Wind Simulation

[Anonymous]

Hi,

I'm new to the wind simulation on robot and would like some advised. I am trying to carryout wind simulation for building and only concerned by max wind pressure and suction on the facade elements.

My steps have been as follows

1. Calc wind speed for the site
2. Import plan of building
3. Extend model to required height
4. frame model with members at corners and roof
5. put cladding over the faces and roof only
6. select wind simulation and generate wind loads
7. select directions
8. input wind velocity as calc above
9. select cladding panel elements only
10. run simulation and calculation

After running calculation I noticed that I have wind load on faces and edges. For example see Y+ I have a wind load pressure on face of 0.88 but on edges i have wind suction of -0.4, 0.36 & 0.57. Why am I getting loads on edges is this correct?

My questions are as follows:-

1. Is my procedure/steps correct?
2. why am I getting load on edges of panels? and why in opposite direction
3. Do I need to split cladding panels into storey heights and add horizontal floors - see model 2
4. When I do this I get different results, why is this different?
5. Is this more accurate?
6. This is for external wind load only and Cpi should be manual calculated and added to this.
7. I have external vertical features to add to panels, can I add and will this be correct if I run simulation? In the code it specifies a Cpe of 2.0 for these.

This appears a very useful tool if I have external features fixed to buildings and I hope it will produce more accurate results.

I have to provide feedback by the end of the week so I would really appreciate some feed back for the above.

The files can be downloaded form the following link.

http://we.tl/v1iTLmhDyM

@StefanoPasquini6790 

May be you could review as I have read many of your posts.

Thanks

[StefanoPasquini6790]

My questions are as follows:- Q: Is my procedure/steps correct? A: It sound so! Q: why am I getting load on edges of panels? and why in opposite direction A: The loading results on claddings it comes from an evaluation of the average values, the load on edges of the panel have the aim to "correct" the uniform loads created for your wind load case. Q: Do I need to split cladding panels into storey heights and add horizontal floors - see model 2 When I do this I get different results, why is this different? Is this more accurate? A: If you divide the claddings you can have a more accurate model, if you compare the reactions in the wind direction, you can have similar or identical results, but the local effects are not correctly evaluated. I suggest you to do a couple of identical models, using a different panels subdivision (few big panels vs many little panels) to better understand what I mean. Q: This is for external wind load only and Cpi should be manual calculated and added to this. A: Yes it is, I made a proposal to autodesk to add the option to calculate automatically the CPi values for building, defining the coefficient and creating two load conditions. Please give kudos to this idea in the Idea Station. Q: I have external vertical features to add to panels, can I add and will this be correct if I run simulation? In the code it specifies a Cpe of 2.0 for these. A: Yes it is correct Cheers _________________________________________ ADSK_Expert_Elite_Icon_S_Color_Blk _________________________________________ image002 Ing. Stefano Romeo Pasquini Progettazione e consulenza strutturale Via don L. Sturzo n?1/a - 65024 - Manoppello(PE) Phone/Fax: +39.085.8561045 Mobile.: +39.328.9589857 e-mail: stefano@pasquiniprogetti.eu e-mail PEC: stefanoromeo.pasquini@ingpec.eu Skype: stefano.romeo.pasquini Web: <> http://www.pasquiniprogetti.eu P Please consider the environment before printing this mail note. Avvertenze D.Lgs.196/2003 Le informazioni contenute in questo messaggio di posta elettronica e/o files allegati sono strettamente riservate. Il loro utilizzo ? consentito esclusivamente al destinatario del messaggio o a persone dallo stesso autorizzate per le finalit? indicate. Qualora riceveste questo messaggio senza esserne il destinatario Vi preghiamo cortesemente di darcene notizia via e-mail e di procedere alla distruzione del messaggio, cancellandolo dal Vostro sistema. DISCLAIMER Unless otherwise expressly stated the information contained in this e-mail is highly confidential and is intended only for the attention or use of the recipient named above. If you are not the intended recipientplease inform the sender by e-mail and delete the message and any attachment from your system. Any use, disclosure or copying of the contents is prohibited according to Italian Law n.196/2003.

[Pawel.Pulak]

Hi,

pressure maps on single panel or cladding are replaced by uniform pressure on it plus UDL loads on the edges of it - total sum equivalent to force resulting from integrating pressure map on the surface of panel or cladding.

