Pachyderm Acoustic

Comment

Comment by David Probst on June 6, 2018 at 5:32am

Hi Arthur,

thanks for the answer, it helped me a lot to be more confident with my work. Also, please don't feel pushed by my questions, I really appreciate all the help you have given me already.

The only thing I don't understand comletely yet is your answer regarding the absorption-coefficient. I do understand the part about why the absorption designer is the right choice and whether or not the surface-structure would matter, but I am still struggling with the impact of the distances beetween each individual hole to the next hole.

I think I had phrased my question a little bit ambigious in the last post. Actually, the whole plate is perforated, but I am talking about the "frequency" of holes along the plate. By that I mean a 1m² plate could have 1000 or 100 or only 5 holes.  I know for example that 1.4 percent of the plate are holes and the rest is massive. With which setting could I compensate for this frequency of holes? Or is there a fundamental flaw in my thought process?

I would love to send you the model and hear your opinion about it, but would prefer to keep that private, since it is not my own creation and I am just testing around with it. Can you accept my friend request here, so I can send you a private message?

Thank you in advance!

Comment by Arthur on May 25, 2018 at 2:31pm

Hi David P,

I have been playing with it a bit more also. One thing I found is that if you make the radius too small, then your result will be poor. Another thing I have found is that the farther away from the sample your sphere is, the lower the magnitude of the scattering will be. This seems to suggest that maybe the base of the hemisphere should always be right at the bottom of the scattering panel.

Regarding whether or not your result is accurate or not, I can't really say. You could send me a copy of your model, though, and I could try it and give you my opinion...

Also, you made the right choice by doing the perforations in the absorption designer, rather than the scattering tool. At present, absorption can't be assigned to the boundaries of the FVM tool. One day I will figure it out, but I will need a lot of time to dedicate to it at once, and family life doesn't allow for that right now... My intuition says that the scattering probably doesn't affect the absorption much, other then that the depth of the perforations will vary with the depth of the panel, and that if you have portions of the panel that are not perforated, the final absorption of the panel will be an area-weighted average of all conditions. (I can't say for sure without studying your design in more detail, though.)

Thanks for giving this experimental tool a shot. I will help in any way I can, within the limits of the many roles I hold.

kind regards,

Arthur

Comment by David Probst on May 24, 2018 at 8:57am

Hi Arthur,

thank you for your detailed answer and sorry for not answering for such a long time, I was just trying to make sure that I did understand everything before I answer back.

Now I do actually get quite reliably the same results and I have no issues with any crashes anymore. The only problem is that my intuition would sometimes suggest really different values. The main surface I am working with(see image) for example only delivers a scattering coefficient below 0.1 , while I would assume that it should be really high. Do you think that these calculations are right or do you have any idea what I might be doing wrong?

Also I was trying to assess the alpha-coefficient for the same pattern with holes and mineral wool behind, but decided to simplify things and just use the absorbtion designer. Do you think that the pattern itself could have a big influence on the absorbtion or is such a simplification probably all right? I was already building some manual tests to integrate its effect but it is a lot more work.
When using the Absorbtion Designer I did not find a value for the perecentage the holes take up, even though I thought, that this should be the most important value? So far I have just been playing around with the diameter of the holes and studying the results.

Thank you!

Comment by Rowan Browan on April 28, 2018 at 10:29pm

Hi Arthur

I seem to be having the same issue at Zaqi below

I recently installed Pachyderm and have an issue using the Pachyderm Mapping Method.

Every time that I try to open the Mapping Method tab it immediately closes my rhino, and offers to send an error report, as though rhino had just crashed. I am running it on a new Dell XPS 13, so I don't think hardware should be the issue?

I have uninstalled then reinstalled Pachyderm, but the error persists.

Can you please let me know if you've had issues like this before/know any possible solutions?

Comment by ZAQI FATHIS on April 27, 2018 at 9:38am

Hi Arthur,

Ok, right now I just want to follow this tutorial (https://vimeo.com/247022069). 

