National Academies Press: OpenBook

Developing Improved Civil Aircraft Arresting Systems (2009)

Chapter: Chapter 10 - Engineered Aggregate Arrestor Concept

« Previous: Chapter 9 - Glass Foam Arrestor Concept
Page 87
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 87
Page 88
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 88
Page 89
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 89
Page 90
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 90
Page 91
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 91
Page 92
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 92
Page 93
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 93
Page 94
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 94
Page 95
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 95
Page 96
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 96
Page 97
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 97
Page 98
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 98
Page 99
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 99
Page 100
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 100
Page 101
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 101
Page 102
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 102
Page 103
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 103
Page 104
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 104
Page 105
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 105
Page 106
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 106
Page 107
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 107
Page 108
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 108
Page 109
Suggested Citation:"Chapter 10 - Engineered Aggregate Arrestor Concept." National Academies of Sciences, Engineering, and Medicine. 2009. Developing Improved Civil Aircraft Arresting Systems. Washington, DC: The National Academies Press. doi: 10.17226/14340.
×
Page 109

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

87 10.1. Concept Description 10.1.1. System Overview An engineered aggregate arrestor concept has been pro- posed. Its primary material is a spherical engineered aggre- gate that has excellent flow properties and resists settling and compaction that are more typical for angular gravels (Fig- ure 10-1). This material would reside in a shallow bed and be covered with a reinforced turf layer. However, the engineered aggregate may also be used without a turf layer, which has been done at four airports in the UK. Other top layer materi- als are possible, such as a thin asphalt skim coat. Figure 10-2 illustrates the two major variants of the engineered aggregate arrestor concept. The difference between the two concepts is the existence of a confining top layer, which can serve several purposes: 1. Prevent aggregate dispersion due to jet blast; 2. Mitigate aggregate spraying during overrun by aircraft tire, thus limiting engine ingestion hazard; 3. Regulate water drainage and potential ice crust formation in winter; and 4. Act as a structural component to prevent lightweight land vehicles from penetrating the arrestor bed. Because the engineered aggregate materials are not greatly affected by water, the bed can be designed to handle precipita- tion in two different ways (Figure 10-3). 1. Drainage Approach. This approach would allow water drainage downward through the bed. In this case, the bed would be designed to prevent standing water within it using normal civil engineering design practices. 2. Waterproof Approach. In the second approach, a compos- ite top layer could be employed to prevent water drainage through the bed. It would likely be composed of a channel- ized or cuspated plastic layer, overlaid with a geo-textile fabric layer, and finally topped with reinforced turf. Pre- cipitation would not seep through the aggregate in this case, but would run off to the perimeter of the arrestor bed. Hence, the bed would be kept dry and the material response during an overrun would likely prove more consistent. 10.1.2. Performance Considerations The dynamic performance issues for the engineered aggre- gate concept are different from those of the crushable mate- rial candidates. During the assessment process, the following aspects were considered: • Dynamic behavior and energy absorption, • The degree of non-linearity of the landing gear loading and danger of failing the nose gear, • Performance in a short-landing situation that involves an aircraft touchdown inside the arrestor bed, • Applicability to arresting a wide range of aircraft types, • Vulnerability to ice crust formation in severe winter envi- ronments, and • Effect of a cover layer on the performance of the aggregate system. 10.2. Modeling and Testing Approach As with the crushable systems, the goal for the performance evaluation was to perform testing that would allow cali- bration of high-accuracy computer models of the engi- neered aggregate concept. Determining the necessary physical tests was driven by the requirements of the modeling soft- ware. Therefore, a modeling method capable of high- fidelity aggregate simulations was determined first, and the required tests for material characterization followed thereafter. C H A P T E R 1 0 Engineered Aggregate Arrestor Concept

10.2.1. Discrete Element Modeling (DEM) Method Modeling aggregates presents a number of complexities when compared with crushable materials because they have both solid-like and fluid-like regions of behavior. Bell et al. describe the complexities from a modeling and simulation standpoint: the fluid-like behaviors cannot be adequately simulated with solid material models (Lagrangian), nor can the solid-like behaviors be adequately simulated with a fluid material model (Eulerian) (39). LS-DYNA, which can simu- late both Lagrangian and Eulerian behaviors, is still not ade- quate for modeling this particular material. The current state of the art for modeling aggregate behav- ior involves the use of DEM codes, which model the aggregate on a particle-by-particle basis. The EDEM software package, produced by DEM Solutions, was selected from among the commercially available DEM codes. Using EDEM, the bed of aggregate was modeled using individual aggregate pieces. The particles move and interact with one another in fundamen- tally the same way that actual aggregate does, with friction between the particles, compaction, fluid-like spraying behav- ior, and momentum transfer. 10.2.2. Required Parameters for Material Models EDEM defines aggregate pieces as rigid spheres, or lumps of rigid spheres grouped together to form irregular pieces. The rigid sphere assumption was valid here since the spheres behave as fairly incompressible particles with respect to the aircraft tire. Particle size, size variation, density, and so on were all defined as expected. The main effort in characterizing the engineered aggregate behavior was in determining the particle interactions with each other and with solid surfaces (i.e., the aircraft tire, bed walls, etc.). This was a fundamental contrast to the crushable material modeling in LS-DYNA, where the main effort was expended in calibrating the material models. Four interaction parameters required definition in EDEM, which are briefly described here: 88 Figure 10-1. Engineered aggregate. Aggregate Bed Aggregate Bed Cover Layer of Engineered Turf Aggregate Arrestor Concept 1: Open Bed Aggregate Arrestor Concept 2: Covered Bed Arrestor Basin Figure 10-2. Engineered aggregate arrestor concepts. Drainage Approach Water-Proof Approach Precipitation Drains Through Bed Layers of Plastic Keep Aggregate Dry Figure 10-3. Aggregate bed methods for handling precipitation and drainage.

