I have been having a problem I noticed when riding a trainer, when riding outside, and even on the work stand. In a nut shell, the gap between the chain and the derailleur cage fluctuates depending on which section of the chain is passing through. This causes rubbing and noise.

I have a Madone 4.5 with Ultegra 6700. 50/34-12/25 setup. When this happens, I am NOT cross-chaining, where noises are expected. For instance, I can be on large chain ring and 2nd smallest sprocket and I can easily see the gap fluctuate from 1-2mm down to 0mm and rubbing occurs.

I am assuming this position is not cross-chaining? Large ring and small sprocket and small ring and large sprocket all rub and make noise. But should large ring and 2nd/3rd largest sprocket and 2nd/3rd smallest sprocket be rubbing?

Is there the possibility my chain is bent, worn, damaged? Could this problem be caused by a bent/untrue chain ring/crank?

It could be the chainrings.

it can also be flexing of the frame too .

Chainring runout of 1-2mm is quite common. The old solution was to find the extreme area and use an adjustable (cresencent) wrench to pry the ring into complicance.

Like the others I'll assume you're talking about the front derailleur. If so, it's either a warped chainring (or the spider) or flexing under load, or a combination of both. Flexing under load is very common (universal?) and varies by bike, acranset, BB and how strong or heavy you are. All bikes do it to a degree, but if you're not strong or heavy you might want check for play at the BB.

Warped chainrings can be straightened if they're bad enough to worry about. I'd say less than 1mm don't bother, but if it's more than that and making good trim difficult or impossible, you might want to improve it, except if the spider or the rings are carbon. The easiest way is to brace a screwdriver blade across the spider at the mounting bolts so you're not pushing on the chainring itself. Then use an improvised crow bar to push the arm in or out using either the BB or the base of the spider as a fulcrum according to which way you're pushing.

Some people also get good results giving the spider arms a shot with a rubber mallet, but this requires a bit of touch, and I prefer levering because there's more control.

BTW- just in case you're talking about the RD, then it's probably worn pulleys, which need replacing.

In addition to the mentioned factors contributing to chain gap within the ft der cage is the chain. many chains have some "back and forth" in their length. If you look along the chain from the rear you'll often see sections of chain links which "curve" one way or to the other side. I look down the length of chains often checking if there was a bent link (from some riding abuse, but that's another topic) and see this waver of chain straightness all the time. Of course the better the chain the more it keeps a straight line. But even expensive chains have had some waviness, IME. Andy.

Thanks for all the tips. Yes, it is occurring at the FD and it really isn't bent too bad. I would say it is not totally noticeable so I will probably end up leaving it alone for now. I just wanted to make sure it wasn't related to chain wear, since the chain isn't that old or have that many miles on it.

I can't imagine a mechanism that would cause a chain itself to wobble at the front derailleur. Although I will agree with Andrew that some chains "snake" back and forth it's doubtful that would show up at the front derailleur to a noticible degree. If that were the case it would occur in time with the chain rotation, which occurs much less often than the cranks rpm. Although the OP says the gap depends on chain position I would guess it's actually crank position. Remember that the chain has 100+ links, so takes at least 2 crank revs to pass through the derailleur.

Assuming that the variation is in time with crank rotation it is a matter of logic and observation that the chain moves in response to the chainwheel moving, and that the chainwheel moves dependent on it's own condition, it's attachment to the BB spindle, BB adjustment and the BB's attachment to the shell. The only thing perhaps not self-evident is that the bottom bracket can flex in relation to the seat tube. Once one knows the possible variables you just eliminate them by a further process of observation and logic.

I appears that this happens without having to exert very high force on the pedals so we can eliminate flexing of the frame, chainwheels/spider or movement of the BB in the shell. After confirming BB adjustment the next place to check would be the chainwheels. If only one chainwheel varies then it's obvious the problem lies with that chainwheel. If not, and if after removing the crank and reinstall after rotating to a different position on the spindle the variation occurs on the same spot then the problem is with the spider to which the chainwheels are attached. These first two are the most common causes.

Next is crank to spindle interface. Usually once the cranks loosen enough to wobble the problem quickly accelerates, but it's good to check tightness. If the amount or location of the variation changes when rotating the cranks to a different position then the problem is the spindle.