Ghostscript wrapper for D:\Digitalizacja\MTS85_t23z1_4_PDF_artyku³y\mts85_t23z3_4.pdf M E C H A N I K A TEORETYCZNA I STOSOWANA i. 4, 23 (1985) THE APPLICATION OF DYNAMIC HOLOGRAPHIC INTERFEROMETRY AND PHOTOELASTICITY TO PROBLEMS IN GEOTECHNICAL ENGINEERING D. C. HOLLOWAY, W. H. WILSON, W. L. FOURNEY, D. B. BARKER Department of Mechanical Engineering University of Maryland College Park, Maryland 20742 USA. Abstract The methods of dynamic holographic interferometry, using a pulsed ruby laser, and dynamic photoelasticity, using a Cranz-Schardin camera, have been applied to the study of explosives and their effects in geologic applications. Specifically, the problem of stress wave generation from a surface detonation and how these stress waves interact with surface trenches was studied via holographic interferometry. This method was also applied to the examination of the shock front emitted from a non-electric detonating cord. Dynamic photoelasticity was employed to examine the role of remote discontinuities in explosive fragmentation, and in the determination of the optimum placement of explosives for gas and oil well stimulation. Introduction Dynamic holography and photoelasticity,can be valuable tools in the study of how explosives can be more efficiently used in the civilian applications of mining, construction, and natural resource stimulation. With the emphasis today on saving costs, improving fragmentation results, improving the quality of the standing geologic material, and reducing immediate and long-term environmental effects there is a need to examine how explosives work to fracture rock. Holography and photoelasticity are convenient full field laboratory methods that provide information on the relative importance of the many blasting para- meters. The results from these tests have been both quantitative and qualitative, and in many cases have been successfully scaled up for field application. This paper will focus on some recent problems that our laboratory has studied. 16* 596 D . C. HOLLOWAY, W. H . WILSON , W. L.  FOURNFA- ,  D .  B.  BARKER 1. A Holographic Study of Dynamic Surface Response to Explosive Loading Within  the last  several  years  there  has  been  an  increased  need  to  con trol  and  reduce ground  vibrations  from  surface  blasting.  As  population  centers  have  spread,  the number of cases  in  which  blasting  must  be don e close  to  or even  within  developed  areas  has increa- sed.  Early  studies  attempted  to  identify  characteristics  of  th e  stress  waves  which  could be correlated  to structural  damage, and  then to  quantify  safe levels  of  blasting  based  on these correlations.  Another  approach  is  to  attem pt  to  isolate  structures  or  entire  surroundings froni  the detonation area. Since most of the stress wave energy  produced from  a detonation is carried  near the surface  by  Rayleigh  type  or R  waves, it lias  been suggested th at  effective isolation  can  be  obtained  by  surrounding  a  blast  area  with  trenches  of  some  depth. For a  more  detailed  discussion  of  th e background  of  this  problem ,  see  H ollową y  and  Wilson (1983).  In order  to  study  this  approach, an d  to  quantify  such  param eters  as  trench depth, location,  and  effectiveness,  we  have  made  extensive  use  of  holographic  interferometry, A  complete  description  of  the  techniques  used  t o  produce  ph otograph s  of  dynamic surface  response  by  holographic  interferometry  can  be  found  in  H olloway  an d  Patacca (1975),  and  H olloway,  F ourney  and  P atacca  (1977).  A  pulsed  ruby  laser,  operating  at 6943  A,  was  used  to  generate  double  exposure  hologram s  of  th e  model  surface.  Typical Q- switched  pulse  duration  was  30  nanoseconds,  with  a  200  microsecond  delay  between pulses.  An  explosive  charge  of  from  70  to  250  mg  of  P E T N   was  deton ated  on  th e model surface  at  a  controlled  time  between  the  two  laser  pulses.  