First of all, it's critical that the cam belts be replaced every 12,000 miles. Space limitations on motorcycles require the use of smaller diameter pulleys that cause the belt to flex more than large pulleys used in automobile engines. Further, Ducati uses a small diameter back-side belt-tensioning idler pulley arrangement that causes the belt to flex in the opposite direction on each revolution. This design approach results in an even greater angle of belt flexing requiring the use of a stronger reinforcement fiber to prevent fatigue failure. The original drive belt design often failed before the first 12,000 mile replacement interval so Ducati switched to a Kevlar fiber reinforced belt.
Itís also important to inspect the idler pulley bearings at each change and also a good idea to replace the belt tensioner locknuts.
Second, Ducati tells us that the proper belt tension is critical to avoid over-tensioning the belts that can lead to premature failure. No one argues with this.
I think itís interesting to note that the Ducati tool for checking the belt tension on the 4V motors has changed for the testastretta. The testastretta engines, as well as all new dual spark engines, use a harmonic tester that can only be used with the Mathesis diagnostic tool. Also, Ducati now specifies different tension specs for the vertical and horizontal cylinders.
The main reason for tensioning in the first place is to prevent the drive belt from jumping a tooth or walking off itís pulleys. I would think that the minimum tension to prevent this is well below Ducatís specs. So why are they so picky about it?
My guess, for what its worth, is that they canít control the manufacturing specs close enough on their supplierís belts. The suppliers are able to supply belts with adequate tensile strength and at the dimensional tolerances needed, but thereís probably a significant variation between suppliers and production lots regarding belt stiffness.
So you just stretch a less-stiff belt a little more to get the proper tension, you say? Well, yes and no. In terms of linear stiffness, yes - it makes little difference. But in terms of bending stiffness, it changes the way the belt vibrates between pulleys, and in particular, the frequency that it vibrates at. It certainly changes the natural frequency of the belt so as to resonate at different engine rpm. The different specs for the horizontal and vertical cylinder belts is the clue here.
So what? I think that Ducati has traced belt fatigue failures to this phenomena, so theyíre trying to control the belts harmonic vibrations instead of just tension.
Of course, itís still possible to cause belt failure by overtensioning it.
One point bears mentioning here. Thereís been a lot written about how to accurately set belt tension. No matter what method you use, what's critically important with any toothed belt is the need to rotate the crankshaft (and the belt) after you set the tension, to assure that the specified tension is the same everywhere along the belt, not just in the part of the belt where you first placed the tension gauge.
If you service Ducati 4V motors for a living, you have set the tension in the belts enough times using the required tension tool and procedure to have become pretty well calibrated yourself as to how tight is tight. So an experienced technician may indeed feel itís a waste of time. They remember how much the belt will deflect between pulleys when under maximum tension when the engine is hot - and how the same belt feels when everything is cooled down. In other words, they have a ďfeelĒ for the spec when they double-check using their calibrated push and pull on the belt.
The rest of us need the tool.