Damage evaluation and protection method of resin pipe for gas conduit subjected to impact load

Abstract

Medium-density polyethylene pipe has been widely introduced to low-pressure gas pipes because of its high flexibility and corrosion resistance. However, many third-party damages due to the impact of heavy equipment have been reported during the construction every year, thus, to prevent the third-party damage, materials such as high-density polyethylene and polyamide have been considered as the new gas pipe candidates. However, their impact resistance capacity under the third-party attack has not been clarified. In this study, static and impact loading experiments were conducted to compare load resistance capacities. As a result, it was revealed that the high-density polyethylene pipe and the polyamide pipe had higher static load capacity and impact resistance than the medium-density polyethylene pipe. By comparing the absorbed energy of the static test and the impact test and calculating the pseudo absorbed energy of the impact test, the evaluation formula judging the safer side of whether the penetration occurred was proposed. Furthermore, as one of the methods to protect the gas pipe, the protective effect of winding a sheet made of reinforced fiber and non-woven fabric was clarified.

Keywords

Gas pipe medium-density polyethylene high-density polyethylene polyamide falling weight impact experiment protective effect

Introduction

City gas is produced at the production plant and then supplied through gas conduits as shown in Figure 1. In particular, the gas produced by vaporizing liquefied natural gas (LNG) is supplied to each household through each conduit as the pressure of the gas is lowered by transformers in the order of high pressure, medium pressure, and low pressure. The installation locations of gas pipes are mainly classified into four: exposed, buried, indoor, and outdoor.

Figure 1.

Supply process of city gas.

The material of the gas pipe is different depending on the installation location and the internal pressure. Steel pipes and cast iron pipes are used for high-pressure pipes and medium-pressure pipes. Regarding the low pressure conduit, plastic pipes such as medium-density polyethylene (MDPE) pipes have been widely introduced in the world. Plastic pipes are weak to external force compared with steel pipes and cast iron pipes and are easily affected by the internal pressure of gas, so the application range of the internal pressure is defined, and plastic pipes are used mainly between low pressure and medium pressure. On the contrary, plastic pipes are more excellent in corrosion resistance, flexibility, and workability than metal pipes, so they can be applied in the ground where the corrosion of metal pipes progresses and follow rapid displacements due to earthquakes and land subsidence.

In the United States, the MDPE gas pipe has been widely introduced since the 1970s. In 1981, plastic pipe accounted for more than 80% of the total annual extension of gas pipe, and the MDPE pipe occupied more than 65% of plastic pipe. Figure 2 shows the amount of distribution mains by material used by the local distribution gas companies (LDCs)in 2015 (Office of Energy Policy and Systems Analysis, US Department of Energy, 2017).

Figure 2.

U.S. Distribution Main Pipeline Miles in 2015.

In Japan, Japan Industrial Standard (JIS) of plastic gas pipe was enacted and the MDPE pipe became applicable to low pressure (less than 0.1 MPa) gas conduit in the 1980s. In the Kobe earthquake in 1995, good earthquake resistance of the MDPE pipe was demonstrated, and the MDPE pipe began to spread rapidly. Nowadays, most of the new low-pressure gas conduits are occupied by MDPE pipes. In addition, it is now possible to apply to medium-pressure gas conduits (0.1 MPa or more and less than 0.3 MPa).

With regard to the accidents of gas pipes, in 2013, a gas explosion accident occurred due to the corrosion of the cast iron gas pipe that was installed in the 1940s in Birmingham, AL, USA (National Transportation Safety Board, 2016), and then the accident occurred almost every day from 2004 to 2014 due to gas leaks caused by the corrosion and factors other than the corrosion (Kelly, 2014). The MDPE pipes are expected to eliminate the accident due to the gas leaks resulting from the corrosion. However, in 2016, third-party damage to the plastic gas pipe buried in the ground occurred in Canton, IL, USA (National Transportation Safety Board, 2018), and plastic gas pipes which are weak to external force compared with metal pipes have not been able to prevent the third-party damage.

Similarly, in Japan, the accidents by the third-party attack have occurred every year, and according to the research of Ministry of Economy, there are 616 cases and 39 people injured from 2013 to 2017 (Ministry of Economy, 2018). Figure 3(a) shows accidents caused by the third party such as heavy equipment. To prevent such accidents, protective measures have been considered; and as one of the protective measures, we prepared a protecting sheet made of reinforced fiber and non-woven fabrics, and covered the gas pipe using the protecting sheet, as shown in Figure 3(b).

Figure 3.

Accident of a gas pipe by a third party and its protection method: (a) image of an accident by a third party and (b) gas pipe covered with a protective sheet.

