dimensional board

[fabiobs83]

hello to everyone, in a while I will have to size some lenses for butterfly valves. given the complex geometry I will rely on a fem analysis but I wanted to ask you if you think I can also make a numerical dimensioning (with calculations) of the simplified geometry (starting point to then perform the fem analysis). the disk will be subject to a pressure only on one side and will have a central passer shaft (to maneuver and that will give further stiffness to the disk). Any advice?
thanks in advance :wink:

 

[meccanicamg]

everything you can and you have to calculate by hand.
as you well should know, the fem is reliable no more than 5% if set well. set a fem well, means to know very well the real behavior, to put strain gauges and to have measured the behavior of the exact object, to have made calculations with accuracy, rays and rajects, theory of flat plates compared to the classic beam model (which is substantially wrong for the application) etc.

I would say that the constraints you put into the fem solutor, mesh, thermal conditions etc will make the accuracy or not of your model.

 

[fabiobs83]

everything you can and you have to calculate by hand.
as you well should know, the fem is reliable no more than 5% if set well. set a fem well, means to know very well the real behavior, to put strain gauges and to have measured the behavior of the exact object, to have made calculations with accuracy, rays and rajects, theory of flat plates compared to the classic beam model (which is substantially wrong for the application) etc.

I would say that the constraints you put into the fem solutor, mesh, thermal conditions etc will make the accuracy or not of your model.

Thank you very much. I know that the fem is much harder than it may seem, so I asked for advice. I have to recover some college books and look for some formulas because from a short research I have found nothing that can serve me. We'll see. .

 

[re_solidworks]

everything you can and you have to calculate by hand.
as you well should know, the fem is reliable no more than 5%....

Your statement leaves me perplexed... I do not make calculations, so we rely on a study that only does that, but the feedback I had has always been excellent. when we performed experimental checks on calculated objects we always found negligible differences. On the contrary, when the checks are done by hand, you cannot grasp many peaks of tension, often due to the deformation of the components that are around the section that is occurring. basically hand calculations we do them and do them the study to which we lean, but only and strictly in the first line, to have a presize.

 

[Max-C]

the fem is reliable no more than 5%

I think this statement is not correct. as re_solidworks we also entrusted ourselves to a study for cfd calculations and the solution they proposed, experimentally, gave the same results of their simulation.
more correct is the fact that the simulations, by completeness, should be accompanied by experimental results in order to have the possible " error" (e.g. cooling fan on heat sink; aerodynamic simulation verified by the wind tunnel etc.).

 

[armfer]

I usually make a first calculation with desaint venant and other to have an order of magnitude of tensions and I change the design.
then, if you cannot oversize, you want to optimize geometries, materials, boundary conditions, steps to the fem. send me a design
on: armfer333@gmail.com . ciao