Take-off segment from above taxiing speed until rudder effective - Tricycle L/G
Quote from chornedsnorkack:So... on speeds above 60...80 knots IAS, aerodynamic forces become effective. On taxi speeds, nosewheel steering can stop weathercocking.
How is a tricycle aircraft kept on a runway in strong crosswinds at speeds above taxi speeds but below the speeds where rudder becomes effective?
[Unquote]
You have quite rightly drawn attention to the gap in my narrative relating to aeroplanes with TRICYCLE L/G – thanks. 
[I can't help with your query on tandem L/G configurations, like the B-52; but presumably the front wheel-set is steerable? Know nothing, sadly, of the (aah) Northrop B-2.]
The awkward segment of the crosswind take-off – between taxiing speed, and the speed at which aerodynamic rudder becomes effective – is worthy of explanation. It's not going to be a brief one, I'm afraid...
What I might have written was something like this:
It may be pointed out that some such airliners, in the past, have had fully-castoring nosewheels with no steering capability. The D.H. Heron (4-engine 15-seater) and Dove spring to mind. Fortunately, they were propeller-driven (see below).
The key difference between the early part of the take-off run and normal taxiing is that the use of differential braking is, strictly speaking, not an option. Until the rudder becomes effective aerodynamically, nosewheel steering makes the major contribution to directional control. On a propeller-driven aircraft, propeller slipstream usually provides useful rudder control from the start (although the slipstream itself induces yaw, unless the propellers are "handed").
No such contribution is made by a jet engine. On jets, therefore, rudder effectiveness is roughly a function of indicated airspeed (IAS) squared.
On any multi-engine aeroplane (except with engines in tandem, or "piggy-back" **) it is possible, to a limited extent, to pre-empt weathercocking – during the period of engine acceleration – by doing a rolling take-off, and gently advancing the upwind throttle(s) slightly ahead of the other(s). However, this slow-acceleration technique will increase the length of runway used, so may not be possible on the day.
AIRCRAFT WITH TILLER STEERING ONLY
Tiller steering is suitable for steering – and to resist weathercocking – at taxiing speeds, and at the beginning of the take-off run. Above about 40 kts ground-speed, however, it becomes progressively more difficult not to over-control in yaw. This problem is increased on a wet runway, particularly if the runway friction is patchy. The painted centreline itself can cause intermittent loss of tyre adhesion. [Not a good point to have an engine failure, particularly on the upwind side, as discussed in previous posts.]
From about 60 kts IAS, the rudder will start to become effective on most large jets. So, for a given crosswind component, it is arguable that the worst case is when the wind is at right angles to the runway – or slightly beyond, giving a tail-wind component. In these situations, there is no headwind to amplify IAS and rudder control in the early stages.
There is no doubt that, on many such older types, there is a period on the take-off when you are too fast for reliable tiller steering, but may not be fast enough for yaw control by rudder in a limiting gust. I am not sure that this was fully taken into account during certification.
AIRCRAFT WITH RUDDER-FINE STEERING
Once lined up on the runway, the pilot removes hand from tiller. On modern jets, the steering-control computer modulates the rudder-fine steering according to the ground speed (GS). Early on, the pilot cannot make full use of any available aerodynamic rudder, because the steering would veer the airplane off the runway. As the GS rises, the nosewheel angle provided by a given deflection of the rudder pedals (which dictate rudder angle) is weaned off, to avoid over-control. This allows the pilot to start to get the feel of using the rudder, but not much.
On the A320, if memory serves, the nosewheel steering is switched off completely above 72 kts GS. At this point, the pilot may well need an increase in downwind rudder to counter weather-cocking using aerodynamics alone. It's worth pointing out that, in the tailwind-component case, particularly at hot-high airfields, the IAS could still be as low as 50 kts.
Hope this helps.
** e.g., English Electric P1B Lightning? 