On the contrary, slow crack growth (SCG) has been considered the main cause of failure for plastic gas pipes (Frank et al., 2012; Krishnaswamy, 2005). SCG is a phenomenon in which the internal pressure of gas acts on the inner surface of the gas pipe in the long term and causes a crack on the inner surface, and the crack grows slowly toward the outer surface. The current plastic pipes are classified under minimum required strength (MRS) calculated from ISO 9080 to guarantee a service life of at least 50 years. In previous research works, lifetime estimation of plastic gas pipe focusing on long-term performance of polythene (PE) pipe and life evaluation method have been proposed (Frank et al., 2009; Kratochvilla et al., 2014; Pinter et al., 2007; Poduška et al., 2016). Other than the failure due to SCG, the failure mechanism of polyethylene pipes affected by ground displacement such as settlement and fault have also been studied (Luo et al., 2015; Zhang et al., 2018). Regarding the third-party attack, Brooker (2005) analyzed the failure behavior of buried X65 steel pipeline under direct excavation load of some dipper-teeth and investigated the relationship between the load of excavation and the dimensions of the pipe and bucket. Liu et al. (2018) studied the mechanical response of buried PE pipeline under excavation load by finite element method (FEM) analysis and found failure procedure of PE pipe under excavation load and effects of excavation position, pipe diameter, wall thickness, and internal pressure. In those researches, although excavation load under static loading condition was studied, the case where the excavation load acted on PE pipe in the short-term under dynamic loading condition was not examined.

However, to prevent the third-party attack, it is necessary to clarify the impact resistance capacity of PE pipes under dynamic load since it is possible that dynamic load acts by swinging the bucket down to the PE pipe in the third-party attack. Therefore, in our previous study, static and impact loading tests had been conducted on the MDPE pipe to clarify the load resistance capacity and effect of the protecting sheet shown in Figure 3(b) (Tamai et al., 2018).

Meanwhile, it has been considered to prevent accidents by increasing the strength of a gas pipe itself, and other polymeric materials such as high-density polyethylene (HDPE) and polyamide (PA) have been listed as new gas pipe candidates. However, their impact resistance capacity under the situation of the collision accidents have not been clarified. In this study, static and impact loading tests were conducted on the MDPE pipe, the HDPE pipe, and the PA pipe to compare load resistance capacities and deformation characteristics. In the impact test, single loading and repeated loading were performed. With regard to the results of the single impact of the HDPE pipe and the PA pipe, the impact resistance capacity against dynamic load was also compared with the results of the single impact of MDPE pipes with the current protecting sheet.

Basic mechanical properties of resin material

The material tests were conducted to understand the basic mechanical properties of resin materials (MDPE, HDPE, and PA) used as gas pipes. Tensile tests were conducted on the JIS K6815-3 corresponding to ISO 6259-3 (1997) under an environment of 23° ± 2°. The tension speed was 100 mm/min. The tensile specimens were made by cutting out gas pipes as shown in Figure 4(a). Figure 4(b) shows the tensile test piece and the installation situation. Figure 5 shows the relationship between the nominal stress and the nominal strain and a representative case of five tests of each material. The average yeild stresses of MDPE, HDPE, and PA were 22.43, 26.41, and 43.73 MPa while the rupture strains were 412%, 294%, and 266%, respectively.

Figure 4.

Tensile test piece: (a) the cutting position of tensile test piece and (b) tensile test piece and installation situation.

Figure 5.

Nominal stress–strain relationship (tensile test).

As a result, the fracture strains of all materials reached about 250%, so it was found that the resin materials such as MDPE, HDPE, and PA had good deformation performance against the tensile load. In the deformation process up to fracture, first, a neck was formed at the central part of the piece after the maximum stress. Second, the neck did not become localized but propagated in the axial direction. Finally, the specimen stretched whitening due to neck propagation and broke (Figure 6). This whitening phenomenon is caused by a craze unique to crystalline polymers (Takahashi, 2013).

Figure 6.

Tensile test pieces after the fracture.

Compressive tests were conducted on JIS K 7181 corresponding to ISO 604 (2002) under an environment of 23° ± 2°. The compression speed was 1 mm/min. The compressive specimens were cut out of gas pipes as shown in Figure 7(a), similar to the tensile test piece. Figure 7(b) shows the compressive test piece and installation situation. Figure 8 shows the relationship between the nominal stress and the nominal strain. It confirmed that the load resistance was small at low compression and MDPE, HDPE, and PA had viscosity. Although they did not have a clear yield point, they hardened after yield (Figure 9).

Figure 7.

Compressive test piece: (a) cutting position of compressive test piece and (b) compressive test piece and installation situation.

Figure 8.

Nominal stress–strain relationship (compressive test): (a) general view and (b) enlarged view.

Figure 9.

Compressive test piece after the test.

