^ 
 
 v-0 
 
 »p-nK 
 
 «" •< 
 
 •I Pa 
 
 
 ' <£ ^ WW* > 
 
 ^ 
 
 <b^ 
 
 ,/ *.•• 
 
 
 w 
 
 
 v>* 
 
 v-o 
 
 
 "J 
 
 ,0' 
 
 *> 
 
 
 W 
 
 ,* % 
 
 ■if 
 
 
 
 . o " « 
 
 W 
 

 o * . „ . o cr 
 
 0* 
 
 e> vP 
 -- * <L. V «► • 
 
 4 «\*SSW. V c° ♦* 
 
 * 4* *« 
 
 <K * 
 
 » o 
 
 
 
 r£- ^ J^lli3$ * ^ & *■* 
 
 . " • " ° A <K 
 
 %>/ -Mfe V**' 
 
 <T7T< G* *o -.". 
 
 ► *r^7* g^ V 
 
 ^ ^. '• • » * ^ <» *'TvT t .0*' *o, •• . » * A 
 
 
 , *S-, t^ 
 
 .^ * 
 
 ^°^ 
 
 
 ■^o* 
 
 
 •*• 
 
 <$ 
 
 .v O * 
 
 6* 
 
 
 4* V o^ * *^fc 
 
 a 
 
 •Pi. ^4 v .« a 
 
 *°* -' 
 
 ,. v »!itf»:% ^. .4° ,*;■ 
 
IC 
 
 8883 
 
 Bureau of Mines Information Circular/1982 
 
 Vibration Qualification of Electronic 
 Instrumentation for Underground 
 Coal Mining Machinery 
 
 By Roy C. Bartholomae, Bruce S. Murray, 
 and Richard Madden 
 
 UNITED STATES DEPARTMENT OF THE INTERIOR 
 
obM* 
 
 Information Circular 8883 
 
 Vibration Qualification of Electronic 
 Instrumentation for Underground 
 Coal Mining Machinery 
 
 By Roy C. Bartholomae, Bruce S. Murray, 
 and Richard Madden 
 
 UNITED STATES DEPARTMENT OF THE INTERIOR 
 James G. Watt, Secretary 
 
 BUREAU OF MINES 
 Robert C. Horton, Director 
 

 
 This publication has been cataloged as follows: 
 
 Bartholomae, R. C 
 
 Vibration qualification of electronic instrumentation for 
 underground coal mining machinery. 
 
 (Information circular ; 8883) 
 
 Includes bibliographical references. 
 
 Supt. of Docs, no.: I 28.27:8883. 
 
 1. Coal-mining machinery— Vibration— Measurement. 2. Electronic 
 apparatus and appliances— Vibration— Testing. I. Murray, Bruce S. 
 II. Madden, Richard. III. Title. IV. Series: Information circular 
 (United States. Bureau of Mines) ; 8883. 
 
 TN295.U4- [TN813] 622s [622] 82-600497 AACR2 
 
jd 
 
 r^/v 
 
 CONTENTS 
 
 Page 
 
 Abstract 1 
 
 Introduction 2 
 
 Data base 2 
 
 Data analysis 3 
 
 Development of vibration qualification tests 6 
 
 Discussion 9 
 
 ILLUSTRATIONS 
 
 1. Horizontal acceleration on main frame of continuous miner during loading... 2 
 
 2. Acceleration level distribution in the 160-Hz 1/3-octave band 4 
 
 3. Cumulative probability of acceleration level in 160-Hz 1/3-octave band 4 
 
 4. Summary of statistics of acceleration levels for underground mining 
 
 machines 6 
 
 5. Summary of swept sine vibration qualification tests issued by the military. 8 
 
 6. Comparison of underground mining machine vibration data with MIL-STD-810B. . 8 
 
 TABLE 
 
 1. 1/3-octave band acceleration level statistics 5 
 
' 
 
