The best method is to steam the ribs and then bend them so that when they dry they will retain their curve and not tend to push the horizontal beams apart. Only a very slight curve should be given and the amount of curvature should be the same for all the ribs.
Some designers construct gliders having flat planes, intending that the pressure of the air underneath the fabric shall produce a natural curve but such a method is exceedingly poor practice and results in a very inefficient machine.
The ribs must be perfectly rigid and the frame of the whole machine strongly trussed so that it cannot possibly be distorted by the air pressure. The following extract from the report of the Smithsonian Institute well illustrates this point.
"This new launching piece did its work effectively and subsequent disaster was, at any rate, not due to it. But now a new series of failures took place, which could not be attributed to any defect of the launching apparatus, but to a cause which was at first obscure; for sometimes the aerodrome, when successfully launched would dash down forward and into the water, and sometimes (under apparently identical conditions) would sweep almost vertically upward into the air, and fall back although the circumstances of flight seemed to be the same. The cause of this class of failures was finally found in the fact that as soon as the whole machine was up-borne by the air, the wings yielded under the pressure which supported them, and were momentarily distorted from the form designed and which they appeared to possess. "Momentarily," but enough to cause the wind to catch the top, directing the flight downward, or under them, directing the flight upward, and to wreck the experiment. When the cause of the difficulty was found the cure was not easy, for it was necessary to make this great _sustaining surfaces rigid_, so that they could not bend.”
The report in question refers to the experiments conducted with Professor Langely’s model aerodrome.
Some experimenters claim that the parabolic curve gives the greatest lift with the least power required for propulsion but it can be safely doubted. The Wright machine is probably the most efficient in existence. Their curve is very nearly the arc of a circle and is not of the parabolic form.