In addition, resin-based materials generally have temperature dependence, so we conducted the material tests on MDPE under different environmental temperature conditions (5°C, 23°C, 40°C). Figure 10 shows the nominal stress–nominal strain relationship. The yield stress at 5°C, 23,°C, and 40°C were 26.86, 23.46, and 17.55 MPa, respectively. MDPE did not have a clear yield point in the compressive test, but we similarly confirmed the temperature dependence. Regarding the strength degradation due to temperature, considering that the glass transition temperature of standard polyethylene is −125°C, it can be said that the glass transition is not greatly involved. When the temperature rises, the crystalline part of MDPE unwinds, and the random part gradually increases, so that flexibility appears and the strength is assumed to have decreased. In this study, we performed experiments under the same temperature condition, so we did not take the temperature dependence into account.

Figure 10.

Nominal stress–nominal strain relationship: (a) tensile test and (b) compressive test.

Outline of static and impact experiment

Targeted pipe specimen

In this study, three kinds of pipes, MDPE pipe, HDPE pipe, and PA pipe, were compared with plastic pipe. Generally, PE pipes have features such as good flexibility, corrosion resistance, and workability. There is a difference in molecular structure between MDPE and HDPE. Although MDPE has branch structures and straight chain structures, HDPE only has straight chain structures. It is also known that an increase in density increases heat resistance and rigidity but decreases elongation and workability. In general, it is said that the PA has a higher impact resistance than the PE, but as it is not quantitatively clarified, it is chosen as the target of this study. Figure 11 shows the cross-section size of plastic pipes. Axial lengths of all the pipes were 500 mm. Regarding HDPE and PA pipes, since there were no similar standard as the MDPE pipe, a similar standard with different ratios of outer diameter-to-thickness was used. The ratio of outer diameter-to-thickness is shown as

Ratio of outer diameter to thickness = \frac{D}{t}

where D is the diameter and t is the thickness.

Figure 11.

Cross section of pipes (unit: mm): (a) MDPE and (b) HDPE, PA.

With regard to the ratio of the outer diameter-to-thickness of the MDPE pipe, the HDPE pipe, and the PA pipe, the MDPE pipe was 13.4, and the HDPE pipe and the PA pipe were 11.0. In a pipe with a small thickness relative to its diameter, a part of the member may be locally buckled when compressed in the axial direction. The ratio of outer diameter-to-thickness is an indicator of the likelihood of local buckling.

Outline of static loading experiment

Static loading tests were conducted to understand the deformability and static loading bearing capacity of MDPE, HDPE, and PA pipes. As shown in Figure 12, the loading jig was used by installing the tip of the bucket of the backward tight quarter excavator (JIS A 8340-4), which is used actually in the construction, on the loading board. Figure 13 shows the installation status of the specimen, loading direction, and the universal testing machine used in this experiment. Measurement items are the load and vertical displacement. The load is measured using a load cell attached to the loading jig, displacement is measured using a high performance contact displacement meter.

Figure 12.

Details of loading jig: (a) the tip of the bucket, (b) size of the tip of the bucket, and (c) loading jig.

Figure 13.

Installation status of the specimen in the testing machine: (a) installation status, (b) loading direction, and (c) universal testing machine.

Outline of the falling weight impact experiment

The impact experiments were conducted using the falling weight impact test apparatus to understand the impact resistance capacities of MDPE, HDPE, and PA pipes. Figure 14 shows the schematic diagram of the gas pipe placed within the impact apparatus. The section sizes of target pipes were the same as shown in Figure 11, and the axial lengths of all the pipes were 500 mm. The eccentricity of the steel weight was able to be controlled by the guide rail in this test apparatus during free fall and at the time of impact. It was confirmed before the start of the experiment that the relation between the impact velocity and the fall height is in accordance with the theoretical value. The impact load was given by free falling of the steel weight from the designated height to the center of the span of the specimen. Instruments imitating the actual bucket tip as shown in Figure 12 are attached to the steel weight. The direction of the loading part was the same as the static experiment and was shown in Figure 13(b). The gas pipes that are actually buried in the ground are elastically supported by the ground. However, in this research, the bottom of the pipe was fixed to the bottom plate with bolts of M10 to grasp the impact resistance capacity of the pipe excluding the influence of the ground and to prevent rotation and jumping up at impact loading.

Figure 14.

Schematic diagram of gas pipe placed within impact apparatus.

Regardless of contact or non-contact type, it is difficult to measure the displacement of locally deformed pipes. Therefore, in this study, the movement of the weight during collision using a laser displacement meter was measured, and the amount of indentation was treated as the displacement of pipes. The impact force of the weight was measured by a load cell. Pipe deformation was recorded by a high speed camera. After the test, the residual displacement on the surface of the impact part and the change rate of pipe thickness were measured. Figure 15 shows the residual displacement on the surface of the impact part and the change rate of pipe thickness.

Figure 15.

Definition of the residual displacement and the change rate of pipe thickness.

