Terrain type, Wind simulation

[Jonsereds]

Hello,

I have start to use the new Wind load simulation in RSA 2015.

Im wondering if you can set the terrain type fore the wind loads in the wind load generator?

(Similar that you can do in the wind and snow 2D/3D generator)

Br. Jesper

[StefanoPasquini6790]

Hi,

this results are very bad...Hope for Adsk voice about it.

The AFD is downloadable only as a student release to this link: http://www.autodesk.com/education/free-software/flow-design

Bye

[YarUnderoaker]

Confirmed. For opposite directions are obtained different results for symmetric structures.

I also noticed that a presure map scale can be normaliized only during simulation process. If you do not normalize scale at the end of simulation  than it looks wrong when you turn on presure map.

[Pawel.Pulak]

Some ADSK voice:)

As concerns the effects in wind simulation discussed by Rafacascudo and StefanoPasquini6790 they can be divided in 3 categories:

1/ Differences between panels and claddings: both these objects are treated in wind simulation not as surface objects but as volumetric objects with some thickness. In case of panels it is simply the thickness of specific panel. In case of claddings it is fixed to 50millimeters. Pressures are generated on faces of these volumetric objects exposed to wind. During the final generation of loads the pressures on faces corresponding to thickness are neglected. So the real geometry, for which the simulation is run considers thickness of panels. It can be seen zooming with pressure maps displayed - as below:

So changing the thickness of panels or using claddings instead of panels changes the geometry of the body for which wind simulation is run and can influence results.

2/ Changing the direction of wind to the opposite one for symmetrical structure can result in change of results. These differences should not be very significant but in general case that cannot be avoided for the calculation engine used.

3/ For the models supplied by Rafacascudo in for instance this post some abnormal results are observed for Y directions of wind. These anomalies (positive pressures where negative pressures should be observed) can be observed also when generating wind for all panels or claddings and not only for selected windward walls. It requires analysis in dev team. The biggest differences (significant ones in comparison to point 2/) for opposite directions were related  to these anomalies. 

Regards,

[StefanoPasquini6790]

Hi Pawel,

As always your answers are clear and on target. Starting from the strange behaviour of the Rafacascudo model, tomorrow I will try to rotate the model of 90°, changing the directions, so Y become X and we can see if there is other surprise to evaluate by the dev team. I will try also to investigate what happens if we push wind only on one facade (for example y+) and then only on the opposite facade.

Greetings and thanks a lot for your professional answers in the forum.

[Rafacascudo]

Hi Pawel,

I was pretty sure , you guys were looking after this! When the answer is not fast ,immediate ,as normally is , some investigation is going on...

I think that what we really need is a clear guidance on what we have to do to get this wind analysis done properly  with results we can trust.

- Opposite directions wind on a symetric structure should give  identical results. If this is not possible due to analysis engine , then which direction should we use ?

- Never came to my mind that the panel thickness would play an important role in the results. Should we use the real facade thicknesses or cladding is enough to get trustable results??

- 0,5% is good enough or lower values will give better results? Discrepancy in results can be huge  comparing 0,5% with 0,25% or 0,1% !!!

- Does the " parent" software , autodesk CFD , gives these same results we see in Robot??

thx!

Stefano , looking forward your test results!!

[rajus]

Hi Stefano, I was wondering if you could please help me in the wind loads simulation using RSA 2015 for Australian Standard AS 1170.2. Before I going into the detail computation, I have modelled a single column (Rectangular hollow section 250mm x 150mm x 10m high. 9mm thick shell). I have used 45m/s wind and generated wind pressure/load for 4 directions. The wind profile is left as default; 1.00 at 0.00m height and 1.00 at 20.00m height. Following is the output from the analysis. Wind Simulation X+ 45m/s pZ(loc) = -0.16 pY(loc) = -0.00 Wind Simulation X- 45m/s pZ(loc) = 0.17 pY(loc) = 0.00 Wind Simulation X+Y+ 45m/s pZ(loc) = -0.19 pY(loc) = 0.33 Wind Simulation X-Y+ 45m/s pZ(loc) = 0.14 pY(loc) = 0.24 I have two questions; 1) Why the wind pressure are not similar between X+ & X- and X+Y+ & X-Y+ wind simulation even though the column section is symmetrical. 2) Is there any way to account for surface roughness of the column/elements during the wind simulation? For example, member with rusted surface or smooth paint will have different drag coefficient and hence different wind loads/pressures? Regards, Raju

[rajus]

Sorry for my previous query, it just became unreadable due to copy-paste from word. Hopefully following will come better.