These UDL loads on edges are used to consider for possible (almost sure) no uniformity of pressure map. In your case these loads on edges are in opposite direction because pressure maps are decreasing close to edges,

As you see such conversion of arbitrary pressure map into uniform pressure plus uniform edge load is very simple. It does not consider precisely local changes of pressure in various zones of the panel or cladding.

Knowing it  it is clear that it is better to split cladding panels in smaller pieces - at least at each story, it may be also useful to define separate ones close to edges of building, where big variations of pressures may be observed.

It seems to explain all your questions except of 6 and 7. As concerns point 6 - yes, in case of closed building (no openings), it considers only external pressure.

As concerns point 7 , additional features cea be modelled and the load on them will be generated too. It is also possible to do not defined them and to add manually the load corresponding to them.

---------------------------------------------
If this post answers your question please click the "Accept as Solution" button. It will help everyone to find answer more quickly!

Regards,

[Anonymous]

Thankyou both for your quick reply.

I am glad I am on the correct way of working this.

I have now prgressed and have a few new questions

1. After reviewing model 2 I noticed that I have wind load UDL and also edge UDL on slab at each level. Is this correct as it seems strange to me?
2. I have also noticed that the cladding appears as a thickness and the corner has large square where wind has been applied. At wind simulation I only slected cladding. Is this ok or does this affect the results? See below
3. 
4. If the edge loads on elavtion account for "correction" then for my design of the facade elements, I assume I design for worse case. example on y+ model 2 I should design for wind pressure = +1.0kN/m2
5. for coner I should design for
6. If this is correct this will be greater than that as calculated according to the code, assuming Cpe = +0.8 (eurocode Uk) qxCpe = 0.75x0.8 = +0.6kN/m2
7. I have progressed the model and have split my panels to each floor height and also split the length into approx 0.2xe according to code. I understand that I can add the feature load separately but the reason for the model is to assess the load on these elements. Can you review again and confirm if the loads are cmodel is correct and loads seem coorect?
8. Is there anyway of highlighting worse case for each panel.
9. I know I can open specific cladding in new window but is there anyway of selecting specific panel and showing load maps without the complete model map appearing.
10. Any additional comments that you can suggest?

Link to updated model

http://we.tl/0IXmIofW8L

Stefano, I do think it would be a grat idea to add internal load and I will add kudos to your idea.

I have now a large tower beside this building and thinking of adding it to see the effect of funneling but I need to know if the above is correct.

Thanks

[Anonymous]

@StefanoPasquini6790

I cannot find the proposal for interanl combination for wind load can you provide a link for it for me to give kudos.

Thanks

[StefanoPasquini6790]

You are perfectly right,

I forgot to write this in Idea Station, but I Wrote it in this post: http://forums.autodesk.com/t5/robot-structural-analysis/wind-loads-generation-on-symmetrical-buildin...

SOrry!

[Anonymous]

Hi just wondering did anyone get a chance to review my questions, I hope to upload my new model on Monday.

Thanks

[Pawel.Pulak]

Explanations/comments to specific points of your post

---

daniels wrote:

1. After reviewing model 2 I noticed that I have wind load UDL and also edge UDL on slab at each level. Is this correct as it seems strange to me?

---

No it is  not correct. I have noticed it in the existing load cases in your model "Building without fins_model 2.rtd" but I was not able to reproduce it when generating new wind simulation for the same file. What parameters have you used when generating these simulation load cases - which objects were selected for simulation, haven't you made any modifications to objects after generating these load cases?

---

@Anonymous wrote:
I have also noticed that the cladding appears as a thickness and the corner has large square where wind has been applied. At wind simulation I only slected cladding. Is this ok or does this affect the results? See below

 

---

The effect like above was observed only in "Building without fins.rtd" and in "Building without fins_model 2.rtd" models. It was not observed in "Building with fins.rtd" model. In first 2 models simulation was run for whole structure, while only for the 3rd one it was run selecting only claddings.

The thickness on the screen capture above is not corresponding to the cladding but to the width of horizontal bar,which was also considered in the simulation.

But returning to the thickness of the cladding - even if the cladding has no thickness in wind load simulation the thickness of 50mm is assumed - see my explanations from this post:

http://forums.autodesk.com/t5/robot-structural-analysis/terrain-type-wind-simulation/m-p/5356745#M27528 

---

@Anonymous wrote:
If the edge loads on elavtion account for "correction" then for my design of the facade elements, I assume I design for worse case. example on y+ model 2 I should design for wind pressure = +1.0kN/m2

---

integrated pressure distribution on the face = total force on the face = face area* uniform pressure on the face + "correction" on edges 

It means that neglecting the "correction" will change the resultant force.

