PERFORMANCE EVALUATION OF 48” DIAMETER ......Deflection gauges, strain gauges, circumferential shortening gauges, and joint separation measurements were used to evaluate the pipe

PERFORMANCE EVALUATION OF 48” DIAMETER CORRUGATED

HIGH DENSITY POLYETHYLENE PIPE

UNDER 315,000 LBS (143,010 kg) RAIL CAR LOADING

Shawn R Coombs, PE

Project Manager/Primary Author

Advanced Drainage Systems, Inc.

2009 Whiskery Court

York, SC 29745

Phone: (704) 451-8844

Fax: (803) 684-2987

Email: [email protected]

Michael Pluimer, PE

Project Review/Supporting Author

Plastics Pipe Institute

105 Decker Court, Suite 825

Irving, TX 75062

Phone: (469) 499-1049

Email: [email protected]

Word Count: 3,784 written; 2,000 photos/Charts/etc.; Total 5,784 - For AREMA Review May 2010

ABSTRACT

The Plastic Pipe Institute (PPI) sponsored minimum cover testing for 48 inch (1200 mm) diameter

corrugated high density polyethylene (HDPE) pipe under heavy haul railroad loading conditions. The

Transportation Technology Center, Inc. (TTCI), Facility for Accelerated Service Testing (FAST), was

used for the project. The FAST train consisted of four locomotives and eighty 315,000 lbs (143,010 kg)

rail cars. The Fast Train cycled over the two buried 48 inch (1200 mm) diameter HDPE pipes until 101

Million Gross Tons (MGT) (91,607,000 Mg) of rail traffic had been reached. The first pipe was

backfilled with fabric wrapped, vibrated in place #57 stone. The second pipe was backfilled with onsite

ASTM Class III backfill at 94% of Standard Proctor Density (SPD). Each buried pipe run had a 58 inch

(1,473 mm) long instrumented section that was joined in place with a split band coupler and an inline

bell/spigot connection. Joints were placed directly under the rail to ensure maximum loading. Deflection

gauges, strain gauges, circumferential shortening gauges, and joint separation measurements were used to

evaluate the pipe and joints. Each pipe had four feet (1.2 m) of cover from the bottom of tie to top of

pipe. At the conclusion of the 101 MGT (91,607,000 Mg) load cycle, minimal pipe deflection and

negligible joint movement was noted. The final results indicate that the HDPE pipe and backfill behave

elastically, and support that the AASHTO Section 12 minimum cover height calculations are inline with

field evaluated pipe performance. A very interesting portion of the test was the video and data cycles of

the heavy haul live loading and the visual evidence of elastic soil behavior. The test verifies that

corrugated HDPE pipe will withstand heavy haul railroad loading over large load cycles.

INTRODUCTION

For years many railroad designers and contractors have used corrugated HDPE pipe under highway

loading conditions with great success. Many of these designers who have successfully used corrugated

HDPE pipe are now using the product to avoid corrosion caused by stray current in electrified rail and/or

corrosion caused by chemical exposure in railroad mining applications. Because their success with

corrugated HDPE pipe, railroad designers are interested in expanding the use of his product. In order for

them to feel more comfortable with corrugated HDPE pipe, they are seeking field verification of the

product’s structural field performance under heavy haul railroad load conditions. As a result of the

railroad industries interest in Corrugated HDPE pipe, Members of the American Railway Engineering and

Maintenance-of-Way Association (AREMA) contacted Plastics Pipe Institute (PPI) Members and asked

them to assist with the evaluation of corrugated HDPE pipe under heavily loaded track. To accomplish

the requested field testing, PPI with the assistance AREMA selected the Transportation Technology

Center, Inc. (TTCI), and their Facility for Accelerated Service Testing (FAST) as the best facility to

perform the tests. The reason TTCI was selected to conduct heavy haul testing is because of their ability

to simulate long-term heavy haul loading conditions at an accelerated rate at FAST. The FAST train is a

13,000 ton (11,791 Mg) train comprising of four (4) locomotives and eighty (80) 315,000 lbs (143,010

kg) rail cars that run at average speeds of 40 mph (64.4 km/h). The load generated by the FAST train is

similar to a Cooper E-80 load.

