Lightning Plays the Stadium

lightning plays the stadium

Grounding a giant structure through its steel isn't so simple when the roof is made of moving parts.

By Andrew Cheng

In the recent creation of dynamic, mobile structures and kinetic architecture, the worlds of architecture and machinery have collided. Thus, a new call for techniques and standards has arisen for proven methodology and applications.

The starting point for this new architecture has been manifested over the last few years in retractable roofs for stadiums. These roof structures have no fixed connections to ground; instead they have thousands of moving contact points. Linkages, spherical bearings, roller bearings, bushings, and wheels have replaced bolts, weldments, and concrete. Given that there is no precedent in moving 14-million-pound civic structures, designers have had to reexamine construction standards.

One item of investigation was proper electrical grounding. Traditionally, for a building or a fixed-roof stadium, lightning rods are grounded to the roof steel, the roof steel connected to columns, and the columns connected to a ground grid. In a static connection, it wouldn't really matter if lightning hits steel and a connection welds up. But if that building were a machine, it is likely that the current will find its way to ground via bearings or the machine's electrical supply. A lightning strike could potentially result in a burr on bearing surfaces, a seized bushing, or the loss of electrical control from the surge.

A mockup in a lab tested a copper shunt ground for the rolling system of the roof over Reliant Stadium in Houston.

In the cases of Enron Field, home to the Houston Astros, and Reliant Stadium, where the Houston Texans NFL franchise will begin play this coming August, grounding became a priority. Those two stadiums feature massive retractable roofs in a city with harsh climate and adverse weather. The roof's mechanization designers, at Uni-Systems Inc. in Minneapolis, decided to study the effects of an electrical impulse through a machine-bearing surface.

Beginning in 1997 with the transporter system requirements for Enron Field, the problem confronted was the issue of getting a proper ground between the top of the stadium and the earth. Enron Field was designed as two conventional buildings, one above, and one below, connected by a five-foot swath of actuating mechanism. This mechanism provided not only actuation of the roof, but also the proper load distribution that made conventional building techniques possible.

The actuating mechanism provided load distribution through a system of independent wheels floating within a steel cavity, supported on all sides by plastic bearings and a urethane suspension spring. This resulted in the upper building (roof) and lower building (rail supports and bowl) being electrically insulated from one another by hundreds of smaller nonconductive contact points. Two issues arose: how to connect these two and how to get a lightning strike from a rolling object to a fixed rail.

Uni-Systems was confident that implementing conventional copper shunt grounding to bypass the nonconductive points would solve one problem. However, once the path was shunted, current was routed to the wheels and bearings. Testing would be required to identify the effects of a lightning strike in a greased bearing.

The test required a facility capable of 3,000-ampere testing for a proper simulation of a lightning impulse. Uni-Systems sent a test assembly to the National Electric Energy Testing Research & Applications Center, or Neetrac, in Forest Park, Ga.

The test assembly consists of two tapered roller bearings, an axle, a spacer, and two end caps installed to a round mechanical tubing machined to house the bearings. The assembly was supported at the end caps and insulated from the support frame by ultrahigh molecular weight, or UHMW, plastic plates. The test frame was also built to simulate loading of the bearings, as it would be in its final application. A 100-ton hydraulic ram applied a radial load of 150,000 pounds to the bearings via a platform welded to the bearing housing. The amperage pulse applied to one side of the frame is taken out on the other side by the bearing.

The test unit was subjected to what was considered a series of worst-case scenarios for a lightning strike. Then the unit was disassembled and inspected for damage. A total of four 3,000-ampere lightning impulses were applied. Each impulse had a measured rise time of 8 microseconds followed by a fall to half value in 20 microseconds.

The assembly was to be inspected both for deterioration and grease consistency. Disassembly of the frame revealed that the axle, the outer races, and the seals had no damage. Also, the grease in the cavity between the retaining rings showed no noticeable change.

The inner races were inspected, and appeared unmarked and able to turn freely on the rollers. The conclusion: The electric impulse of a magnitude of 3,000 amperes had no adverse effect, short or long term, on the wheel bearings. It was thus acceptable to use the wheel, and the greased bearing, as part of the grounding path for lightning protection.

Reliant Stadium combines the disciplines of civil and mechanical engineering, and looms large enough to beg for the lightning in Houston.

Later, in 2000, new problems arose for the Reliant Stadium transporter system. This configuration has a linkage system between the spanning roof trusses and the supporting transporter wheels. This maintenance-free hinge included a spherical bearing with a liner of very thin polymer coating.

The concern was that the high current from a lightning strike would penetrate the coating and damage it permanently. The previously designed shunting system would not alleviate concerns. Even if most of the current would bypass the bearing, would the residual heat be detrimental to the polymer lining?

Uni-Systems returned to Neetrac to test a new assembly, consisting of a plain spherical bearing with fibriloid liner, an axle, and spacers, installed to an eye bracket machined to house the bearing and a clevis bracket. The eye and clevis brackets were bolted together through a pair of 1-inch-thick end plates with threaded rods. Two shunt wires were mounted between the clevis and eye brackets. The assembly was insulated from the end plates by 3-inch-thick UHMW plastic plates.

As with the previous test, the objective was to subject the stadium test unit to what was considered a series of worst-case scenarios. A series of four impulses were administered at 5 kA, 10 kA, 15 kA, and 20 kA. After each test, the assemblies were taken apart. The axle was removed and the clevis bracket lifted off to provide access to the bearing. The inner race was turned and rotated to expose the inner surface of the outer race where the fibriloid liner was located.

After each test, in all cases, the inner race was free to move and turn, and the bearing's performance was not affected.
The tests showed that with shunt wires installed, a current with up to 20,000 amperes passing through the hinge has no adverse effect—either short or long term—on the hinge bearing. Thus, the hinge bearing assembly can be used as part of the grounding path for lightning protection.

Andrew Cheng is the senior mechanical engineer for Uni-Systems Inc. in Minneapolis.