1. Coefficients of Static Friction – a standard friction def- inition based on the contact force between particle/ particle and particle/surface. 2. Coefficients of Rolling Friction – defines the resistance of the particles to tumbling/rolling past one another or over a surface, which could be due to shape irregularities (jagged- ness) or other material properties. 3. Coefficients of Restitution – defines the degree of rebound that particles have following a collision. 4. Particle Bond Strengths – defines a bonding force between particles simulating glue or adhesion-type forces. Here this parameter was used to define the bonding of particles together for simulating a turf layer. These parameters drove the selection of suitable physical tests of both the engineered aggregate and reinforced turf materials. Naturally, this collection of tests differed substantially from those of the crushable material candidates. 10.3. Testing Effort 10.3.1. Density and Dimension Measurements The density and dimensions of the engineered aggregate were measured for several hundred randomly selected aggre- gate particles. The particle size was found to follow a bounded normal distribution, with a diameter mean and standard deviation of 0.348 and 0.047 in., respectively. The particles were essentially round, with an average max/min diameter ratio of 1.24. The density of the actual material, on a per-particle basis, was 94.9 pcf, while the effective density of the continuum aggregate was 52.1 pcf. This leads to nominally 45% void space in the aggregate bed. The high void percentage and sphericity of the particles lead to good drainage through the bed and effectively limit the settling of the material over time. The density is about three times greater than that of typical EMAS cellular cement (∼18 pcf). The higher density leads to an increased sensitivity to speed in an arresting application; higher speeds will lead to higher loads on the landing gear. 10.3.2. Angle of Repose Angle of repose measurements (Figure 10-4) were performed by overturning and raising a 5-gal bucket of the engineered aggregate material, allowing the material to spill and settle into a pile. The sides of the resulting heap settled to the aggregate’s natural angle of repose. The test was repeated several times for the dry material, and again for wet material, which had been fully soaked and drained. The dry and wet angle of repose measurements averaged to 18.7 and 19.4 degrees, respectively. The higher angle of the wet samples results from increased wet particle adhesion. This physical test eventually helped to calibrate the coeffi- cient of rolling friction in the DEM model. 10.3.3. Hydrostatic Triaxial Tests Standard hydrostatic triaxial compression tests (Figure 10-5) were performed on the engineered aggregate at various con- fining pressures in accordance with ASTM D2850. Cylindrical samples of the aggregate were enclosed in a flexible membrane and subjected to various hydrostatic confinement pressures. The specimens were then compressed uniaxially until the spec- imen bulged and deformed out of its original cylindrical shape. The aggregate behaved much like a soil sample, which was in accordance with expectations. Values obtained from these tests included the elastic modulus of the material, the inter- nal angle of friction, and the effective particle-to-particle cohesion. These tests further illustrated that the behavior of the material was consistent with essentially non-deforming particles and standard aggregate/soil behavior, thereby con- firming that DEM modeling of the material was appropriate. 10.3.4. Pendulum Tests 10.3.4.1. Test Apparatus In addition to the small-scale laboratory tests, a larger-scale one-wheel bogy test was conducted using a large pendulum test apparatus. The pendulum test apparatus featured a heavy 4,400-lb mass that hung from an overhead support frame, giving it a swing arc of 24.5 ft. The mass was hoisted to the desired height and then released; the speed of the mass was controlled by the release height. The pendulum mass was fitted with a strut and wheel assembly, which Figure 10-6 illustrates in an exploded view. The strut was instrumented with three load cells that meas- ured loads at the connections. These three connection loads were resolved into orthogonal vertical and drag loads on the strut and wheel assembly. To reduce the number of variables in the design, a rigid aluminum wheel form was used rather than a pneumatic tire. 89 Figure 10-4. Angle of repose test for engineered aggregate material.

After cutting the initial rut, the pendulum continued to swing back and forth until eventually coming to rest. Load measurements were only considered for the initial pass through the material. The final resting position and resulting rut are illustrated in Figure 10-9. Figure 10-10 shows the average drag and vertical forces during the pendulum swing for the different permutations of the test matrix. The error markers atop the bars indicate a 90% confidence interval, as an indication of the spread in the test results for the three repetitions of each case. The drag force is defined as the rearward load on the strut, which would decelerate the aircraft (Figure 7.9). The vertical force is defined as the upward load, which would support the air- craft from sinking further into the aggregate. Overall, both the speed and the moisture conditions affect the outcome, but speed has the greatest effect. Reducing the speed from 17.2 to 8.5 mph created a 65% reduction in the drag force. Wetting the aggregate increased the drag force by 26%. These findings were consistent with initial observations and assumptions regarding the aggregate as an arrestor medium. The imparted landing gear load is highly dependent on the speed of the aircraft overrun and somewhat dependent on the moisture content of the aggregate. The results of these tests were subsequently used to help calibrate the coefficients of static friction and restitution for the aggregate material. 90 Figure 10-5. Hydrostatic test specimen for engineered aggregate pre-test (left) and post-test (right), 10 psi confining pressure. The diameter and width were 14.8 and 5.5 in., respectively. These proportions were based on a nominally one-third scale B737-800 main-gear tire, which has a diameter and width of 44.5 and 16.5 in., respectively. Although this wheel size was smaller than full-scale, it was substantially larger than the typ- ical particle size; this ensured the continuum-type behavior of the aggregate that was necessary for characterization. Below the pendulum assembly, a bed of aggregate was con- structed with a length of 22 ft, a width of 8 ft, and a depth of 22.5 in. Figure 10-7 shows the overall pendulum apparatus with the aggregate bed beneath it. 10.3.4.2. Test Matrix For the test series, the pendulum was set to swing such that the wheel penetrated to a depth of one-half diameter, or 7.4 in. Two swing heights were used to produce speeds of 17.2 and 8.5 mph. For the high speed tests, both wet and dry beds were used. The wet bed was well soaked with water and allowed to freely drain prior to every test. Three repetitions were conducted for each permutation tested, as shown in the test matrix of Table 10-1. 10.3.4.3. Test Results Figure 10-8 gives an action sequence of photos for the pen- dulum strut cutting a rut through the material.

10.3.5. Tensile Turf Tests 10.3.5.1. Overview and Assumptions The reinforced turf material used a sandy soil mixed with small swatches of reinforcing material. Grass seed was planted in this soil base and grown to maturity. The resulting turf layer had changing properties through its thickness, with grass on top and sandy soil on the bottom. Characterizing the reinforced turf required non-standard testing procedures. Several different test methodologies were considered, including punch tests, tensile tests, and tearing tests. At the small scale, the soil and reinforcement swatch mixture presented problems of heterogeneous properties, which eliminated some of the punch test concepts from con- sideration. To ensure a generally homogeneous behavior, the specimen size needed to be fairly large. For tire overruns, it was assumed that the material would essentially act as a confining membrane layer atop the aggre- gate. It was further assumed that the through-thickness homogeneities could be neglected in the DEM models, where the turf layer would be represented using a homogeneous layer with equivalent membrane properties. This layer of par- ticles would use defined particle bonds to create a membrane- type behavior. With that end in mind, a simple tensile test was selected as the best all-around means of determining membrane per- formance. It produced results that could be transferred to bond strength values in the DEM modeling code. While the tests sought to characterize the material prop- erties for the turf layer, this was undertaken with fairly modest expectations for exactness. In actual application, the use of a turf layer presents several factors that could be difficult to control. The strength of the layer would be dependent on: • Type of soil base used (sandy, clayey, etc.), • Amount of reinforcement per unit area, • Type of grass used (regionally dependent), • Age/maturity of the grass, and • Seasonal rainfall levels. Therefore, the tests sought only to establish a nominal strength level for the turf layer, which in practice could either be stronger or weaker for different arrestor beds, at different times of the year, and at different ages. 10.3.5.2. Tests Conducted The turf specimens began as 4-ft square samples, which were 4.6 in. thick with a nominal areal density of 30 psf. From these squares, test specimens were cut out in dog-bone shapes, which narrowed to an 8-in. wide neck that was 8 in. long (Figure 10-11). The total specimen length was 14.25 in. These specimens were clamped in a special jig and attached to a ten- sile testing device. The specimens were then pulled apart axi- ally, and the load history was recorded. Three specimens were tested overall, with a typical failure strength of 120 lbf at elongations of 4 in. 10.3.6. Environmental Tests Environmental tests were considered for the engineered aggregate and turf materials. However, after consideration of the nature of both materials, as well as the body of existing research information, it was determined that such testing would not be included in the research effort. A simple set of environmental tests was not useful, and a comprehensive set of tests was not affordable within the current effort. 91 Figure 10-6. Pendulum test device with one-wheel bogy.