So my computer is windows x64, with the latest Rhino 5 SR 14 and Grasshopper.  Another tab works fine for me.

Comment by Arthur on April 27, 2018 at 4:55am

Hi Zaqi,

I can help, But I would need to know exactly what you did.

Before that, though... Is there anything unusual about your computer? Is it an apple?

Arthur

Comment by ZAQI FATHIS on April 27, 2018 at 1:38am

Hi Arthur, 

I have installed the latest Pachyderm v.02. I tried to use Pachyderm mapping method option, but it always makes my rhino crash. Do you know why is that happen?

Cheers!

Comment by Arthur on April 26, 2018 at 6:08am

On meshes- pachyderm won't read them. Use surfaces only. Pachyderm will make the mesh as it needs it based on the surface model.

It shouldn't crash the main simulation... Just exit out before running, but it might crash the FVM models. I'm not sure.

Arthur

Comment by Arthur on April 26, 2018 at 6:04am

Hi David,

Thanks for starting the conversation.

Scattering has a definition in an iso standard. You should probably look up ISO 17497. For the purposes of geometrical simulation, the definition you are using is correct. The problem is that sound doesn't actually travel in rays. It's a simplification of sound propagation.

The threshold you set assigns the density of the mesh used. You may still get data above that threshold, but you shouldn't trust it.

There are several reasons why it is marked experimental.

1 I used the correlation scattering coefficient, which is not standardized, but I suspect may give a more applicable result for geometrical modeling. (Suspect only... I'm not actually certain)

2 in some cases, it gives very plausible results, but it is very sensitive to a) sample placement and model measurement space alignment b) the scale of the mesh (which was very alarming).

I will look at it again sometime. In the meantime, if you feel inclined, give it a look yourself. Maybe you will see something I missed.

Comment by David Probst on April 26, 2018 at 5:12am

Hi Arthur and Pachyderm Community,

first of all I want to say thank you for providing us with free and open-source acoustic-simulation software. I am a student from Munich currently examining the acoustical effects of more complex, parametrical shapes; Without Pachyderm that would be very, very hard. However, I am running into a few problems and it would be awesome if you could find a few minutes to help me.

I am mainly using Pachyderm to try to find out the Alpha-Values and the Scattering-Coefficient of certain structures. Talking about the scattering analysis, I know that the feature is experimental, but I still think you might be able to clear a few things up. Generally, I understand the "Calculate up to"-Property as a frequency-treshold up to which the programm will calculate the correlation-coefficient. However, when I perform the calculations, Pachyderm does not respect these thresholds.

500hz: Pachyderm calculates 0 - ~7000hz

1000hz: Pachyderm calculates 0 - ~12000hz

2000hz: Pachyderm calculates 0 - ~30000hz

4000hz: Pachyderm calculates 0 - ~55000hz

To add to the confusion the values of the scattering coefficient will also change dependent on the threshold. I will attach all the files, the results and so on for these tests. I tried this with various structures, always had only one object with an assigned material in the center and the box did completely surround the object. Also just on a side note, the program crashes for me, whenever I am performing the analysis with a Mesh (Windows 10, Rhino5).

Furthermore, I could not find a common definition of the scattering coefficient. To me the definition "percentage of rays that is not reflected geometrically (+/- 10°)" makes a lot of sense; Does Pachyderm follow a similiar definition?

Since I am pretty new to CAD in general and definetly not an expert at all in neither acoustics nor simulation I have attached all the files that I used. Even though I have been testing around quite a bit, it might just be a really stupid mistake. I also did the tests with more complex surfaces (which I am actually interested in), but since the problems do already show for this simple rotated surface it might be a good place to start. I know this is a really long text, but I already tried to shorten it as much as I can, leaving a lot of other questions open. Thank you for any help!

ScatteringTests

All the best,

David

• ‹ Previous
• 1
• 2
• 3
• 4
• 5
• …
• 7
• Next ›
• Page