In this study, the impact condition in this experiment set conditions similar to that occur in an actual accident. To set conditions, it was necessary to clarify the equivalent mass and impact velocity of the weight falling on the gas pipe. The following two methods were considered to estimate the equivalent mass. (1) The only backhoe bucket, shown in Figure 16, acts directly as equivalent mass and (2) the equivalent mass is determined in consideration of the inertia efficiency of the whole mechanism of the front part including the arm part. In the latter method to be considered in previous studies (Kubodera, 1964), the impact stress to the ground is calculated by vibration analysis assuming a model in which the front part is mechanically simplified, and the equivalent mass is shown below

m = \frac{J - m^{'} l^{' 2}}{l^{2}}

where J is the inertia efficiency of the front part around the boom foot; m is the bucket mass; and m′ is the mass of the front part excluding the bucket.

Figure 16.

Image of backhoe.

Assuming a backward ultra-small swing type backhoe with a bucket capacity of 0.45 m³ widely used in actual gas construction, the equivalent mass was found to be 240 kg. Considering that the mass of the bucket part was 100–220 kg when the assumed bucket capacity was about 0.1–0.45 m³, it was found that the equivalent mass calculated by the latter method was mostly the same as the mass of the bucket part. Therefore, in this study, the weight mass was 220 kg as a considerable maximum mass. Next, the impact velocity in single impact loading and constant velocity repetitive impact loading was based on 2 m/s in consideration of the following. Regarding the impact velocity, drilling speed varied depending on the skill of the backhoe operator. Considering that it was generally said that the drilling speed was 0.38 m/s (Ming, 1989) and a hearing survey result that no bucket was swung down from a height higher than 50 cm, the impact velocity was estimated to be 0.38–3 m/s. Based on the above, experimental cases were shown in Table 1.

Table 1.

Experimental cases.

Loading method	Details
Single impact	2, 2.5, 3, 3.5, 4 m/s If the pipe penetrates, do not conduct experiments at higher velocity
Repetitive impact	Continue to impact with the same specimen at the same impact velocity (2 m/s) until the pipe penetrates

Experimental results and observation

Static load resistance characteristics

Figure 17 shows the relationship between load and displacement. The common thing to the three types of pipes is that there was no clear yield point. The maximum loads of the MDPE pipe, HDPE pipe, and PA pipe were 14.92, 19.97, and 34.47 kN, respectively. The maximum displacements for the MDPE pipe, HDPE pipe, and PA pipe were 105.4, 82.51, and 69.07 mm, respectively. The load bearing capacity of the PA pipe was larger than that of the MDPE pipe, but the PA pipe stretched more hardly than the MDPE pipe. When the displacement was about 20 mm, the increase in displacement of the whole pipes got larger. When the displacement was about 40 mm, the increase in the rate of load of the whole pipes got larger again. Figure 18 shows the comparison of the deformation situation at a static load of 10 kN. At the static load of 10 kN, the displacements of the PA pipe and the HDPE pipe did not reach 20 mm. On the contrary, the displacement of the MDPE pipe was more than 40 mm, and the local deformation of the loading part was progressing. From this comparison, it was found that the pipes were totally deformed against the static load when the increase in the displacement was small, and the increase in the displacement got large because the local deformation of the loading part was superior. After the weight penetration, as shown in Figure 19, whitening was confirmed on the rear surface of the loading part. It was considered that this change in the material due to whitening affected the change in slope of the load and displacement relationship. From the above, it was confirmed that the whole pipes had a good performance and deformation process against the static load.

Figure 17.

Comparison of the relationship between load and displacement: (a) MDPE pipe, (b) HDPE pipe, and (c) PA pipe.

Figure 18.

Comparison of deformation situation at static load 10 kN: (a) MDPE pipe, (b) HDPE pipe, and (c) PA pipe.

Figure 19.

Comparison of deformation situations at weight penetration.

Impact response and damage state at single impact

Figure 20 shows a comparison of the impact force response at each impact velocity. Figure 21 shows the maximum deformation situation at impact velocity of 3 m/s taken with a high-speed camera. The maximum impact forces at impact velocity of 2 m/s for the MDPE pipe, HDPE pipe, and PA pipe were 12.6, 20.2, and 29.3 kN, respectively. The maximum impact force of the HDPE pipe was 1.6 times larger than that of the MDPE pipe, and PA pipe was 2.3 times larger than that of the MDPE pipe. Regarding the impact duration, the impact duration of the HDPE pipe was 0.58 times larger than that of the MDPE pipe, and PA pipe was 0.46 times larger than that of the MDPE pipe.

Figure 20.

Comparison of impact force response. Impact velocity at (a) 2 m/s, (b) 2.5 m/s, (c) 3 m/s, (d) 3.5 m/s, and (e) 4 m/s.

Figure 21.

Comparison of maximum deformation situation at impact velocity 3 m/s: (a) MDPE pipe, (b) HDPE pipe, and (c) PA pipe.

