fiddlerman.com/wp-content/forum-media-uploads/cdennyb/2012/05/http___www.andreasgrutter.nl_bow-couch_bow-couch.pdf

springer said

Come on Dennis give us a technical answer.

Well...

All of the above is a great place to start. The technical thing that's happening is this:

As the bow is pulled across the strings with, lets say X1 amount of tension, and the humidity is H1 and not going to change (say 50% relative) the constant amount of rosin and the constant speed of bowing are both unchanged (just for the argument here). The pressure of the hair against the string(s) is changing constantly and we'll call that V1 (for variable)... and now the explanation...

As the V1 is fluctuating, it causes a wrap-up on the string as it travel across it, rolling the string slightly as it glides across the string. As this happens it causes the X1 to change as well very rapidly, like dozens if not hundreds of times per second, and the string also vibrates at whatever frequency the note is that's being played is at. Now for the fun part... as the bow hair is wrapped around the string slightly and released at that incredible speed which also will match the string vibrational speed on occassion the grab part will suddenly 'let go' of the string and the bow will shoot forward or backward whatever the direction is that it happens in, and slightly bounce until the hair (coated with rosin) comes back to the string and grabs once again, repeating the cycle all over bunches of times per second, the number is directly proportional to the speed of the bowing motion.

The ways to stop this from happening is to make sure there isn't "too much" rosin build up on the hair strands (which tend to wrap up more on the string) and make sure the string has little if any build up on them as well. As the humidity and temperature go up, like in summer... many change rosins to get away from that 'sticky' problem and thus stop or greatly reduce the stick and grab mechanisim from happening.

Also, the X1 (hair tension) is a factor that the player will constantly change from session to session, thus one time it bounces the next it doesn't) loose hairs will wrap up easier and faster than tight hairs will...but tight hairs will be difficult to grab and release in their normal fashion to create a vibratory note in uniform value which is the "balancing act" that every player of strings has to contend with.

Trying to play lightly is more prone to a squeak or high pitched squeal and playing agressively will tend to forcefully drag the hair across the strings creating a harder vibration which is usually always lacking in a bow bounce.

This is pretty much the simple version of the dynamics, and I've left out a lot of technical stuff because you would need a good foundation of elementary physics and math to comprehend the actual variables and their relationship to each other.

The variables are in the range a couple of dozen and the player introduces another dozen or so. The method and causeation of bow bounce is very complex both in physical and auditory details.

Here's a nice 227 page doc from a bow maker. It makes for very interesting reading and addresses this issue in some detail.