A crystal deflec- tion of three micro-inches may produce as much as 100 volts.9 0 10 -r-eIAencY (cps)---& J. ooo Conventional Crystal Accelerometer Fig. When the accelerometer is subjected to an input a mechanical strain is applied to the crystal. A typical response curve for a crystal accelerometer is shown in Figure 1. The ability of the transducer to faithfully reproduce an input is related to the frequency response of the accelerometer. I i (. The input to be considered in this paper will be a shock rather than a vibration input. 10 millisecond half sine wave shock pulse.

This paper will present a method of calculating the error of a crystal accelerometer due to its frequency response being less than unity in the low frequency range.r ST t1o P . For example. 2 Input and Output Wave Forms for Accelerometer with Frequency Response less than Unity in the Low Frequency Range.-r. where the gain is independent of frequency. 2). 4 A Acc s. arecmns. Above and below this mid-band region the gain is For this reason it is always desirable to The dependent upon the frequency.- 0 "1" -r Fig. 1) with respect to frequency has a useful region called the mid-band range. faithfully reproducing the input.3 The response of a crystal accelerometer (Fig. The inputs considered will be the three types of shock pulses previously mentioned. operate the accelerometer at frequencies within this mid-band range. question then arises as to how satisfactory the accelerometer data would be for certain types of shock excitation where the frequency content of the pulse could be below the mid-band range. . if the frequency content is below the mid-band range (gain less than unity) and the input is a square wave the output of the accelerometer will not be a "square" In this example the accelerometer would not be square wave (Fig.

000 cyps. the number of frequencies which must be sunmed to duplicate the pulse. dt Eq. E. The mathematical tool which is used to transform from the time domain to the frequency domain is the Fourier Integral (Equation 1. however shock measurements are primarily associated with the low frequencies and this paper will deal only with frequencies between 0-10. 'Van Valkenburge.000 cps however many amplifiers associated with the accelerometer have a rapge of 0-10.1)1. 1.1 This matlematical tool will enable the engineer to determine the frequency content of any nonperiodic pulse. J. 1955. Englewood Cliffs. N. NETWORK ANALYSIS...CHAPTER II BROCK INPUTS TRANSFORMD TO THE FREQUENCY DOKAIN The shock input is a function of time (nonperiodic) "F(t)" and must be transformed into a function of frequency "F(f)". It is recognized that there are a number of suitable anplifiers whose range may well exceed 50. Prentice-Hall Inc. This equation is analogous to the Fourier Series which is used for a periodic input function. p..1 .000 cps. . M. by this it is meant. The frequency range of most crystal accelerometers is 3-50. This is neces- sary to determine what effect the frequency response of the accelerometer wlli have on the input signal.000 cps. F(L) - f F(t) eJt.

ej• t dt + A 0 . F(f) in given by F(w) J -T AO 0 0. 3) is given by F(t) and - A0 - 1 (t T Eq.*e tdt+ f e-'a1tdt - I IPGO T TT 2 .1. 2. into Equation 1. 000 + 00 -%- Square Wave Shock Pulse Fig.5VM WAVI BOCK PULSE A.1 F(t) = 0 every•here else Substituting F(t) . 3 The time function of a square vave (Fig.A0 where "T" is the period of duration of the pulse.

2 Substituting 2. A-I. the final form of Eq. 2. A0 Sin •. shown in Fig. 2.3 is the frequency in cycles per second (cps).3 is Fig. 4) is between 0 and T and zero (0) everywhere else.3 is the equation for the frequency spectrum curve for a square wave shock pulse of duration "T".•tT The frequency spectrum for a half sine wave is obtained by again using Equation 1. 4 Half Sine Shock Pulse The time function F(t) for a sine wave (Fig.6 jW! F(W) F() - T 2j = - AO Sin WT Eq. AO _L(srn f Eq.2 is F(f) where "f" - 14 - 21f and rearranging terms.1 as follows: . HALF SINE WAVE SHOCK PULSE The graph of Equation 2. Equation 2.

j(l +W)t) d -eId T A A 0 (J(~. e (CT Jyg SC -JTT 1I + eW(iT J)T ) _1~T + W()W ) P 23I W) FAw 0A (I W)e"-ej1 j )eJ -W ..- jF(t) f() -09 0-jw dt T F~~i -f A0 si~n 'T e-J~t dt 0 F(w) A0 2 e-J dt T ~j 2 AS 0 f (ej(!-Aw)t e e.T! ___ ___ )t J(2!+ W)t T F(w)- 2j ~A ~ A0 i +e ) -~W) A + W) + - F ~ J(2 W)(.