 Bnona 
 
VIBRATION QUALIFICATION OF ELECTRONIC INSTRUMENTATION 
 FOR UNDERGROUND COAL MINING MACHINERY 
 
 By Roy C. Bartholomae, ! Bruce S. Murray, 2 and Richard Madden 3 
 
 TJ-i 
 
 ABSTRACT 
 
 An accurate characterization of the vibration environment and a 
 vibration qualification test derived from it will be a very useful tool 
 for manufacturers of instrumentation for use on underground coal mining 
 equipment. Recognizing this, the Bureau of Mines sponsored a study 
 wherein vibration levels were measured on mining equipment to form a 
 basis for developing the required vibration test. The data base was 
 composed of 160 samples taken at different positions on a variety of 
 underground machinery. The data were analyzed and presented in a format 
 typical of military vibration qualification tests. The form was shown 
 to be virtually identical to the swept sine test envelope specified in 
 MIL-STD-810B for tracked vehicles. 
 
 Supervisory electrical engineer, Pittsburgh Research Center, Bureau of Mines, 
 Pittsburgh, Pa. 
 
 ■'Senior engineer, Bolt Beranek and Newman, Inc., Cambridge, Mass. 
 3 Department manager, Bolt Beranek and Newman, Inc., Cambridge, Mass. 
 
INTRODUCTION 
 
 The desire for increased production 
 of coal, together with the growing empha- 
 sis on health and safety in underground 
 coal mines, means that additional elec- 
 tronic instrumentation will be needed on 
 underground mining machines. To insure 
 
 operation of such 
 the harsh environ- 
 
 long-term reliable 
 
 instrumentation in 
 
 ment of underground 
 
 must consider the 
 
 at the design stage. 
 
 the instrument designer is a vibration 
 
 mines, manufacturers 
 
 effects of vibration 
 
 A useful tool for 
 
 qualification test, preferably one de- 
 signed to meet an existing military stan- 
 dard, because many commercial labora- 
 tories are already set up to perform 
 testing to meet military standards. 
 
 This Bureau of Mines report de- 
 scribes the development of a vibration 
 qualification test and compares the test 
 with similar ones used by the armed 
 forces. 
 
 DATA BASE 
 
 The first step in developing the 
 vibration qualification test was to cre- 
 ate a data base. Under U.S. Bureau of 
 Mines Contract H0155113, data were taken 
 on 30 mining machines of 8 different 
 types, including continuous miners, 
 loaders, cutting machines, track jeeps, 
 face drills, shuttle cars, roof bolters, 
 and scoop trams. Vibration measurements 
 
 were made at a variety of positions on 
 each machine, resulting in a total of 
 160 samples in the data base. Data were 
 recorded on magnetic tape and then ana- 
 lyzed in 1/3-octave bands in the fre- 
 quency range between 3.2 and 500 Hz. A 
 typical sample of data taken on a conti- 
 nuous miner is shown in figure 1. 
 
 -10 
 
 o E 
 
 ^ < OJ 
 
 o = ^ 
 
 o "J 
 
 55 c_, ja 
 
 in < -° 
 
 t— 
 
 1 o — i 
 
 UJ z uj 
 
 2 <S. ^ 
 
 ° m ^ 
 
 -20 - 
 
 3.15 6.3 12.5 25 50 100 200 400 
 
 ONE-THIRD OCTAVE BAND CENTER FREQUENCY, Hz 
 
 FIGURE 1. - Horizontal acceleration on main frame of continuous miner during loading. 
 
 800 
 
DATA ANALYSIS 
 
 Data from each of the samples were 
 combined to form a 160-sample data set in 
 each of the 23 1/3-octave bands between 
 3.2 and 500 Hz. The following analysis 
 was used to convert the data base into a 
 usable vibration qualification test. 
 
 Qualification tests are customarily 
 based upon the highest vibration levels 
 encountered in service. The highest 
 level can be estimated as that level 
 which is not exceeded either more than 
 1 time in 100 (99 +h percentile) or more 
 than 1 time in 1,000 (99.9 +h percentile). 
 Our 160-sample data base is clearly too 
 small in itself to determine these low- 
 probability occurrences. If, however, 
 the 160-sample data base is representa- 
 tive of a larger well-defined data base, 
 we can determine either or both of these 
 percentiles with confidence. Fortu- 
 nately, we are able to show that a normal 
 distribution curve well represents the 
 data. We illustrate this by constructing 
 an amplitude histogram from the samples 
 in each 1/3-octave band and comparing it 
 to the normal distribution. The histo- 
 gram was formed by grouping the data in 
 ten 5-db steps in the acceleration level 
 range from -40 db re 1 g (0.01 g) 4 to +5 
 re 1 g (1.78 g) . 
 
