Hi. I am an architect, and in last few months, am having a great difficulties at my project in calculating the stress from wind on my roof.
I heard that Solid works can do this with a great accuracy.
So are there any tutorials about this?
My roof is pretty unstandard - two curved surface.
I must admit I know nothing about Computational Fluid Dynamics/
Thank you for all the answers.
In Australia this kind of analysis is normally undertaken by structural engineers. Have you spoke to any structural engineers on the matter. They would normally design to the required statutory regulations ie. codes. If the codes don't cover the shaped roof in question CFD may be useful but would probably need to be validated using physical testwork or whatever else the engineer or code determines is necessary.
There may be engineers on this site who can use CFD to determine the pressure distributions on the roof resulting from the wind.
SolidWorks is primarily a package developed for the design of mechanical plant and equipment and is often not efficient when used for doing architectural/structural work.
thank you for reply.
The problem is that I do not own any other CFD software (like Ansys CFD maybe?) nor I will do in the next couple of months, because simply I do not have a money to buy it. So I will have to do it with Solid works.
I am a little bit of a structural engineer myself, so I do this kind of stuff. The problem is that the roof is so unstandard - curved, that there is no other way to calculate a wind impact, then to make an simplifiction. But even in this case it is difficult to establish the right sipmplified shape. That's why I need Solid Works
I hope some of those engineers you mentioned will give me some kind of help.
An example of my roof can be seen in the attachment
I disagree with Derek Bishop's assertion that implies that SolidWorks Flow Simulation is not suitable for this type of work. Flow sim would do as good a job as Fluent or any other general purpose CFD code for this type of external aerodynamic work with one key caveat. The only issue you may have with Flow Sim is that its turbulence model for the computation of drag or tangential forces loses accuracy above 6- 8 million R'number. Not sure about the scale of your building hence the uncertainty on that one but it is likely to be an issue if it is big. The purpose of the computation is to get a get a building permit I would assume. Selection of an appropriate safety factor could probably do the trick. Given that there are a large number of uncertainties, not the least of which is a new building is erected nearby in 5 years time getting overly precise doesn't make much sense in practical terms as far as precision and accuracy is concerned. The object seems to me is build in adequate margin to enable a robust structural design over the structures expected life time. There are specialized codes for these kinds of calc's that are used at firms that specialize in this type of work like RWDI (in Guelph Ontario Canada) where modelling of the planetary B'layer is accounted for as is the surrounding landscape. Depends what your building is worth and the environment it is going into would probably determine what tools and what effort go into examining the loads on the building.
Bill, you need to read my posts a bit more carefully. I stand by my comments having used the software on a few small architectural projects. In many cases and in many ways SolidWorks is not an efficient tool for doing architectural and structural work. I work with structural engineers all day and they reassure me there are much better and more efficient packages for doing structural analysis and modelling than SolidWorks. There are exceptions and I've used SolidWorks in one such case with some good results. There are some companies developing software for this application in partnership with SolidWorks (eg. SolidAce BuildWorks). It will be interesting to see how they go.
Have you used Flow Simulation to model wind pressures on a large buildings and structures? If you have and found it suitable, then great. I'd love to hear about your experiences. Having been completely confused by the Australian wind codes in carrying out some preliminary structural analysis and somewhat stunned by the limitations of the cases presented in the code as a basis for design I believe that a CFD package that can give some half decent answers would be useful.
Here are a few small architectural jobs I've worked on in SolidWorks:
Hi Derek,
I reread your post. I took the comment "SolidWorks is primarily a package developed for the design of mechanical plant and equipment and is often not efficient when used for doing architectural/structural work." as a comment on Flow Simulation as opposed to SolidWorks since it was a Flow Forum question. I do agree that SolidWorks as a CAD package has some serious productivity challenges in the area of civil or building type structures and the beam analysis capabilities has a vast array of somewhat subtle technical short comings in this area as opposed to say STAAD or maybe even S-Frame. On the upside it can be quite fast to get an analysis model but it isn't all that much good when you can't get it to model the reality you want, with the subtleties you require. Rumor has it that cleaning this up is on the agenda for the 2012 release. Though if history is any guide it will leave one wanting but no doubt will be better.
My comments were directed solely at Flow being able to predict a decent approximation of the pressure distribution on the roof/structure which it can do with out a great deal of difficulty. Transferring the loads to the structural package is another matter and has some issues in that it does not currently support shell elements/surfaces, which can make it a bit tricky to get the loads in a structural model of this type.
I hadn't noticed it was in the CFD forum and as you note my comments related to other aspects of the design as the OP seemed to make reference to these. Anyhow, I'm interested to hear of your experiences modelling wind loads on large structures.
I cold list a number of areas where SolidWorks falls down in regard to this type of work. My observation is that the problems result from it not being a core area of the software and that in turn results from it being primarily a tool for mehanical designers and engineers designing relatvely small items of mechanical equipment and plant. From where I sit there are a few things that could change to make it a useful tool in some of these other areas, but I guess there are limits to how much any package can be reasonably expected to achieve.
Hi Derek,
there are efforts being made to make SWX a better design tool for "mechanical buildings and mechanical equipment integration into buildings". In the 2011 release the addition of a "grid" is being implemented to support that particular referencing scheme typically used in the design and detailing of buildings. Obviously, the addition of certain "features" like walls, which may appear as lines in a drawing or slabs in a 3D assembly but when looked at in detail and are composed of various "parts" like framing, drywall, insulation etc are not going to be done for 2011- I don't even know if they are planning them but it would seem like a logical thing to do if you want to make some head way in that space. Also the drawing package would need to support the particulars of a typical building drawing scheme like all the wall details one one page, the door details on another - we wouldn't want the civil guys to get out of the habit of flipping all over drawing sets to get a handle on some detail like say an indoor water feature. While you could do this sort of drawing now it would fall in the full pain in the you know where area - at least at my skill level. I would surmise that their strategy on this front is something like, well let's see if we can lure the civil guys to some tool connected with us (Draftsight), which in addition to getting us connected with the right users in a positive manner - giving the thing away - it also has the potential of inflicting damage on our main competitors major cash generator in this area. By the time we get some traction going in that area the need (demand really) for 3D in that space may well have emerged and by that time we hopefully will have added enough horsepower to (SWX or even CATIA) be a contender in the BIM space. Construction is the worlds largest industry and the way it is currently practiced leaves a lot of room for improvement compared to the mechanical tools (sensibly adopted to the practice of buildings/civil structures which isn't done yet), in my humble opinion. I agree it is a different kettle of fish and work needs to be done. However, the advantages of 3D, assuming the complexity problem can be managed (not a bad assumption in my view, given enough time), are quite compelling. It will drive significant changes to the supply chain as the opportunity to charge for coordination error driven change will fall quite a bit. Thus the cost of building will drop accordingly. Obviously other advantages emerge as well such as leveraging 3D data in analysis and to some extent manufacturing.
Well enough of that. As far as wind modeling of structures with Flow sim goes - as a first approximation it is probably a pretty decent tool. I have done a hockey rink flow simulation the goal of which was to identify low velocity areas where snow would accumulate which in turn would be used to assess snow loading distributions. Essentially is a uniform load distribution a sensible assumption? Does non-uniformity lead to any sort of issues? You can design a structure to take a fairly substantial load which is evenly distributed but will have serious issues if the load is along way from uniform even at reduced magnitude. Same would apply for pressures. In my experience which I have to admit is pretty limited when it comes to building codes, is that the codes through out some loading like 40 PSF or the 100 years storm numbers, which may or may not be anything close to reality. Which is why engineers in this area, when things get expensive (like building a large tower), might do some flow modeling to gain insight in specific areas, but when it comes to validation they are going to head into a boundary layer wind tunnel and run some measurements. The 40 PSF number or a Fow simulation result is probably good enough for a solar panel array, a house, a factory, a billboard....whatever. You can see how well this works, and it isn't just a load specification problem, it can be a construction problem as well, when roofs blow off in storms, trailer parks get trashed etc. But it isn't very often a big building goes down - they may lose some glass panels but even that isn't all that common.
Got to go.
Hi, I'm in Korea and mechanical enginneer on CFD Field.
First of all, If you want to solve your problem, you must think the calucation condition on your problem.
For example, Ambient temperature, wind speed and direction(X,Y,Z), humidity and so on.
Take advantage of this oppouinty to collaborate wity international partners. ^.^;;
If you want to detail inforatiom, please send e-mail for me. (hhkim504@yahoo.co.kr)
Best regards!!
A large "computer" wind tunnel model can be created on Gambit and then the mesh generation resulted by Gambit can be exported to Fluent (ANSYS). Take a look at those two applications they can handle large models, I use those to analyze wind loads on large membrane tension structures. Simulation will take for ever to analyze a model of that size (unless you can use a scale model). Remember that to simulate wind load on buildings you must consider any sorrounding structures.
To analyze large buildings formed by complex surfaces sorrounded by other buildings, normally you build an scale model of the buildings placed over load cells sensors and place them on a wind tunnel .
Thank you for the answers.
I am aware that for a precise soultion I must involve all the surrounding buildings, trees and objects. But I do not want to do that. I just want to make a simple wind pressure calculation on my roof, with no surrounding objects.
Are there any FAQs or tutorials about this?
Mauricio Martinez-Saez wrote:
To analyze large buildings formed by complex surfaces sorrounded by other buildings, normally you build an scale model of the buildings placed over load cells sensors and place them on a wind tunnel .
In Australia this kind of analysis is normally undertaken by structural engineers. Have you spoke to any structural engineers on the matter. They would normally design to the required statutory regulations ie. codes. If the codes don't cover the shaped roof in question CFD may be useful but would probably need to be validated using physical testwork or whatever else the engineer or code determines is necessary.
There may be engineers on this site who can use CFD to determine the pressure distributions on the roof resulting from the wind.
SolidWorks is primarily a package developed for the design of mechanical plant and equipment and is often not efficient when used for doing architectural/structural work.