5 is shown .5 in the equation for the frequency spectrum curve for a half sine vave shock pulse of duration 'T". oT -.j Sin WT -(T T F(W) - T ~ W)- A xT (w) - C(+CooWT- j SinWT) The phase shift vill be neglected because the magnitude of the frequency content is the primary factor in this paper.4 and rearranging terms the A T Coo T E 2.8 F(CU) - A w ((T:JWT 1+CsoWT. WT)y 7(w) F(•) ~ - (T WCsL4J TW••'. (T 0 +CooW-T 41 CJ) . S•TW) + Co. in Fig. A-2.5 7(f) u :SL- I Equation 2. 2o2 Substituting final form is W - 21f into equation 2.. The graph of Equation 2.2 14•+ 2 2 Coo WT A x T r2 7(w) . -x2 (TW) 2 c . WT) 2 + (Si.E 2 Eq. ( Therefore.

- A.6 into the transform Equation 1.9 THIANIJLAR WAVE SMOCK PULSE Fig. 2. • (e-j IT -l) .t 0 t (T Eq.6 Substituting Equation 2. T T( T e-j WT WT + (e )2 ( -j jT -j F(•Q .Te'j JT] . 5 Triangular Shook Pulse The time function of a triangular wave shock pulse (Fig.1 the resulting frequency spectrum equation is given by T f A t e-j WTt 0 LO A(t e-jWt S0 + i T e-j Wt d 0 AO . 6) is A0t F(t) and F(t) = 0 everywhere else.

8 .10 F(Uj) - Ao ~ + T)e W1 F(•) - AO [(1 + (Coo W T) .(Cos 2xfT + 2xfT Sin 2xfT) Eq._ 2TTSinWT +T1 A Jd T2 Coo WT .WT Sin WT A t 0= T 2 -2 .7 Substituting F(A) W) = 2xF the final form is 2 A2TA2 2 T2it(•- T7 /T + (21fT) 2 .T Sin W) -T ) P((u) - AO [(Cos WT -i) + T Sin WT] + J(T Cos WT - - Sin WT Again neglecting phase shift =A F( •(•) 1 2 (Cos WT-I) + T Sin WT] (T Cos WT Sin WT]2 W T AO A r W1T W-To WT + 2 Coo W+2 +j2 l + T2 sn ••.(Cos WT + WT Sin WT) Eq. 2.8 is the equation for the frequency spectrum curve for a triangular wave shock pulse of duration "T". 2. is shown in Figures A-3 and A-4. The graph of Equation 2.8 Equation 2.

Cuvove o f~~'~~ 3 (C'ý. A-2.' ) (cs)-/-/ #C ~ Fig. The absolute areas bounded by these At the low frequencies curves are used to determine the induced error. This means the absolute area under the frequency specThe trum curve at the transducer output is less than at the input. and A-4. the response of the accelerometer is less than unity. 7). therefore. 6 Product (Curve C) of the Frequency Spectrum (Curve A) for a 10 Millisecond Square Pulse and Frequency Response (Curve B) of the Accelerometer . A-3. curve at the output is the product of the frequency spectrum of the input pulse times the frequency response of the accelerometer (Fig.CHAPTER III LNDUCED ERROR DUE TO MM FREUNCY RESPONSE BEING LESS MAN UNITY In Chapter II the three types of shock inputs were transformed into the frequency domain and the resulting frequency equations were plotted in Figures A-i. those frequencies in the input signal in the low range will be distorted by the transducer.