Figure 22 shows the image of the deformation process from collision of the weight to bounce back. In the process from Figure 22(a) and (b), the pipe was totally deformed against the impact, and this deformation process corresponded to the first load rise in the load–time relationship. Then, in the process from Figure 22(b) and (c), the increase of the load got constant in the load–time relationship, and the local deformation of the collision part began. When the local deformation progressed and became as shown in Figure 22(d), the area directly under the collision part was further expanded locally because the collision part of the weight was flat against the pipe. The result showed that the back surface of the collision part was deformed to “w-shape.” After the weight bounced, the deformation on the backside was not restored as shown in Figures 22(e) and 23.

Figure 22.

Image of deformation process from collision to bounce back.

Figure 23.

Cross section picture of the MDPE pipe after the collision (impact velocity 2.5 m/s).

Figure 24 shows a comparison of the displacement response at each impact velocity. The HDPE pipe and the PA pipe showed smaller displacement than the MDPE pipe. According to the above results, the PA pipe had the hardest and hardly deformable quality, followed in order by the HDPE pipe and the MDPE pipe.

Figure 24.

Comparison of displacement response. Impact velocity at (a) 2 m/s, (b) 2.5 m/s, (c) 3 m/s, (d) 3.5 m/s, and (e) 4 m/s.

Table 2 shows the summary of the damaged conditions of pipes. In Japan, when the change rate of pipe thickness exceeds 20% in the quality evaluation of gas pipes, the pipe is judged to be “damaged” and it is decided to replace (The High Pressure Gas Safety Institute of Japan, 2013). Therefore, in this study, the damage condition was judged by same way. From a mechanical point of view, the “damaged” is a state in which residual deformation on the surface of the impact part occurred along with a white region called a craze due to plastic strain. The velocity of the HDPE pipe and the PA pipe at which the penetration occurred was higher than that of the MDPE pipe, and both the HDPE pipe and the PA pipe were not penetrated at impact velocity of 3 m/s, which was the maximum velocity that could occur in accidents.

Table 2.

Summary of damaged conditions of pipes (at single impact).

Material	Impact velocity (m/s)	Initial pipe thickness (mm)	Pipe thickness after test (mm)	Change rate of pipe thickness (%)	Damaged condition
MDPE	2	12.3	9.84	20	Damaged
	2.5		7.16	42	Damaged
	3		–	–	Penetration
HDPE	2	14.6	12.92	12
	2.5		10.92	25	Damaged
	3		8.68	41	Damaged
	3.5		–	–	Penetration
PA	2		13.88	5
	2.5		13.7	6
	3		12.58	14
	3.5		11.96	18
	4		–	–	Penetration

MDPE: medium-density polyethylene; HDPE: high-density polyethylene; PA: polyamide.

Proposal of the evaluation formula of penetration based on energy theory

Relationship between the existence of the penetration and the absorbed energy

It is necessary to prevent the accident by measuring the impact resistance capacity of the gas pipe by impact experiments. However, it takes much time to conduct impact experiments for each gas pipe type, size, and collision condition. If it is possible to judge the existence of the penetration without conducting impact experiments based on collision conditions and basic material properties which were obtained from static tests and material tests, future quality control of gas pipes will be simplified. Therefore, in this section, we considered and proposed the evaluation formula of the existence of the penetration based on energy theory.

Figure 25 shows the comparison between the absorbed energy of the static test and that of the impact test. The red line is the absorbed energy of the static test, and this energy was calculated from the area enclosed by the load–displacement graph shown in Figure 17. The black point is the absorbed energy of each impact velocity in the impact test. As shown in Figure 26, it was known that the velocity of the weight was zero at the maximum deformation, so the absorbed energy in the impact experiment was calculated by the sum of the potential energy and kinetic energy with the maximum deformation as the reference height. However, in the case where penetration occurred, the deformation just before penetration was treated as the maximum deformation. The velocities at the time of penetration of MDPE, HDPE, and PA were 1.1, 0.9, and 0.1 m/s, respectively, and were not strictly zero. However, the effect of them is less than 10% in the absorbed energy and does not affect the results described later. Furthermore, to establish a simple and safe evaluation method for estimating penetration, the influence of these velocities was decided to be ignored. Figure 25 shows that as a result, penetration occurred when the absorbed energy during the impact test exceeded the static absorbed energy. Therefore, in the experiment conducted in this study, the existence of the penetration can be judged by comparing the absorbed energy of the impact test and the static test.

Figure 25.

Comparison of the absorbed energy between the static test and the impact test: (a) MDPE pipe, (b) HDPE pipe, and (c) PA pipe.

Figure 26.

