Learning Forum

[lborroto]

Hi,

This is probably a subject that has been discussed a few times. Still, sometimes I see information somewhat contradictory from different sources.

I am aware of different methods for modeling pre-stressed bolts in ANSYS i.e solid body, beams, bonded areas and recently saw something about including joint technology in the latest software version.

I am currently using beams (with bolt radius) connecting two areas of clamped parts. When checking the results I take the shear, axial and bending moment to determine von mises stress and the factors of safety for each individual bolt.

I would like to ask why some times analysts decide to ignore the bending moment and determine the von Mises stress using only axial and shear forces. I have seen this assumption in customer specifications to conduct stress analysis as well.

In the model I am working on I ran a couple of tests increasing the friction (to 0.5 from 0.1) between faying surfaces of clamped parts and the bending moments reported for beams were reduced more significantly than axial and shear forces.

The main question is: Is it reasonable to drop the bending moment from the hand calculation?

Thanks you,

Lino

[Vigneswaran Sridharan]

Hi The deformation of beams is represented by Shear forces and Bending moment. Shear forces contribute to Shear stress developed on the cross-section of the beam.
Whereas, the Bending moment is caused by normal stresses otherwise called Bending Stresses. So, you will still need Bending moment for calculation of the Bending stress of Beam.
But, when I say Von-Mises stress, it's equivalent stress used in Finite Element Analysis to find the yielding of the material (Ductile material).
This failure theory is independent of the Hydrostatic stress which is responsible for volume change. Here, the Shear stresses causing the shape change are responsible for yielding.
Hope this helps.
Vigneswaran

[lborroto]

I think one important aspect to take into account here is friction between clamped plates. Having sufficient friction will work against bolt shank contacting the walls of the bolt holes. However this type of contact (bolt hole wall vs bolt shank) is not modeled in ANSYS when per-stressed beams are used to model bolt connections.
I think the beam model in ANSYS considers the beam constrained in two translation and one rotation in one end (X,Y,R) and one translation and rotation in the other end (X,R). This would explain extracted results are: axial, torque and moment.
But the question remains: Is it OK to discard the bending moment results?
Thank you Lino

[Rameez_ul_Haq]

,if a beam is undergoing axial forces as well as bending (due to shear force), then you have to take both the bending moment and axial forces into account. Bending happens due to equal and opposite forces act on a cross section with a certain distance between them in solid bodies, and when this force is multiplied with this distance, you get the bending moment at that cross section. Hence, at that cross section, the net axial force is still zero. However, if you additionally add an axial force to that solid body, then that can treated separately and stresses can found independently (like axial stress due to axial force and bending stress due to bending) and then they can be superimposed onto each other as long as the material still lies in the elastic range of the stress-strain curve. As you would have probably guessed, the net axial force in that cross section will still be just equal to the axial force that you added additionally, since the axial forces due to bending has already cancelled each other out. However, the effect due to the bending still remains.
This asserts that you have to always take bending moment into account when calculating Von-Mises stress (or in ANSYS, its called Equivalent stress). Since superimosing (i.e. adding the axial stress due to axial force and maximum bending stress due to bending) will result in sizing more sensible.

[lborroto]

Thank you Rameez for your answer.
I am trying to bring more the discussion to the context of bolt connections. As I mentioned earlier, I have seen customer specifications requesting the use of reported axial and shear forces in bolt checks and ignore the bending moment.

[Rameez_ul_Haq]

that might be because the customer knows the length of the bolt is so small that the stresses generated due to bending moment will be almost negligible as compare to generated due to axial force and shear force. Just to simplify it a little bit, thats it. But in reality, to be more accurate, bending moment should always be kept an eye on because of the reason I mentioned in my previous comment.

[lborroto]

Rameez, that sound reasonable but in this case the customer is asking without knowing bolt sizes, lenghts and materials we will use. I agree with you in including the bending moment in determining the axial stress and vM stress to compare vs bolt allowable stress.
von Mises stress would be:
SvM = sqrt(Saxial^2+3*Tshear^2).
Where axial and shear stresses are:
Saxial = (Faxial/A)+BM*(d/2)/Ibolt)
Tshear = Fshear/A

thanks

[Rameez_ul_Haq]

,Simple Solution: Just do as the customer say, and provide an alternative solution too which includes the bending moment affect as well. Have a brief discussion about pros and cons of if bending is neglected and when bending is included. This will depend on if the sizes are significantly different from each other or almost similar. Let the customer decide. End of discussion

[lborroto]

Well Rameez the problem is that If I include the bending moment in calculations then many bolts won't pass. FOS < 1
So, before putting more (or bigger) bolts I would like to get a better understanding of the situation. Hopefully the subject will attract others and they will add to the subject.
There is lots of information on the methods that can be used to model bolt connections but I am not finding anything about the pass/fail criteria.

[Rameez_ul_Haq]

,but you are sizing the bolts depending upon the axial forces, shear forces, bending moments (and maybe torque as well). So the sizes are dependent on these values. If you only consider the axial and shear forces, then size is going to be smaller than the size if you also take into account the bending moment (and maybe torque).
I mean its a simple subject, I don't know why are you getting confused. It proceeds like this: You model the bolt as line body in the ANSYS (if you are only interested in finding the max forces and moments at any cross section of the bolt), then you determine the diameter of the bolt depending upon these values. You have to use an appropriate failure theory; for ductile materials you can use Tresca or Von-Mises but for brittle materials you can use Columb-Mohr or Modified Mohr. I usually use Von-Mises for ductile and Modifield Mohr for brittle since they agree to what we see in reality. You can also use a FOS as a part of your sizing and I usually use 1.5.
You can also use solid bodies to model the bolt in your model in ANSYS but that doesn't bring any benefit unless your purpose of analysis is not to size the bolt but analyze an already sized bolt that it will fail or not.

[lborroto]

Rameez, I totally agree with your reasoning. I am confused is because I don't know what is behind that practice of ignoring the bending moment. I will do some research and come back hopefully with good answers. There are some aspects of the assembly I can't discuss here but I am trying not to modify it for several reasons. Thanks.