 The total number of samples in each 
 histogram equals the total number of re- 
 cordings (that is, 160). In cases where 
 
 ^Acceleration level 
 (acceleration in g's). 
 
 data fell below the noise level of the 
 measurement system, these data were 
 grouped with the -40 db re 1 g (0.01 g) 
 data. 
 
 As an example, the histogram for the 
 160-Hz 1/3-octave band is presented in 
 figure 2. The number of occurrences in 
 each 5-db step is given by the ordinate 
 to the left, and the associated probabil- 
 ity of occurrence is given on the right. 
 
 The cumulative percentage distribu- 
 tion is plotted on normal distribution 
 paper in figure 3. Since a straight line 
 fits the data very well, the data can be 
 considered normal. Note that deviations 
 from the line at the end points are not 
 significant; that is, at lower levels, 
 data below the noise floor were grouped 
 with the -40 db re 1 g data, and at the 
 high end there is only one sample greater 
 than db re 1 g. The mean for this sam- 
 ple is given by the 50 +h percentile as 
 -22.5 db, and the standard deviation is 
 10.5 db. 
 
 Extrapolation based upon a normal 
 distribution with a mean of -22.5 db re 
 1 g and a standard deviation of 10.5 db 
 gives the 99 +h and 99.9 +h percentiles 
 (rounded to the nearest 0.5 db) as 
 
 ALg9 # 9 = P + 3.1a = 10 db re 1 g 
 
 and AL99 = \i + 2.33a = 2 db re 1 g. 
 
 equals 20 1og<|Q 
 
 
40 
 
 CO 
 
 111 
 
 O 30 
 
 LU 
 
 cr 
 
 tr 
 
 D 
 O 
 O 
 
 o 
 
 Li. 
 
 o 
 
 DC 
 LU 
 
 CD 
 
 20 - 
 
 10 - 
 
 
 I 
 
 I I 
 
 I 
 
 i 
 
 
 
 I ■ 
 
 i 
 
 • i 
 
 I 
 
 
 I 
 
 i 
 
 
 
 ■50 
 
 - 10 
 
 20 1 
 
 3 
 Q- 
 
 o 
 
 LU 
 
 (5 
 < 
 
 I- 
 Z 
 LU 
 
 o 
 
 DC 
 LU 
 Q- 
 
 -40 
 
 10 
 
 -30 -20 -10 
 
 ACCELERATION LEVEL, db re 1 g rms 
 
 FIGURE 2. - Acceleration level distribution in the 160-Hz 1/3-octave band. 
 
 -30 -20 -10 
 
 ACCELERATION LEVEL, db re 1 g rms 
 
 FIGURE 3. - Cumulative probability of acceleration level in 160-Hz 1/3-octave 
 
 band. 
 
A similar analysis was performed for 
 data in each of the 1/3-octave bands, and 
 in each case the distribution was normal. 
 The mean, standard deviation, and 99 th 
 and 99.9 +h percentile values for each 
 1/3-octave band are listed in table 1 and 
 plotted in figure 4. It is important to 
 remember that these values describe the 
 statistics of the samples from each 
 1/3-octave band and that they do not 
 describe an expected spectrum shape. 
 
 Rather, the data provide an overall enve- 
 lope upon which to base qualification 
 tests. Also plotted in figure 4 are the 
 maximum values recorded in each of the 
 1/3-octave bands. Note that a majority 
 of these values fall between the 99 th 
 and the 99.99 th percentile values, as 
 expected from the 160-point sample size, 
 thus providing a limited validation of 
 the approach. 
 
