I have a wire spool assembly that I am using Inventor Simulation within Inventor to do an FE analysis on. The assembly was modeled and analized using Inventor 2013. My model is a large round plate, 83" outside diameter. I am using the Simulation add-in to automatically create a mesh of the flange. I have varying loads starting at the base of the flange, near the spool barrel, and lessening radially outward though all of the loads are uniform pressures. To apply these loads I constrained "load bands" to the inner face of the spool flange. I edited the contacts where these bands contact each other to sliding/non-separating contacts so that they slide in relation to each other but stay normal to the spool flange face. My fixed point is the face of the spool barrel at the center of the spool. After running the analysis I get "high" stresses at 90 degree increments about the round face of the flange.
My question is, why do I get stress concentrations at every 90 degrees on the surface when I am applying uniform pressures to the solid round plate? Is this a function of the automatic mesh within Inventor? How can I interpret my results confidently? Is there something that I can do to eliminate this?
Thank you in advance,
Autodesk Design Suite 2013
Inventor 2014 Ultimate
Windows 8 Pro
Note that most of the readers of this group are Simulation Mechanical users, so there may not be many users of Inventor Simulation reading this thread. If you do not get the ideas you need from here, you might want to post your message on the Inventor Discussion Group.
I agree that the high stress at 90 degree angles is not correct. Do the displacement results look correct? My guess is that you have a statically unstable model which the solver was not able to fully take into account. In particular, is the "sliding/no-separating" contact what happens in real life? I cannot picture in my mind what the design looks like, so I do not know if this is the correct contact or not. Also, I think the slide/no separation contact will let the surfaces slide an infinite distance but not separate (nor penetrate) in the perpendicular direction. So unless you have some type of restraint that prevents that type of motion, this may be the source of the instability.
Keep in mind that linear stress simply solves the equation F = k * displacement for the displacement results. Or stated another way, it solves F/k. If k = 0 or approximately = 0, the solution can be wrong regardless of whether there is a force in that direction or not.
A picture of the flange (and maybe the cross-section) showing how the contact is arrange may be helpful.
16 years experience with Simulation Mechanical
I had this posted in the Inventor section but moved it here. I thought this forum would be more specific to an FEA versus the Inventor forum.
What I have is a wire spool with side flanges of 83" OD. The OD of the barrel the flanges are welded to has a 26" OD. The 1.75" steel wire rope is applying a load to the side flanges as it is spooled onto the wire spool. My flanges are modeled each as a single part as they would be in real life. My outward loads applied by the wire lessens in the radial direction as more wraps of wire rope are spooled. Given that information I modeled "load rings" the width of the wire OD, 1.75", and very thin then applied one "ring" for each wrap of wire contacting the flange. The rings have bonded contacts to the flange but have sliding/non-separating contacts at contact surfaces with respect to each other, the same as each successive wrap of wire rope may move laterally in real life. Ultimately, the loads applied, as calculated pressures, are forming a lamp shade out of each flange.
When I assemble stiffeners to the the outside of the flange to prevent deformation, I see high loads every 90 degrees about the flange in the stiffeners. I refined my mesh enough today that I could actually see a "cross" shape in the mesh that was generated by the software which leads me to believe this is just the software.
I was hoping someone from Autodesk would help out with this and maybe give some incite as to how the software meshes a round form, especially one this large with respect to the mesh size.
Inventor 2014 Ultimate
Windows 8 Pro