What values to enter for Lambda for 'Acoustic Parameters' for 'Rectangular Fitting Zones'?

56 views
Skip to first unread message

Mark Kubis

unread,
Jun 6, 2019, 11:29:23 AM6/6/19
to I-Simpa User Forum


Hi,

I'm using the SPPS part of version 1.3.4.

Lambda is the 'mean free path' in metres.

For a Sabine room it is given by Lambda = 4*Volume/Surface area

However, the rooms I will be modelling will not be Sabine because the absorption coefficient of the surfaces will not be the same.

I have found that in my model, as expected, values of SPL at a punctual receiver and RT change with different values of Lambda and this occurs regardless of whether the 'Diffusion Law' is selected as Uniform or Lambert.

Any suggestions as to how to calculate or what value of Lambda to use would be most appreciated.

Mark

UMRAE Official

unread,
Jun 7, 2019, 3:57:13 AM6/7/19
to I-Simpa User Forum
Dear Mark

Thank you for using I-Simpa.

The mean free path, you are referring (fitting zones)  is used when you consider fittings (scattering objects) in a zone/room. This is different from the classical room free path that is used in the classical theory of reverberation (Sabine).

If you use SPPS in a room without fittings, you don't need to define a fitting zone, then you don't need to define a mean free path.
If you need to consider a fitting zone, the mean free path to use must be calculated according your own information about the fitting zone: λc = 4/ (nc x qc), where nc=Nc/Vc where Nc is the number of obstacles/fittings ; Vc the volume of the fitting zone, qc=sc/4 with sc the surface of the scattering object.

I'm agree that the current I-Simpa documentation (https://i-simpa-wiki.readthedocs.io) is not well detailed about this aspect.

You can find more information on SPPS physical model implementation here :

Bests.
J. Picaut

Mark Kubis

unread,
Jun 7, 2019, 5:00:23 AM6/7/19
to I-Simpa User Forum
Hi Judicaël Picaut,

Thanks for your prompt reply and clarification. Just to make sure I understand what to do, imagine there are 2 fitting zones (i.e. 2 scattering objects) in the room. One is object X, the other is object Y.

In this example am I correct in thinking that:

(i) For object X:

Nc=2
Vc= (Volume of X) 
sc= (Surface area of X)

The result of the λc calculated using the above values is the mean free path which is the value of lambda entered into the 'acoustic parameters' of the 'fitting zone' for object X. All frequencies will have the same value of lambda.

(ii) For object Y:

Repeat as in (i) where Vc=2 again, but using Vc and sc values for Y

Thanks for your help in advance.

Regards
Mark




UMRAE Official

unread,
Jun 7, 2019, 5:17:27 AM6/7/19
to I-Simpa User Forum
The fitting zone is seen as a statistical point of view, considering many (many) obstacles in a given zone. All obstacles are similar, with the same shape and size (i.e. scatterring surface sc) and acoustic absorptions (alpha_c). All these objects are localized in a volume Vc; the number of obstacles are nc, and must large enough to consider that there is a diffusion process in the fitting zone.

See the example below. Each black disk is an obstacle. The volume Vc is the whole volume containing all the obstacles.


In your case, if you have only 2 obstacles, you should define them in the 3D scene as real object. In this case, you will simulate the real acoustic interaction between them (absorption, reflection; note that SPPS do not consider edge scattering at the moment).

diffusion.png


Reply all
Reply to author
Forward
0 new messages