Appendix E

Report to: Shakespeare

Verification Tests of Distribution Size Fiberglass Poles

July 1993

1. INTRODUCTION

The Shakespeare Products Division contracted Engineering Data management (EDM) of Fort Collins, Colorado, to conduct independent verification testing of the structural capabilities of distribution-sized fiberglass utility poles. EDM has over 11 years of experience with full scale product testing and development for the utility industry. Testing was conducted at the Colorado State University Engineering Research Center on July 13-14, 1993.

At the request of Shakespeare, EDM developed a structural testing plan in cooperation with Montana Power Company and Southern California Edison to evaluate the field performance of fiberglass poles. The performance of these fiberglass poles is reported and evaluated relative to their wood pole counter parts. Since the design of single-pole structures is typically controlled by the wind load bending capacity, it was determined that full scale bending tests needed to be performed to evaluate the bending capacity of the poles. Special consideration also needed to be given to the performance of the fiberglass pole at framing connections. These considerations led to the development of a testing program which included full scale bending tests, vertical load tests and torsional load tests. The destructive bending tests were performed on full-length
sections of 40-ft. fiberglass poles. The vertical and torsional load tests were performed on sections cut from the tops of poles after they had been tested in bending. The vertical connection capacity was evaluated utilizing both standard wood and steel framing supplied by utilities. The torsional connection capacity was assessed using wood framing to evaluate broken conductor performance.

2. BENDING TESTS

Test Procedure

Bending tests were performed following the procedures specified in ASTM D 1036, "Standard Methods of Static Tests of Wood Poles".(Note 1)  A plan view of the test facility is shown in Figure E1, and photographs of poles under load are provided in Appendix EB.

Practical considerations require that a pole be tested in the horizontal position. The pole butt support and clamping system used for the bending tests consists of a reinforced  concrete base containing wood faced saddles and a 50-kip jack. The 12 ton concrete base is rigidly attached to an 18 inch thick reinforced concrete floor. The jack is used to clamp the base of the pole through two steel saddles with wood bearing surfaces. The curved shape of these saddles approximates the circular cross section of the pole.
__________________
(Note 1) 1993 Annual Book of ASTM Standards, Volume 04.09 - Wood, American Society for  Testing and Materials, 1916 Race St., Philadelphia, PA.

While this clamping system is rigid in the bending plane, the lack of support in the area of the neutral plane approximates a low density or loosely compacted soil condition.

Prior to testing, each pole was placed in the concrete base so that its groundline coincides with the edge of the support saddle, as shown in Figure E1. The poles were supported at approximately mid-length by a steel pipe support to eliminate the dead weight effect and bending outside the plane of loading. Sliding friction between the pole and pipe supports was reduced by use of a lubricated Teflon strip between the pole and the support.

Each pole was loaded through a cable attached to the pole at 2 ft. below the tip. Load was applied through a 10-kip electric winch with a variable speed control to maintain the desired deformation rate. The winch is attached to a frame that is supported by rollers which permit the assembly to align itself during loading, thus compensating for the longitudinal component of the pole tip deflection and ensuring that the loading direction remains perpendicular to the original pole axis.

The transverse deflection of the pole was measured by a helical potentiometer. Load and deflection data were collected at 1/3 second intervals via a computerized data acquisition system. These data were later analyzed to determine the bending capacity and elastic modulus of each pole.

Destructive Tests

Five 40 ft. (Class 4 wood equivalent) Shakespeare fiberglass poles were tested to destruction using the above test configuration. Of these five poles, three were in the standard configuration (Specimens #1, #4 and #5) and two (Specimens #2 and #3) had their groundline sections inserted into steel culverts with a structural foam placed in the annulus between the culvert and the fiberglass pole. These latter two tests were performed to model an actual field foundation proposed by Montana Power Company for use in a remote location requiring helicopter placement and excavation through blasting.

During testing the poles were oriented such that the bolts for the steps would fall on the extreme compression and tension faces. This orientation results in the maximum reduction in section modulus due to the bolt holes in the bending plane. Load was applied to the poles at a rate so as to develop the standard strain rate of approximately 0.001 in./in./min. used in testing wood poles.

The first test of a standard pole (Specimen #1) resulted in a below groundline transverse crushing failure of the butt section of the pole in the wooden cradle used to secure the pole for testing. The maximum load sustained by the pole for this mode of failure was 1610 Ibs. This type of failure occurred due to compression forces transverse to the pole's longitudinal axis. `As this failure did not occur in a bending mode, the internal (hoIlow) portion of the pole was stiffened to prevent crushing and the pole was retested. The specimen was stiffened through blocking the internal portion of the pole to prevent transverse crushing by the clamps. For this test both the groundline and butt sections were blocked to insure a true bending test of the pole by avoiding transverse crushing of the hollow pole in the clamps. Upon retest, pole specimen #1 held 2730 Ibs. which exceeds the class 4 load capacity of 2400 Ibs. specified in the American National Standards Institute (ANSI-O5.1) Specification for Wood Poles. (Note.2)

Prior to testing, specimen #4 was stiffened at the butt with no reinforcing placed in the groundline region. In this configuration the pole held a maximum load of 1730 Ibs. and failed at groundline. As a result of this failure location, there was no ability to stiffen the groundline portion and retest this specimen. The final standard pole, specimen #5, had blocks placed inside the pole at groundline and at the butt. In this configuration the pole achieved a load of 4400 Ibs. which approximates the class 1 requirement for wood poles of 4500 Ibs.

