John Gaertner and I made 2 sets of tools. The first form block was made from a 3" wide piece of 3/8" aluminum and was a copy of the shape of Cliff's tool I photographed. We quickly found a couple things. First the inside curve, that the bottom roller follows to allow putting a load on the forming roller, must be a very smooth curve or the roller stops on every low spot. Next we found the spring back of 5052-H32 was much more that the old tool's curve would form to the airfoil shape needed. The curve came out very flat.
Also on the first tool we used about a 30 degree angle on the edge of the roller instead of 45 degrees like on the old tool. It needs about 45 degrees to get the bend started. The outer tips of the roller start bending the metal down so it can fit into the channel shaping area of the roller. The last thing we learned was the spacing between the rollers had to increase so that the angle relative to the surface of the form block was tipped more from vertical as on Cliff's tool. If the angle is to close to perpendicular you can't get enough leverage to form the channel.
For our next effort John got a 4" wide piece of aluminum. This allowed us to decrease the radius of the curve to better allow for the spring back. If I did it again I would use a 6" wide piece because we ran out of room at the ends. The wider the piece the more it costs so we tried to keep it narrow since we had no idea how quickly we would get the shape right. This tool took some reshaping but we got it to work by making one pass and then turning the part end for end so we could use the tool to add a little bend in one area. The second roller incorporated what we learned from the first and we made the main roller a little larger in diameter. At that point we could make parts but the engineer in me was not willing to stop experimenting so John had his tool and I went off excited to try out more ideas.