ARMY RESEARCH LABORATORY

ARMY RESEARCH LABORATORY Instrumented Projectiles for Measuring Impact Forces to Characterize Ballistic Behavior of Fabrics and Composites Dahsin Liu, Guojing Li and Dan Schleh Final Report October 15, 2015 Prepared by Liuman Technologies, LLC 2727 Alliance Drive Lansing, MI 48910 Under contract W911QX-13-C-0025 UNCLASSIFIED 1

CONTENT ABSTRACT. 3 I. SUMMARY OF PROJECT INVESTIGATION... 4 II. FINAL DESIGN AND DELIVERABLES 11 III. FUTURE DEVELOPMENT.. 12 IV. APPENDICES 1. User s Manual - Horizontal Impact Testing Method... 14 User s Manual - Vertical Impact Testing Method.. 22 2. Testing Results. separated A. Testing Results of IP21-IP27 - Vertical Impact B. Testing Results of IP21-IP27 - Horizontal Impact C. Testing Results of IP33-IP39 - Vertical Impact 2

ABSTRACT Measuring impact-induced contact force history is critically important to understanding the nature of impact events and the associated damage processes of materials/structures involved in the impact. In order to investigate the impact-induced contact force history involved in free impacts, such as ballistic impact and high-speed crash, this study developed an innovative instrumented projectile. Both instrumented projectiles with and without an innovative geometry were constructed, tested and compared. An independent optical method was also used to justify the measurements from the innovative instrumented projectile. Experimental results confirmed the effectiveness and accuracy of the innovative instrumented projectile in measuring impactinduced force history. 3

I. SUMMARY OF PROJECT INVESTIGATION 1. Background and Motivation Drop-weight impact testers (DWIT) have been commonly used in laboratories for measuring force histories involved in low-velocity impact [1]. The measurements from DWIT are usually based on a metallic cylinder mounted with a pair of electrical resistance strain gauges. When dropped onto a target specimen, the metallic cylinder experiences a length change which, in turn, causes a change of the electrical resistance in the strain gauges. The change of resistance can be converted into the impact force involved in the impact event with an external data acquisition and signal processing (DAQ&SP) system. The success of DWIT in measuring low-velocity impact force raises interest in using a similar concept to measure the forces involved in highervelocity impacts, such as free projectile impacts and crash impacts. By taking advantage of the capability and efficiency of electrical resistance strain gauges in recording time-dependent strain histories, this study looks into building, testing and validating a resistance strain gauge based instrumented projectile (IP) for measuring impact-induced force history due to higher-velocity free impact. 2. Development of a Strain Gauge Based IP When materials and structures are subjected to impact loading, complex strain wave propagation, rebounding and overlapping can take place in them, resulting in disfigurement of the initial impact-induced strain wave. Subsequent vibration and damage process can take place in the materials and structures involved in the impact event and further complicate the measurement of the impact-induced strain wave history and subsequent force history. Since the initial impact-induced strain wave bears the critical element to understanding the primary behavior of the materials and structures being impacted, it is of great interest to isolate the initial impact-induced strain wave from the complex wave combining impacting, reflecting and rebounding elements. This study is focused on identifying the initial wave segment measured by the electrical resistance strain gauges. 3. Instrumented Projectile In this study, an aluminum projectile [1] as shown in Fig. 1 was prepared. It consisted of a thick cylindrical neck and a hollow cylindrical body. They were joined by a special cylindrical 4