Two poles were tested after they had been encased in steel culverts according to Montana Power Company procedures. The use of these culverts restrained deformation of the butt section and forced-the poles to fail in bending at or above the groundline. Consequently, these poles demonstrated ultimate capacities significantly higher than some of the standard poles. Test specimen #2 demonstrated an ultimate capacity of 4120 Ibs., well above the 2400 Ib. capacity specified in the ANSI 05.1 specification for class 4 wood poles. This specimen failed 128 in. above groundline. The second encased pole (Specimen #3) performed similarly with an ultimate capacity of 4530 Ibs. and failed at groundline. The ultimate bending loads attained by all five poles are provided in Table E2 along with a brief description of the failure mode.

The results of the bending tests may be compared to ANSI-05.1 wood pole class load values. These class loads are provided in Table E1. Load vs. Deflection graphs are provided for each bending test in Appendix EA.

Table E1. ANSl-05.1 Designated Loads for Wood Poles

Table E2. Bending Capacity Results for Shakespeare Fiberglass Poles

** AGL = Above Groundline
     BGL= Below Groundilne

Table E3: Effective Fiberglass Pole Elastic Modulus Values for Direct Comparison with ANSI 05.1 Wood Poles

**Values must he adjusted using appropriate conditioning factors:
  1.) Kiln Drying                             0.90
  2.) Boullonizing                            0.90
  3.) Steam conditioning                 0.95 (Southern Pine Only)

Appendix C to the ANSI 05.1 specification gives mean Modulus of Elasticity (MOE) values for various species of wood poles. These elastic modulus values are based on ANSI minimum class dimensions for each species. To compare the stiffness of the fiberglass poles to the stiffness of wood poles, it is necessary to calculate the MOE of the fiberglass poles based on actual deflections and the ANSI 05.1 minimum dimensions for a given species.

Stiffness information for the fiberglass poles is provided in Table E3. The values in Table E3 are presented to allow comparisons to be made between the fiberglass poles and wood poles made from the most common wood pole species: Douglas Fir, southern pine, and westem red cedar. The effective ANSI-based Modulus of Elasticity (AMOE) values are calculated from the actual deflections of the poles, and consider the tested pole to be a solid pole with ANSI 05.1 minimum class dimensions. This approach allows direct comparisons between the effective AMOE of the fiberglass poles and the published modulus of elasticity values for individual species of wood poles in the ANSI 05.1 wood pole specification.

As the ANSI 05.1 specification gives different minimum class dimensions for various species of wood poles, three different effective AMOE values are calculated for each fiberglass pole to allow a direct comparison with ANSI-based MOE values for Douglas fir, southern pine, and Westem red cedar poles. The modulus of elasticity values specified by the ANSI 05.1 specification for wood poles are also provided in Table E3 to allow comparison to the effective modulus of elasticity values from the fiberglass pole tests. It is important to note that adjustments to the ANSI MOE values, detailed in the standard, are required if the wood poles have been kiln dried, boultonized or steam conditioned.

The Modulus of Elasticity values presented in Table E3 indicate that the fiberglass poles are nearly as stiff as wood poles. A wood pole with ANSI southern pine minimum class dimensions would exhibit the average deflection of the fiberglass poles if it had an elastic modulus of 2.36 x 10^e psi. The average elastic modulus value given in the ANSI 05.1 specification for southern pine poles with ANSI minimum dimensions is 8 x 10^6 psi. The average fiberglass pole is therefore 88% (2.36/2.68) as stiff as a southern pine pole. Similar calculations indicate that on the average fiberglass poles are 116% as stiff as western red cedar and 70% as stiff as Douglas fir.

3. VERTICAL LOAD TESTS

To evaluate the capacity of connections made to the fiberglass pole, a test section ten feet long was removed from a pole which had been previously tested to destruction in bending. Although the pole had been previously tested, failure occurred near the groundline region leaving the top  portions undamaged. To simulate field conditions the pole section was clamped in a loading frame in a vertical position. Vertical loads were applied to this top portion of the pole via a braced wood crossarm. The test was then repeated using an unbraced steel arm.

When testing the wood arm system, an unbalanced vertical load was placed on the wood arm by applying load to only one side of the crossarm. This load was gradually increased to a maximum of 2,000 Ibs. After testing was completed, the crossarm was removed from the fiberglass pole and the connection was inspected for local deformation. No deformation of the fiberglass at the bolt hole was observed.