 TABLE 1. - 1/3-octave band acceleration level statistics 
 
 1/3-octave band 
 
 Acceleration level, db re 
 
 1 g rms 
 
 center frequency, 
 
 Mean 
 
 Standard 
 
 99th 
 
 99.9th 
 
 Hz 
 
 
 deviation 
 
 percentile 
 
 percentile 
 
 3.2 
 
 -18 
 
 9.5 
 
 +4 
 
 +11.5 
 
 4.0 
 
 -20 
 
 9 
 
 +1 
 
 +8 
 
 5.0 
 
 -21 
 
 11 
 
 +4.5 
 
 +12 
 
 6.3 
 
 -21.5 
 
 11.5 
 
 +5.5 
 
 +14 
 
 8.0 
 
 -23.5 
 
 8.5 
 
 -4 
 
 +2.5 
 
 10 
 
 -25 
 
 10.5 
 
 -.5 
 
 +7.5 
 
 12 
 
 -25 
 
 9.5 
 
 -.5 
 
 +4.5 
 
 16 
 
 -26 
 
 10.5 
 
 -1.5 
 
 +6.5 
 
 20 
 
 -25 
 
 9.5 
 
 -3 
 
 +4.5 
 
 25 
 
 -23.5 
 
 8.5 
 
 -1.5 
 
 +3 
 
 32 
 
 -24 
 
 10 
 
 -.5 
 
 +7 
 
 40 
 
 -24.5 
 
 13 
 
 6 
 
 +15 
 
 50 
 
 -24 
 
 11 
 
 +1.5 
 
 +10 
 
 63 
 
 -25 
 
 12.5 
 
 +4 
 
 +13.5 
 
 80 
 
 -23.5 
 
 11 
 
 +2 
 
 +10.5 
 
 100 
 
 -23.5 
 
 10 
 
 
 
 +7.5 
 
 125 
 
 -22.5 
 
 10 
 
 +1 
 
 +8.5 
 
 160 
 
 -22.5 
 
 10.5 
 
 +2 
 
 +10 
 
 200 
 
 -22 
 
 10 
 
 +1 
 
 +9 
 
 250 
 
 -20 
 
 10 
 
 +3.5 
 
 +11 
 
 315 
 
 -17 
 
 11.5 
 
 +10 
 
 +18.5 
 
 400 
 
 -17.5 
 
 12 
 
 +10.5 
 
 +19.5 
 
 500 
 
 -17.5 
 
 11.5 
 
 +9.5 
 
 +18 
 
LU 
 > 
 LU 
 
 < 
 
 DC 
 LU 
 
 LU 
 O co 
 
 2 E 
 
 < 
 
 H 
 O 
 
 o 
 
 Q 
 
 oc 
 
 I -20 
 
 h- 
 
 LU 
 
 -30 
 
 99.9th 
 percentile 
 
 99th 
 percentile 
 
 KEY 
 Highest measured levels 
 
 Mean 
 
 i I ■ ■ I 
 
 3.15 6.3 12.5 25 50 100 200 400 
 
 ONE-THRID OCTAVE BAND CENTER FREQUENCY, Hz 
 
 FIGURE 4. - Summary of statistics of acceleration levels for underground mining machines. 
 
 800 
 
 DEVELOPMENT OF VIBRATION QUALIFICATION TESTS 
 
 Virtually all the vibration qualifi- 
 cation tests conducted in the United 
 States are based on Military Standards 
 (MIL-STD) developed over many years. 
 These standards reflect experience in a 
 vast number of environments. It is rea- 
 sonable, therefore, to base any vibration 
 tests upon the military standards, at 
 least initially, and amend them later if 
 necessary to meet particular require- 
 ments. Another major benefit of the use 
 of military standards relates to test 
 equipment performance and limitations. 
 
 The development of these standards as 
 viable test procedures has been closely 
 linked with the attainable performance of 
 readily available shakers and associated 
 hardware. Thus, specification of MIL-STD 
 levels will assure that the tests are 
 experimentally possible. 
 
 With these considerations in mind, 
 we reviewed four pertinent military stan- 
 dards to determine their suitability 
 for the underground mining equipment. 
 
The vibration qualification standards 
 reviewed were MIL-STD-167, MIL-STD-810B , 
 MIL-STD-810C, and MIL-E-5272-C . 
 
 Although many other standards in- 
 clude vibration test procedures, these 
 four provided data for a wide range of 
 applications. MIL-STD-167 provides data 
 on shipboard vibration levels; MIL- 
 STD-810 is the U.S. Air Force environmen- 
 tal test procedure for flight and ground 
 vehicles; and MIL-E-5272-C is concerned 
 with environmental testing of equipment 
 destined for aircraft, ship, and missile 
 applications. 
 