Neither material degenerates under environmental condi- tions in a manner analogous to crushable materials, such as the currently used cellular cement. Both the turf and aggre- gate are durable to freeze–thaw, rain, and heat. Obviously, turf can degrade over time during periods of drought, which would lead to impacts on strength as discussed in Section 10.3.5. However, characterizing such impacts would require a very substantial body of tests using multiple hydration con- ditions, soil types, and grass types. Perhaps the most relevant issue is that of ice crust forma- tion on an arrestor bed during winter conditions. Rainfall dur- ing freezing conditions could lead to water penetration and freezing in the upper layer of the aggregate and turf. However, freezing at deeper levels may be self-limiting due to insulating properties of the aggregate bed. Prior research by Rogers dis- cusses cold-weather testing information that demonstrates the potential for aggregate arresting beds to freeze over and become ineffective in harsh winter environments (13). In the event of such freezing, we can deduce several areas of performance impact. First, the particles would effectively be adhered to one another, which would create a solid block of material that an overrunning tire would likely break apart. After breakage of the ice bonds, however, the void spaces between particles would still be filled with ice and water, increasing the rolling friction of the particles and decreasing their effective flow around the tire. This would likely result in 92 Figure 10-7. Overview pictures of engineered aggregate pendulum test setup. Number of Tests Conducted High Speed Low Speed Dry Aggregate 3 3 3 Wet Aggregate - Table 10-1. Pendulum test matrix for engineered aggregate.

increased drag and/or vertical loads on the tire. Overall pen- etration of the tire could prove similar to that in a normal bed, or it could be diminished. However, predictability of the total performance impact is difficult since the degree of such crusting would be variable. To assess such effects on performance, a larger test bed would have to be constructed and subjected to artificial rain and freeze–thaw conditions. Further, this would need to be followed for the two different drainage and waterproofing approaches to the arrestor bed. In the former, the ice crust could penetrate below the turf layer; in the latter, it should be confined to the turf alone. Larger testing as described above was beyond the scope of the current effort. However, it would be advisable if the 93 Figure 10-8. Pendulum wheel passing through engineered aggregate material (upper left is first frame, lower right is last frame). Figure 10-9. Post-test view of rut in engineered aggregate created by pendulum wheel. 0 200 400 600 800 1,000 1,200 Drag Avg Force (lbf) HS Dry HS Wet LS Dry Vertical Avg Force (lbf) Figure 10-10. Plot comparing strut forces for high-speed dry, high-speed wet, and low-speed dry conditions.

engineered aggregate concept is to be transitioned into a fieldable system. 10.4. Modeling Effort The modeling effort sought to replicate the important behav- iors of the engineered aggregate arrestor system concept. The aggregate and turf simulations were conducted using the EDEM software, which is a DEM code, as discussed in Section 10.2.1. 10.4.1. Particle Sizes and Shapes 10.4.1.1. Size Selection In the DEM method, the model particles are often larger than the actual real-world particles. Because DEM modeling is a numerical method, various levels of fidelity are possible. High fidelity simulations will use a large number of smaller particles, while low fidelity simulations use a reduced number of larger particles. If the particles chosen are too large, accu- racy suffers. Conversely, if the particles are too small, simula- tion times rapidly increase to impractical levels. The general goal for such models is to select a particle size that gives suffi- cient accuracy while keeping simulation times short. During the course of the model development and calibration, care was taken to select particle sizes that struck the right bal- ance of accuracy and time efficiency. For the aggregate models, particle diameters were as small as 0.348 in. (the actual aggre- gate size) and as large as 2 in. in diameter. The size required for good accuracy was highly dependent on the size of the tire or structure that was interacting with the material. Larger wheels passing through the material permitted larger particles to be used with negligible decrease in accuracy. 10.4.1.2. Size Distributions Section 10.3.1 discussed the particle size variation of the engineered aggregate, which followed a bounded normal dis- tribution. This normal distribution was important to repli- cate in EDEM because it affects the packing density and the void ratio of the particles. As such, all aggregate simulations used the same normal distribution regardless of size scaling. The turf material was represented using uniformly sized particles in EDEM. 10.4.1.3. Shapes The average maximum-to-minimum diameter ratio of the aggregate was 1.24, which is close to that of a perfect sphere (1.0). As such, all aggregate particles were modeled as simple spheres. The irregularities in shape were essentially accounted for through the coefficient of rolling friction for the aggre- gate, which is discussed further in Section 9.4.2.1. 10.4.2. Calibration to Physical Tests The aggregate and turf models were calibrated to match the physical tests as closely as possible. Properties of density, par- ticle size distribution, and so on were simply entered into the software. However, the particle interaction properties of the aggregate and particle bonding properties of the turf required iterative determination. This section briefly summarizes the outcome of the long calibration process. 94 Figure 10-11. Reinforced turf material (left) and dog-bone specimen in tensile test jig (right).

