This is my understanding, although it only partly answers the question.
As stated above V2 and VR are based on functions of the stall speed and Vmca. On most aircraft the stall speed function is usually dominant and Vmca only controls the speeds at light weights and when the density altitude is low leading to high asymmetric thrust in the engine out case.
In the area where stall speed controls VR and V2 then the EAS of these speeds should remain unchanged for altitude. Because we read the speeds as indicated airspeeds (IAS) not EAS we should find a small rise in the IAS of VR and V2 as density altitude increases. To get specific, as JT suggests, the L1011-1 flap 10º graphs illustrate just this, at least with VR. The wierd thing is that V2 goes the other way, it decreases as density altitude increases. Now why would that happen?
[edited to get 'increases' & 'decreases' right. Doh!]
Last edited by Alex Whittingham; 11th April 2002 at 09:06.