Hi Stefano,

I was wondering if you could please help me in the wind loads simulation using RSA 2015 for Australian Standard AS 1170.2. Before starting any complex model, I have modelled a single column (Rectangular hollow section 250mm x 150mm x 10m high. 9mm thick shell). I have used 45m/s wind and generated wind pressure/load for 4 directions. The wind profile is left as default; 1.00 at 0.00m height and 1.00 at 20.00m height. Following is the output from the analysis.

• Wind Simulation X+ 45m/s

        pZ(loc) = -0.16

        pY(loc) = -0.00

• Wind Simulation X- 45m/s

       pZ(loc) = 0.17

       pY(loc) = 0.00

• Wind Simulation X+Y+ 45m/s

       pZ(loc) = -0.19

       pY(loc) = 0.33

• Wind Simulation X-Y+ 45m/s

      pZ(loc) = 0.14

      pY(loc) = 0.24

I have two questions;

1. Why the wind pressure are not similar between X+ & X- and X+Y+ & X-Y+ wind simulation even though the column section is symmetrical.
2. Is there any way to account for surface roughness of the column/elements? For example, member with rusted surface or smooth paint will have different drag coefficient and hence different wind loads/pressures?

Regards,

Raju

[StefanoPasquini6790]

Hi rajus, this is my answers

 1.   Why the wind pressure are not similar between X+ & X- and X+Y+ & X-Y+ wind simulation even though the column section is symmetrical.

This issue is probably solved in next hotfix or in next versions of Robot. Is it clear that the problem is caused by the wind speed, try to run an analysis with 20 m/s wind speed and make the same comparison, for more accuracy set your units to Newton instead of KiloNewton.

2.   Is there any way to account for surface roughness of the column/elements? For example, member with rusted surface or smooth paint will have different drag coefficient and hence different wind loads/pressures?

You can do this only with a coefficient that define your wind speed, I don't know your wind design code so I can't be more accurate.

Only one notation, don't expect from wind simulation the perfect design code loads, this is a FEM analysis and the accuracy of the mesh can cause a little differences in load generation. Before you make a structural design you have to make also a comparison between the design code wind loads and the wind simulation ones, Only if the wind simulation loads are conservative you can use it in your model!! Be careful

Hope I help you, greetings and good luck

[rajus]

Hi Stefano,

Thank you for your promt response. After your feedback I have run few models with different wind velocities (45m/s, 20m/s and 10m/s) with force unit in kN. The wind load results from opposite wind direction becomes comparable as we reduce the wind speed. However they are not comparable for symmetrical wind direction even with the lower wind speed. Please find attached extract from my three model. I have also attached a model with 20m/s wind simulation. In the spreadsheet, I have added resultant wind load myself for diganol wind based on vector combination.

As you have mentioned there is some issue with the current version of the Robot (or may be it is limited to higher wind speed only?). If that is the case, It looks like we need to wait unit Autodesk fixes the problem. But I wonder why Autodesk would use an algorithm which will produce different results for symmetrical structure with symmetrical wind direction or is there something I am missing.

You also mentioned in your earlier post regarding the white paper from Autodesk in relation the Wind simulatio and wind tunnel test; do you have a handy link to this documentation.

Regards,

Raju

[AhmedZA]

---

@rajus wrote:

As you have mentioned there is some issue with the current version of the Robot (or may be it is limited to higher wind speed only?). If that is the case, It looks like we need to wait unit Autodesk fixes the problem. But I wonder why Autodesk would use an algorithm which will produce different results for symmetrical structure with symmetrical wind direction or is there something I am missing.

read this thread

http://forums.autodesk.com/t5/robot-structural-analysis/wind-loads-generation-on-symmetrical-buildin...