---

@Anonymous wrote:
Is there anyway of highlighting worse case for each panel.

---

No, there is no such feature

---

@Anonymous wrote:

I know I can open specific cladding in new window but is there anyway of selecting specific panel and showing load maps without the complete model map appearing.

---

There is no "automatic" way to limit the display of pressure maps - but it is possible to set the general clipping planes - as in this video:

---------------------------------------------
If this post answers your question please click the "Accept as Solution" button. It will help everyone to find answer more quickly!

Regards,

[Anonymous]

@StefanoPasquini6790

@Pawel.Pulak

Sorry for annoying you both again but I have been questioned on the wind simulation in Robot. As you are aware I have been trying to model the triangular building with fins. I wasn't sure if I was getting correct results so i have tried to model the building as a rectangle without any fins. I found I was getting strange results and in some cases higher than the code values which I thought was very strange. (code used is BS EN 1991-1-4 in UK)

I have attached the model and some PDF's marked with some questions. I have concentrated on 2 directions +X and + Y. The questions I have are as follows:-

1. Why is the display colour for the simulation not symmetrical? As you can see it means the suction on the corners are not symmetrical
2. Comparing code values to robot values I get in some cases higher than what I was expecting. See attached for both X direction and Y direction. I was expecting lower values than the code especially for the wind pressure. see Y direction

4.  On the Y direction the max wind suction is not as expected - near the front corner. Can you please explain?

5.  In the wind Load on Y direction the pressure scale is very different if you review the color layout.Can you please explain?

Link for robot model:

http://we.tl/RaovKQoV7X

I have attached the code

Can you please review?

[Anonymous]

I should have noted I am only selecting external panels when running simulation and inputting wind velocity. This is what all is based on so far.

However when I calculate with wind pressure - I get very different results for wind suction and visual scales. Can you confirm why?

[YarUnderoaker]

May be becouse of that http://forums.autodesk.com/t5/robot-structural-analysis/terrain-type-wind-simulation/m-p/5356083#M27...

[Anonymous]

Unfortintely, no, this does not help.

[Anonymous]

Unfortintely, no, this does not help.

Any other ideas anybody??

[Anonymous]

Hi @Pawel.Pulak Is Robot looking into this??

[Pawel.Pulak]

Apologies, can you enable the link to download the file again.

The link http://we.tl/RaovKQoV7X given 2016-01-19 already expired.

Regards,

[Anonymous]

Thanks @Pawel.Pulak

See below link

http://we.tl/t1NxZwciVx

[Anonymous]

@Pawel.Pulak

Any update?

[Pawel.Pulak]

Apologies for delay.

I have made some tests and collected some info. I will post it on the forum tomorrow.

Regards,

[Pawel.Pulak]

First thing is to know what assumptions are used and what effects can be considered in the wind load simulation available in Robot.
This module uses the same calculation engine as Autodesk Flow Design (3D simulation part of it). Basic assumptions and properties of it are available there.

As it can be seen in this description it is a transient flow solver which also considers vortex shedding and turbulences.
Knowing about it the effects observed for your model (symmetrical, rectangular, box-like building) may be explained easier.

1/ Not symmetrical maps of pressure on front faces of symmetrical, rectangular model for wind directions X+, Y+, X-, Y-.
It seems related to non-symmetrical discretisation in voxels by the calculation engine. Dev team works to improve it.

2/ Pressure values different than code ones, max pressures not at the front corner of walls parallel to wind direction and no symmetry of pressures between walls parallel to wind direction

Code values are the values averaged for typical, stationary situations.
Flow calculated in Robot is transient, changing in time and depending on geometry and wind speed this variability may be permanent (not converging). For instance vortex shedding in various locations may result in almost permanent disturbing the symmetry of pressures.
The automatic criterion to stop simulation at deviation factor= 0,50% was adopted for simlicity as the efficient one for the majority of situations. But there are situations, where the rate of flow changes is very irregular and this criterion does not result in stopping the simulation at the most representative step.
In such situations it is possible to run simulation with "Manual" setting instead of "Automatic" and to use "Generate loads now" button to generate the loads for the flow state selected by the user.

I hope these explanations help to some extent.

Best regards,