Calculation of absorbed energy in the impact test.

\begin{array}{l} Absorbed energy at impact test = Kinetic energy at collision \\ + Potential energy = \frac{1}{2} m v^{2} + m g h \end{array}

However, it was clarified whether penetration occurred under the condition that the mass of the weight was different, so to grasp whether the penetration occurred under the condition that the mass of the weight was different, the comparison of the absorbed energy between the static and impact test was conducted in case where the mass of the falling weight was 100 kg. Table 3 shows the experimental cases used for the comparison, and the impact velocity was set in two cases (3 and 4.4 m/s), taking kinetic energy into consideration so that the presence or absence of penetration could be divided.

Table 3.

Experimental cases used for the comparison.

Mass of weight (kg)	220	220	100	100
Impact velocity (m/s)	2	3	3	4.4
Kinematic energy (J)	440	990	450	968
Damaged condition	–	Penetration	–	Penetration

Figure 27 shows the comparison in the case where the mass of the falling weight was 100 kg, and it was confirmed that the penetration occurred in an experimental case where the absorbed energy of the impact test was bigger than that of the static test. From results shown in Figures 25 and 27, it was considered that in the range of the actual bucket weight, the presence or absence of penetration can be judged from the absorbed energy of the static and impact test.

Figure 27.

Absorbed energy–impact velocity relationship (the mass of the falling weight was 100 kg).

Estimation of absorbed energy of the impact test from the static test results

In section “Relationship between the existence of the penetration and the absorbed energy,” it was confirmed that the penetration occurred when the absorbed energy of the impact test exceeded that of the static test, and the existence of penetration can be judged from the absorbed energy of the static and impact test. However, since the absorbed energy of the impact test was calculated from the results of the test where the presence or absence of the penetration had been identified, it had no meaning that the absorbed energy was used for the determination of the existence of the penetration. On the contrary, if the absorbed energy of the impact test can be estimated without conducting the impact test, it can be used to judge the existence of the penetration. Therefore, in this section, it was considered that the absorbed energy of the impact test was estimated without the impact test and used to judge the existence of the penetration.

In the estimation of the absorbed energy, first, it was necessary to determine the maximum displacement h as shown in Figure 26. Figure 28 shows the comparison of the maximum displacement between the static test and the impact test. As the impact velocity increased, the displacement tended to increase, but in any case, the displacement of the impact test did not exceed that of the static test.

Figure 28.

Comparison of the maximum displacement between the static test and the impact test: (a) MDPE, (b) HDPE, and (c) PA.

Although the maximum displacement of the static test was larger than that of the impact test, the maximum displacement h′ of the static test was applied to the maximum displacement h of the impact test shown in Figure 26, and the pseudo energy absorbed at the impact test was calculated as shown below

\begin{matrix} The pseudo absorbed energy = Kinetic energy + The pseudo absorbed energy \\ = \frac{1}{2} m v^{2} + m g h^{'} \\ h^{'} : The maximum displacement of the static test \end{matrix}

Figure 29 shows the comparison between the actual energy and the pseudo energy absorbed at the impact test. Because the maximum displacement of the static test was larger than that of the impact test, the pseudo absorbed energy also became larger than the actual energy. As for MDPE, the difference in the displacement between the static test and the impact test was large, so the energy error was also large. The largest error was 19.4% at impact velocity 2 m/s. However, because the maximum displacement of the impact test approached that of the static test as the impact velocity increased, the error decreased and was 7.1% at impact velocity of 3 m/s. From the above, the pseudo energy overestimated the actual absorbed energy a little, but the pseudo energy could be used at least for evaluation on the safety side for determining the existence of the penetration.

Figure 29.

Comparison between the actual absorbed energy and the pseudo absorbed energy of impact test: (a) MDPE, (b) HDPE, and (c) PA.

Number of impacts to penetration velocity at repetitive impact

Figure 30 shows the relationship between the residual displacement on the surface of the impact part and the number of impacts. Although the MDPE pipe and the HDPE pipe differed in the residual displacement on the surface of the impact part after the first time, the increment of the residual displacement after the second time showed similar tendency. Even if the number of impacts increased, the residual displacement of the PA pipe gently increased compared with the MDPE pipe and the HDPE pipe. In addition, the MDPE pipe penetrated with two impacts and the HDPE pipe penetrated with six impacts. The PA pipe was not penetrated even when the number of the impacts reached 40 times, so it was found that the PA pipe had an excellent impact resistance performance against repeated impact loading at an impact velocity of 2 m/s or less.

Figure 30.

The residual displacement on the surface of the impact part—the number of impacts relationship.

Comparison with gas pipe with protecting sheet

Outline of protecting sheet

As shown in Figure 31, the protecting sheet made of reinforced fiber and non-woven fabric has been used to protect the given MDPE pipe. The reinforced fiber is made of polypropylene and is bonded to non-woven fabric by resin adhesive. With regard to the use of the protecting sheet, it is standard to wind the protecting sheet two turns around MDPE pipes with non-woven fabric inside based on the past construction results.

