hexagon logo
Register Sign in
  • State Not Answered
  • Replies 27 replies
  • Subscribers 3 subscribers
  • Views 269 views
  • Users 0 members are here
Options
Related

Modeling of Bolts

ron.kunz
Dear all,
I have a big assembly and want to model bolts in the complete assembly. What is the best and accurate way to model this.
Cbar analysis is accurate enough to analyse the bolt forces. Boltstress is of secundair interest.
Are there examples for these bolts?
In addition bar orientation for this CBAR property, how can this best be modeled?
 
Thanks in advance
  • 0
    @Ron Kunz​ You don't have to do anything to define the X axis. The element's connectivity defines the X-Axis: it always goes from node 1-2 (including offsets). The Y and Z axes are defined on the Patran properties form.
  • 0
    Thanks all.
    Regarding the bar orientation because the bolts are circular like Vivek mentioned one can for all bolts have an input for the bar orientation of: <1. 1. 1.>
     
  • 0
    I have created a model of a solid plate with a hole and with rbe2 utility and added a bar with properties of M12. No pretension added. When I add a force of 9kN to the centerpoint of the spider to simulate a tension and analyse the model I get for the axial force a value of 1620N. I would expect the axial force equal to the force I have specified?
     
     
  • 0
    Where that 9kN goes depends on the relative stiffness of the bar to the plate (aka the load path). (Remember, the R in RBE == Rigid, so there is no flexibility in this connection. The center and edges move together.) The bar/bolt carried 1620N, the remaining 7380N went into the spider, then to the solid elements, and was eventually reacted at constraints. Without seeing the geometry and other loads, it's hard to comment on the load path. I suspect you have a lot more load of the 9kN load going into the plate than you realize.
    To diagnose, request MPC Forces, or Grid Point Balance Table if you want to visualize the load path in Patran (with the Freebody tool).
    Also, if you want to correctly model a preload, your need more than 1 CBAR to do it. You need at least 2 CBARs plus the BOLT entity. Input the preload force, then lock the bolt to hold. (You can still use RBE2 or 3 for the spider.)
     
    The rigidity of the RBE2 spider is the reason some prefer to use RBE3 to model this type of connection (it has flexibility).
     
  • 0
    Thanks Ken,
    Regarding modeling the preload, are there some examples to find?
    What do you mean with BOLT entity? You are also mentioning 2CBARS, so that means you divide the bolt in 2 parts in axial direction?
     
  • 0
    Ron, there is a bolt preload example in the Nastran Demonstration Problems Manual (Chapter 23, Bolted Plates). I extracted and attached an excerpt for you. Note: this example uses solid elements for the plates and bolt. You can model the bolt with bars/beams using the same control node process to define the preload (with a force or displacement).
    Also, this uses SOL400 to simplify the process of locking the preload force as a displacement. You can do the same with SOL 101, but it takes a couple of runs. First you run with the preload force only -- to get the required displacement. In the second run, you replace the preload force at the control node with that displacement, and add your other loads in the same subcase (pressure in the example).
    Attached Files (1)
  • 0
    Thanks Ken.
    In that example SOL400 is used and a contact. Because I don`t have SOL400 I should replace this with linear gaps or linear contact otherwise the two plates can "go through" eachother. But if there is a big assembly (so a lot of prestressed bolts) can one also model this with a couple of runs?
    And another question is: if like you said the SOLID bolt is replaced by a beam, this should be one beam and only two rigids? So a rigid in the topplate and a rigid at the small plate connected with a beam?
    Thanks
  • 0
    You need 3 nodes to define the preload condition on the BOLT entity (or as an MPC). Review the figures on page 4 of the attachment (labeled pg 427) for the details. How you do this is up to you. You can split the beam into 2 elements with coincident nodes at the split. Or you can have coincident nodes at the end of one beam and the center of your RBE2 spider. Create a control node at a convenient location (coordinates don't matter) and use it when you create the BOLT between all 3 nodes. You apply the preload (as a force or displacement) to the Control Grid: GRIDC (the first one).
     
    To demonstrate the behavior, I created (and attached) a simple example w/ 4 four subcases:
    SC 1: Preload bolt only with a force
    SC 2: Preload bolt only with a SPC (determined from SC 1 displacement)
    SC 3: Pressure only (no bolt preload)
    SC 4: Pressure plus SPC preloaded bolt
    (My example doesn't have any contact bodies - just a preloaded bolt connected to a flat plate.)
     
    The Bolted Plates PDF example shows preload tension in the bolt creating compression on the plate faces. I always use contact bodies to model contact. There is no need for linear gaps. Model the assembly however you wish.
    Attached Files (1)
  • 0
    Thanks Ken,
    I have Patran version 2017.03. I get these messages:
    -Creating Nastran Database using SOL nfimp_v22: Failed but continue
    then:
    -Fatal erros encountered trying to read input file. he import process was aborted.
     
    Also you mention, you use contact bodies to model contact. However I don`t have SOL400.
    Only SOL101 so I cannot use contact bodies.
    Can you comment on this?
    Thanks.
  • 0
    Flexible or Rigid Contact Body in 2D and 3D (New Form) specified by BCBODY1. Quick reference guide describes that flexible or rigid contact body in 2D or 3D can be used SOLs 101, 400, and 700 only. So I am not sure what you mean when you say you can not use contact bodies in 101.
     
    During bdf import into patran a log file would have been created which will possibly point to the problem which may have prevented the input file from being read into patran.
© 2025 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement