A Bridge Maintenance Case Study

By Julio Banks

Abstract

A Bridge Maintenance Frame required a FEA (Finite Element Analysis) for a client with the objective of reducing the lateral deflection of the frame from 7” to 3/4”.  I produced the solution within 10 hours using a combination of Mathcad, IronCAD, FEMAP and NEi Nastran FEA software.

This study was required to understand the structural adequacy of a mobile frame and catwalk designed to be used for bridge maintenance and repair by a crew of maximum two occupants.  Its design needed to meet Occupational Safety and Health Administration (OSHA) requirements. The FEA showed that the lateral deflection was reduced from 7” to 3/4” as desired.

Description of the Problem

A bridge maintenance frame was arbitrarily modified by the fabricator. The base of the frame was doubled in width. This is a typical example where no thought was given to the response of the platform such as lateral deflection due to the increase in width of the platform. The original fabricator maintained the single 2x2x0.124 HSS made from steel. During the modification study, the shop proposed the use of double the original 2x2x0.124 HSS without any offset. I promptly demonstrated that a 6” minimum offset would be required to take advantage of the parallel-axis theorem use of the square-influence of the offset dimension multiplying the cross sectional are of the HSS (See Figure 1)

Figure 1

Proposed Solution

Manual calculations using Mathcad indicated that the solution was to offset the vertical members from the bottom of the platform to the top and around the C-shape of the lateral portions of the frame at minimum, six inches. The final geometry is shown in Figure 1.

Method of Solution

The solution was made most effective by the use of the combined application of Mathcad (for manual calculations), IronCAD (for solids modeling), FEMAP (for FEA Pre- and Post-processing) and NEi Nastran (for FEA solution) for the structural deflections and stresses. I have used Mathcad, IronCAD, FEMAP and NEi Nastran for over twelve continuous years and find such combination of analytical tools to be the most cost-effective method of solution especially for solo-consultants.

Mathcad Manual Analyses

The first step in a structural analysis is to determine, at a global level, the geometry and material to be used for a given set of loads. Mathcad is the best cost-effective means of achieving cost-effective and prompt solutions. Once the most promising design candidates are selected, then the generation of detail geometry proceeds to be performed using IronCAD solids modeling.

IronCAD Solid Modeling

IronCAD, which I believe is the best solid modeling software, easily allows me to create a full 3D solid model of the frame. Additionally, IronCAD is especially useful for analyst engineers since it allows the “slicing” of 3D surfaces and solids to accommodate the FE (Finite Elements) entities (nodes and elements). With its short learning curve and ease of use, I can shorten the project time by spending more time on the analysis versus the design aspects.

NEi Nastran Finite Element Analyses (FEA)

In order to effectively ensure the requirements mentioned in the abstract above, I used NEi Nastran FEA software to run a structural analysis.  The mesh consisted of plate and beam elements for the walking surfaces and tubular elements for the supporting structure. A steady-state gravity load was applied, in accordance with OSHA requirements, to represent the weight of the occupants. Lastly, fixed constraints were placed on the supporting structure to represent real world conditions. The FEA model shown in Figure 1 was created using IronCAD and Figure 2 shows the FEA solution.

The model was analyzed with the objective of reducing the lateral deflection of the frame from 7” to 3/4”.  By using the NEi Nastran FEA software, I was able to adjust localized structural properties (such as moment of inertia by offsetting a new HSS member) to produce reactions at the supports, as well as beam and plate stresses and deflections. With these tools in hand, both the customer’s and OSHA’s specifications were met before any prototypes had to be manufactured.

Figure 2

Having used the software for over 14 years, I believe that NEi Nastran is truly reliable software allowing me to make a commitment to my consulting client and deliver without any delays.

Conclusion

This case can serve as a cautionary case due to the ease of use of current FEA software at affordable prices. It is recommended that, at a minimum, companies should have their design reviewed by either in-house consultant or independent consultants to ensure that a FEA user has not made critical errors during the FEA modeling and interpretation of results. This warning is especially of greatest importance if the design is mass produced such as consumer products with potential harm to the users in the event of a structural failure.

About Julio Banks

Julio C. Banks, MSME, PE & CGC is the owner of J. C. Banks Construction, LLC established in 1993. Julio has had direct experience in the analysis and design of thermo fluids (incompressible & compressible) flow as well as structural engineering consulting.  He has worked in diverse industries such as Fossil and Nuclear (Stone & Webster), Military-grade Weapons (General Electric), Turbojet  (Pratt & Whitney) and Turboshat (Pratt & Whitney and Siemens-Westinghouse), Airframes (Northrop Grumman, and Piper Aircraft), as well as commercial and residential construction including General Contracting. Numerical Analysis and Solution Methods is his passion and he considers the subject of computational algorithms a hobby.

J. C. Banks Construction, LLC
P. O. Box 880187
Port Saint Lucie, FL 34988-0187
USA
Phone: 1.772.204.8888
Fax: 1.772.204.8899
Email: BanksJ (at) asme (dot) org

==============

If you wish to contribute to Deelip.com please click here.

