Composite Steel/Wood Space frame analysis

Discussion in 'Software' started by lawrencepeckham, Nov 1, 2006.

  1. lawrencepeckham
    Joined: Nov 2006
    Posts: 2
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: UK

    lawrencepeckham New Member

    I am working as structural analyst on a large epoxy/wood project. We have laminated mahogany frames and deck beams which are reinforced by a stainless steel space frame in mast/chain plate/keel area. I am looking to analise deflections under jacking loads of the space frame. Could anyone advise on suitable software/books for this analysis.

    Thanks,

    Lawrence
     
  2. alpamis34
    Joined: Jul 2006
    Posts: 63
    Likes: 0, Points: 0, Legacy Rep: 2
    Location: A city that never sleeps-Istanbul

    alpamis34 Senior Member

    Ansys
     
  3. lawrencepeckham
    Joined: Nov 2006
    Posts: 2
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: UK

    lawrencepeckham New Member

    Unfortunately is a bit beyond our budget, do you know any specialist yacht FEA programs that are a little cheaper?
     
  4. CGN
    Joined: Jan 2003
    Posts: 547
    Likes: 9, Points: 18, Legacy Rep: 138
    Location: MX

    CGN Senior Member

    have a look at Cadre analytic maybe it can do the job or Fesawin?
     
  5. Murdock
    Joined: Aug 2006
    Posts: 24
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Brown & smelly Water nowadays...

    Murdock Junior Member

    FEA for free... yes, just check what fits your needs at
    http://www.itspanish.org/modules.php?name=Downloads
    on the download window, go for "Finite", & get 8 results.
    It's an engineering forum no piracy allowed, all free & legal...
    hope it helps
    (I'm the mod there MDP)
     
  6. Andrew Mason
    Joined: Mar 2003
    Posts: 397
    Likes: 18, Points: 18, Legacy Rep: 206
    Location: Perth, Western Australia

    Andrew Mason Senior Member

  7. nero
    Joined: Aug 2003
    Posts: 624
    Likes: 13, Points: 0, Legacy Rep: 112
    Location: Marseille, France / Illinois, US

    nero Senior Member

    Structurix is simple straight forward and makes simple picts with displacements with color coded stress.
    http://mac.softpedia.com/get/Math-Scientific/Structurix-X.shtml
    This is the Mac OSX version.
    It does 3D beams so you could use it for a space frame.

    So far I have used it to design a hoist and a boat trailer in steel. It is free and very simple to learn. Tech support is available and quick.
     
  8. darbikrash
    Joined: Nov 2006
    Posts: 7
    Likes: 1, Points: 3, Legacy Rep: 12
    Location: Southern California

    darbikrash Junior Member

    I'd be wary of "free" or low cost FEA programs. Much of the value of modern FEA codes is in it's longevity, with many years of testing and refinement correlating analytical models with physical test results and validating the code.

    Also, the real cost of FEA programs is not the cost of the software, it's how much it costs you to learn how to use it correctly. This is not something you are going to learn in a few evenings with a users manual.

    Some of the challenges you are going to face include some rather tricky aspects of FEA, and I would suspect you are squarely in the middle of a large, non-linear problem. Consider:

    - You will have material non-linearities with the mahogany multi-layer laminates. The epoxy/wood inteface will have to be modeled as a multi-layer laminate, with simple shear, compressive, and tensile properites, all which will be different in each case.

    - You will have boundary non-linearities as the model will likely rely heavily on contact analysis to transfer the load from the space frame to the wood/epoxy structure. Most codes have quite a bit of trouble handling contact analysis.

    - The pre and post proecessor will have to have specific functionality for multi-layer laminates, including directional oreientation and allowance for the epoxy properties. The element library will have to have robust support for these types of materials.

    These considerations would indicate that a more advanced solver be used, the previous suggestion for ANSYS is a good one, also look at NASTRAN/PATRAN, and in my opinion the best tool, MSC.MARC.

    You might consider partnering with a software reseller that can help with the current project, and provide some training and support so you can do subsequent projects with decreasing reliance on outside help.

    John
     
  9. nero
    Joined: Aug 2003
    Posts: 624
    Likes: 13, Points: 0, Legacy Rep: 112
    Location: Marseille, France / Illinois, US

    nero Senior Member

    There appears to many levels of FEA programs for their complexity and their precision.

    If one can break out individual parts or systems then a free simple more or less spreadsheet can provide close results. The more one asks of the software the better FEA program is needed. Structurix gives a starting point of how the force is carried and what size members will carry them. Not being a mesh based graphical solver eliminates a lot of the complexity. Down side is it doesn't make cool pictures.
     
  10. alidesigner
    Joined: Nov 2006
    Posts: 187
    Likes: 6, Points: 0, Legacy Rep: 93
    Location: Australia

    alidesigner Senior Member

    FEA can be a great tool but if you want to analyse a space frame structure, and depending on your saftey factors, you may not need it. If you want plate stresses then you will need it but for space frame deflection analysis you can start with simple 2d beam theory. The fact that you have composite beams shouldnt scare you. Any good Mechanics of Materials book will have a section on dealing with composite beams using the transformation method. By comparing Youngs modulus and section properties of each material you develop an equivalent section and analyse it as any other beam. (You should so this as a first step anyway to cross check the results of any FEA or beam software analysis). The next best method would be frame analysis using a 3D beam analysis program. They are easy to use and you can create a model very quickly. But if you are trying to deisgn the most lightweight and efficient structure then an FEA model that includes the shell plating will give you the best results, provided that you know how to drive it - garbage in = garbage out.