Figure 31.

The constitution and the installation situation of protecting sheet.

In the previous study [16], the falling weight impact experiment was performed on the MDPE pipe with a protecting sheet under the same conditions as section “Outline of static and impact experiment.” The result showed the protecting sheet did not reduce the impact force of the weight and the displacement of the weight but reduced the residual displacement on the surface of the collision part. In other words, the protecting sheet prevented the progress of failure from the surface of the collision part and made full use of the deformation performance of the MDPE pipe. In this section, the falling weight impact experiment was performed on the MDPE pipe with protecting sheets under the same conditions as section “Outline of static and impact experiment.” The impact velocity was 3 m/s. Moreover, the impact resistance capacity and the change rate of pipe thickness of MDPE pipe with protecting sheet were compared with those of the HDPE pipe and the PA pipe.

Results of experiments

Table 4 shows the comparison of the damaged condition of pipes. Although the change rate of the thickness of the protected MDPE pipe exceeded 20% and was judged to be “damaged,” the protecting sheet was able to prevent the penetration. Figure 32 shows the comparison between the unprotected MDPE pipe and the protected MDPE pipe. In the relationship between the load and time shown in Figure 32(a), the peak value of the load of the unprotected MDPE pipe was 15.00 kN, and that of the protected MDPE pipe was 20.15 kN. The penetration of the unprotected MDPE pipe occurred at the time of 0.028 s, and the load decreased rapidly at the same time. That is why the position of the peak load was largely different from the protected MDPE pipe, but the initial rising parts of the load–time relationship showed the same tendency. Regarding the relationship between the displacement of weight and time shown in Figure 32(b), the peak value of the protected MDPE pipe was 89.05 mm and was bigger than the value at when the unprotected MDPE pipe was penetrated. The protected MDPE pipe had more deformability than the unprotected MDPE pipe. From these results, it was confirmed that the protecting sheet was working properly.

Table 4.

Comparison of damaged condition of pipes.

Material	Impact velocity (m/s)	Initial pipe thickness (mm)	Pipe thickness after test (mm)	Change rate of pipe thickness (%)	Damaged condition
MDPE	3	12.3	–	–	Penetration
MDPE + Protecting sheet		12.3	9.74	21	Damaged
HDPE		14.6	8.68	41	Damaged
PA		14.6	12.58	14

MDPE: medium-density polyethylene; HDPE: high-density polyethylene; PA: polyamide.

Figure 32.

Comparison between MDPE pipe and MDPE pipe with protecting sheet: (a) load–time relationship and (b) displacement of weight–time relationship.

Figure 33 shows the comparison of the load and the displacement responses among three kinds of pipes (protected MDPE pipe, HDPE pipe, and PA pipe). The maximum loads of the HDPE pipe and PA pipe were 22.3 and 31.58 kN; and maximum displacements of the HDPE and PA pipes were 56.65 and 40.55 mm, respectively. The protecting sheet did not reduce the load and the displacement as mentioned above, so the correlation among three kinds of the pipes did not change. On the contrary, the change rate of the pipe thickness for the protected MDPE pipe was 21%, for the HDPE pipe was 41%, and for the PA pipe was 14%. It was confirmed that the MDPE pipe with the protecting sheet had superior impact resistance capacity than the HDPE pipe in terms of quality evaluation of gas pipes.

Figure 33.

Comparison of load and displacement responses: (a) load–time relationship and (b) displacement of weight–time relationship.

Conclusion

In this study, the results are summarized as follows.

Through the static and impact experiments, it was revealed that the HDPE pipe and the PA pipe have higher static load capacity and impact resistance, although the deformation performance is inferior to the MDPE pipe. This is easy to understand from the material test results.

Under the impact condition that could occur in an actual third-party attack, the MDPE pipe and HDPE pipe was penetrated and damaged, respectively, the PA pipe was not damaged. So, the PA pipe has a higher impact resistance against a single impact. In addition, the repeated impact resistance performance was relatively higher than the other pipes.

By comparing the absorbed energy of the static test and the impact test, and calculating the pseudo absorbed energy of the impact test, the evaluation formula judging on the safety side whether the penetration occurred was proposed.

It was found that the protecting sheet made of reinforced fiber and non-woven fabric was hardly effective for reducing impact load and absorbing impact energy, but has the effect of suppressing local deformation leading to penetration. The MDPE pipe with the protecting sheet had superior impact resistance capacity than the HDPE pipe without the sheet in terms of the quality evaluation of gas pipes. On the contrary, it was confirmed that the PA pipe without the sheet had an impact resistance equal to or better than the MDPE pipe with the sheet.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Hiroki Tamai

References

Brooker

(2005) Experimental puncture loads for external interference of pipelines by excavator equipment. International Journal of Pressure Vessels and Piping 82(11): 825–832. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S0308016105000931 (accessed 5 August 2019).