CALCULATED FIELD PERFORMANCE

Many studies have been done to determine the performance characteristics of corrugated HDPE pipe in

field applications. However, no in field structural testing has been done for corrugated HDPE pipe under

heavy railroad loading conditions. The intent of the test at TTCI was to pay greater attention to the

details of pipe performance using instrumentation and see how it compared to the previously calculated

performance values. The current industry practice for calculating corrugated HDPE pipe performance is

found in the American Association of State Highway and Transportation Officials (AASHTO), Load

Resistance Factor Design (LRFD), Bridge Design Specifications, Section 12 – Buried Structures and

Tunnel Liners (Section 12). By using set values for corrugated HDPE pipe established in AASHTO M

294 - Standard Specification for Corrugated Polyethylene 12 to 60 inch (300 to 1500 mm), Section 12

allows designers to evaluate how corrugated HDPE pipe is expected to perform under Cooper E-80 loads.

Prior to setting up the TTCI FAST Test, Table I1 was generated by Bill Vanhoose, Advanced Drainage

Systems, Inc. to evaluate which pipe should be installed at FAST and field tested.

TABLE I: Minimum Cover for ADS N-12, N-12 ST, and N-12 WT Pipe with E-80 Loading, ft (m)

Diameter in (mm)

Class 1 Class 2 Class 3 Compacted 95% 90% 95%

4 (100)

2 (0.6)

2 (0.6)

2 (0.6)

2 (0.6)

6 (150) 8 (200)

10 (250) 12 (300) 15 (375) 18 (450) 24 (600)

30 (750) 3

(0.9) 3

(0.9) 3.5

(1.1) 3.5

(1.1)

36 (900) 3.5 (1.1)

3.5 (1.1)

3.5 (1.1)

3.5 (1.1) 42 (1050)

48 (1200) 4

(1.2) 4

(1.2) 4

(1.2) 4

(1.2) 54 (1350) 60 (1500)

Calculation Notes:

1. Results based on AASHTO LRFD Section 12 structural calculations as shown in the Structures

section of the ADS Drainage Handbook. Calculations assume no hydrostatic pressure and a

density of 120 pcf (1926 kg/m3) for overburden material.

2. Installation assumed to be in accordance with ASTM D2321 and the Installation section of the

Drainage Handbook.

3. Material must be adequately “knifed” into haunch and in between corrugations. Compaction and

backfill material is assumed uniform throughout entire backfill zone.

4. Compaction levels shown are for standard Proctor density.

5. Live load calculated at standard E-80 loading which simulates 80,000 lb/ft railway load plus

impact.

Once the pipe table was complete the PPI project team along with AREMA Committee 1, Subcommittee

4 reviewed the data and agreed upon the pipe diameter should be used in the TTCI FAST Test.

PIPE SELECTION

Selecting the appropriate size pipe from among the most commonly manufactured corrugated HDPE

pipes was somewhat of a challenge. By comparing the calculated acceptable minimum fill heights (Table

I), to the resulting safety factors produced by the calculations, and then evaluating field performance

history of various pipe diameters, it was determined that a 48 inch (1200mm) diameter pipe would be the

best pipe diameter to test at TTCI. It represented the largest diameter that is commonly manufactured by

the greatest number of PPI member manufacturers. Also, when the 48 inch (1200mm) pipe is compared

to the performance characteristics of Advanced Drainage Systems, Inc.’s (ADS) 60 inch (1500 mm)

diameter pipe, the ADS 60 inch (1500) corrugated HDPE pipe out performed the 48 inch (1200 mm)

diameter pipe selected for the test.

MEASURED VALUES

From the AASHTO and ASTM Standards there are many performance indicators in determining if HDPE

pipe is performing adequately in the field. From the many indicators available, deflection and strain are

the most commonly used and simplest to field verify.

Deflection

Based on AASHTO Section 30 – Thermoplastic Pipe, the recommended allowable deflection is 5% of the

pipe’s actual diameter. An additional 2.5% deflection is allowed if a designer reviews and verifies that

the system is structural sound. This allows for a total permitted deflection of 7.5%, before standards

recommend that the excessively deflected product is removed. To measure deflection in the TTCI FAST

Test, 5 inch (127 mm) string pots were mounted inside the pipe test sections and oriented to measure

vertical, horizontal and diagonal deflections in real time while the FAST train passes over the top of the

pipes. Diagonal deflection normally isn’t a concern, unless construction loads is excessive, but we added

the diagonal deflection as redundancy to ensure the pipes maintained their proper shape. We also wanted

to measure if there was any negative effect caused by a pressure wave generated by the heavy freight train

as it passed at 40 mph (64.4 km/h) over the pipes. Photo I shows the deflection gauges in place.

