Further considerations on the simulation results lead to think that – as always in simulation – the results are limited to the modelled physics. We still don’t understand the presence of a low frequency component (100 Hz) only found in Repercussion Bessel vented snares testing , where model shows a “Helmhotz resonator” behaviour for standard vented snares at 50 Hz (see Shell depth , Venting & admittance) not heard nor seen in the testing…
Could this be coming from a lack in our model ? In our case, we considered infinitely rigid shell and studied the behaviour of two stretched membranes coupled by a cavity – practically this means that the transfer of energy from the stuck top membrane to the bottom one would be done solely by the air within the shell.
In reality, mechanical waves are transmitted to the shell structure, and transferred to the bottom membrane. As the stiffness of the shell “axially” is considerably higher than the one of the air in the cavity, and as velocity of sound within material is higher than in the air, it is very likely that in reality, the sound wave coming from the top head hits a membrane that is already in motion due to structure borne transfer…
Does it mean that when we split the shell in halves we kill this transfer ? what consequences for the sound ?
Let’s get back to testing where various shells material and construction were tested in the lab with calibrated stick hit, and vibration of top and bottom head recorded : let us look at the ratio of those over the range of the 6 snares tested.
It seems there is a global trend that makes sense : intuitively, a steel shell is stiffer along its axis (that is traction/compression waves), then the fairly thick and stave built oak shell that would channel the sound wave faster to the bottom head, then the repercussion 10 slots which is the only prototype that actually was NOT splitting completely the maple shell (only slots), then the standard snare that features U shaped beams of aluminum to hold the two half shells apart, finally the Orchestral snare, with a fully split shell and tubelugs of small cross section to space the two 12mm maple shell.
Close up of the Repercussion 10 slot vent
Close up of the Orchestral snare
Inside the standard radial vented snare prototype
Finally, the mahogany snare, which would mean that this type of material either damps the energy or that the relative thin shell bending takes over the axial transmission.
Sound velocity for various material used for snare shell / In comparison with our testing : Steel is the fastest, the Maple before Oak, but in our case Oak is thicker and built in staves / for reference sound in air is 340m/s so the waves in the shell material would travel at least 10x faster.
Conclusion :
“at first glance” there might be some correlation between the shell presumed stiffness and the level of vibration seen on the bottom head, however material and construction seem to play a stronger role than the fact of splitting the shell, and the effect on the perceived sound is yet to be documented. Keep in mind that radial venting does enhance top head vibration and sound as presented in The snare was killing batter head sound !