hexagon logo
Register Sign in
  • State Not Answered
  • Replies 16 replies
  • Subscribers 3 subscribers
  • Views 201 views
  • Users 0 members are here
Options
Related

Examples for Linear Gaps modelling.

ron.kunz
Dear all,
I would like to study some examples with linear gap elements. We don`t have SOL400 available and to have a work around for i.e. objects touching or near eachother.
Could you provide wtih some examples or give location of study material? Thanks.
  • 0
    Hello Ken,
    Your last remark on this topic: "Choice of contact model (linear gaps or linear contact bodies) should not affect calculated bolt forces, assuming all other model details are the same. "
    Can you comment on the difference between linear contact and non-linear contact.
    The theory says this with SOL101:
    "SOL 101 contact or linear contact is, by nature, a nonlinear solution process. It requires solution iterations to satisfy both the equilibrium and contact conditions under certain tolerances. The user does not have to specify NLPARM card"
     
    With SOL101 does it really specify only initial contact and afterwards the shearforces are really transmitted true bolts (in case of bolts) and not through contact?
    Can someone explain me the difference between SOL400 and SOL101 for contacts if in both cases it is in fact non-linear?
     
    Thanks in advance.
     
     
     
  • 0
    There isn't any difference in contact in 101 and 400. The difference is linear vs nonlinear assumptions (small vs large displacements) .
    The confusion comes from the iterative solver used for linear contact and for nonlinear analysis. A linear contact problem cannot be solved directly, so requires an iterative solution. Nonlinear analysis also uses an iterative solver. In other words, linear contact isn't nonlinear (like materiel non-linearity, large strain and large displacement ), but does require an iterative solution.
    The big differences are the linear assumptions in 101. Remember, everything in 101 is linear: linear materials, small strain elements, small displacements and small rotation assumptions. Users frequently overlook the influence of the small displacement assumption. It is a limitation with linear contact .
    Remember, with small displacements nodal locations are not updated during the solution (displacements are not added to the coordinates). With contact, this means nodes/elements on opposing contact bodies should remain "relatively close" (because the contact patch isn't updated). In other words, with 101 you can model the weight distribution of a box on a plate, but you cannot slide that box sideways to another location on the plate. (The contact is assumed to occur at the original contact patch on the plate.)
    This is the big difference compared to 400: You can model the large sliding with 400 because it includes large displacement and rotation affects. But, the difference is due to small vs large deformation, not the contact.
     
    Also, regarding, NLPARM...that is an old way to specify iterative solver options. NLSTEP was introduced several years ago, and has many advantages over NLPARM. For example, the CTRLDEF field is used to easily set 101 convergence parameters (LCPERF or LCACCU).
  • 0
    Regarding memory and disk usage, linear contact does't require a lot more H/W resources than a standard linear analysis. In other words, if you can run a linear analysis with 860000 elements and 740000 nodes on your desktop, you can probably run with linear contact too. The big difference isn't RAM or disk usage, it's the number solves required it iterate to a converged solution. With contact you might have 10, 20 or 100 passes thru the solver, so the job will run a lot longer.
    This is where linear contact has a significant advantage over linear gaps. Adding 20,000 linear gaps to your model would require significant scratch space.
  • 0
    I was told the Contact Table in Patran can be used with SOL101 but in fact is made for use with SOL400.
    In the Contact Control Parameters we can change the friction parameters. My question now is, if this will affect also an analysis with SOL101 or that a change in these parameters does not affect the solution at all due to the SOL101 limitations.
    Thanks.
  • 0
    Not sure where you got that info about "Contact Table made for SOL400".
    Contact can be used in several Nastran sequences. Touching contact is available in 101 and 400, and glued contact can be used in 101, 103, 105, 107-112, 200 and 400. Nastran uses the same BDF entries to define contact for all sequences.
     
    My previous post explained the difference between contact in 101 and 400 -- it's all about small vs large deformation assumptions.
     
    Also, the contact table is an older way to define contact (inherited from Marc and SOL600). Contact Pairs were added to Nastran and Patran 2013, and many cases are an easier way to define contact (especially when you have a lot of contact bodies).
    Given the number of contact questions you have, I highly recommend this Nastran course: NAS133: Contact Analysis using MSC Nastran and Patran (with Contact Tables or Contact Pairs). It is available as a 16 hour online course.
     
© 2025 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement