Degree of Freedom for Supports on 3D (Strand7) 1

[Totobolz]

Good day to all FEA experts.

I am currently working in a Box with a 6 plate-type supports that is bolted on beams.(Attach is the model of the Box).
I am quite confused on assigning the restraint/constraint for the degrees of freedom of the nodes.
I understand that the support will be fixed type since the bolt and nut will restrict the translational movement(Dx, Dy and Dz)
In case the bolt holes in the supports are slotted. What is gonna be the degree of freedom?

Thank you very much in advance for answering my query.

Link

 

[GregLocock]

if the bolts are done up then until they slip it's an XYZ constraint, in a slotted hole. if they slip you are into non linear. good luck.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[FEA way]

If you want to account for the effect of slotted bolt holes, you will have to use solid model for plates and bolts (simplified of course) and include contact between these parts.

 

[jhardy1]

I would provide the following basic support as a start:

DZ restraint at all 6 supports
DX restraint at 3 supports down one long side; X translation free on the other long side
DY restraint at the two mid-span supports, one on each side; Y translation free at the other 4 supports

This should be fine for a basic linear elastic global analysis, but you may need to add some more sophisticated restraint modelling if you are doing anything more detailed.

http://julianh72.blogspot.com

 

[Totobolz]

@GregLocock , noted. apparently stresses are not high enough to cause the bolts to slip, linear is enough. thanks btw.

@FEA way , I'll try a solid model if I still have the time but I do get your point. bolt size is already overdesigned. thanks.

@jhardy1, I'll try the 2 freedom cases you suggest. Just want to verify the reaction forces for the support plate stress and deformation. thanks for the tips.

 

[FEA way]

*jhardy1: Your idea seems really interesting. Could you say a few words about the reasoning behind this approach ? Asking out of curiosity - I'm still learning how to combine displacement restraints in various cases.

 

[Totobolz]

Just want to add that the box will be in full liquid condition(worst case scenario) with an SG of 2.0. So definitely the support is the concern. though overstress(linear analysis) has already been noticed in the bottom plate(additional stiffeners or larger flat bar is the option for the redesign).

 

[jhardy1]

@FEA way:

I'm not sure why you think my suggested restraint case is "interesting" - to me, it's the logical default way to restrain this sort of model.

To clarify, I was describing a single restraint case, not "2 freedom cases" as Totobolz seems to have interpreted it. Basically, you provide vertical support at all 6 support points, and then you also need to globally restrain the model laterally and longitudinally (and also make sure you don't allow it to spin around the Z-axis while you're at it).

In general, I would suggest NOT over-restraining laterally or longitudinally for at least the initial linear static analysis (i.e. DON'T apply lateral and longitudinal restraint to all 6 supports), as this will allow spurious equal and opposite stresses and strains to become "locked in" between pairs of restraints, which won't happen in the real-life situation due to over-size and / or slotted bolt-holes, laterally flexibility of the supporting beams / corbels, etc. For example, the design intent will be that loads due to internal pressures from contained material will be resisted by internal hoop tension and plate bending behaviour, not by equal and opposite thrusting against the lateral restraints; if you over-restrain the model laterally, you will "lose" some of the lateral loads into the lateral restraints, and consequently under-report the loads carried within the steel structure.

Even where such stresses and strains are feasible, they will tend to be self-limiting due to flexibility of the supporting structures, and you would probably ignore these effects for the primary global design case. Where you do want to model these effects more precisely, you need to be very careful with how you model the interfaces - friction, flexibility, free travel until a bolt engages against the side of the bolt-hole, etc all need to be taken into account.

http://julianh72.blogspot.com

 

[glass99]

jhardy is right - A slotted hole is usually modeled as a released degree of restraint unless you have reason to believe there is too much friction. You do not need solid models or anything fancy. If its linear static analysis it usually doesn't make much difference either.