10.4.2.1. Angle of Repose The angle of repose test was replicated in EDEM using a bucket of identical size, filled with particles. The bucket was then removed, and the particles were allowed to fall and set- tle into a heap (Figure 10-12). Simulations were conducted for several scaled particle sizes ranging from the actual size of 0.348 in. up to 1 in. in diameter. Each model had nominally 35K particles and 43K particle-to-particle contacts. The coefficient of rolling friction for the aggregate was determined using this method, and it was found to be depen- dent on the particle diameter. Therefore, in all subsequent simulations, the coefficient of rolling friction was matched to the scaled particle size to ensure accurate results. The overall accuracy of the predicted angle of repose was within 1⁄2 degree (Figure 10-13). 10.4.2.2. Hydrostatic Triaxial Tests The hydrostatic triaxial tests were not replicated in EDEM due to software limitations. EDEM does not support deform- able structure modeling or pressure load applications. Both of these features would be required to replicate the elastic membrane that surrounded the triaxial test articles and the applied hydrostatic pressure. The hydrostatic triaxial tests nevertheless revealed that the aggregate performed normally and had no unusual character- istics that might make the use of EDEM inappropriate. The tests also provided data for the shear modulus of the aggre- gate material, which was a parameter required in EDEM. 10.4.2.3. Pendulum Tests An EDEM model was constructed to replicate the pendu- lum tests and was used to calibrate the remaining aggregate material parameters (Figure 10-14). The pendulum strut was simplified in the model to include only the features that inter- acted with the aggregate in the actual test. The aggregate bed was 16 ft in length, 5 ft in width, and 22.5 in. in depth. The model bed width was less than the actual bed width (8 ft) in order to reduce the particle count, but was still wide enough to prevent substantial boundary effects. The strut followed an arced path approximating that of the actual strut, with the same 1⁄2-diameter penetration depth into the arrestor bed. The wheel was set to a constant rotation rate approximating the observed rotation rate of the wheel on test video. Several particle sizes were attempted, and an error conver- gence study was undertaken. It was determined that a 1.0-in. particle diameter was required for the relatively small pendu- lum strut/wheel assembly. The particle size error was esti- mated to be less than 6%. The model had a total of 315K particles and 781K particle contacts. Figure 10-14 and Figure 10-15 show the pendulum model replicating the high-speed test. The shading indicates the rel- ative particle speeds as the one-wheel bogy passes through the arrestor bed. Figure 10-16, by contrast, shows the compres- sion force on the particles, indicating the region of high load- ing in front of and beneath the wheel. The pendulum tests were used to calibrate two critical parameter sets for the aggregate: the coefficients of restitution and the coefficients of static friction. Both parameters must be defined for particle-to-particle and particle-to-surface interactions. These interactions had a substantial effect on the predicted loads for the pendulum strut as it passed through the aggregate bed. After calibrating these parameters, the pendulum model matched the test results for both the high- and low-speed tests for the dry aggregate bed. Figure 10-17 shows a graphical comparison of the simulation and test data, where the error bars indicate the scatter of the actual test data. The deviations from the true test data for the low- and high-speed tests are 3.1% and 3.0%, respectively. Recalling the speed dependence manifested during the physical tests (Figure 10-10), it was important that the aggre- gate model be capable of capturing this effect. It was clear from the calibration results in this section that the model did in fact do this with high accuracy. This lends confidence to the later predictions made for even faster overrun situations, at speeds as high as 70 knots (81 mph). 10.4.2.4. Turf Tests The turf tensile test was replicated in EDEM using a layer of particles joined together with a bond contact definition (Figure 10-18). The bonded layer was then pulled apart using simulated clamps at either end. The geometry of the tensile specimen is not an exact match to that of the physical tests due to the rigid geometry limitations of EDEM. However, the loaded area reflects the 8-in. wide region used in the physical test specimens. The model had a total of 946 particles, 2,126 contacts, and 1,937 bonds. Figure 10-18 illustrates the tensile loading of the model turf. The failure of the specimen is progressive, and as the turf is stretched an increasing number of inter-particle bonds are broken. The bonds were defined in terms of stiffness and fail- ure strength, in both normal and shear directions. Iterative combinations of these four parameters were tried until the overall turf strength and elongation at failure matched the actual test specimens. Other material properties (shear mod- ulus, Poisson’s ratio, etc.) were taken from nominal values for sand or the aggregate model, depending on appropriateness. The tensile strength of the model turf was within 1% of the test data average, while the energy absorption was within 11% of the test data average. However, the spread in the test data itself was +/– 7% and 11% for these quantities, respectively. 95

96 Figure 10-12. Angle of repose simulation in EDEM.

Given the highly variable nature of the turf as a material, this accuracy was deemed sufficient for the needs of the evaluation. 10.4.3. Tire and Arrestor Simulations Using the calibrated aggregate and turf material models previously described (Section 9.4.2), a large-scale arrestor model was created in EDEM to simulate overruns by aircraft tires. Figure 10-19 illustrates the model with a 36-in. depth and a B737-800 main-gear tire (Goodyear H44.5x16.5) at 50% penetration depth. 10.4.3.1. Arrestor Bed Models The arrestor bed models for the aircraft tire simulations were 8 ft wide and 25 ft long. Versions of the arrestor bed model were created with different depths in 6-in. increments, from 6 to 36 in. The deeper the bed, the more particles were included. However, the turf layer was always 4.6 in. deep, since this was a fixed layer thickness. All bed depths cited include the thickness of the turf layer. Compared with the pendulum model, this new model was longer, wider, and deeper, giving a larger overall volume of aggregate. As such, larger particles became a practical neces- sity to keep simulation times efficient. A particle size conver- gence study using the B737 main-gear tire showed that a 2-in. diameter particle could be used with an estimated particle size error of less than 6%. For the 36-in. depth, the model had a total of 149K parti- cles, including 138K aggregate particles and 11K turf parti- cles. The particles formed 372K contacts, and the turf layer had 15K inter-particle bonds. 97 Figure 10-13. Angle of repose measurement. Figure 10-14. Overall view of engineered aggregate pendulum model, velocity fringe plot.

While most simulations were conducted on a turf-covered bed, some were also run using an aggregate-only bed. The comparison of results is discussed in Section 10.4.4.4. 10.4.3.2. Tire Models EDEM does not inherently support deformable tire model- ing, which necessitated a different approach be taken than that of the LS-DYNA crushable arrestor models. The model tires were rigid forms that matched the inflated dimensions of the respective tires. During overruns, the tires maintained this undeformed shape. In reality, the tires would form flat regions on the bottom and front faces, as is shown in the LS-DYNA crushable material models. To account for this discrepancy, a corrective calculation was undertaken in the MATLAB Arrestor Prediction Code. For the EDEM arrestor models, the important measured components were the depth and width of the rut created, and the corresponding loads that it produced. 98 Figure 10-15. Side cutaway view of engineered aggregate pendulum model, velocity fringe plot. Figure 10-16. Side cutaway view of engineered aggregate pendulum model, force fringe plot.

