AFMG Network Forum

Gradient · Member

I am comparing measured data with EASE analysis data for an Exhibition Hall of 83000m^3 and am finding that the EASE predictions are showing an increased RT time of 2.5 sec over the measured data; from 500 Hz onward. I have modeled everything in the room except for the rafters in the ceiling. I tried various scattering values but that does not seem to adjust the profile of the RT line to match the measured line. Any ideas? I am using Aura and EASE 4.3 with Very High resolution.

At first i was only using global scattering values, but then i switched to guessing what each material would be. But this seems to be opening a can of worms because its hard to isolate one material when your guessing 4 others. 90 % of the walls are gypsum and its a carpet floor. Any help on how to get accurate RT predictions?

Lastly, when modeling rooms with dropped ceilings I am finding the measured results to show a huge dip in the 500-2000 Hz range for RT times. Ease cannot replicate this result. Any ideas why?

PS: In Aura Response, should Schroeder times be less then Eyring/ Sabine? I thought that Eyring/Sabine were when you have 100% scattering, and this is never the case.

Tom.

opacheco · Member

I have a similar results too.....but lower RT than mrasurements RT values!!

Opacheco.

opacheco · Member

I have a similar results too.....but lower RT than mrasurements RT values!!

Opacheco.

AFMG Pedro Lima · Forum Moderator

Dear Gradient,

Can you please send your Project to pedro.lima (at) afmg.eu along with a description of the problem so that I can have a closer look at it and try to provide you some assistence?

Please remember to pack your Project before sending through EASE > File > Pack Project.

Best Regards,
Pedro

opacheco · Member

[quote="AFMG Pedro Lima"]Dear Gradient,

Can you please send your Project to pedro.lima (at) afmg.eu along with a description of the problem so that I can have a closer look at it and try to provide you some assistence?

Please remember to pack your Project before sending through EASE > File > Pack Project.

Best Regards,
Pedro[/quote]

Pedro,
Could you do some comments about this unmatch data(measurements/simulated) in order to know what was the problem? or in a more general disscus, could you make a check list for these kind of problem please?.....I am having similar troubles too!

AFMG Pedro Lima · Forum Moderator

Dear Opacheco,

There may be numerous problems in such process, which would vary even more with each the Project in question. These might be concernced to measurements and also to your model. It would take some time to elaborate such check list, really.

That's why I asked you to send the Projects and measurements, if possible. That way I can take a brief look at them, trying to find some obvious mistakes, so that we can discuss the problematic points more directly.

Best Regards,
Pedro

opacheco · Member

Pedro,

Thanks for your response and time!

I can to send to you some data in order to get some feedback in this topic....please, give me a email for that!......but more than get a particular response, I would like to share our experience (or more specific, your experiences and knowledge!!) in this particular case of study with all the EASE comunity here because this topic is realy a good theme to analyse (and very commun question on measurements versus modeled results).

I will wait your comments
Thanks a lot for your time!
Opacheco.

AFMG Pedro Lima · Forum Moderator

Opachecco,

The e-mail address is above (pedro.lima (at) afmg.eu).

Pedro

opacheco · Member

Pedro,

I got it!...please give a couple of days in order to send all these data to you!....

Thanks a lot anticipated
Opacheco.

AFMG Pedro Lima · Forum Moderator

To people who is keeping track of this post:

"Hi Oscar,

There was only the EASE Project attached. Do you also have the raw measurement files? It would also be nice to know the location where they have been taken.

Well, there are a number of inconsistencies in your drawing. It’s likely that they are the cause for the unnatural RT behavior. I will try to point some of them here but you should understand that I cannot take a deep and through view to your project, ok? Also, it seems there are far too many details for an acoustical model. This would have as consequence a huge increase in the calculations times in general, especially for AURA and evaluations involving Ray Tracing. You must distinguish between an Architectural Model and an Acoustical Model. The second will typically have far less details, as you need to be straight forward in your analysis and be careful with the calculation times.

• You should only use Two-Fold faces for barriers or elements that you know will have both sides of the Face exposed to the sound. The pillars on both sides of your drawing should not be Two-Fold faces, as their internal area are not exposed to sound. This will also affect the total volume, as their internal volume would actually be subtracted. There are also other places with similar conditions.

• There are rooms other than the main room. This is critical. There are rooms above the main entrance of your church that there are not part of the main acoustical space. This will only increase the room volume and contribute negatively to your simulation. EASE is a program based on statistical acoustics assumptions and will not take into account transmission from one room to other. I could not be sure due to the high volume of details, but it seems this is also the case with the altar. This is especially important if Eyring and Sabine RT simulations are being performed other than AURA and Raytracing.

• There are Faces Coating the wrong Faces. The Coating Face should be Coated over the Face that is covered by it. Please refer to the EASE tutorial for more information about Coated Faces (Is Coat Of in its properties window). There are also Faces that are not being Coated, while they should. As the top rectangular windows nearby the entrance.

Oscar, it seems you are in the right way but there are still a lot of modeling issues that you must address before going on. Please allow me to advise you some extra literature to help in the process: Vorlander M. – Auralization, Fundamentals of Acoustics, Modelling, Simulation, Algorithms and Acoustic Virtual Reality.

I hope this helps.

Pedro"

opacheco · Member

Pedro,

I appreciate so much your time and effort in order to comments about my acoustics errors and point of view architectural versus acoustical. Well let me to re-check that for all this kinds of mistakes and I will comment to you then!!......I will check the literature too!.

If you think about some anothers comments more please let me to know!

Thanks for your time too!
Opacheco.

opacheco · Member

Pedro, I forgot...I will check the measurement again, I will send to you when I have it ready.

Thanks a lot again!
Opacheco

Ron Sauro · Member

I have been researching this for a very long time now and I think our research may help here.
1st... All of these simulation programs use the absorption coefficient for calculation of the total absorption of a closed room. In our research papers we have shown that area alone cannot be used to multiply the absorption coefficient to get total absorption...
The edge effect is the problem here... In these programs the assumption is made that an absorption coefficient cannot be above 1.0.
The way that absorption is measured does not support that theory.
It is not at all unusual to get coefficients way above 1.0.
This means that if a particular frequency is looked at that might have a coefficient of 1.7 and it is limited to 0.999 you are underestimating the total absorption of that surface by a very large percentage and there for will always over estimate the RT of the room....
My papers are on the front page of my website at
www.nwaalabs.com
Ron Sauro
NWAA Labs, Inc

thomas · Member

Another point for the comparison of measured and simulated RT of large halls
is the important influence of temperature and humidity.
You have to simulate with the conditions of the measurement.
Or you adjust your measured RT to the temperature/humidity values,
you use in EASE (standard: 20°C/60%).

And: you can find real RT's longer than predicted by any formula derived
from statistic-acoustic assumptions because the mechanism of reverberation
in a room can be more complicated, than in the formulas assumed. Or, at
least, different. Some mechanical solutions to increase the reverberation
time in concert halls or opera houses work therefor.

Best regards
Thomas