Quote of ssg’s original scenario:
Assume you have 5000 ft extra runway past your balanced field length, your light, cool weather, everything is in your favor...you accelerate past V1 twards V2, and you get an engine fire...
Would you fly it off the ground or try to stop it?
[Unquote]
Not sure if you’re a pilot, ssg, but you certainly made us all examine our principles and understanding of take-off performance. No bad thing. But there are at least two factors in your scenario that you seem to have overlooked.
1) Once VR has been called and rotation initiated, there is a major handling issue, as others have commented. [Granted V1 can be 30kts below VR.]
2) Bringing the aircraft safely to a halt on the runway with an engine fire is not necessarily non-fatal. If an engine fire occurs before V1, we must also assume thrust failure; so performance theory suggests continuing is not normally an option (unless someone wants to propose that on another thread). In that case we must settle for extinguishing the fire at a standstill − backed up by the fire service. Think of the B737 at Manchester, when a light crosswind blew the flames into the fuselage. After V1, even assuming the crew has a choice, taking the fire into the air may be the safer way of handling the fire. Trouble is, there are engine fires and engine fires. Perhaps other Posters will comment?
The scenario you have chosen is predictable at the planning stage. So the decision can be made then, by choosing V1 = VR (but NOT V2). When there is so much surplus runway, a big range of possible V1s is available (between BOAC’s Vgo and Vstop); provided your company’s performance manual or computer is up to the job… Having said that, you may decide to choose a low V1 if departing an airfield with no engineering support, for example, in a 4-engine aeroplane.
Once a V1 has been briefed, there has to be a very extreme failure to justify flouting it. The above performance calculation is always based on the failure of ONE engine, as in your scenario. A more difficult case would be the failure of more than one (before VR). Amending your scenario to the recognised fire and/or failure of 2 engines on a light DC10 before VR, I think most of us would hope to make the decision to close the throttles. In a 4-engine jet, it's not so clear-cut (we used to practise it in the VC10 simulator, but the second one didn’t quit until we had rotated).