Specimens  tested  were  granite blocks  or  cast  blocks  of  H ydrostone,  a  product  of  U n ited  States  G ypś um  which  when cured has  a  specific  gravity  of  1.69  and  a  cpmpressive  .strength  of  9,750  psi  (67.2 mPa), I n  each test,  the charge  was  located  80  mm  from a  trench  of  rectangular  cross  section cut into the model surface.  Trench depths of  10 mm an d  18 m m were used. The illumination ,:  u i Fig.  1, R  and RP  wave patterns  on a  H ydrostone model, 90 f/.s  after  detonation.  Lines  7, 2  and 3  used in plots in F ig. 2. Points A  through G  (except  C) show  maximums, minimums and inflection  points of the vertical  surface  displacement T H E  APPLICATION  OF  DYNAMIC... 597 and  viewing  angles  were  arranged  so  that  the  fringe  pattern  in  the  hologram  represents a topological  map  of  displacement  normal  to the model  surface  at the instant  of the  second laser  pulse;  Thus,  we  were  able  to  study  the  overall  effect  on  vertical  ground  motion in the waves crossing  the  trench.  Figure  1 shows  the  fringe  pattern  90 ;J.S after  detonation, for the case of an  18 mm trench  depth.  The  radial  lines marked  I and  2 and  a line marked  3 along  the  trench  have  been  used  in  the  analysis of  the  surface  displacements  presented in  Figure  2.  At  the  far  left  along  line  1  (no  trench)  are  seen  remnants  of  a  shear  wave recognized  by its typically asymmetric pattern. Nearer the source, the peak positive displace- ment of the R wave is located at point A. The region of steepest displacement gradient then follows, as can be seen from the very tight clustering of the fringes. This is also the region of highest particle velocity. A minimum or valley is reached at point B. The surface then moves upward again. The air shock front is reached at point C, and all surface inter- ferometrie information is lost inside this radius. Moving inward along line 2 from the right, we first notice a disturbance which has no counterpart on line 1. A small positive peak is reached at point D, followed by a local minimum at E. This displacement profile is very similar to that of an iJ-wave, This wave shows up consistently from test to test. Analysis of the tests show its velocity to be the same as that of an R-wave, and it is found to have been located at the trench just as the /'-wave passed the trench. This R type wave (precursor to the main i?-wave which follows), is created by the P-wave, and will be called an RP-wave. Further in along line 2, the relative peak of the transmitted i?-wave is found at point F, and its minimum at G. As can be seen from Figure 2 there is a reduction of amplitude and an increase in wavelength for the transmitted wave; this has been found to be a general result. UJ u s 1 DISPLACEMENTS ALONG THREE LINES A B • • • • " ., 160 R  WAVE  POSITION TRENCH DEPTH'ISmm ­ N O  TRENCH —TRENCH .­• ALONG  FAR  COGf OF  TRENCH 240 260 Fig. 2. Wave shapes 90 ;xs after detonation, along lines 1, 2 and 3 of test shown in Figure 1. Wave position is relative to load location. Distance i\ is a repeatable feature in the R wave Along line 3 on the far edge of the trench, there is at the lower end a very weak upward motion in the head of the wave followed by a downward displacement which is greater than the deepest part of the iJ-wave on line 1. We believe that there is another R-type wave traveling on the vertical side of the trench that is contributing to the R-wave displace- 598 D.  C.  HOLLOWAV,  W.  H .  WILSON,  W.  L.  FOURNEY,  D .  B.  BARKER ments  on  the  horizontal  surface.  This  observation  points  to  the  inadvisability  of  locating a  structure  within  an R  wavelength  of  an  isolation  trench. An  easily  identifiable  feature  of  the i?