    So before you spend time and $$$ on FEA first consider if you really need it.
     
  11. darbikrash
    Joined: Nov 2006
    Posts: 7
    Likes: 1, Points: 3, Legacy Rep: 12
    Location: Southern California

    darbikrash Junior Member

    Great points, I agree totally.

    There is a tendency to believe that very complex problems can be solved quite easily by just buying some FEA code and "running the numbers". It is likely that more failures are developed from this belief than if people would stay away from FEA altogether.

    The decision to use FEA is a simple one, estimate the cost to build an physical prototpye of sufficient fidelity to allow meaningful testing, and compare this with the cost of having an experienced analyst perform the test in a computer simualation.

    It's not always cheaper to do FEA, sometimes a physical model can be cheaper. FEA performed by a qualified analyst in the traditional method on a high end solver is very, very expensive.

    And certainly, hand calcs are the prerequisite of any formal analysis.

    Having said this, there are accuracy limits to what can be acheived with hand calcs, and a question arises if the output of this effort is sufficient to validate the design. The anwer to this is often industry specific. In aerospace, or any (failure) risk averse field, basic beam sizing efforts (manual or computer generated) are insufficient to justify release of a design. The cost liability of a field failure is much too high.

    To be sure, a first order approxiamtion can be acheived using the methods described previously, but the value of FEA is not neccesarily in using the computer to size beams or plates. In the traditional use profile, it is more useful to determine localized hot spots and as a visual aid to optimizing the design. Additionally, things like Charpy notch sensitvity can be evaluated at key load connections and more detailed failure modes can be explored, rather than just global sizing excersies.

    Newer approaches to FEA no longer use this traditional approach, as it is not predictive, rather is almost forensic as it is only applied after the design is near completion. New topography FEA tools (such as Altair's Optistruct) use an inverted use profile, e.g. the FEA is performed first, before the design begins, and the FEA ouptut drives the design.

    Using the example of a space frame, this is done by modeling a very coarse envelope which encompasses the available space in the assembly that the frame can occupy. The shape is a primitive, rectangular or as appropriate. This 3D envelope volume is meshed, constraints (boundary conditions) applied, material properties applied, and lastly the loads applied. Then, significantly, solution constraints are applied which might force the solution to consider that only 1/4" thick plate walls be used, for example, or a minimum bend radii, or a maximum deflection, or a direction of draw specified (for a molded part), or whatever is appropriate for the design intent and manufacturing method.

    The solver will then return a solution which will "turn off" the elements that are not directly contributing to the structure, and the resultant file will be a coarse, but geometrically complex, mesh "template" that is then given to the designer as an overlay for detailing. Global beam and plate sizing, attach points, maximum deflection critieria, max stress, etc, etc are now determined before the design even begins. If the designer follows the template closely, we have reasonable assurance that the resulting design will comform completely to the design loads.

    These types of studies are very, very fast, due to the simplicity of the primitive input model, sometimes only taking a few hours if the loads and constraints are available.

    A good text as a primer to FEA is:

    "Building Better Products with Finite Element Analysis" by Vince Adams and Abraham Askenazi.

    For more information on topology and topography analysis see:

    http://www.altair.com/software/hw_os.htm

    John
     
    1 person likes this.
  12. Andrew Mason
    Joined: Mar 2003
    Posts: 397
    Likes: 18, Points: 18, Legacy Rep: 206
    Location: Perth, Western Australia

    Andrew Mason Senior Member

    lawrencepeckham asked about analysing a space frame, which can be done easily in a frame analysis program such as Multiframe without resorting to the complexities of full blown FEA.

    No point using a sledge hammer to drive in a thumbtack.
     
  13. darbikrash
    Joined: Nov 2006
    Posts: 7
    Likes: 1, Points: 3, Legacy Rep: 12
    Location: Southern California

    darbikrash Junior Member

    This does not seem quite so simple as you make it out to be, your response seems flippant and typical of software vendors that trivialize tough problems and present entry level solutions as viable alternatives.

    In my business I see many instances of "casual" users who need to solve problems that at first blush are not particularly difficult, and they underestimate the complexity of the problem and look for a low cost software application to help with the solution. The logic is they do not run up against this type of problem with enough frequency to "justify" the proper simulation tools (or the training to use them). This tendency is fueled by the software industry who would have us believe that anyone with a PC and $1500 worth of code can do structural analysis, much like the CAD world operates.

    I think the question was geared towards suggesting "suitable" software and reference textbooks, and I have tried to provide some discussion towards this end. I think there are some comments in the responses that add some insight into the subject of computer analysis, and list some of the functionality that may well be required to arrive at a satisfactory solution.

    If the original poster would like to provide some more detail on the specifics of the proposed analysis, perhaps some more refined recommendations would be possible.

    John
     
  14. Andrew Mason
    Joined: Mar 2003
    Posts: 397
    Likes: 18, Points: 18, Legacy Rep: 206
    Location: Perth, Western Australia

    Andrew Mason Senior Member

    John

    I suggest you examine the capabilities of Multiframe before you label either my response as flippant or the software as entry level.

    And I also suggest you get a few more than 3 posts behind you before you start name calling on this forum.
     
    1 person likes this.

  15. Mikey
    Joined: Sep 2004
    Posts: 368
    Likes: 7, Points: 0, Legacy Rep: 75
    Location: Bangkok, Thailand

    Mikey Senior Member

    John,
    Welcome to the form, but...
    I suggest that you use the search button before posting like you did, you will find Andrew's posts to be both professional and very helpful, something to take into consideration when posting replies
    Mikey
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.