from the 44.5-in. and 27-in. tire models using width correc- tion factors. The tire path through the arrestors was based on a pre- scribed motion using a fixed forward speed and rotation rate. As the tire passed through the bed, a steady-state loading resulted. The vertical and horizontal loading at this state was measured. Unlike the LS-DYNA arrestor models, the tire spin could not be released to settle at a steady self-rotation rate in EDEM, Because the tires were non-deforming, the tire library was simplified down to three tires instead of the five tires used in the crushable models (Table 10-2). The 44.5-in. and 49-in. main-gear tires of the B737 and B747 were very close to the same size. Similarly, the 29-in. main-gear CRJ-200 tire and 27-in. nose gear B737 tire were very close to the same size. The ruts each tire created would, therefore, be similar. As such, predictions for the 49-in. and 29-in. tires were scaled 99 0 100 200 300 400 Fo rc e (lb f) Low-Speed Test (8.5 mph) Test Simulation 0 200 400 600 800 Fo rc e (lb f) High-Speed Test (17.2 mph) Test Simulation X Fo rc e Av g Z Fo rc e Av g X Fo rc e Av g Z Fo rc e Av g Figure 10-17. Engineered aggregate pendulum model data comparison to test data. Clamp Motion Loose Turf Particles After Bonds Broken Figure 10-18. Turf tensile test model showing turf particles (dark) and bonds (light). Turf Layer with Bonded Particles Aggregate Layer with Loose Particles Loose Turf Particles After Bonds Broken Figure 10-19. Model of combined tire and engineered aggregate arrestor system showing aggregate (light) and turf (dark) particles.

since EDEM does not have inherent dynamics calculations for the rigid tire parts. Instead, it was held at a constant pre- scribed rate, assumed to be the ideal rate of spin that the tire would have on a hard surface with no slippage at the given forward speed. This unfortunate necessity affected the pre- diction accuracy because the tires can exert a forward “driv- ing” type of torque due to their constant rotation rate. The impact of this effect was mitigated by prescribing a low tire-to- aggregate friction of 0.10, which limits any fictitious driving effects. The impact of such driving was deemed a second-order effect in the simulations. Future simulations using the aggregate could potentially link EDEM with a dynamics code for improved accuracy; such enhancements were, however, outside the scope of the current effort. 10.4.4. Batch Simulations Using the arrestor bed model, large batches of simulations were conducted to generate substantial bodies of data for a wide range of overrun conditions. This data was then assem- bled into “metamodels” for uploading and use by the APC. 10.4.4.1. Methodology Batch simulations were conducted for each tire with three open variables: • Speed, from 10 to 70 knots (Speeds below 10 knots were impractical due to the long simulation times required for a tire to travel the required minimum distance. Loading at speeds below 10 knots was based on the extrapolated meta- model data fit.); • Bed depth, from 6 to 36 inches (Figure 10-20); and • Penetration into the bed, from 10% to 100% of bottom- ing depth. Batch simulations for the EDEM arrestor models were more involved than for the LS-DYNA models because the optimization software, LS-OPT, could not automatically cre- ate parameterized variants of the models. Consequently, the input files were created manually prior to running, using experimental design points specified by LS-OPT. After the simulations were finished, the measured load data was then extracted manually and assembled into a form that could be read back into LS-OPT. The time-consuming nature of this manual approach led to a simplified process: simulations were done in batches of 50 for each tire in a single iteration. No additional add-on runs were undertaken, which was often done for the LS-DYNA batches in an attempt to increase the accuracy of the final data set. 10.4.4.2. Summary Tables of Metamodels The output from the batch simulations was assembled and uploaded into LS-OPT, where metamodels were constructed for the drag and vertical load forces. Metamodeling is analo- gous to fitting a curve through experimental data, except it is applied to multi-dimensional data sets. Here, the data sets are 100 Aircraft Landing Gear Tire Designation Modeling Method Main Gear H29x9.0-15 Scaled from H27 Data CRJ-200 Nose Gear R18x4.4 Modeled in EDEM Main Gear H44.5x16.5-21 Modeled in EDEM B737-800 Nose Gear H27x7.7-15 Modeled in EDEM Main Gear H49x19-22 Scaled from H44.5 DataB747-400 Nose Gear H49x19-22 Scaled from H44.5 Data Table 10-2. Tire library simplification for engineered aggregate arrestor models. Bed Depth Bottoming Depth Penetration Depth Figure 10-20. Depth definitions for engineered aggregate bed models.

four-dimensional, including speed, depth, penetration, and load (either vertical or drag). The metamodels used for these responses were RBF networks, which can effectively capture non-linear behaviors including multiple concavity changes across the data set. Table 10-3 summarizes the fit quality for the metamodels. The RMS error was typically below 5%, and the R-squared value was typically above 0.99, indicating good fit quality with minimal noise. In the table, two outlier points were eliminated from each data set such that the metamodels were constructed using 48 of the 50 original experimental design points. Reasons for the outliers generally were data measurement issues arising from a failure of the run to reach steady-state conditions prior to the tire exiting the arrestor bed. 10.4.4.3. Parameter Sensitivities Using the metamodels, it is possible to review how sensitive the landing gear loads are to the different variables of speed, bed depth, and penetration percentage. The unusual feature for the engineered aggregate system is that the response is very sensitive to forward speed. Practically speaking, this means that the engineered aggregate system will exert more load on the landing gear when the aircraft is travelling at high speed, and less at low speed. Figure 10-21 shows a surface plot for the metamodel. The drag force in this case becomes stronger (more negative) where the speed increases. For a deep penetration along the left-front edge of the plot, the loading at 70 knots is 34,700 lbf, and at 10 knots it is only 2,400 lbf—a factor of 14 difference. 101 Tire Experimental Points Points Used Response RMS Error R 2 Drag 3.52% 0.999 H44 50 48 Vertical 4.20% 0.996 Drag 4.38% 0.998 H27 50 48 Vertical 8.38% 0.982 Drag 7.59% 0.993 R18 50 48 Vertical 5.00% 0.987 Table 10-3. Metamodel accuracy summary for engineered aggregate/turf arrestor bed. D ra g Fo rc e (lb f) Speed (knots) Penetration Depth Ratio Figure 10-21. Metamodel drag load surface plot for 44.5-in. tire in an 18-in. deep arrestor/turf bed.