-wave is the portion of the wavelength between the maximum upward and the deepest positions. The highest particle velocities occur in this part of the .R-wave. This distance has been labeled as v\ in Figure 2 and is used in normalizing the results with trench depth, so that the effect of the trench can be measured. Peak particle velocities were calculated for fourteen tests, comparing J?-waves on each side of the trench. The quantity (K/)-trench)/(F/>no-trench), where Vp is peak particle velocity, is plotted in Figure 3 as a function of the ratio of trench depth / / over fractional, wavelength ij (#/?/). The linear fit to the data in the region of analysis is good, with a least squares correlation coefficient of .995. Figure 3 shows that trench effectiveness in R wave attenuation increases with the ratio, trench depth over wave length. Thus, for a given trench depth, the higher frequency com- PARTICLE  VELOCITY  REDUCTION  v> TRENCH  DEPTH  (NORMALIZED) . 6 5 . 4 . ­ . 3 . 2 • . 1 0 .2 .4 .6 Fig. 3. Least squares linear fit for 14 tests, showing ratio of peak particle vertical velocity vs. Hjrj. H is depth of trench, and ij is portion of wave from maximum to minimum. Typical ?j length is shown in Figure 2 ponents of the .R-wave are more effectively attenuated. The effect of the trench on the i?-wave is to reduce the overall displacement and to broaden the wave, thus lowering its frequency and reducing its particle velocity. Our study also has identified other effects of the trench on the full wave system which should be considered in design of screening trenches. The distance of the trench from the source and trench depth are important factors, with some trade-offs involved. The P-wave, upon crossing the trench, has part of its energy converted to a new i?P-wave, attenuating as l/r2, while the i?P-wave generated will attenuate only as l/j/V. Although increasing trench depth improves i?-wave blocking, it also increases the amount of P-wave energy the trench will intercept and convert to RP-wave energy. Finally, it also was observed that a high amplitude iMype wave propagates on the vertical surface of the trench. 2. Holographic Interferometry Applied to the Study of an Explosive Shock Front Holographic methods are being used in our laboratory to study the behavior of shock fronts from various explosive detonators and initiator systems. Important parameters to be studied are the speed of the shock front, the pressure distribution behind the shock, and T H E  APPLICATION  OF  DYNAMIC... 599 the shape  of  the  shock  as  it  advances.  The  advantage  of  holographic  techniques  is that  we nain  information  which  can  be  related  to  all  of  these  topics.  Figure  4  shows  a  spherical shock  front  generated  95  j^sec  earlier  at  the  end  of  an Ensign-Bickford Nonel detonating tube. Nonel tube is a 3 mm diameter plastic tube coated internally with a thin film of high explosive. It is used to initiate detonations from remote locations by propagating, an explosive shock along the inside of the tube. If is manufactured by Ensign-Bickford using Fig. 4. Gas shock 95 s after being formed at end of Nonel detonator tube proprietary and patented processes, under license from the Nitro Nobel Co. of Sweden, its inventor. As the explosively driven shock front propagates away from the end of the tube, the gas behind it is compressed, altering its index of refraction. Thus, the optical path is changed due to both a pressure rise and a change in the explosive gas composition. The combined effect produces interferometrie fringes in the holographic image. By integra- tion through, the thickness of the spherical shock volume, we will be able to produce a pressure map of the shock intensity. Holographic interferometry can be used in a similar way to measure the strength and shape of other detonation system components, such as blasting caps and detonating cord, and to illustrate the high speed effects of these detonators on nearby components such as coupling blocks or other explosives. 