Selected product reference excerpt. Search a specific planning question for relevant details.
# Offset Tension Rod in Aluminum Pergola Beams > **Shop and engineering record.** This is the method used inside long aluminum tubes when deflection, not stress, is what limits the beam. Fiberglass beams get their crown at pultrusion. Aluminum tubes do not. A stainless rod set below the center of each tube does that job after the rafters and purlins are on, and the nuts stay reachable so the pull can be checked later. The Bedford summer house (Katonah, New York) is the worked example below. Project numbers live in `Martha Engineering.xlsx`. Do not copy the pounds or the foot-pounds onto another span without rerunning that sheet. ## Overview One 5/8-11 rod of 18-8 stainless (304) runs inside each aluminum 2×8, low in the tube, unbonded. Removable 1/2 inch 6061 plates sit on the cut ends of the tube. Nuts on the outside pull those plates against the wall ends, so the rod goes into tension and the tube goes into compression. The rod is not on the center of the tube. Compression below the center hogs the beam upward. Jack after the rafters and purlins are on, and stop when that hog has taken out the dead-load droop. Later snow or wind still bends the beam, but it starts from flat instead of from a sag, and the rod picks up some extra force as the bottom fiber lengthens. The rod does not make the aluminum much stiffer. EI barely changes. The benefit is the counter-moment, rod force times the distance below center. On a deflection-governed beam that is enough to matter. On the Bedford 24 foot tubes it took the beam from 94 percent of the deflection limit to 64 percent, and it took the rest sag from 0.351 inch to flat. Nothing is welded. A weld at the end of a 6061-T6 tube softens the temper in the heat-affected zone, and the load path does not need one. ## Key Concepts - One unbonded 5/8-11 rod of 18-8 (304) inside each tube of a double aluminum beam, not one rod for the pair - The rod sits low. A longer moment arm does the same hog with less force - Plates are 1/2 inch 6061, with a relief 1/8 inch deep and as wide as the wall, so they center on the tube and come back off - Nylon washer against the aluminum, stainless washer under the nut, Lanacote on every aluminum-to-stainless contact - Jack with the roof weight already on the beam. Match the measured rise. Do not chase a high torque number - The useful torque on the Bedford beam is about 32 ft-lb at a nut factor of 0.18. Eighty foot-pounds can reach or pass the rod's working table once Lanacote makes the threads slippery - Start on the stock 1/8 inch aluminum wall. The engineering workbook says whether that tube is strong enough. If it is not, upgrade the wall to 1/4 inch. The rod is not a substitute for that upgrade - Stresses stay far below the 35 ksi yield of 6061-T6, so the beam springs back. The rod cuts droop. It is not there to stop a yield that these loads do not cause - Cold weather relaxes the pull, on the order of 20 lb per degree, because aluminum shortens more t
Reference documents may contain historical pricing, earlier configurator links, or technical claims needing review. Use quote tools for current estimates and returned links for designs. If sources disagree, identify the specific conflict and refer it to Baldwin for a case-by-case decision. Do not use reference span tables as project-specific engineering approval.