shoulder. The nose had a length of 50mm and diameter of 7mm. The body was 85mm long with an external diameter around 13.5mm and a thickness of 1mm. The external surface of the cylindrical neck of the projectile was equipped with a pair of electrical resistance strain gauges on the opposite sides while the cylindrical body housed a microprocessor and a battery, as shown in Fig. 1, to record the strain history during impact. Hence, the stand-alone IP is able to record the strain history which can later be downloaded onto an outside computer for calculation of forces involved in impact events. Fig. 2 shows an experimental setup. It includes a gas gun barrel, a guiding tube, triggering elements and a long target bar. The measurement from the long target bar can be compared with that from the IP. If they match well, the ability of the IP to eliminate the wave rebounding can be confirmed. 3.1 IP Without a Specially Designed Shoulder To begin with, an IP consisted of a tube with a regular cylindrical shoulder, i.e. without a specially designed shoulder, was prepared. A long solid bar of 3.5m, similar to those commonly used in split Hopkinson s pressure bar (SHPB), was also installed with a pair of strain gauges and used as a target bar. Fig. 3 shows the strain history recorded by both the long target bar and the IP without a specially designed shoulder when the latter was fired from the gas gun and impacted the long target bar. The two measurements do not seem to have any correlation due to the multiple rebounding waves overlapped in the short IP. It should be pointed out that the wave recorded in the long target bar only included the initial impact-induced strain history. That is, there was no rebounding wave overlapped with the initial strain wave due to the length of the bar. However, the IP without a special geometry recorded the overlapped wave combining the initial impact-induced wave and subsequent reflected ones. 3.2 IP With a Specially Designed Shoulder Since the initial impact-induced force history bears the most fundamental characteristics of an impact event, there is a need to isolate the initial impact-induced strain history from overlapping with the subsequent strain waves reflected from the ends of the short IP. In order to achieve such a goal, this study takes an optimization approach to achieve a special geometry for the shoulder of the IP via multiple trial-and-errors. Fig. 1 shows an IP with an identical thickwalled cylindrical neck and an identical hollow cylindrical body as the IP without a special 5

geometry. However, its specially designed shoulder is capable of eliminating the majority of rebounding waves. Similar microprocessor and battery used in the IP without a specially designed shoulder was also installed in it. The IP with the specially designed shoulder was then used for impacting the same long bar used for the IP without a specially designed shoulder. Fig. 4 shows that the force histories from the IP with the specially designed shoulder and the long bar are almost identical to each other. The capability of the special geometry in maintaining the initial impact-induced wave is thus confirmed. In other words, the IP with the specially designed shoulder was able to isolate the initial strain history and to prevent it from being distorted by the rebounding strain waves. This result demonstrates that the IP with a specially designed shoulder is equivalent to a long bar. 4. Validation of Force Measurement To justify the force history measured by the innovative instrumented projectile, IP, a ballistic impact of the IP into a thin aluminum specimen was performed. The experiment included an IP impacting on one side of the specimen and measurement of the out-of-plane deformation on the other side with an optical method so-called fringe projection (FP) [2,3]. Fig. 5 shows the top view of the experimental setup. The non-impacted surface of the specimen was painted white to increase optical contrast for dynamic measurements. A grating (lines with constant distance) was installed in front of the light source and shone onto the back surface of the painted aluminum specimen. When the projectile impacted the specimen, it caused deformation to the specimen, resulting in the distortion of the projected straight lines. Contours of the out-of-plane deformation of the specimen could then be identified from the comparison between the projected straight lines and the distorted lines. By taking the out-plane deformation twice, an acceleration history, and subsequently the force history of the projectile, can be obtained. Fig. 6 show the comparison from the IP and the FP. They seem to agree quite well. 5. Conclusions When a projectile hits a target, impact-induced waves can form in both the projectile and the target. If the projectile is short, there will be significant wave rebounding between the ends of the projectile, resulting in a very complex wave overlapping. This causes the initial characteristics of 6

the impact-induced wave, which is critically important to understanding the impact-induced event in the structure, to be obscured. With a careful design, it is possible to isolate the initial impact-induced wave in the projectile from the remaining rebounding waves. In this study, a projectile with a specially designed geometry is developed and shown to significantly reduce the reflection of the wave propagation from the end of the projectile. Hence, it can be used for measuring the initial wave propagation generated at the onset of impact. References 1. Li, G. and Liu, D., Instrumented Free Projectile for Impact Testing, Proceedings of the XI International Conference and Exposition on Experimental and Applied Mechanics, Orlando, FL, June 2-5, 2008. 2. Gulker, B.G., Lureau, R. and Liu, D., Investigation of Impact Response of Composites with Projection Moiré Enhancement, Experimental Mechanics, 54(1), 35-43, 2014. 3. Schleh, D. and Liu, D., Fringe Projection in Horizontal Impact, Structural Durability and Health Monitoring, 9(3), 201-216, 2013. 7

