Skip to main content
J
JonathanHodgson13-Aquamarine
13-Aquamarine
April 4, 2016
Solved

Getting fasteners to work

  • April 4, 2016
  • 1 reply
  • 8485 views

So, inspired by Roland Jakel‌'s presentation, I thought I'd revisit an old analysis and try to re-run it using the fastener functionality.

Unfortunately, the analysis ran with "The fastener is invalid" warnings, and resulted in no loads (despite specifying preload).

The model is slightly unusual in that there is a deliberate gap which the bolts should pull down once tightened, so I closed the gap to make the surfaces coincident, but this made no difference.

The clearance holes are chamfered on both sides, so I suppressed these chamfers (and reinstated the gap) then redefined the fastener references, only to be told that "The fastener references cannot lie on the contacting surfaces."  Well, a) there's a gap so technically the surfaces aren't contacting, and b) how else am I supposed to select the references, if I can't select the open edge of the tapped hole?

On a related note, it tells me that I can't put a 6 mm fastener into my M6 tapped (Ø5) hole - should I be modelling my tapped holes at the major diameter, or is the root diameter of the screw thread appropriate?

    Best answer by rjakel

     

    Hi Jonathan,

     

    could you provide an image of your assembly or the model itself? There may be several reasons why you obtain various errors when using the Simulate fastener feature, so I can just speculate. First of all some remarks:

     

    • You must bolt exactly two different parts together – three or more (e.g. two flanges and additional washer(s)) don’t work, neither does 1 single part bolted with itself, like a clamp
    • If there is a gap between the flanges in contact (what you state), the algorithm that detects the contacting surfaces cannot find the flange interstice, and a warning should be issued
    • Unusual part geometry may also prevent the algorithm from detecting contacting surfaces at the flange interstice correctly. Also, a volume region applied there may be responsible for that. You should obtain a GUI message what’s up.
    • Yes, fastener (threaded) holes must have at least the nominal fastener diameter, you have to change the Pro/E-model accordingly.
    • You may have used wrong references or wrongly applied “bolt” or “screw” type fasteners. E.g., you cannot use the cosmetic surface used as thread representation as reference

      

    I guess a volume bolt and the preload element would be a better choice to solve your problem, but I have to see the model to judge.

     

    According to my experience, the fastener feature in Creo 2 works relatively robust now. I filed some cases leading to SPRs approx. two years ago, but those should be fixed in the meantime. I remember there was (and may be still is, I don’t remember exactly) a problem if you mix fasteners with linear and contact interfaces in one model, but this should not be responsible for the trouble you are facing.

     

    From Altran side, we offer a special 3-day workshop how to do fastener analysis with Creo Simulate: This consists of a 1-day workshop in nonlinear contact analysis and a 2-day-workshop in fastener theory and application. This may be helpful for users that want to become familiar with this topic. Analyzing fasteners with a FEM code correctly is more difficult as many users believe.

     

    Best regards,

     

    Roland

     

    1 reply

    R
    rjakel1-VisitorAnswer
    1-Visitor
    April 4, 2016

     

    Hi Jonathan,

     

    could you provide an image of your assembly or the model itself? There may be several reasons why you obtain various errors when using the Simulate fastener feature, so I can just speculate. First of all some remarks:

     

    • You must bolt exactly two different parts together – three or more (e.g. two flanges and additional washer(s)) don’t work, neither does 1 single part bolted with itself, like a clamp
    • If there is a gap between the flanges in contact (what you state), the algorithm that detects the contacting surfaces cannot find the flange interstice, and a warning should be issued
    • Unusual part geometry may also prevent the algorithm from detecting contacting surfaces at the flange interstice correctly. Also, a volume region applied there may be responsible for that. You should obtain a GUI message what’s up.
    • Yes, fastener (threaded) holes must have at least the nominal fastener diameter, you have to change the Pro/E-model accordingly.
    • You may have used wrong references or wrongly applied “bolt” or “screw” type fasteners. E.g., you cannot use the cosmetic surface used as thread representation as reference

      

    I guess a volume bolt and the preload element would be a better choice to solve your problem, but I have to see the model to judge.

     

    According to my experience, the fastener feature in Creo 2 works relatively robust now. I filed some cases leading to SPRs approx. two years ago, but those should be fixed in the meantime. I remember there was (and may be still is, I don’t remember exactly) a problem if you mix fasteners with linear and contact interfaces in one model, but this should not be responsible for the trouble you are facing.

     

    From Altran side, we offer a special 3-day workshop how to do fastener analysis with Creo Simulate: This consists of a 1-day workshop in nonlinear contact analysis and a 2-day-workshop in fastener theory and application. This may be helpful for users that want to become familiar with this topic. Analyzing fasteners with a FEM code correctly is more difficult as many users believe.

     

    Best regards,

     

    Roland

     

    J
    JonathanHodgson13-AquamarineAuthor
    13-Aquamarine
    April 5, 2016

    Hi Roland,

    Thank you for taking the time to reply.

    I think my problem is the "exactly two parts" requirement - I have three, and although they're not in series like a bolt and a washer, I suspect this is the root cause of my problem.  It looks like this is not an appropriate model to experiment with fasteners!

    fastener.png

    I'll go and try out the preload element, though, as that could be useful to understand.

    Thanks again,

    Jonathan

    3
    346gnu13-Aquamarine
    13-Aquamarine
    June 14, 2017

    Jonathan,

    That is something being preloaded. We so that all the time, usually bearings.

    Roland,

    I 'bolt' lots of bits together and it behaves well (don't know about Creo 4 yet). We just need springs to stop parts from being squeezed out like a bar of soap.

    Just thoughts ...

    Terms & ConditionsCookie settingsAccessibility statement