 The review consisted of evaluating 
 the vibration test specifications in each 
 of the Military Standards against the 
 measured and predicted vibration levels 
 of underground mining equipment. Fig- 
 ure 5 presents a summary of the vibra- 
 tion spectra specified for swept sine 
 tests in the MIL-STD's reviewed. The 
 spectra shown for MIL-E-5272-C and MIL- 
 STD-167 are the highest levels indicated 
 in that particular specification. How- 
 ever, in the case of MIL-STD-810B and 
 MIL-STD-810C, we have presented the most 
 appropriate spectrum based on the 
 description in the standard "tracked 
 vehicles." Note that these spectra are 
 presented with the measure of vibra- 
 tion amplitude being "displacement (peak- 
 to-peak) inches" rather than "db re 1 g 
 rms. " 
 
 The basic expression used to con- 
 vert vibration amplitude from acceler- 
 ation (in db re 1 g rms) to displacement 
 
 (in inches peak to peak), assuming that 
 the vibration is sinusoidal, is given by 
 
 D.A. - g*$ 1C/20 
 
 where D.A. = displacement (peak-to-peak), 
 inches, a = acceleration (db re 1 g rms), 
 f = frequency (Hz), and g = 386.4 
 (in/sec 2 ). 
 
 This expression reduces to 
 
 27.68 
 
 D.A = 10 a / 20 x 
 
 77- 
 
 As an example, we calculate the dis- 
 placement produced by the 99.9 th percen- 
 tile acceleration at 160 Hz from the data 
 in table 1: 
 
 D.A. = 1010/20 x 
 
 27.68 
 (160)2 
 
 = 0.0034 in. 
 
 Figure 6 presents a summary of the 
 vibration levels expected on underground 
 mining machines, the displacement values 
 being derived from the data given in 
 table 1. The open circles represent the 
 vibration amplitude that will only be 
 exceeded 1 time in 1,000 and the closed 
 circles show the amplitudes that will be 
 exceeded 1 time in 100. We have also 
 plotted the MIL-STD-810B level for 
 tracked vehicles, and it is seen to be a 
 good match over virtually all the fre- 
 quency range except below 8 Hz . The 
 other specifications plotted in figure 5 
 do not provide a suitable match to the 
 plotted points. 
 
o 
 
 
 co 
 
 DO 
 
 CD 
 
 O 
 
 O 
 
 T— 
 
 sz 
 
 on 
 
 CD 
 
 i 
 
 > 
 
 Q 
 
 01 
 
 U) 
 
 .* 
 
 i 
 
 _J 
 
 o 
 
 CO 
 
 o 
 o 
 o 
 
 0) 
 
 c 
 
 .1c 
 u 
 o 
 
 E 
 
 .E o 
 ■ — oo 
 E O 
 
 I 5 °? 
 
 2d 
 ?2 
 
 ° (- 
 
 — O 
 
 > 
 
 O 
 
 z 
 
 LU 
 
 =) 
 
 o 
 
 LU 
 
 DC 
 
 O 
 
 o 2 
 
 o T 
 
 i_ o 
 o • r 
 
 o -9 
 U > 
 
 LU 
 
 O 
 
 o 
 o 
 
 o 
 o 
 
 S9L|0U! '(*V3d-01->lV3d) lN3W30VndSia 
 
 o 
 o 
 o 
 
 a 
 
 o- 
 
 c 
 o 
 
 2 ^ 
 
 > .■£ 
 
 <u ~ 
 
 .E E 
 
 W <D 
 
 4- _C 
 
 Q. *" 
 
 3! >- 
 
 ^ s 
 ts 
 
 a — 
 E w 
 
 I " 
 
 CO « 
 
 LU 
 
 ct: 
 