In a similar manner, the vertical capacity of the connection between the fiberglass pole and a steel davit arm (gooseneck) was tested. Load was uniformly increased until the steel arm failed. Inspection of the connection to the fiberglass pole indicated minimal deformation of the fiberglass at the bolt hole. The maximum vertical load sustained by this structural configuration was 3650 Ibs. Load versus time graphs for both vertical load tests are provided in Appendix E-A.

4. TORTIONAL LOAD TESTS

To evaluate the effects of torsion on connections between the fiberglass pole and framing hardware, a wood braced crossarm was attached to a 10 ft. long section removed from the top of one of the bending test poles. The base of the pole section was clamped in the load frame so that the pole was vertical and a torsional load was applied through the unbalanced horizontal loading of the wood crossarm. This loading configuration simulates the torsional load applied to the pole due to a broken conductor.

The fiberglass pole withstood a torsional moment of approximately 6,075 ft.-lbs. resulting from a 2480 lb. load on the crossarm which extended a distance of 2.45 ft. beyond the pole's longitudinal axis. Application of further load resulted in excessive deformation of the bolt which connected the crossarm to the pole. A maximum horizontal load of 2550 Ib caused rotation of the pole in the load frame. Appendix E-A contains the load versus time graph for this test. Inspection of the connection between the crossarm and the pole indicated that minimal deformation occurred in the fiberglass portion of the connection and that deformation of the framing was due to the bending of the through bolt connecting the crossarm to the pole.

5 SUMMARY

Verification testing of the Shakespeare fiberglass poles under the structural testing plan detailed within indicates that their being performance meets or exceeds the requirements specified for wood poles. Since the below groundline clamping system used in the testing of wood poles does not confine the pole completely about its circumference, internal stiffening of the hollow groundline and butt sections of the fiberglass poles was necessary to achieve bending failure. Such stiffening may not be necessary for poles encased similar to Montana Power Company's specific application or in highly compacted or heavy soil conditions.

To evaluate connection capacity, vertical and torsional load tests were performed. These tests indicate that the fiberglass portion of the connection has adequate structural capacity for use with the framing components tested. In both cases, failure of the hardware occurred before deformation of the fiberglass at the connection was observed.

_______  

Appendix E-A

Bending Test

Load vs. Deflection Graphs

Vertical and Tortional Test

Load vs. Time Graphs

	Appendix E - Page 11 Graph: Bending Test 001
	Appendix E - Page 12 Graph: Bending Test 01R
	Appendix E - Page 13 Graph: Bending Test 002
	Appendix E - Page 14 Graph: Bending Test 003
	Appendix E - Page 15 Graph: Bending Test 004
	Appendix E - Page 16 Graph: Bending Test 005
	Appendix E - Page 17 Graph: Vertical Load Test Wood Crossarm
	Appendix E - Page 18 Graph: Vertical Load Test Steel Davit arm
	Appendix E - Page 19 Graph: Tortion Test

August 31, 1993  Mr. Lynn Derrick  V.P. Sales and Marketing  Shakespeare Co. 19845 Highway. 76  Newberry, SC 29108  Dear Lynn:  I am providing a letter report, 35 mm slides and 1/2" VHS video of the fiberglass crossarm testing. As is illustrated in the load-time curve, the maximum load failure was 4,790 pounds.  As shown in the video, every attempt was made to test the crossarm in cantilever bending with a tension brace opposing the load. However, due to the extreme load encountered in the brace due to the brace angle, failure of wood braces, steel brackets and even the crushing of the pole in the clamp was encountered prior to reaching the load necessary to fail the arm in bending. As a result, the crossarm was supported by a cable placed vertically and attached to the test frame. This test configuration provides a successful means of attaining true bending failure in the fiberglass arm.  As expected, the arm failed in local buckling at the through-bolt connection to the pole. The slight drop in load shown on the plot illustrates down slotting of the through bolt in the crossarm. For comparison purposes, a 5 x 5 (actual dimension) solid-sawn Douglas fir crossarm with a 3.5' lever arm (distance from through bolt to load application) and 7400 psi average fiber strength would theoretically fail at approximately 4280 pounds of load. Thus, the fiberglass arm would support more load and likely exhibit significantly less material variation.  It is important to note that typical distribution-size wood crossarms are 3-1/2 x 4-1/2 x  8'. The size and uniform load capability of the 5 x 5 fiberglass arm make it well suited to resist longitudinal loads equally as well as vertical loads. Once utility load requirements are identified, fiberglass crossarms can be easily designed for dead-end and heavy-angle applications.    Sincerely,  Robert F. Nelson  Project Manager

[Pages 22 to 24 of the Appendix E report contain 1 graph each, accessible as independent graphics by clicking the icons below - Webmaster]

	Appendix E - Page 22 Graph: Shakespeare Fiberglass Crossarm
	Appendix E - Page 23 Graph: Class 4 Bending Strength
	Appendix E - Page 24 Graph: Class 5 Bending Strength

---

Shakespeare Composite Structures

19845 Highway 76, Newberry, SC 29108 · 803.276.5504 Fax: 803.276.8940
www.skp-cs.com