  • Mook

    Use of NEi Nastran seems to be overkill to an extreme for this application given that several popular structural analysis programs which include built-in AISC section libraries and design code checks are available at a fraction of the cost of general purpose FEA. Furthermore, general purpose FEA products require time(manhours) to manually define section properties for each of the common structural steel and concrete shapes when those libraries are already built-in to structural software products. Using Risa, SAP2000 or Staad stand-alone without “solid modeling”, that model could be built, analyzed and designed in under an hour, maybe 1/2 hour if well dimensioned drawings were available, using products that cost considerably less (Risa and SAP2000 offer versions to handle this problem at $2,000 or less, Staad is probably not much more). Furthermore, the engineer could design not only for deflection, but also obtain AISC code checks as well which is not an option with general purpose FEA products. FEA has its place, but in this example, it's not the best choice of tool IMO compared to numerous structural analysis software options.

    As for the comment on using IronCAD, why would solid modeling be required in such a design? This is a classic beam/line element problem

  • Mathemagician

    Dear Mook, please note that I have been in the Algorithms
    studies since 1979 and my hobby is computational methods
    using Mathcad as my primary tool of analysis. That is, I have
    31 years of continuous mathematical analysis using FEA and
    closed-form (manual solution) and my nickname is “The
    Mathemagician”

    I would agree with you if the article was intended to promote
    NEi Nastran for the sole purpose of solving such a relatively
    simple structural problem.

    Please note that the objective of the article was simply to
    promote the effective use of relatively inexpensive software.

    Based on your comments, I will have Deelip Menenzes
    authorize me to post additional solutions using NEi Nastran
    to demonstrate that the FEA software is capable of solving
    a “Broad Spectrum of Thermo-fluids and Structural” problems.

    I will answer your criticism one item at a time and wish to
    hear from you what is your opinion of my response:

    Issue:
    “Use of NEi Nastran seems to be overkill to an extreme for
    this application given that several popular structural analysis
    programs which include built-in AISC section libraries and
    design code checks are available at a fraction of the cost
    of general purpose FEA. ”

    Answer:
    It is not an overkill based on the entire scope of the project.
    You have voiced an opinion without sufficient data to arrive
    at a rational conclusion. If the scope of the problem would
    have been a simple deflection, then I would have used VisualAnalysis which I do have a license
    and would be more appropriate. Again, I would not have used
    NEi Nastran nor would have purchased such a General FEA
    for the relatively simple problem as the subject case study.

    Issue (Corrected Spelling in Bold-type):
    Furthermore, general purpose FEA products require time (man-
    hours) to manually define section properties for each of the
    common structural steel and concrete shapes when those
    libraries are already built-in to structural software products.
    Using RISA, SAP2000 or STAAD.Pro stand-alone without “solid
    modeling”, that model could be built, analyzed and designed
    in under an hour, maybe 1/2 hour if well dimensioned drawings
    were available, using products that cost considerably less
    (RISA and SAP2000 offer versions to handle this problem at
    $2,000 or less, STAAD.Pro is probably not much more).

    Answer:
    Again, if the only purpose of the Finite Elements Design
    and Analysis every statement made would be correct,
    however, the model was build for general expansion and
    no profit loss neither due to budgetary time excess, nor
    was there delays due to my choice of the analytical tools
    being utilized in the solution of such a “relatively simple”
    structural problem. Incidentally, on the lighter side of life,
    Risa means “Laughter in Spanish (my native language) which
    makes me think of RISA as a joke although I am not saying
    that it is not a serious FEA software.

    Issue:
    Furthermore, the engineer could design not only for deflection,
    but also obtain AISC code checks as well which is not an
    option with general purpose FEA products. FEA has its place,
    but in this example, it's not the best choice of tool IMO
    compared to numerous structural analysis software options

    Answer:
    I am quite sure and confident that the choice of tools
    and duration of the solution is the correct method. Not
    once in my 31 years of engineering career had I had one
    single design review that was found to be flawed, not a
    single one. Furthermore, not once have I had a project
    cost overrun due to my misuse of allocated time. Therefore,
    I am happy to respond to your concerns with a great smile
    knowing that this day could not have come any sooner as I
    have prepared throughout my career to know what I am talking
    about before I air my concerns about another engineer.

    Issue:
    As for the comment on using IronCAD, why would solid
    modeling be required in such a design? This is a classic
    beam/line element problem

    Answer:
    Incorrect conclusion without sufficient information, which
    was not available in the article.

    I have used IronCAD since its inception and find it the most
    cost effective tool in creating solid models. For over 15 years
    I have never had the need to change from IronCAD to any
    other successful contender. This is my strongest endorsement
    ever !

    Again, amigo, my hobby is optimization solutios using
    numerical methods programming with MathCAD. Additionally,
    I have written over 870 of Mathcad pages for Knovel Corporation in 2008 in less than six (6) months and published
    a PTC World Event paper from such a body of work. If you wish
    to engage in an amicable discussion off-line, please use my e-
    mail provided.