Figure 1 An instrumented projectile. target bar impact point IR emitter & collector for oscilloscope triggering guide tube triggering tube gas gun barrel Fig. 2 Assembly of gas gun, triggering tube and target bar. 8

10 Force (kn) 5 0 0.0 0.5 1.0 1.5 2.0 2.5 Bar IP -5 Time (ms) Fig. 3 Comparison of force histories between the results from an IP without a special geometry and that from a long bar. 6 Force (kn) 4 2 0 0.00 0.50 1.00 1.50 2.00-2 Time (ms) Bar IP Fig. 4 Comparison of force histories between the results from an IP with a special geometry and that from a long bar. 9

Fig. 5 Experimental setup of fringe projection technique. Force (kn) 3 2.5 2 1.5 1 0.5 FP and IP Force Histories 0 0 0.2 0.4 0.6 0.8 1 1.2 Time (ms) FP 4th Order Curve Fit Force IP Filtered Force Fig. 6 Comparison between IP (instrumented projectile) and FP (fringe projection). 10

II. FINAL DESIGN AND DELIVERABLES Up to date, eight PIs have been delivered to ARL. After discussions, the design of four additional IPs with a tube diameter of 0.5 has been finalized and is currently under manufacturing. They will be delivered within four months, no later than late February of 2016. The overall design of the 0.5 IP is given below. Liuman Technologies will also provide at least one-year service for the final deliverables. 11

III. FUTURE DEVELOPMENT 1. The Problem measuring dynamic forces Measuring force and deformation histories are perhaps the most fundamental activities in mechanical investigations. Many deformation-based measuring techniques, such as electrical resistance strain gages (RSG) and optical methods, like laser interferometry and digital image correlation, have been successfully used for measuring deformations caused by static loading. For measuring deformations associated with dynamic loading, optical methods in conjunction with a high-speed movie camera have been commonly used due to their capability of whole-field measurements. Once the deformation fields are obtained, they may be converted into associated force fields, with the use of Hooke s law and corresponding material properties, to complete the mechanical investigations. Though being a point-by-point technique, RSG is much simpler in practice and more tolerable to environmental effects than optical methods. However, the strain wave due to dynamic loading can bounce back and forth between the ends of the structure on which the RSG is mounted. The overlap of the incoming and rebounding strain waves can then result in a completely distorted strain history. Consequently, the initial strain wave which bears the most important information of the dynamic loading can be lost. Though highly demanded in many engineering applications, no reliable technique for measuring dynamic forces is available until the introduction of an instrumented projectile by the author and his associate very recently. It is the objective of this proposed study to extend the recently developed technique to several other testing techniques to improve their accuracy and applications. 2. The Innovation an instrumented projectile (IP) When the nose of a cylindrical rod impacts on a target, a strain wave will be generated in the rod. Subsequently, the strain wave will propagate to the tail of the rod. Once reaching there, the wave will rebound from it and propagate back to the nose. The wave will continue to bounce back and forth between the two ends of the rod until it is completely attenuated. If the rod is 12

short, numerous wave propagations will overlap one another and result in a completely distorted wave form which is very different from the initially impact-induced one. With a special geometry design at the tail of the rod, it is possible to stop the strain wave at the tail without rebounding back to the nose. The initial strain wave caused by impact can then be isolated. The instrumented projectile (IP), based on geometrical manipulation at the tail, has been developed recently by the author and his associate in this SBIR project. The short IP (about 6-inch long) consists of a solid neck and a hollow body. A pair of RSG s is mounted on the opposite sides of the neck. A data acquisition board and a battery have also been installed inside the IP s body. The IP has been used for impact testing. Experimental results have shown the similarity of the wave propagation from the short IP and that from a 6-foot long cylindrical bar (which can separate the incoming wave from the rebounding one) when the IP and the long bar collide with each other. The wave propagation in a short rod without a special geometry is completely different from that in a long cylindrical bar when the two collide with each other. However, the wave propagation in a short rod with a special geometry is almost identical to that in a long cylindrical bar when the two collide with each other. The ability to eliminate the wave rebounding based on the special geometry design is thus demonstrated. Accordingly, the initial strain wave which bears the most important information of impact-induce force history can be clearly identified by the short instrumented projectile. 3. The Research improving experimental accuracy Based on the innovation of the wave elimination technique and the success of the IP design, more applications based on the IP to improve engineering dynamic measurements will be explored. Some tasks are listed below and will be investigated in this research program. (1) Drop-weight impact tester (DWIT) has been commonly used in both academia and industries due to its simplicity. However, the issue concerning wave propagation has been completely neglected in the design of the DWIT currently available in the market. It is the first task of this research study to comprehensively evaluate the effect of wave propagation involved 13

