Sweep has the same stabilizing effect as dihedral, except its not as pronounced. I beleive the rule of thumb is 5 degrees of sweep has the same effect as 1 degree of dihedral.

From : Don Stackhouse

This is another "rule of thumb" (boy, do I hate those!). I've heard 3 degrees of sweep is equal to one degree of dihedral.

The problem is that both of those numbers are right, and both are also wrong. The fact is that the amount of roll effect due to sweep depends on the cosine of the sweep angle, and (more importantly for this discussion) on the lift coefficient. If Cl is zero (such as in zero-G flight, knife edge, verticals, etc.), then there is zero dihedral effect from sweep.

If Cl is high (such as when flying slowly on final approach, or during high-G maneuvers), then the dihedral effect is large.

The effect on roll is linearly proportional to lift coefficient. BTW, the effects of sweep on yaw stability are proportional to the square of the lift coefficient.

Altogether, this can cause some problems. In particular, if your model attempts to use rudder-only (not ailerons) for roll control, then your control response will vary from pretty strong at high lift conditions, to utterly non-existent at high speed.

To complicate things even further, consider that both ailerons and rudder+ dihedral continue to roll the airplane in the same direction whether inverted or upright. However, the effects of sweep reverse their direction in inverted flight. It would be like someone flipped the servo reverse switch on your roll servo just as you passed through inverted flight. To do a slow roll with rudder + sweep, you would have to reverse the direction of the rudder during the second half of the roll to keep the roll going in the same direction it had in the first half. Of course you would never get into that problem if you were trying to do a true slow roll. The roll would stop and refuse to go any further when you reached the first 90-degree point in the roll and the G's went to zero. You could of course pull some positive G to keep the roll going, but then it would be a barrel roll, not a slow roll.

Then there's the whole problem of dutch roll and spiral stability. These two are controlled by the balance between vertical tail and the wing's dihedral. Too much dihedral and you get positive spiral stability, but you also get dutch roll. Too little wing dihedral and the dutch roll goes away, but you get spiral instability (i.e.: in a turn, the plane wants to steepen up the bank and wind itself into a "graveyard spiral", all by itself). Flying a plane with spiral instability can be mentally very fatiguing, like herding cats or trying to nail jelly to a tree. It's always trying to wander off to someplace other than where you want it to go.

The new quirk that sweep introduces to this dutch roll/spiral stability problem is the fact that its dihedral effect varies with airspeed and "G" loading. At low speeds such as during final approach to landing, the plane gets lots of extra dihedral effect from the sweep, which can cause some pretty significant dutch roll problems. That's a very bad thing if you're low-and-slow. However, at high speed, the sweep's dihedral effect goes away, and that can leave you with significant spiral instability problems.

Is all this just theory? Nope, I've seen it firsthand, in a swept-wing UAV we were working on a few years ago. We were able to successfully design around these problems, but essentially all of them raised their ugly heads in one way or another.

The nice thing about sweep is that it works the same when the plane is inverted, whereas diherdal is a destabilizing when the plane is inverted.

Then there's the whole business about the "lift valley" in the center of a swept wing, that tends to unload the root and transfer more of the lift distribution out to the tips. This makes the tips prone to some nasty tip stalling tendencies. To make matters worse, if a tip stalls, that loss of lift (because it is further aft than the root) tends to make the plane pitch up even farther, worsening the stall. Also, if you do manage to get the root to stall first, the spanwise airflow typical of a swept wing tends to make the stall propagate rapidly out to the tips, making the whole wing stall with the root. This is one of the reasons why some swept wing airplanes have large stall fences, vortilons, or notched leading edges to try to keep the stalled area from spreading out to where it can blank the ailerons.

For more in-depth discussion of swept wing effects on subsonic aircraft, I recommend finding a copy of "Tailless Aircraft in Theory and Practice" by Nickel and Wohlfahrt, published by AIAA. It has one of the better discussions of these issues, with much of it still in reasonably plain (plane?) English.

No question that swept wings can look "cool". They have very meaningful effects in delaying the onset of shock-related problems when trying to fly at transonic Mach numbers (i.e.: near the speed of sound). They can also be the foundation of some very effective ways to eliminate the need for a tail. However, they can open a whole "Pandora's box" of new problems of their own. The designer had best be wary.

Don Stackhouse
DJ Aerotech