Photo I: Placement of strain gauges in test section of 48 inch (1200 mm) pipe

Strain

AASHTO - Section 12 indicates that the maximum allowable strain is 5%. During the TTCI FAST Test

AREMA Committee 1, Subcommittee 4, members wanted to verify that E-80 loading wouldn’t cause

long-term strain in excess of 5%. To ensure this didn’t happen strain gauges were placed at the 12:00,

1:30, 3:00, 6:00 and 9:00 positions. At the 12:00, 3:00, and 9:00 positions strain gauges were placed in

the inside crown of the corrugations (Photo II). The pipe wall was cut away from the inside of the pipe to

allow access for gauge placement. In Photo II you can see the silicone suction cup that was used to help

hold the gauge in place as the bonding agent cured. At the 12:00, 1:30, 3:00, 6:00, and 9:00 positions,

strain gauges were placed on the inside wall of the pipe directly below the valley between corrugations as

shown in Photo III. The strain gauges placed at 1:30 to determine if there was excessive strain due to an

anticipated train pressure wave. In a normal highway loading test, strain would normally never be tested

at 1:30. As illustrated in Photo III, a protective flexible coating was placed over the top of the strain

gauges to prevent moisture from damaging placed gauges. Overall the strain gauge placement and testing

was flawless.

Circumferential shortening

Circumferential shortening is the uniform compression of the pipe wall resulting in shortening of the

pipe’s circumference. It is a phenomenon that typically occurs in deep fill applications. In the TTCI Test

we wanted to see if the excessive loads produced by the FAST Train would cause the corrugated HDPE

pipe to shorten circumferentially. The shortening of the pipe’s circumference isn’t restricted or even

normally a measured value, but since it was so inexpensive to add a few additional string pots to the test

we decided to proceed and collect this data.

Photo II: Strain gauge being placed on the inside crown of pipe

Visual inspection

As with all storm drainage pipe materials, after the 48 inch (1200mm) HDPE pipes were installed, they

were visually inspected and then periodically visually inspected several more times throughout the

duration of the test. The visual inspection was accomplished through the use of hand held digital video

cameras, digital static cameras. TTCI, ADS and PPI staff also entering the pipe and tried to find evidence

of poorly shaped or damaged pipe, specifically signs of flattening, buckling, racking, and/or cracking.

AASHTO recommends that a post installation visual inspection be conducted on all storm drainage pipes.

Post installation inspection has become a key criterion to fight against installer error and/or post

installation damage. During the test, pipe joints were hand measured to determine if there was

movement in the joints.

Photo III: Strain gauge being placed on the pipe’s inside wall over the corrugation valley

TEST SECTION CONFIGURATION AND INSTALLATION STANDARD

Once the appropriate pipe size and gauge type and placement was determined, tested section length, joint

types and backfill selection came next.

Test Section Length

For the purpose of the TTCI FAST Test, it was assumed that the maximum load on the pipe would be

located directly under each of the track’s rails. With the rails set at 59.5 inches (1511 mm) apart a 58 inch

(1473 mm) long pipe test section was selected for gauging and testing. This length of pipe was the best

sice that ensured maximum loading over the weakest part of the pipe, the joints. The 58 inch (1473 mm)

length allowed the pipe to be easily cut in the pipe valley between corrugations. It also allowed a pipe

joint to be placed as close to center of rail as possible.

Joint Selection

Because of concern over joint separation due to transverse soil movement or pipe settlement, the railroad

industry traditionally have used extremely good pipe connections to help hold the storm drainage system

together. The movement pipe that causes joint separation is typically tied to the use of poor bedding and

backfill. In order to see if joint pull apart strength really makes a huge difference in pipe performance, at

TTCI we reversed the normal philosophy and used good backfill with non-constrained standard joints.

The typical watertight bell and spigot joint was used for one side of the test section and a typical plain end

pipe with a fabric wrapped split coupler connected with nylon zip ties was used on the other end of the

test section.

Bedding and Backfill

After reviewing the minimum covers listed in Table I, AREMA and PPI agreed that the test sections at

FAST should be backfilled with a fabric wrapped, vibrated in place #57 stone and an onsite ASTM Class

III soils compacted to ASTM 95% Standard Proctor Density. Both of backfills are considered to be good

backfill, which is what should be used when installing storm drainage pipe under track.

Final Placement of Test Sections

Pipe 1, which is the pipe that was closest to the bridge at FAST, was approximately 60 feet (18.3 m) long

with a 58 inch (1473 mm) long 48 inch (1200 mm) diameter section of pipe located directly under the

rails. Pipe 1 was backfilled with local #57 Stone meeting the requirements of an ASTM Class I soil type.