 ID 
 O 
 
 S8L|0U! '(*V3d-0±->lV3d) lN3W30VHdSia 
 
DISCUSSION 
 
 The deviation of the standard below 
 8 Hz is most likely due to the normal 
 displacement limitations of commercial 
 shakers. This is an example of a speci- 
 fication being tailored to fit the avail- 
 able equipment so that the testing can 
 be performed economically. It is likely 
 that the tracked vehicles considered 
 by MIL-STD-810B have low-frequency vibra- 
 tion amplitudes much like those plotted 
 in figure 6. Since the displacement- 
 limited curve satisfactorily tests the 
 equipment of these vehicles, we may con- 
 clude that it will do the same for the 
 equipment on underground mining machin- 
 ery. Therefore, we recommend that the 
 MIL-STD-810B vibration tests curves for 
 
 category f equipment 
 curve W. 
 
 be employed, using 
 
 It should be noted that MIL-STD- 
 810B, issued in June 1967, is not the 
 latest issue of this standard; it is used 
 because MIL-STD-810C, issued in 1975, 
 specifies a different vibration test for 
 components on tracked vehicles that does 
 not adequately simulate the expected 
 levels on mining equipment. The specifi- 
 cation set in issue B is equivalent to a 
 ±4 g level from 9 to 500 Hz. The speci- 
 fication in issue C is ±1.5 g from 5.5 to 
 30 Hz and ±4.2 g from 50 to 500 Hz. The 
 reason for the reduction in levels below 
 50 Hz is not known. 
 
 INT.-BU.OF MINES, PGH..PA. 26133 
 


 
 W 232 
 
~ 
 
 ■■ 
 
 ■ 
 
 ^~ 
 
 ™ 
 
 r_— — _ 
 
 
 
 * > . 
 
 
 
 
 
 
 »°^ 
 
 -i*"' 
 
 
 ^o N 
 
 
 
 A c o " ° 
 
 ^ 
 
 iV 
 
 
 
 > 
 
 kV 
 
 
 
 
 ^ .'J^t' ^ 
 
 ...... 
 
 1 ", **b 
 v*Cr 
 
 
 <<;• 
 
 
 
 
 
 •> 
 
 ; ^ V % '-ISP* ^^ °-^w ; ^ v % '-3K* j^^\ >~ 
 
 A^' 
 
 
H°* 
 
 
 JlOv\ 
 
 
 
 ^ % 
 
 
 o „ o 
 
 
 ^** a* 5 "** 
 
 ^r 
 
 
 *>*o* 
 
 
 H°* 
 
 C, vP 
 
 ** ,** /a&- ^ ^ 
 
 *^ A^ •' 
 
 O vP^ ° 
 
 
 ^V 
 
 
 ^\ s • • 
 
 C u ♦ 
 
 ■» o 
 
 v ,.\ww> q* j, 
 
 <<**, ^l!?!3l5* vP C, 
 
 * ** ** • 
 
 C, <P 
 
 r > 
 
 
 ' » . o ' A ^P- 
 
 C, vP 
 
 
 1 ** ^ • 
 
 
 «5°^ 
 
 
 v^ 1 
 
 
 O vP^ c 
 
 " V* , \ 
 
 .' .0' %> -. 
 
 ^5^ 
 
 ■» o 
 
 ***** 
 
 <^ vP 
 
 ^ -P-N. 
 
 .^.•a5it.S /*&k>* /s&kS ^tifc* * 
 
 \ v .. <U ' " ° ' A 
 \P 9 * 
 
 ► ** -^ . 
 
 ^ ^. -: 
 
 ^ t ^ 
 
 
 ^0^ 
 
 
 <e» v 
 
 
 #++. -. 
 
 
 ^^ 
 
 <. *'TVi« ,0 
 
 e> vP 
 
 
 '%<? 
 
 
 ^ 
 
 
 4 > 5 • • 
 
 c. »p 
 
 tP vP» 
 
 N 
 
 > ^ 
 
 *^* 
 
 c,«P 
 
 ^ AUG 82 
 
 N. MANCHESTER, 
 'Sfe^ INDIANA 46962 
 
 ^>V 
 
 "W 
 
 /.C^^ o 
 
 
 • A>*^. - 
 
 ... ^ * e "°° ^ 
 
 ,v %^ :j 
 
 
 
 v.^ *^v^\ ^ ^ y- 
 
,'^OFCOHGRBSS 
 
 002 959 8216