3. A Dynamic Photoelastic Investigation of Fracture Initiation from Flaws Driven by Explosively Generated Stress Waves In order to gain a basic understanding of the various fragmentation mechanisms in rock blasting, an experimental program was conducted with two dimensional polymeric models. Dynamic photoelasticity permitted the crack propagation behavior to be studied 600 D.  C.  HOLLOWAV,  W.  H.  WILSON,  W.  L.  FOURNEY,  D.  B.  BARKER in  detail.  The  use  of  the Cranz-Schardin camera to record dynamic photoelastic fringe patterns, stress wave propagation, and crack propagation has been previously described; sec for example Riley and Dally (1969). The particular example presented in this paper describes work conducted to study the initiation of cracks from flaws outside of the irnme- diate borehole vicinity, and the contribution of these remote cracks to the overall fragmen- tation pattern. For a more detailed description, see Barker and Foiirney (1978). The fragmentation model was 305 mm in diameter and 6 mm thick, made from a sheet of a brittle birefringent polymer known as Homalite 100. Approximately 250 mg of PETN was tightly packed into a 6 mm diameter borehole in the center of the model. Pressure containment within the borehole was obtained with steel end caps and a lever arm arrange- ment that did not significantly obscure the visibility of the experiment. The explosive was detonated while the model was in a light field polariscope, and a Cranz-Schardin multiple spark gap camera recorded the dynamic photoelastic fringe patterns. Figure 5 is a photo of the model 65'fisec after the explosive was detonated. The fringes of the outgoing P-wave are just about to reach the model boundary. The dark vertical shadows extending upward from the center of the model are the arms which hold the pressure containment caps over the borehole. At each grid intersection in the model small 6 5 ps Fig. 5. and 6. Stress waves in artificially flawed photoelastic model. '1 ime after deionaiion is shown. Note crack formation and growth at each flaw ' : 135  jus . &  > • . . • INITIATION COMSIHEO Pl» AMD S COMBINED PS  ANO S ^ ^ ZONES PP AND PS BORE O.Z ' 0 4 — — • " i . . . . . 0* BARRIER HOLE 1 & ^ I s anANCMIIIO IN BOREHOLE f VfAVELENftTH • 0 s rjj:;jt  LACS  r » » CRACKS FREE SUdrACE • Fig.  7.  Schematic  showing  regions  in  which  stress  waves  can  initiate  cracks  if  a  natural  flaw  is  present [601] 602 D .  C.  HOLLOW AY,  W.  H.  WILSON,  W.  L .  FOURNEY,  D .  B.  BARKER surface  flaws  exist. These were made  by lightly  tapping  a sharp  knife  into  the model when it was prepared.  In this frame  we see that  cracks  from  the flaws  close to the borehole have been  initiated  by  the  outgoing  wave  systems.  These  cracks  are  propagating  in  a  radial direction. In Figure 6,  135 |*sec after  detonation, it can be seen that cracks at all of the small flaw  locations  have  been  initiated  by  outgoing  and  reflected  stress  waves.  By  studying the  full  sequence  of photographs  from  the experiment  it was possible  to determine  which wave system  initiates the flaws  and how the stress waves control  the direction  of the cracks. Figure  7 summarizes  the regions  where  flaws  are initiated  by the  outgoing,  reflected, and combined  wave  systems. 4.  A  Dynamic Photoelastic Investigation of Gas and Oil Well Stimulation  by  Means  of  Explosives From  the  very  beginning  of  our  oil  and  natural  gas  industry,  owners  and  operators have attempted  to recover  a greater  percentage  of oil  and gas from  their  wells. If the reser- voir is not highly fractured, then drainage is poor and much of the oil and gas remains trapped in the rock. An early technique to fracture the formation used nitroglycerin explo- sive which was lowered into the producing zone and detonated. The results were mixed, but in many cases recovery was enhanced. Analysis done since the oil shortage of the early seventies has shown however that explosives can damage the well bore and actually impede the flow of gas or oil because of fine material trapped in the