The response is flat (unity gain) above 40 cps. For all curves the bounded area is found by integrating the equation for the frequency spectrum curve (2. 161 last California Boulevard./ -fI&t¶ueLo Cy cre s)- re Fig. 7). California. integration was accomplished on the Bendix G-15D digital computer by using Trapezoidal Rule which is the numerical integration or addition of small incremental areas along the frequency axis. Pasadena.1.(area Aner curve C) (area umder curve A) The transducer that vas chosen as a typical piesoelectric accelerometer was the Endevco model 2213 manufactured by the Nndevco Corporation. 8) for the unit is based on a total capacitance of 500 nfd (accelerometer and cable) and an amplifier input impedance of 100 megoha. . The general fora in This (Yi 2 )( f) where Yi and y¥ 1 are the heights above the frequency axis and / f is the frequency increment along the axis (Fig.3) between zero and 10.12 The calculated percentage error as used in the paper is given by error 100 (area under curve A) . The frequency response (Fig.. Frequency Curve Divided into Small Incremental Areas .000 cps.

13 0R - ___ .

14 RROR CALCULATION RESULTS For a square pulse the computed area is 1.027238 100 (A0 ) (loss in area) = 1.000 and 10.000 cps.027238 (A 0 ).084288814(A 0 ) (loss in area)9 k0. The loss in area due to the frequency response being less than unity below 25 cps is 0. The area with distortion is the summation of the incremental areas times the gain factor for each respective frequency.25 and 25 cps. = 9.7111 . This is the difference between the total area and the total area as a result of distortion.8145% The percentage error for the half sine wave shock pulse of 10 milliseconds duration (using the sase approach as for the square wave) is 100 0.83149729(Ao) (total area) .000 cps. Likewise the percentage error for a triangular wave shock pulse of seven milliseconds duration is 100 ( 0. The percentage error for the 10 millisecond square wave shock pulse is /0.501086 (A0 ) (total area) 1.034355(A ) (loss in area) 0.8988% .25 cps between 0.501860 (A0) using increments of .92572(A 0 ) (total area)] 3. 5 cps between 25 and 3. and 20 cps between 3.

the purpose has been to emphasize this particular source of error.5. smaller than for the therefore.CLAPIM IV CONCWSIONS The accelerometer error due to the frequency response of the Endevco unit is 1. The area under the curve for the half sine wave pulse rapidly decreases for increasing frequency due to the 4f2t 2 term in the denominator of Equation 2.8145% for a 10 millisecond square wave shock pulse. While the results of this paper cannot be used in a quantitative sense. the total area for the half sine wave is square wave. The error for the square wave is small becau-e most of the area under the curve is about 25 cps where the gain is unity. .7111% for a 10 millisecond half sine wave shock pulse. However. it could represent a sizeable amount of distortion. It is generally neglected in practice but as has been shown. 3. the areas below 25 cps are nearly the same thereby making a larger error for the half sine wave. The large error for the triangular wave is due to the fact that most of the area in the low frequency range where the gain is less than unity. and 9.8988% for a seven millisecond triangular wave shock pulse.

APPENDIX .

0to ------~oo Rc "26O 06 1 3Cooo II Fig A-i Frequency Spectrum for a Square Wave Shoot Pulse for a 10 Millisecond Duration .to q0o COO Frequency (cps) --.

ooo Fig A-i Frequency Spectrum for a Square Wave Shoct Pulse for a 10 Millisecond Duration .Frequency (cps) r1a !•o I:6oo 2 .

Frequency (cps) Fig A-2 Frequency Spectrum for a Half Sine Wave Shook Pulse for a 10 Millisecond Duration tii .

.•' •~• Frequency (cps) -o Fig A-2 Frequency Spectrum for a Half Sine Wave Shock Pulse for a 10 Millisecond Duration 2"'- .

200 ape .5- 10 t"oo L+oO 3oa Frequency (cps) goa tI Fig A-3 Frequency Spectrum for a Triangular Wave Shoal Pulse for a 7 MIlliseoond Duration between 5 and 2.

zoo 300 +O0 ooo Frequency (cps) goo-- Fig A-3 Frequency Spectrum for a Triangular Wave Shock Pulse for a 7 uillisecond Duration between 5 and 2.200 ops .

.5" 3.• •z.25 and 6 cpa ."g Frequency (ope) Fig A-4 Frequency Spectrum for a Triangular Wrve Shook Pulse for a 7 Millisecond Duration between 0.

The half sine pulse which is often used in the simulation of shock events can be described in the frequency domain.

The half sine pulse which is often used in the simulation of shock events can be described in the frequency domain.

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue listening from where you left off, or restart the preview.

scribd