in the DWIT by comparing the result from the DWIT with that from the IP (which preserves the initial impact-induced wave). The goal of this study is to demonstrate the importance of eliminating the wave propagation in correctly measuring dynamic forces. (2) Split Hopkinson s pressure bar (SHPB) SHPB has been commonly used for characterizing dynamic constitutive relations of materials at high strain rates which are required for modeling the performance of materials and structures subjected to high-velocity loading. SHPB is usually organized by aligning two long bars (such as 1.5m each and mounted with strain gages) with a gas gun while the specimen to be tested is situated between the two long bars. When an impactor is shot out of the gas gun and collides onto the first long bar, a strain wave will be generated. The wave will then propagate through the bar into the specimen before entering the second bar. Based on the wave propagations in the two long bars measured by electrical resistance strain gages mounted on them, the constitutive relation of the specimen material can be established for the high strain rate performed. Since the IP can record the incoming strain wave without the rebounding one, it can be used to replace the second long bar. The length of the SHPB can then be reduced to about half. The shorter SHPB should prove to be more convenient for laboratory operations. (3) Ballistic impact and Taylor s impact (a) The development of instrumented projectile (IP) is aimed at measuring the force history of ballistic impact (BI). A very rigid material is hence required for making the core of the IP. (b) If, however, the property of a material at high impact velocity is of concern, Taylor s impact (TI) should be performed. In a conventional TI test, the concerned material is made into a projectile and shot onto a rigid target. Based on the deformation of the projectile and phenomenological equations, the dynamic properties of the material used for the projectile can be estimated with phenomenological equations. For a more accurate TI test, the concerned material may be made into the core of an instrumented projectile for Taylor s impact, hence instrumented Taylor s impact (ITI). Besides advancing the dynamic testing techniques given above, this research program will also explore other applications of the IP. 14

4. The Dissemination commercializing the advanced techniques In addition to advancing the technologies, disseminating them is another goal of the proposed project. To disseminate the research results, besides publishing journal articles, it is to commercialize the following innovative instrumented testing facilities for high strain rate and high velocity tests. (1) A more accurate drop-weight impact tester The ones currently used in engineering and research institutes neglect the wave propagation involved in the DWIT testing. (2) An instrumented ballistic impactor The ones currently used are non-instrumented and are not useful for scientific investigations. (3) A shorter Hopkinson s pressure bar The ones currently used are twice longer than the proposed one and are not convenient for labs with restricted space. (4) An instrumented Taylor s impactor The ones currently used are based on post-test measurements instead of in situ measurements. It is the ultimate goal of this study is to eventually achieve significant advancement in dynamic material characterizations and structure tests in both academia and industries. 15

Horizontal Impact Testing Method Instrumented Projectile (IP) Comparison Testing Gas Gun For IP validation testing the test window needs to be identified for the IP and the constituent validation device, which in this case is a bar with a strain gage attached. The term test window refers to the duration between two points in time during which the event of interest occurs. The test window must be determined relative to the trigger signals of different data recording devices. A dummy IP should be used to find the test window for the bar. A triggering signal for the data acquisition device and the oscilloscope needs to be established. In the case of the bars, IR emitter and collectors are used. As the IP passes between and prevents the IR light from the emitter from being seen by the collector, a change in voltage in an associated circuit occurs. This voltage change is monitored and is used as the trigger for the point in time onwards from which data should be collected. The IR sensors are placed at a point where the IP passes them prior to impact with the end of the bar that the strain gages are attached to. Bar Impact Point- End of Bar IR Emitters and Collectors for Oscilloscope Triggering IP Guide Tube IP Triggering Tube Gas Gun Barrel Figure 1. Gas Gun Bar interface Multiple tests with the dummy IP striking the end of the bar should be run to identify the appropriate test settings on the oscilloscope that will maximize the number of data points of the data collected that represent the event in question. Specifically, this is done by locating the test window and then reducing the Time/Div setting on the oscilloscope so that it is small enough to 16