fractures, and the residual compressive hoop stresses. Other modern methods have also evolved for well stimulation. These include hydraulic fracturing with numerous combinations of liquids, gases, and liquefied gases. In general these processes are quasi-static; only a few fractures are formed, and their direction is controlled by the in-situ stresses. Propellants also have been used, since it is possible to control to some degree the rate of chemical reaction, thus reducing the well bore damage. See for example Cuderman and Northrop (1984). However, propellants are 10 to 20 times more expensive than explosives. Because of these deficiencies there has been renewed interest in developing a stimulation method that would provide an optimum pressure pulse in terms of peak pressure, total duration and time to peak pressure. In our laboratory we have investigated placing the explosive away from the zone Leman cap 12.7mm dig grooved T H E  APPLICATION  OF  DYNAMIC... 603 to  be  stimulated  (thus  minimizing  damage)  and  using  the  well  bore  geometry  and  end conditions  to  tailor  the  explosive  pressure.  For  further  background  see  Fourney,  Barker, and  Holloway  (1981)  and  Fourney,  Holloway,  and  Simha  (1984). The techniques  developed,  called  stem  induced  fracturing,  have  been  studied  using  the Cranz-Schardin camera and dynamic photoelasticity. A typical model used in the testing is depicted in Fig. 8. The models, made from 102 mm thick PMMA, were placed in front of the camera so that fractures initiated at the tips of notches along the borehole length would be seen as a plane. Stem induced fracturing utilizes a highly decoupled charge placed in the bottom of the borehole. Small piezoelectric transducers were used to monitor the pressures within the borehole and within the propagating fractures, such as at locations A and B. Four frames from a typical 16 frame test are presented in Figure 9. PMMA, commer- cially known as plexiglas, is a birefringent material. The camera was fitted with circular polaroids to permit areas of high stresses to be identified during the testing. The black fringes in Figure 9 are isochromatics. For the test shown, a 250 mg charge was used. Frame 5, Figure 9a, shows the fracture location 90 p.s after detonation (the fracture is the dark 'FRAME 15 , .;;; ; Fig. 9. Sequence from stem induced fracturing test 604 D.  C.  HOLLOWAY,  W.  H.  WILSON,  W.  L.  FOURNEY,  D.  B.  BARKER bladder-shaped area near the bottom of the model). Notice the two long straight fringes located in the upper third of the model. These show the P-wave in the PMMA which is being generated by the gas shock wave traveling up the borehole from the detonation. Each fringe makes an angle of 40° with the borehole wall, from which it is calculated that the ratio of shock wave speed to P-wave speed in the PMMA is 1.19. Note that there are two bright areas in the borehole, one at the bottom where the charge was detonated, and one near the top, at the stem. The bright area at the stem is due to an increase in pressure and temperature, caused by the shock wave reflection from the stem. Light is emitted due to ionization of, the gas at that location. The camera used to record the photographs is such that light emitted from any area within the field of view will be visible in all 16 frames, so it is not possible to determine at what time th& light flash occurred. By the time Frame 6 (Fig. 9b) was recorded, just 23 (is later at 113 (xsec, a fracture is seen to have initiated at the stemming area and has grown considerably. At this time the charge area fracture continues to grow, but at a much slower rate. By 191 [zsec (Fig. 9c) the fracture that originated at the stem is larger than the charge area fracture. Frame 15 (Fig. 9d) shows the stem area fracture beginning to engulf the arrested charge area fracture at 296 ĵ sec after detonation. Figure 10 is a sketch showing how the fracture grew from frame to frame. The times at which each frame was photographed and the velocities computed from the crack front 1 2 3 k 5 6 7 •a 10 12 14 Id 12.S 54.0 57.0 80.5 89.5 • 112.5 1*0.0 . 