The drag load was found to be dependent on speed and speed- squared terms: 10.4.4.4. Effect of Turf Cover Layer The presence of the turf cover layer produced a mild decrease in the rate sensitivity as compared with a bed of plain aggregate without a cover layer. Table 10-4 compares the rela- tive sensitivities of the two systems to the aircraft speed (H44.5x16.5-22 tire). It may be possible to further enhance this mitigating effect with design improvements to the cover layer. Another advantage of the turf layer is a predicted reduction in the aggregate ejecta thrown from the tire rut. Figure 10-22 shows a comparison for an 18-in. deep bed with a 44.5-in. tire at a deep penetration level (99%). In the left-hand illustra- tion, the aggregate particle flyout is shown with the turf layer omitted. The right-hand image is the analogous case for a bed with no turf cover layer. The maximum aggregate particle Drag A Speed B Speed C= ⋅ + ⋅ +2 velocities are similar in both cases (∼1,000 in./s), but the vol- ume of debris appears to be reduced where the turf layer is present. If the ejecta plume is mitigated in this way, it could reduce the risk of ingesting aggregate particles in the aircraft engines. However, the reality of this predicted behavior requires con- firmation through physical testing. 10.4.4.5. Effects of Penetration Depth The bi-layer construction of the turf-covered aggregate bed can lead to situations where the tire creates a tunnel beneath the turf surface. Figure 10-23 illustrates such a situation, where the turf layer is stretched, but not broken, above the small 18-in. diameter tire. In actuality, the landing gear ver- tical strut member would sever the turf layer, preventing a complete tunnel. Additionally, bed designs of this relative depth would typically not be feasible because they would overload the landing gear. 10.4.4.6. Data Transformation The final metamodel data for each tire was converted for use by the APC. LS-OPT was used to extract nominally 9,000 data points from each metamodel and export them into tab- ular form. A MATLAB matrix conversion program was writ- ten to map this data into multi-dimensional matrix form that could be quickly accessed by the APC. For the two tires not explicitly modeled, the appropriate response surface data was scaled to produce an approximate metamodel (Section 10.4.3.2). 102 System Drag Force Vertical Force Aggregate/Turf -14 20 Aggregate Only -20 23 Percent Advantage for Aggregate/Turf 30% 13% Table 10-4. Comparison of speed sensitivities for engineered aggregate systems with and without turf cover layer. Figure 10-22. Ejecta plumes of engineered aggregate for beds with (left) and without (right) turf cover layers.

10.5. Arrestor Performance Predictions 10.5.1. Scope of Simulations Using the APC, a separate optimal arrestor was designed for each of the three trial aircraft: CRJ-200, B737-800, and B747-400. Subsequently, an optimal mixed-fleet arrestor was designed as a compromise best-fit for all three aircraft. All simulations assumed a turf-covered arrestor bed with dry aggregate. If a drainage-type arrestor bed were used, the aggregate would be intermittently wet and dry, and longer arrestor designs would be required. It was determined through experimentation that the aggre- gate arrestor design functioned best as a fully recessed bed, with the top of the turf level with the runway, and the remainder of the material placed in a basin below grade. The predicted stopping distances for the three aircraft are given in Table 10-5. All arrestment predictions assumed the following: • 50-ft setback distance; • 50-ft gradual decline to the maximum bed depth; • 70-knot starting speed for the aircraft; • No reverse thrust; • Braking factor of 0.25 before and within the bed; and • Arrestor bed loads based on interaction with tires, neglect- ing strut and axle components. Arrestor beds were designed for two different nose-gear loading criteria: 1. Limit Load Criterion, where the drag load applied to the nose strut cannot exceed the limit load for the nose gear (FAR Part 25.509); 2. Ultimate Load Criterion, where the drag load applied to the nose strut cannot exceed the ultimate load for the nose gear; Since the ultimate loading criterion permits higher loads on the strut, deeper beds and shorter stopping distances resulted from those cases. 10.5.2. Performance for Test Aircraft Table 10-5 lists best-case arrestor designs for each aircraft taken individually. Each arrestor bed listed uses a different material strength and depth that are optimized for the design aircraft. Generally, a range of acceptable strength and depth combinations was available. Compared with the similar EMAS design cases on the right (provided by ESCO), the dis- tances are comparable if the ultimate loading criterion is used. Table 10-6 shows the compromise design case with the best arrestor design for all three aircraft. The CRJ-200 controls the bed depth in this case, while the B737 controls the bed length. With the material strength and depth as specified, the B747 103 Figure 10-23. Tunneling of tire below turf surface. Nose-Gear Limit Load Criterion Nose-Gear Ultimate Load Criterion Current EMAS, Optimal Designs Arrestor Bed Design Depth (in.) Bed Length (ft) Depth (in.) Bed Length (ft) Depth (in.) Bed Length (ft) Single Aircraft CRJ-200 9.6 495 14.1 310 20 258 B737-800 13.5 462 19.5 361 20 287 B747-400 29.1 568 34.7 517 26 495 Table 10-5. APC predicted 70-knot stopping distances for engineered aggregate arrestor system.

would require 922 feet to decelerate from 70 knots. Per typical design practice for an EMAS, the bed length may be specified such that all aircraft satisfy the minimum 40-knot exit speed requirement. For comparison with the other alternatives, the bed designs in the table assume a 400-ft length. At this length, the B737-800 and B747-400 would have maximum exit speeds of approximately 63 and 39 knots, respectively. The former sat- isfies the requirements of AC 150-5220-22a, while the latter is just below the 40-knot minimum. 10.5.3. General Observations Figure 10-24 and Figure 10-25 give sample output plots from the APC for the B737 arrestor bed case for the limit and ultimate design cases, respectively. The setback for the bed is given in negative x-distance values, with the arrestor bed beginning at x = 0. As shown, the deceleration and nose gear drag loading are highest just after entry into the bed, when the nose wheel is located at 70 ft into the bed. Here the bed depth has increased to a maximum, but the forward velocity is still high. Due to the rate dependent nature of the loading, the drag load peaks here and tapers steadily as the aircraft decelerates. From a standpoint of mechanical efficiency, this behavior is less desirable than an ideal constant deceleration rate. As the aircraft speed decreases, the arrestor becomes ever less efficient, stretching out the end of the arresting process. From a standpoint of safety, this behavior requires a suit- able design criterion to be developed with regard to design exit speeds. If, for example, a bed has been designed for a par- ticular aircraft at a 70-knot overrun speed, the landing gear loading would reach a maximum redline value at about 70 knots. If that aircraft were to overrun the arrestor at 80-knots in an actual event, damage or failure of the landing gear could occur. A second possible scenario would involve a short landing, when the aircraft touches down in the arrestor bed. A suitable criterion could require higher design speeds to ensure a margin of safety; however, the current EMAS advi- sory circular does not specifically contain such a provision. 10.5.4. Braking Effects The assumed braking factor of 0.25 is applied to the main gear as a coefficient of friction, which helps to slow the air- craft. However, this may not be a valid assumption for the engineered aggregate systems. Because the aggregate is loose and rolls easily over itself, it may not provide a good base beneath the tire for braking. This “bottoming” behavior, when the material is compacted in a thin layer beneath the tire, will require more experimentation to better understand and accurately represent within the APC. However, in the absence of data to the contrary, a braking factor of 0.25 is likely still a reasonable, conservative assumption. 10.5.5. Short Landings Short landings involving an aircraft touch down inside the arrestor bed were not simulated. However, the potential for short landings presents an additional issue for consideration. The basin geometry of the arrestor concept would force the aircraft to roll up the decline slope in the reverse direction from normal, acting as a ramp that would cause a strong load to the landing gear. This issue could be eliminated by only partially recessing the bed, as in the ideal EMAS design cases. Retaining walls or berms would be required for confining the aggregate and allowing the bed to maintain its shape in an above-grade orientation. With a nominal bed thickness of 20 in., it is unlikely that this would present a significant obsta- cle to implementation. 10.6. Estimated System Cost and Upkeep 10.6.1. Installation Process The engineered aggregate concept would require excava- tion of an arrestor bed basin with a depth nominally equiv- alent to that of an EMAS bed. This basin may or may not require paving before being filled with aggregate. However, the below-grade nature of the basin would require drainage from the bed to be included in the design using standard roadway engineering practices. The basin would be filled with aggregate using earth-moving heavy equipment. If a reinforced turf layer is used, the turf would be grown ahead of time, then cut into segments and placed atop the bed with heavy equipment. The loose aggregate solution offers the advantage of construct-in-place simplicity that could produce installa- tion cost savings over a traditional EMAS. It reduces site preparation and eliminates block manufacturing, place- ment, and joint sealing. Additionally, the durable nature of the engineered aggregate means that the bed filling process will not require special care to be taken; this will further 104 Nose-Gear Ultimate Load Criterion Bed Dimensions 14.1-in. depth 400 ft long Aircraft Exit Speed (knots) Stopping Distance (ft) CRJ-200 70+ 335 B737-800 63 400 B747-400 39 400 Table 10-6. Fleet design arrestor bed for glass foam arrestor system.

