Simulation

[garryson09]

Good Day. I want to know if there's a feature in Fusion 360 simulation to give a coating to the material other than its physical materials? i want to simulate a component with steel as a material and chromium as its coating

[johnswetz1982]

This is probably more of a design forum question, @seth.madore . Does it really matter? I wouldn't think that a material with a few molecules thick coating would behave any differently than the bulk material. 

[garryson09]

thank you. Yes, actually it matters. some coating like black oxide improve the corrosion resistance and fatigue life of a steel. Im planning to investigate this phenomenon using fusion 360

[John_Holtz]

Hi @garryson09 . I agree with @johnswetz1982 . The answer is no.

• No, it is not practical to simulate a coating in the analysis. Even if you make a physical coating and give it some realistic thickness, the coating will either not mesh or will give wrong results because the mesh is super thin.
• No, the coating has virtually no effect on the simulation (unless you are coating something that is the dimension of a spider web). 

I doubt that you are simulating the corrosion in Fusion, so the coating does not add any value to the simulation for corrosion. You are not simulating fatigue in Fusion, so the coating does not add any value for that. You might perform a static stress analysis and then use the stress results to do a hand calculation for the fatigue life, but the stress is not affected by the coating. The coating affects the life because it changes the endurance limit. That is part of the hand calculation; it is not related to the simulation.

[garryson09]

Thank you so much sir for the reply, actually i'm about to do the coat using a physical solid with very thin dimension. But i think the meshing will not work because its super thin. Also, can you elaborate the portion where the coating will take effect in a spider web dimension? is there other way?

[John_Holtz · Accepted solution]

Hi @garryson09 

What I was trying to say is the coating will not affect the results of the stress analysis unless the thickness of the part is on the same order as the thickness of the coating. In other words, if the real part is the size of a spider web and you add a coating to it, then the coating will affect the stress analysis. If the real part is approximately 1 mm thick or thicker, then the coating will have no affect on the stress analysis.

You should perform the stress analysis on the part without the coating. Neither the thickness of the coating nor the material properties of the coating will have any effect on the stress analysis. The reason is there is nothing in the analysis that is a function of the surface finish, at least not to the molecular level that is involved. Any effects that are caused by the coating need to be included in the hand calculations that you do, using the stress results as the input to the hand calculations.

Try the analysis without the coating and let us know what you get.

[garryson09]

Thank you so much sir! appreciate your help. 🙂

[mj_reich]

@John_Holtz This is an old thread, but how about simulating the stress of chrome plated ABS? Structural plating of plastic is broadly used, and I would like to simulate and compare chrome plated ABS with standard ABS. Is there a method to do this now? Or is it is the same answer as the one above? Thanks!

[TheCADWhisperer]

@mj_reich 

What would the size of the mesh tetrahedrons be?

How many tetrahedrons would Fusion need to calculate?

Linear Elastic Static Stress Analysis has the following limitations.

Orthotropic material (so you would have to analyze as an assembly of two components, the base and the coating.

No relatively extreme aspect ratio (you have a really thin vs relatively wide area - so fails this limitation).

No relatively "rubbery" or "brittle" material properties (the Chrome plate might violate the brittle limitation).

Slowly applied loads (no impact loads).

Linear Elastic Static Stress Analysis does not measure fatigue.

[John_Holtz]

Hi @mj_reich 

Fusion has not changed in this regard.

To do the calculation properly, you would model the chucky part (ABS in your case) using solid elements and model the plating using shell elements. Shell elements have area but no thickness; the thickness is a numerical input. (This makes shell elements ideal for thin pieces.) Essentially, the shell elements "paint" the surface of the solid, just like the coating.  Fusion can do the solid elements but not the shell elements. You could do the analysis using Inventor Nastran.

Regarding the displacement and stress in the plated model, you can estimate the effect of the plating as follows:

• The solid has a stiffness and the plating has a stiffness which can be estimated by hand. For example, if the cross section is in pure tension, the stiffness is Area*(Modulus of Elasticity)/Length = A*E/L. If the cross-section is in pure bending, there is another formula relating load/displacement = stiffness.
• You can treat the stiffness like springs. Since the solid and plating displace the same amount, they act like springs in parallel.
• The equivalent stiffness of two springs in parallel = the sum of stiffness = (stiffness of solid + stiffness of coating). From the area and modulus of elasticity, you can estimate how much the equivalent stiffness changes with the coating.

If the stiffness of the coating is insignificant compared to the solid, you do the stress analysis in Fusion to get the stress, displacement, and strain. My guess is the strain in the coating is the same as the strain on the surface of the solid. (You should confirm if this is true if you decide to do the analysis.)Thus, you can do a hand calculation to get the stress in the coating using the strain results from Fusion.

John

[mj_reich]

Thank you both for the help! It looks like it may take some playing around and some napkin math, but on the topic of surface stress with a different objective; how would I bias a generative design to increase surface area and optimize a chrome plated ABS part for strength? Is there a way to configure a generative design to do this?