allow for sufficient data resolution, but large enough to allow for the entire desired event to be captured. It is best to begin with sampling a large time window and then reducing it once the event of interest has been identified. For the testing conducted, a total of 200,000 data points were taken per test with the sampling rate being 10M/s. The IP was loaded to a depth of 5 as measured from the end of the triggering tube. Gas pressure to drive the projectile was held constant at 24psi. All three passageways around the gas gun holding chamber were open during testing. The depth of loading and the pressure used to drive the IP have an effect on the velocity of the IP and the subsequent force at impact. When need be, these parameters should be changed conservatively as to minimize the potential for damage to the IP and the bar as well. Once data is captured using the data acquisition card in conjunction with the National Instruments Scope program, the data needs to be saved and processed. The file should be saved as a LVM file type and only the channel producing data from the bar needs to be saved. The time data should be saved as well or at least the sampling frequency so time data can be generated later. Gas Gun Operating Procedure 1) Load projectile through the opening of the gas gun. The depth of loading should be a controllable variable. Using a steel cable will allow for the loading depth to be controlled. 2) Make sure all values for the three avenues that gas can travel from the gas gun holding chamber are open (handle is in line with pipe or tube when open). 3) Partially open the value on the compressed air tank. 4) Use the Intake button on the control box to allow air into the holding chamber of the gas gun. 5) Use the Vent button to reduce pressure in the holding chamber to the desired level. 6) Set the NI Scope program ready to trigger. Using the Intake and Vent buttons on the control box will causes the Scope to trigger and record data due to the spike in voltage of the triggering channel caused by the cross talking of electrical signals. 7) Gas can be released from the holding chambers into the barrel of the gas gun to drive the projectile by pressing the Fire button. 17

Bar Data Processing Data from the bar represents a signal from a Wheatstone bridge with two active arms. Each active arm represents a strain gage adhered to the exterior of the bar. Voltage is supplied to the Wheatstone bridge circuit to facilitate measurement. The signal from the strain gage is also amplified by the gain to allow for the data acquisition card to more readily see the data. The gain in this case was 100. The recorded voltage and time data from testing need to be converted to a force. 1) Locate the data in the time span of interest. It is often easiest to locate the data by plotting the raw data allowing the times of interest to be determined and noted. 2) Isolate the data of interest. It is advantageous when working with large data sets to only perform operations on only the necessary data 3) Shift or zero the voltage data. This can be done by taking the average of the voltage values for a region of the signal prior to the impact event and subtracting the average value when no loading is present from every data point recorded. Given the sample size, a few thousand samples were averaged together to arrive at the zeroing factor. 4) Once the voltage values are zeroed, each value should be divided by the gain (in this case 100). 5) The strain calculation. The sensitivity,, of the Wheatstone bridge circuit should be calculated according to (eq. 1) where is the excitation voltage (2.01volts), is the gage factor associated with the strain gages (2.150), is the number of active arms (2) and (1) is the ratio of resistances R2 to R1. = (Eq. 1) 6) Multiply the zeroed voltage values by the sensitivity according to (Eq. 2) to arrive at the strain. = (Eq. 2) 7) Calculate the force according to the following equation, Eq. 3. = (Eq. 3) In Eq.3, is the cross sectional area of the bar, is Young s modulus of the bar material and is the measured strain. For the bar positioned horizontally and impacted with the gas gun, the 18