You also mentioned in your earlier post regarding the white paper from Autodesk in relation the Wind simulatio and wind tunnel test; do you have a handy link to this documentation.

Wind Load Simulation Whitepaperand wait  Stefano to give you more instructions

---

---

[rajus]

Hi Ahmedeffect,

Thank you for the link, I will definately go throught it and do some other research as well. Do have any example to use this sort of simulation into the design based on any wind design code. We want to use the design for Australian code AS 1170.2. However if we can get an example or step by step procedure for American code ASCE 7-05, then we can translate into other codes.

Also, I look forward to any other commets/suggestion to my previous post.

Thank you,

[AhmedZA]

Hi
Sorry right now I am working on RSA2014 and I don’t have any verification example regarding to this issue so I would suggest you to ask StefanoPasquini6790  - Pawel Pulak or as rafal said (soon )

Let me add a few points to what has been raised as we know there is some different between analytical method based on code and wind tunnel and according to robot help file WS acts as wind tunnel because Codes doesn’t consider (the aerodynamic effect of the actual shape of the structure - the influence of adjacent buildings and topography - detailed wind directionality effects - aeroelastic interaction between the structural motion and airflow )
So I would suggest to give us new white paper and it contains on comparison between analytical method (CODE ESTIMATES) and Wind tunnel (WS) PREDICTIONS as shown below

[rajus]

Hi ahmedeffect,

Thank you for your reponse and comparasion of the results. It appears that the Robot WS is closer to the codal provision. Howevere I am still interested to know why the wind loading for symmetrical structure gets different on symmetrical wind direction? May be we need some help form Robot team.

Thank you,

[RSAJunior-]

Hi @StefanoPasquini6790, I have tried with both my model and yours, but every time the max pressure on element it is 0.45 kPa. 

Cal you explain better what do you mean for "The trick, as I've explained in the previous post, is to set the pressure value to 1065 N/mq (as shown in the following pic), set this value to 1 and refer all the curve value to this."

In the CNR-DT 207/2008 I am able to calculate the force to be applied (forza su ciascuna falda) and the total pressure (Pressione complessiva sulla copertura). The results are different than RSA. Can you help me please?

[RSAJunior-]

@StefanoPasquini6790 Can you also share with us a model where you have calculated the value in both RSA and by hand, because the difference pressure as I wrote before between RSA and CNR-DT 207/2008 is huge!

Thanks!

[mortenx83]

why is not possible to input Terrain type 1.2.3.4?

[kshephman]

Hi Stefano

Thanks for this guidance, I do have a query though. I had initially thought that the wind pressure calculation should exclude only the projected are of the structure so that the pressure acting on the structure then gives the force in RSA. However, when calculating wind load, there are usually certain drag/force factors involved that are related to the shape and dimensions of the structure.  You stated that the pressure should be calculated upto about twice the height of the structure but in the upper half of this calculation, there aren't any applicable drag/force factors. This made me think that perhaps you are only considering a product of the velocity pressure and an applicable gust effect factor in that calculation. First of all am I thinking correctly in this regard and secondly if I am, how do you then incorporate the drag factors into the load generation so as to get the actions stipulated by the code?

Regards

[alvaro.conde]

Hello, 

I know it is a long time since nobody posts in this topic, but after reading all the forums on wind simulation I still haven't got it. I have an excel spread**** with the exposure factor (Ce (z)) for all the 25 stories of my building, as well as a height twice as high (197 m in this case). I also have a column with the velocity factor (=Ce^1/2, as explained in previous posts).  I know the basic wind velocity (vb) and the basic wind pressure(qb).

My questions are:

Are the two alternative inputs (wind velocity/wind pressure) the basic wind velocity/pressure?

If I were to use the method recomended by @StefanoPasquini6790 by writing as an input the basic pressure in kPa instead of the wind velocity, in which way is that better than writing the velocity?

And third and last, as I enter the wind profile (in terms of wind pressure), do I just have to enter the Ce (z) for each height or is it necessary to scale it down so that it is equal to 1 in the base/top??

I've seen posts here setting the maximum value to 1 so that all the rest are greater than 1 and others setting the value on the ground level to 1.

Thank you very much