    In closing:

    “A fool ends at the beginning”

    “When the only tool one has is a hammer every
    problem resembles a nail”

    “Fortune favors the prepared mind”

    “To the simple-minded, there is always a solution that
    seems simple but it is ultimately wrong”

    Peace

    The Mathemagician

  • Mook

    Dear Mook, please note that I have been in the Algorithms
    studies since 1979 and my hobby is computational methods
    using Mathcad as my primary tool of analysis. That is, I have
    31 years of continuous mathematical analysis using FEA and
    closed-form (manual solution) and my nickname is “The
    Mathemagician”

    Sounds impressive

    It is not an overkill based on the entire scope of the project.

    In your original problem description, you wrote that the problem with the structural design was “lateral deflection”. So in that context, could you please elaborate on the “scope” of your project which would exceed the capabilities of structural programs such as Risa, SAP2000, Staad or Visual analysis which, unlike programs like NEi Nastran, have built-in libraries of structural shapes + code compliance checking?

    Since you're dealing with AISC shapes including HSS members, any structural design such as this would need to comply with either AISC design requirements which each of the above mentioned structural programs offers.. or possibly AASHTO. You seem to dispute this. Why would code compliance not be an important factor in this particular design?

    I have used IronCAD since its inception and find it the most
    cost effective tool in creating solid models

    This does not address my original question regarding use of IronCAD. My question was why would “solid modeling” be used in a design of this nature? I could understand if you were modeling a structural connection in detail for checking local stresses.. but for a design of steel frame deflection check? I'm honestly curious as to how Ironcad would be applicable in this design.

    “To the simple-minded, there is always a solution that
    seems simple but it is ultimately wrong”

    Is this your way of giving me a backhanded insult?

  • Mathemagician

    Since you're dealing with AISC shapes including HSS members,
    any structural design such as this would need to comply with either
    AISC design requirements which each of the above mentioned
    structural programs offers.. or possibly AASHTO. You seem to
    dispute this. Why would code compliance not be an important
    factor in this particular design?

    The model required lateral deflection and the tools available to
    me were utilized to solve such a “relatively simple” problem.
    Since load interaction (utilization) for instance was much less
    than unity, then simple “3D Beam Elements” are an adequate
    representation of the physical model. As time progresses, I will
    provide additional case studies using the same set of tools in order
    for the interested reader to properly assessed and appreciate the
    diversity of problems that can be handled with the same set of
    tools. Please note that the tools I am advocating are both affordable
    and effective for which I have made a commitment to continue
    to use them for over 14 years.

    I have used IronCAD since its inception and find it the most
    cost effective tool in creating solid models

    This does not address my original question regarding use of
    IronCAD. My question was why would “solid modeling” be
    used in a design of this nature? I could understand if you were
    modeling a structural connection in detail for checking local
    stresses but for a design of steel frame deflection check? I'm
    honestly curious as to how IronCAD would be applicable in
    this design.

    IronCAD has a unique tool called the Tri-Ball. Please research
    it and then let me know what other 3D CAD software has the
    ease-of-creating 3D objects for FEA simulation. I am open
    to discuss the Tri-Ball at length with you including speaking
    with you evenings and weekends.

    My dear Mook, it is my motto that “Sarcasm is a poor substitute
    for substance” That means that if I was sarcastic, then it is a
    measure of “lack of substance”. It was not intended for the
    truism to be an insult but just a fact of life. Before I entered
    Tufts University Graduate School, my analytical world revolved
    around the 2D space, steady-state events and ODE (Ordinary
    Differential Equations). Post-Graduate School Education and
    training, I think in 3D Space, Transient-state, nonlinear material
    and geometric (like fishing-rods) behavior and PDE (Partial
    Differential Equations). It has been my experience that many
    people air their opinion as though their solution is both correct
    and simple such as is the case of “Global Warming”. I tell you
    that when people talk to be about global warming I ask simple
    questions such as “what is your understanding of latent heat
    and thermal capacitance of materials” or “how many thermal
    events simulations have you executed or papers you have read”.
    When their answer is “I don’t know”, then I respond neither
    I know whether Global warming should be ignored J

    Mook, if you have MathCAD v14 or you know of someone who
    does, have them download the ODE Handbook (free) and you
    will find three (3) out of the eleven (11) cases given are mine.
    One is in vibration, another one is in fluids and the third one
    is in Electrical circuit all simulating transient behavior using
    the OdeSolve command in Mathcad.

    In conclusion, your comments and concern are valid and due to the
    compactness of my case study report some information could not be
    included for the sake of brevity. I apologize if you interpreted
    my proverb as an insult.

    Let us have fun discussion simulation solutions amicably to
    enhance our understanding of CAE. Philosophy = Phileo +
    Sophia = Love of Wisdom (or Knowledge). I love both teaching
    and learning and “The only think I am sure I know is that I do
    not know enough”