16I.V 190.5 2 U . 5 216.5 271.5 322.5 Chargu ,;rcn 530* 350 •: • 209 317 330 U 3 ; 128 Stun are, .V..':* 4 5 ! • 442 :, 556 314 733 107 Velocity jirob/ittly low airtcn fracture was not UW > In'.iiMi r.:_ the v.cm. inHtmr.t the Vise i-mtv.i s t - r 1 Fig. 10 positions are indicated. Numbers on the sketch represent the frame sequence and the straight lines indicate directions along which the velocity measurements were made. Notice how quickly the charge area fracture slows and arrests, while the stemming area fracture continues to propagate at a somewhat erratic but high speed. The high velocity between Frames 10 and 12 occured because the fracture reached the free edge of the model. The reason for the low velocity between Frames 12 and 14 is not clear. In many similar tests TH E  APPLICATION   OF  D YN AMIC...  •   605 using  P M M A  and  geologic  materials,  this  stem  induced  fracturing  machanism  has  been shown  to  be  very  efficient  in  initiating  and  advancing  fractures. The  im portan t  poin t  of  this  work  is  th at explosives  can  be  efficiently  used  to  generate long  fractures  from  well  bores,  which  in  turn  connect  natural  fractures,  thus  increasing the  yield  from  a  well. .  :  Acknowledgement The  authors  wish  to  thank  the N ation al  Science  F undation  and  the  U .S.  D epartment  of  Energy for  partial  sponsorship  of  the work.  Some  of  the work  by  Mr.  Wilson  was  performed  in  partial fulfillment  of  the  requirements  for the degree  of  D octor of  Philosophy  at  the U niversity  of  Maryland. References 1.  D. B.  BARKER,  and  W. L.  FOURNKY, Photoelastic Investigation of Fragmentation Mechanisms. Part II- Flaw  Initiated N etwork, Report  to U .S. N ational  Science  F oundation, Photomechanics  Laboratory, Mechanical  Engineering  D ept.,  U niversity  of  Maryland  1978. 2.  i. F. CUDERMAN, and D . A. N ORTH ROP, A Propellant- Based T echnology for  Multiple Fracturing  W ellbores T o Enhance  Gas Recovery: Application  and Results in Devonian  Shale, Proceedings, SPE/ D OE/ G RI Unconventional  G as  Recovery  Symposium,  Pittsburgh,  P a., pp.  77- 86,  1984. 3.  \V. L.  FOURN EY,  D .  B.  BARKER,  and D . C.  HOLLOW AY,  Model  Studies of  Explosive W ell  Stimulation T echniques,  I n t. J.  Rock  Mech.  Min.  Sci. & G eomech. Abstr.,  Vol. 18, pp. 113- 127, 1981. 4.  VV. L.  FOU RN EY,  D . C.  H OLLOW AY,  and  K. R. V.  SIMHA,  Model  Investigation  of  Borehole  Pressure Distribution,  Proceedings,  SPE/ D OE/ G RI  U nconventional  G as  Recovery  Symposium,  Pittsburgh, Pa.,  pp.  497  -  506, 1984. 5.  D . C.  H OLLOW AY, and A. M .  PATACCA, Application  of  Holography  to a Study of  W ave  Propagation  in Rock, Report  to  N ational  Science  F oundation  by  the Photomechanics  Laboratory,  Mechanical En- gineering  D epartment,  U niversity  of  Maryland,  College  Park,  M d.  1975. 6.  D . C.  HOLLOWAY,  W. L.  FOU RN EY,  and A. M.  PATACCA,  A  Study  of  Surface  W ave Propagation  by Holographic  Interferometry, Experimental  Mechanics,  Vol.  17, N o . 8,  August,  pp. 281- 289, 1977. 7.  D . C.  HOLLOWAY,  and  W. H .  WILSON ,  Experimental Investigation  of  Dynamic Surface Response  to Explosive  L oading,  Proceedings,  F irst  International  Symposium  on Rock  Fragmentation  by Blasting, Lulea,  Sweden,  pp.  605- 624, 1983. S.  W. F .  RILEY,  and J. W.  D ALLY,  Recording  Dynamic Fringe Patterns with a  Cranz- Schardln  Camera Experimental  Mechanics,  Vol. 9,  N o. 8, pp. 27- 33,  1969. P  e 3  IO  M e n P H M E H E H H E  T O JI O rP AO H H E C K O fl  H H T E P *E P O M E T P H H   H JSJISL   H CCJIEJtOBAH I'W  .HHHAMITqECKI- IX 3 AJ W1 TEOTEXH H KH B  paQoTe  o6cy>i< êHo  npHMenenne  roJiorpacbtfraecKoii  HHTepdpepoiweTpHH  H  cboToynpyrocni  RIIH iiccae,30BaHHH  B3ptiBou  H  Bo3Hin