Frank

Hutar

Pinter

(2012) Numerical assessment of PE 80 and PE 100 pipe lifetime based on paris-erdogan equation. Macromolecular Symposia 311(1): 112–121. Available at: https://onlinelibrary-wiley-com-443.web.bisu.edu.cn/doi/full/10.1002/masy.201000096 (accessed 5 August 2019).

Frank

Pinter

Lang

(2009) Prediction of the remaining lifetime of polyethylene pipes after up to 30 years in use. Polymer Testing 287: 737–745. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S0142941809001044 (accessed 5 August 2019).

Kelly

(2014) Look out below: Danger lurks underground from aging gas pipes. USA Today, 23 September. Available at: www.usatoday.com/story/news/nation/2014/09/23/gas-pipes-cast-iron-deaths-explosions-investigation/15783697/ (accessed 5 August 2019).

Kratochvilla

Frank

Pinter

(2014) Determination of slow crack growth behaviour of polyethylene pressure pipes with cracked round bar test. Polymer Testing 40: 299–303. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S0142941814002232 (accessed 5 August 2019).

Krishnaswamy

(2005) Analysis of ductile and brittle failures from creep rupture testing of high-density polyethylene (HDPE) pipes. Polymer 4625: 11664–11672. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S0032386105014473 (accessed 5 August 2019).

Kubodera

(1964) Impact stress of dragshovel front. Hitachi Review [in Japanese]. Available at: www.hitachihyoron.com/jp/archive/1960s/1964/10.html (accessed 5 August 2019).

Liu

Zhang

Xia

, et al. (2018) Mechanical response of buried polyethylene pipelines under excavation load during pavement construction. Engineering Failure Analysis 90: 355–370. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S135063071730969X (accessed 5 August 2019).

Luo

Shi

, et al. (2015) Numerical simulation of strength failure of buried polyethylene pipe under foundation settlement. Engineering Failure Analysis 48: 144–152. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S1350630714003495 (accessed 5 August 2019).

10.

Ming

(1989) Study on optimum design of hydraulic excavator. PhD Thesis, Kyoto University, Kyoto, Japan. Unpublished material [in Japanese]. Available at: https://repository.kulib.kyoto-u.ac.jp/dspace/handle/2433/138433 (accessed 5 August 2019).

11.

Ministry of Economy (2018) Prevention of gas pipe damage accident in construction work [in Japanese]. Available at: www.meti.go.jp/policy/safety_security/industrial_safety/oshirase/2019/2/310226-01.html (accessed 5 August 2019).

12.

National Transportation Safety Board (2016) Pipeline accident brief: Birmingham public housing gas explosion. Available at: www.ntsb.gov/investigations/AccidentReports/Pages/PAB1601.aspx (accessed 5 August 2019).

13.

National Transportation Safety Board (2018) Pipeline accident brief: Third-party damage by sure shot communications to Ameren natural gas distribution system, canton, Illinois. Available at: www.ntsb.gov/investigations/AccidentReports/Pages/PAB1802.aspx (accessed 5 August 2019).

14.

Office of Energy Policy and Systems Analysis, US Department of Energy (2017) Natural Gas Infrastructure Modernization Programs at Local Distribution Companies: Key Issues and Considerations. Washington, DC: Office of Energy Policy and Systems Analysis, US Department of Energy, pp. 8–9.

15.

Pinter

Haager

Balika

, et al. (2007) Cyclic crack growth tests with CRB specimens for the evaluation of the long-term performance of PE pipe grades. Polymer Testing 26(2): 180–188. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S0142941806001784?via%3Dihub (accessed 5 August 2019).

16.

Poduška

Hutař

Kučera

, et al. (2016) Residual stress in polyethylene pipes. Polymer Testing 54: 288–295. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S0142941816304512 (accessed 5 August 2019).

17.

Takahashi

(2013) Modeling of craze evolution behavior for crystalline polymer and its application to FE simulation for fracture prediction. PhD Thesis, Keio University, Keio, Japan. Unpublished material [in Japanese].

18.

Tamai

Yoshitake

Sonoda

, et al. (2018) An experimental study on impact resistance capacity of polyethylene gas pipe and effect of its protecting sheet. Journal of Structural Engineering A64: 847–854 [in Japanese].

19.

The High Pressure Gas Safety Institute of Japan (2013) Technical Data on Polyethylene Pipes for Gas (Rev. ed.). The High Pressure Gas Safety Institute of Japan [in Japanese].

20.

Zhang

Xiao

Liang

(2018) Mechanical behaviors and failure mechanisms of buried polyethylene pipes crossing active strike-slip faults. Composites Part B: Engineering 154: 449–466. Available at: www-sciencedirect-com.web.bisu.edu.cn/science/article/pii/S1359836818309788 (accessed 5 August 2019).