The #57 stone was wrapped with a Propex 1201 geotextile fabric to prevent migration of fines into the

select backfill. To compact the #57 stone it was vibrated in place with a jumping jack in approximate 15

inch (381 mm) loose lifts after the pipe was set on loose bed of #57 stone. Refer to Drawing I for the

proposed schematic of the FAST test section configuration.

Pipe 2, approximately 20 feet (6.1 m) from pipe 1, was placed with same configuration as pipe 1, except

that the pipe was set on a ASTM D2321 Class III onsite soil and then compacted to 94% standard proctor

density (SPD) using a jumping jack. Soil lifts were installed in 8 inches (203 mm) loose lifts.

Once stone or soil backfill reached 6 inches (152 mm) over the top outside of the pipes, backfill was

placed with a remote control vibratory roller compactor in 8 inch (203 mm) loose lifts to 18 inches over

the top of the pipes. At this point a large loader and asphalt grade vibratory roller were used to complete

compaction of the onsite ASTM Class III soils to finish subgrade. The same heavy equipment was used

to compact the sub ballast as well. Compaction tests verified that the on site Class III backfill was placed

at 94% SPD along the sides of pipe 2 and at 98% SPD over the top of the pipes to finish subgrade.

Finally 9 inches (229 mm) wide, 102 inches (2591 mm) long wooden ties at 19.5 inches (495 mm) on

centers were placed on 12 inches (305 mm) of AREMA #24 ballast that was vibrated in place using

typical ballast placement equipment. Photo IV and V show the placement of final backfill and the

equipment used to vibrate ballast into place.

Photo IV: Placement of ASTM Class III soil over TTCI FAST 48 inch (1200 mm) test sections

Drawing I: TTCI FAST proposed 48 inch (1200 mm) Test Sections

ON SITE SOILS

A sample representing the FAST onsite soils was provided by Joe LoPresti the FAST Project Manager.

The onsite soil type was non plastic silty sand, Unified Classification SM. Practical size gradation100%

passing #4 sieve, 99.8 passing #10 sieve, 94.6 passing #20 sieve, 78.5 passing #40 sieve, 54.5% passing

#100 sieve, and 36.6 passing #200 sieve.

Photo V: Equipment used to place vibrate AREMA #24 ballast into place

PIPE SAMPLE TESTING

As a part of the test, remnant samples were saved from the cut ends of the tested samples and taken to

Texas Research Institute (TRI) for verification of AASHTO M 294 material property verification.

TEST DURARION AND DATA COLLECTION

After the pipe was placed and the tracks were set to grade, the FAST train commenced operations over

the test sections. Operations continued until 101 MGT (91,607,000 Mg) had passed over the pipe. Two

dynamic data sets were taken during the test. The first was taken just after the first MGT (907,000 Mg)

passed over the pipes and the second after 96 MGT (87,072,000 Mg) passed over the system. Both data

sets lasted for 24 hours. In addition to the two dynamic data sets, two static tests were recorded during the

test. The first static test was taken in the lab to zero out the instruments and the second static test was

taken just after construction.

To evaluate the effect of parked rail cars over the test samples, TTCI parked two 315,000 lbs (143,010

kg) rail cars over the test sections such that two axles were set directly over the top of the buried pipes.

The parked cars sat over the pipes for approximately 3 months.

Multiple visual inspections were made of the inside of the pipe, and joint movements were hand measured

throughout the test.

TEST RESULTS

Pipe Material Properties

The pipe test samples taken to Texas Research Institute met the requirements of AASHTO M 294 testing

standard. The pipe test samples were from separate production runs yet had similar material properties.

Deflection

Deflection was measured in the 0 degree (vertical), 45 degree (diagonal), 180 degree (horizontal), and 315

degree (diagonal) locations. The maximum deflection values listed below2 are based on an actual

measured diameter of 47.6 inches (1,209 mm) pipe.

Maximum horizontal deflection after construction = 0.7 inches (17.8 mm) (+1.5%).

Maximum horizontal deflection after 96 MGT = 0.6 inches (15.2 mm) (+1.3%)

Maximum vertical deflection after construction = 0.1 inches (2.5 mm) (-0.2%)

Maximum vertical deflection after 96 MGT = 0.5 inches (12.7 mm) (-1.05%)

Maximum deflection due to 96 MGT loading alone = 0.065 inches (1.7 mm) (0.14%)

Deflection results are all well below the allowable AASHTO Section 30 recommended construction limit

values of 5.0% deflection. As anticipated, the diagonal deflections were much less than vertical or

horizontal deflection values.

