Natural Science Forum / Physics / Acoustics / August 2006

-snip-

My observation of "thump" noise is that it is at frequencies down to and
through infrasound. One of the first steps is to summarize the resonant
modes of this coupled system comprising the mechanical resonance modes
of the joists and of the room modes of the receiver room below.

The joists have their fundamental bending mode around 5-8 Hz, but also
have higher modes thouugh infrasound and into the low end audio
frequencies that are conventionally measurd and reported.

residential rooom modes have their first resonance around 20-40 Hz,
depending on room dimensions and the local air sound velocity.

There is no way I know of the stiffen the joists sufficiently to make
the first joist mode above audio frequencies, so we are stuck with
damping as the only other strategy, and this as limited effectiveness
with practical materials (I have see data on dry sand used between
sleepers as one one such effective damping measure).

So we are stuck with mechanical modes of the joist system as an option
for mode suppression. Hopefully the stiffening so effected will reduce
the velocity of response, and that the nailed joints will supply added
damping, all aimed to reduce the transfer of vibation evergy into the
room air below.

Hopefully we can uncover some such data here.

More empirical data needed.

    Angelo Campanella

From Jobs I attended and tested and those where I recieved reports from in
my last employment, there was some low frequency improvement from stiffening
the floor in this method, particularly in improving the low frequency end of
the Impact and airborn transmission tests (Dntw & Lntw). However the
measurable improvement was often lower than the percieved benefit from
qualitative reports where the lower room was booming from footsteps on the
floor above.

Passing some of the load to the adjacent joists has the effect of reducing
the deflection at mid point from a footstep on the floor above. The
increased stiffness in the floor may bring the noise level to below an
annoyance level or prevent excitation of room modes making it less
uncomfortable for those living beneath.

Installing noggins is well known in the building industry and it is fairly
common to be already installed (though maybe with smaller section timber) in
large spanned rooms in many victorian properties.

Of course, floor vibration is dimension dependent and a lightweight timber
floor unsupported over too wide a span will deflect and may be noticably
uncomfortable even where a standard sound test would not flag up an issue
due to the low frequencies.

Floors that have passed an Lntw test and even a Dntw test may have low
frequency vibrations that cause uncomfortable feelings in the room below,
though this is not picked up in standard sound insulation tests. In fact
excitation of this trouble boominess, can reveal it to room modes in the
room downstairs rather than excessive excursion of the floor. The nature of
the sound tests and the lower limit of such do not look at spot frequency
response deliberately to reduce the effect of room modes on measurement,
however in wide timber floors low frequency vibration may be possible which
in a weighted band measurement is not significant, though
psycho-acoustically is a real annoyance.

There are often practical limitation of conversions of some listed buildings
in this area. Certain features need to remain and retention of existing
structure is required to avoid damage to listed period features. Without
re-building the ceiling/floor etc, or where head room is not high enough to
allow reinforcement there is a possible issue of "boomy" floor noise where
large floor spans are unsupported and unavoidable. Adding noggins or
blocking can improve this to some extent if the beams are not exposed.

I hope this is of interest and of use.
Regards
Richard Price