material is stainless steel 17-440C with young s modulus being 210 GPa, and the bar having a 0.25 diameter. 8) For this application units of ms for time and kn for force work well for constructing plots of the data. 9) Compare the peak force values with the load limit of the IP in use, 8kN for aluminum IP s and 60kN for steel IP s. If the measured load falls outside of this range reduce the speed of the IP until acceptable force values are achieved. 10) Replacing the Dummy IP with a Functional IP 1) See Manual-v3-052714 for directions and specifics on using the IP and accessing the data it records. 2) Once the force levels associated with impact have been determined to be acceptable replace the dummy IP with a functional IP 3) Follow the steps under Gas Gun Operating Procedure to perform testing with the addition that he light source connected to the trigger tube must be activated prior to testing but after loading the IP into the gas gun. After a test is run, it is good practice to turn off the IP trigger tube. 4) Using a functional IP with a different mass than the dummy IP may cause the testing window to shift. The testing window associated with the IPS s will need to be identified with each specific test condition (mass of IP, loading depth in the barrel and gas pressure). To do this without knowing the velocity of the IP it is recommended to keep the flying time short or zero and the sampling rate lower so as to prolong the time span over which data is collected. The flying time is the delay between when the IP sensors are triggered and data begins to record. It should be noted that sampling rates that are too low may not allow for the testing window to be identified. The flying time and testing window should be increased to the point where the sampling frequency of the data is as high as possible thus allowing for good data resolution. For the aluminum IP s flying time was around 25ms with the sampling frequency at its maximum setting of 1MHz with the associated sampling occurring over roughly 3.7ms. For the steel IP, flying time was around 32ms with the IP sampling at it s maximum capability of 1MHz. 19

5) Following a test the IP data should be extracted, saved and examined to ensure the test event was captured. IP Data Processing 1) The IP data should be placed in the same workspace as the bar data as it will aide in plotting both data sets together for comparative purposes. 2) The IP force data will need to be shifted or zeroed so that only the change in force is conveyed. This can be done by taking the average of several hundred data points prior to the event of interest and then subtracting this average from every available force data point. 3) IP force and time data points should be converted to match the units of measure of the bar data (kn and ms). Data Comparison IP and Bar Plot the IP data and the bar data on a single graph for a given test and shift the data sets along the time scales to get the onset of the initial loading peaks to coincide with one another. Note: The data sets from the Bar often contain several data points prior to the onset of loading at very high force values relative to the rest of the data sets. These points are presumed to be erroneous and thus are not considered to be a realistic representation of the loading occurring. Additionally, the signal of the Bar data from the gas gun loading has demonstrated shifting around the time of the impact event that does not coincide with the physical phenomena occurring. The Data from the bar should be considered carefully and with skepticism. Figure 2 provides an example of compared good data and Figure 3, a plot with unrealistic behavior in the bar data. The suspect portion in Figure 3 is the sudden and large loading of 15kN applied at the onset of loading. Despite testing conditions remaining constant, a large change in the magnitude of the load occurred between tests. Also, the loading begins as an impulse and that is uncharacteristic of the testing conditions. 20

Force (kn) 8 6 4 2 0 0.0 0.5 1.0 1.5-2 -4 Bar IP -6 Time (ms) Figure 2. IP and gas gun bar force data. 25 20 15 Force (kn) 10 5 0 0.0 0.5 1.0 1.5 2.0-5 -10 Bar IP -15 Time (ms) Figure 3. IP and gas gun bar force data with suspect bar data. 21

Vertical Impact Testing Method Testing involving the vertical drop bar is a little more simplistic than the horizontal gas gun bar. Instead of using compressed gas to propel a projectile into the end of a bar, gravity is used. The same basic procedure and much of the details as well required to conduct gas gun testing are the same with the vertical drop bar. The gas gun testing procedure should be followed except as noted otherwise.instead of loading a projectile into the barrel, the projectile should be dropped from a consistent height through tube properly aligned to direct impact onto a rod. Guide Tube IP Triggering Tube IR Emitters and Collectors for Oscilloscope Triggering Impact Point- End of Bar Figure 3. Vertical Drop Bar Interface Different types of rods were impacted in the vertical drop and horizontal gas gun tests. The rod being impacted on the end is now 0.375 in diameter and is Aluminum with a Young s modulus of 73.1 GPa. The flying time of the IP s was around 26ms and the sampling rate for the IP was 1 MHz. For the vertical drop bar data, a wheat stone bridge with two active arms was used with all 22

resistors having resistances of 350Ω. The voltage supplied to the Wheatstone bridge was 2.01V. The gain was 100. The gage factor for the strain gages was 2.150. 4 3 2 Force (kn) 1 0 0.0 0.5 1.0 1.5 2.0-1 Bar IP -2-3 -4 Time (ms) Figure 4. Example of data comparison between IP and Bar data from a vertical drop bar test. 23