105 Figure 10-24. Limit criterion engineered aggregate arrestor design plots for B737-800 showing speed (top), deceleration (middle), and nose-gear load (bottom). -20 0 20 40 60 80 Sp ee d [kn ot ] a nd D ep th [in .] Aircraft Velocity and Bed Profile Aircraft Speed Upper Surface of Arrestor Lower Surface of Arrestor -100 0 100 200 300 400 500 0 0.2 0.4 0.6 0.8 1 Nose-Wheel Location [ft] -100 0 100 200 300 400 500 Nose-Wheel Location [ft] -100 0 100 200 300 400 500 Location [ft] D ec el er at io n [g ] Aircraft Decleration Aircraft Deceleration -10 0 10 20 30 40 Fo rc e (ki p) Landing Gear Forces - NOSE STRUT Nose Gear Drag Nose Gear Limit Load Nose Gear Ultimate Load

106 Figure 10-25. Ultimate criterion engineered aggregate arrestor design plots for B737-800 showing speed (top), deceleration (middle), and nose-gear load (bottom). -20 0 20 40 60 80 Sp ee d [kn ot ] a nd D ep th [in .] Aircraft Velocity and Bed Profile Aircraft Speed Upper Surface of Arrestor Lower Surface of Arrestor 0.2 0.4 0.6 0.8 1 1.2 Nose-Wheel Location [ft] Nose-Wheel Location [ft] Location [ft] D ec el er at io n [g ] Aircraft Decleration Aircraft Deceleration -10 0 10 20 30 40 Fo rc e (ki p) Landing Gear Forces - NOSE STRUT Nose Gear Drag Nose Gear Limit Load Nose Gear Ultimate Load -50 400 0 50 100 150 200 250 300 350 -50 4000 50 100 150 200 250 300 350 -50 4000 50 100 150 200 250 300 350

speed the installation process. However, turf preparation and placement would be additional tasks not required for the current EMAS design. 10.6.2. Cost to Establish System A preliminary estimate was made for the cost to establish an engineered aggregate arrestor system. It must be noted that the cost estimate from this section is only a basic approxima- tion for the purposes of comparing the different arrestor alter- natives. The cost estimate is based on a mixture of information from the manufacturer, the airport survey, and FAA Order 5200.9. To develop a more accurate estimate of the costs to install such a system, it is recommended that a detailed cost quote be sought from a firm qualified to undertake an instal- lation effort. Where possible, the methodologies used were consistent with the prior survey information collected regard- ing the existing EMAS (Section 3.5). The costs may be broken into two major categories: site preparation and installation. The site preparation costs were estimated for two cases. The engineered aggregate arrestor would use a basin for the arresting materials rather than a flat runway-type surface as is used for the current EMAS. The bot- tom of the basin could either be paved or earthen. Drainage, excavation, and leveling would be required for either option. Assuming that a full-paved surface is not provided under the bed, the cost for site preparation was assumed to be reduced by half; this value was used for the lower-bound cost estimate for the system. If a full-paved surface is provided, identical to that of an EMAS, then the preparatory costs were assumed to be the same as for an EMAS; this provided the upper-bound cost estimate. The installation cost estimate was separated into specific materials and general installation labor needs. Because these costs were specific to the engineered aggregate arrestor con- cept, they do not have a direct connection to any prior EMAS data. Discussions with the manufacturer produced cost esti- mates for the engineered aggregate, reinforced turf cover layer, and geo-textile/geo-plastic layers. Where applicable, materials included freight costs for trans-Atlantic shipping. The labor costs were based on estimates from the manufac- turer established from similar installation efforts. Finally, the site preparation and estimated EMAS costs were computed in two ways: (1) assuming average survey costs from this research, and (2) assuming FAA Order 5200.9 costs. The final cost estimates for both options are given in Table 10-7 and Table 10-8, respectively. Using the survey cost assumptions of Table 10-7, a 300-ft arrestor bed would cost between 30% and 43% less than the current EMAS. If the Order 5200.9 costs are assumed, the cost advantage drops to between 5% and 12% (Table 10-8). In addition to the tables in this section, longer-term life-cycle issues could also be considered. FAA Order 5200.9 includes a standard 10-year replacement interval for an EMAS, which translates into present-value life-cycle costs. Such a replace- ment could arguably be unnecessary for this arrestor concept (Section 10.6.4). Eliminating the assumed 10-year replacement could effectively trim about $2.6M off present-value life-cycle costs (based on the EMAS replacement cost estimates of the survey). 10.6.3. Maintenance Maintenance for the engineered aggregate concept would be relatively simple, and should be limited to standard grounds- keeping measures for the protective turf layer. Drainage of the area to prevent standing water is required and periodic inspec- tions would be advisable to ensure that no issues arise due to seasonal weather changes. Due to the lack of joints, blocks, and degradable materials, many protective measures used in cur- rent EMAS construction would not be necessary. Therefore, the predicted maintenance needs are lower than for that of the existing EMAS. 107 Engineered Aggregate System Cost Category Lower Bound Upper Bound Current EMAS Site Preparation $ 1.08 $ 2.16 $ 2.17 Installation $ 3.61 $ 3.61 $ 6.03 Cost to Establish $ 4.69 $ 5.77 $ 8.19 Percent of EMAS 57% 70% Table 10-7. Estimated costs to establish engineered aggregate arrestor, 150 x 300 ft, assuming survey average costs for current EMAS, units of millions USD. Engineered Aggregate System Cost Category Lower Bound Upper Bound Current EMAS Site Preparation $ 0.34 $ 0.68 $ 0.68 Installation $ 3.61 $ 3.61 $ 3.83 Cost to Establish $ 3.95 $ 4.29 $ 4.50 Percent of EMAS 88% 95% Table 10-8. Estimated costs to establish engineered aggregate arrestor, 150 x 300 ft, assuming Order 5200.9 costs for current EMAS, units of millions USD.