Strain

The maximum strain (compressive) caused by construction loading was 7,300 microstrain (0.73% strain).

The maximum strain (compressive) caused after 96 MGT (87,072,000 Mg) from the FAST train

combined with construction loading was 8,800 microstrain (0.88% strain). These values are well below

the AASHTO Section 12, 5% recommended strain limit.

Circumferential Shortening

The maximum circumferential shortening from construction was 0.4 inches (10.2 mm) (0.27%), and after

96 MGT (87,072,000 Mg) circumferential shortening was 0.8 inches (20.3 mm) (0.54%). Although

circumferential shortening doesn’t have a set limit, when compared to set strain and deflection limits, the

measured values are very acceptable.

Joint Separation

Minor movement was noted in the joints; however, maximum joint movement was below 0.25 inch (6.35

mm). With temperatures varying over 80˚F (44.4˚C) outside the pipe, this could have been caused by

pipe elongation and contraction due to thermo expansion. All joints looked and were performing

extremely well with no notable visual variation from the beginning to end of the test.

Elastic Soil Reaction

Figure I2 illustrates dynamic strain and deflection during 40 mph (64.4 kmph) train operations, evidencing

how the buried pipes at TTCI FAST reacted elastically. As load is applied, the structure compresses and

then as load is removed it rebounds and eventually recovers to its original state. Through the strain and

deflection data in can also be inferred that the soil and track also reacted elastically.

Figure I: Dynamic strain and deflection during 40 mph (64.4 kmph) train operations

Ride Quality

Locomotive engineers who operated the FAST train during the test period reported that ride quality over

the pipes was satisfactory. No track geometry maintenance was required at the test site at the conclusion

of the test.

Static Parked Rail Car Loading

There was some movement noted at the track level while the rail cars were parked over the HDPE pipe;

however, when the train was removed the rail rebounded back to its original location. This appeared to be

an elastic soil reaction due to the rail car over burden, and not pipe degradation.

CONCLUSIONS

Corrugated HDPE pipe per AASHTO M294 performed as anticipated thought the TTCI FAST Test and

proved excellent service for the 101 MGT (91,607,000 Mg) Test with no signs of significant system

degradation. Test results indicate that when AASHTO M294 pipe is installed per ASTM D2321

guidelines with calculated fill heights based on AASTHO Section 12, corrugated HDPE pipe will perform

extremely well in railroad applications and is a viable material for storm drainage conveyance for Cooper

E-80 loading conditions. From the FAST Test it also appeared the buried HDPE pipe and soil backfill

exhibited elastic when exposed to heavy loading conditions. With soils reacting elastically it is

recommended that in areas where backfill is likely to be saturated and subject to high loading conditions

that only non-plastic backfill be used around all gravity flow storm drainage systems.

ACKNOWLEDGEMENTS

I would like to thank the testing team at TTCI for their professional and friendly services they provided

during the test. Their level of expertise evidenced throughout the testing procedures clearly elevated the

quality of data obtained during testing. I am also grateful Members of AREMA Committee 1,

Subcommittee4 for their input on testing protocol.

REFERENCES

1. Bill Vanhoose, PE, Advanced Drainage Systems, Inc., Structures and Materials Engineer, January

2009, Data table contained in final FAST Letter Report No. P-09-052 Protocol.

2. Corrugated Polyethylene Pipe Testing under 315,000-Pound Cars at FAST

Letter Report No. P-09-052, Prepared for Plastic Pipe Institute, by Joseph A. LoPresti

Transportation Technology Center, Inc., Revised January 14, 2010.

LIST OF TABLES, DRAWINGS, FIGURES, AND PHOTOS

Table I: Minimum Cover for ADS N-12, N-12 ST, and N-12 WT Pipe with E-80 Loading, ft (m), Pg. 4

Drawing I: TTCI FAST proposed 48 inch (1200 mm) Test Sections, Pg. 11

Figure I: Dynamic strain and deflection during 40 mph (64.4 kmph) train operations, Pg. 15

Photo I: Placement of strain gauges in test section of 48 inch (1200 mm) pipe, Pa. 6

Photo II: Strain gauge being placed on the inside crown of pipe, Pg. 7

Photo III: Strain gauge being placed on the pipe’s inside wall over the corrugation valley, Pg. 8

Photo IV: Placement of ASTM Class III soil over TTCI FAST 48 inch (1200 mm) test sections, Pg. 10

Photo V: Equipment used to place vibrate AREMA #24 ballast into place, Pg. 12