[Daniel Prates]

Originally Posted by TorpX

At the risk of coming off as onery, I feel I have to point out a few things here.

The "range is irrelevant" is often repeated, but is really misunderstood. It is true that you can compute the lead angle for a firing solution, knowing the speed, track angle and bearing, without knowing the range, but that does not mean the range is irrelevant. (Also, it has nothing to do with the sub being on the "beam". The range dropping out of the equation for the lead angle holds only for a zero-gyro angle solution, whatever the positions of target and sub.) The firing solution for a 4,000 yd. shot may have the same angles and speeds, as a solution for a 2,000 yd. shot, but they are not the same. A small error in the data that would get you hits at 2,000 yds., would likely produce misses at 4,000 yds.

The thing often glossed over is the question of how would you obtain the speed or course without knowing the range? Speed might be estimated by observing the bow wave or by 'seaman's eye', but this would require visual observation anyway. Timing by wire also requires visual observation. The only other way I can think of that would not require any range info, is to use propeller counts. I think in practice RL captains relied mainly on calculations of distance/time (i.e. plotting). Another thing is that it would be almost impossible to determine the target's course without either knowing the range at several points, or accurately observing the AoB. This brings you back to visual (or radar) observations again.

Only a very few "sonar-only" attacks were made by US submarines. As far as I know, none were successful.

Ha! My point exactely.

Another key issue to consider is that you should always try to fire from as close as possible. All margins of error decrese as you near in (except for your personal safety margins, but that is another issue). If we are talking about a lonely merchant, it is fairly easy to lurk ahead of him and wait till he is less then 2000, maybr 1500 meters away. And in that situation, a couple of statimeter readings are all you need for a "good-enough" range (and speed!) estimate.