Compared with crushable materials, there are no material breakdown concerns regarding freeze–thaw exposure. How- ever, it is presently unknown how substantially ice may affect the inter-particle friction of the aggregate; changes to the fric- tion could alter the dynamic arresting response. Additionally, dirt entrainment over time could solidify the bed if the drainage approach is used for handling precipitation. Therefore, the waterproofed option is recommended to minimize the mainte- nance and performance impacts of water, soil, and ice. 10.6.4. Replacement All materials included in the engineered aggregate concept are expected to last for the 20-year life cycle prescribed in the EMAS Advisory Circular, if not longer. It is not anticipated that replacement of the bed would be required after 10 years, as anticipated for an EMAS in FAA Order 5200.9. 10.6.5. Repair After an overrun event, the rut areas would require repair. The aggregate remaining in the ruts would be reusable, and fresh aggregate would likely need to be added to bring the beds back to level condition. The damaged cover layer, if used, would need to be removed from the rut areas and replaced with fresh material. If the cover layer is turf, it is possible that replacement turf could be planted at another area of the facil- ity on initial installation of the system. When replacement needs arise, this turf could be harvested for immediate use to prevent delays that would accompany growing fresh turf. 10.7. Transition to Fielded System In order to transition the engineered aggregate concept to a fielded system, the following additional development steps may be advisable. 10.7.1. Cover-Layer Design Several cover-layer design concepts are possible: 1. Thin sealant, 2. Asphalt skim coat, 3. Reinforced turf layer, 4. Reinforced turf with geo-textile fabric, and 5. Turf/geo-textile/plastic composite cover layer. Only concept three was modeled during the research effort. Additional testing and modeling would be required for whichever method is selected since the membrane behavior of a cover layer will impact the dynamic mechanical perfor- mance of the arrestor bed. The cover layer performance should be further characterized under frozen conditions because it will be exposed to such conditions even where the aggregate layer of the system is waterproofed. 10.7.2. Bottoming and Braking Dynamics The research performed has identified that tire bottoming and braking in the aggregate material require further study. The current modeling method of the APC makes simplifying assumptions regarding both phenomena sufficient for a concept-level evaluation. However, to ensure accuracy of design predictions, some additional tests would be required. One method could involve using a one-wheel bogy apparatus fitted with brakes and a load measurement system that is towed through the material. 10.7.3. Aggregate Ingestion While the presence of a cover layer appears to offer mitiga- tion of aggregate spray, it is unclear as to the level of residual hazard remaining. The spraying gravel from the nose-wheel may or may not pose an ingestion hazard for the aircraft engines. If an ingestion hazard is present, it is unclear what prac- tical risk this presents with regard to engine damage, potential fire, and so on. A limited test series using a one-wheel bogy apparatus may be able to answer the issue of ingestion likeli- hood. Discussions with aircraft manufacturers or other FAA safety personnel should provide insight into the practical risks in the event that ingestion occurred. 10.7.4. Requirements and Standards Due to the speed dependence issues, development of a prac- tical design criterion for the design exit speed may be advisable to ensure a factor of safety for the landing gear. 10.7.5. Full-Scale Testing A full-scale aircraft overrun test of an engineered aggregate arrestor bed is advisable because this concept represents a substantial departure from the current EMAS design in terms of mechanical loading and the materials used. 10.8. Summary Of the candidate systems evaluated, the engineered aggregate arrestor concept is most dissimilar to the current EMAS tech- nology. It uses a hard spherical aggregate that primarily under- goes displacement rather than the compaction that occurs with crushable foams like cellular cement. The arresting bed would be constructed with a shallow basin of the material topped with a reinforced turf cover layer. Since the engineered aggregate material is not a porous foam, it is durable against moisture and other environmental factors, 108

including water immersion. However, while the material itself does not degrade with such exposures, the mechanical response changes, exhibiting an increase in arresting loads when wet. To maintain predictable arresting performance, the material would require the use of protective plastic geo-membranes and typical drainage provisions. The material characteristics indicate that long service life is possible, potentially eliminating the standard 10-year replacement interval assumed in FAA Order 5200.9. Installation of the system would likely be simpler and less expensive than for the current EMAS since placement of blocks and sealing joints is unnecessary. Heavy equipment would place the material in the bed basin and top it with pre- grown reinforced turf segments. Geo-membrane and geo- textile layers, as applicable, would be placed and joined man- ually. The arrestor basin could be constructed with or without paving, which could result in preparatory cost reduction. The cost to establish such a system would be nominally 57% to 70% of the survey cost of the existing EMAS. Life- cycle costs could be reduced due to longer bed life. Mainte- nance needs appear to be simplified, requiring standard grounds-keeping measures, but no block or joint repairs. The APC predictions for the engineered aggregate arrestor show that the deceleration varies throughout the arrestment process, exhibiting a strong dependence on aircraft speed. This is not a preferable characteristic; however, functional arrestor designs are still feasible. Arrestor bed lengths for individual aircraft were nominally 15% longer than for the current EMAS technology due to the speed-dependent prop- erties of the arrestor. Additionally, the multi-aircraft design case for the concept showed the weakest performance among the three candidate systems. Bed designs that were safe for the smallest aircraft did not exert a strong deceleration load on the largest aircraft. The speed dependence of the engineered aggregate bed would require development of a new design criterion. Over- runs exceeding the rated exit speed or short landings into the bed could result in overloads to the landing gear if the speed dependence is not considered in the design process. Arrestor designs that include provisions for these events would neces- sarily have gentler decelerations and longer arrestment dis- tances than the design cases cited in this research. Transition to a fielded system would require finalizing a composite turf cover-layer design and calibrating a predictive model to match the response. Characterization would be advisable for the soil layer under various freezing conditions to assess the impact on arresting dynamics. Additionally, investigation should be made regarding the basin geometry to determine whether above- or below-grade construction is preferable. Full-scale testing is advisable for evaluation of the complete system. 109

Next: Chapter 11 - Aggregate Foam Arrestor Concept »
Developing Improved Civil Aircraft Arresting Systems Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s Airport Cooperative Research Program (ACRP) Report 29: Developing Improved Civil Aircraft Arresting Systems explores alternative materials that could be used for an engineered material arresting system (EMAS), as well as potential active arrestor designs for civil aircraft applications. The report examines cellular glass foam, aggregate foam, engineered aggregate, and a main-gear engagement active arrestor system.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!