FEA for Biomedical Testing Standards

[Jabberwocky]

Hello all, my company is looking into developing some in-house FEA capability. We use SolidWorks, so I am thinking that COSMOSWorks or neiWorks would be best for us at this point. As a smaller company, we do not have any dedicated FEA/FEM people so the program needs to be relatively easy to use.

I have two questions on that line:

Do you feel that we can get reasonable analysis out of these "lighter" programs? Assuming of course the operator is competent.

Does anybody know of any computer simulation programs that are set up to mimic ASTM standard tests? (specifically in the biomedical field for my purposes)

I realize you can try almost anything with any program, but if a certain program was better suited I'd love to hear about it.

 

[ProbaSci]

Biomed is a huge field that encompasses several different specialties including structural, thermal, fluids, electronics, chemical, and biologics among others. I deal primarily with the structural/mechanical aspects of medical devices and think you will find most packages are capable of simulating ASTM type standard tests, though the assumptions that you need to make will obviously vary with the limitations of the software and operator. For example, if one wants to simulate testing of a single component hip replacement, the peak load can be applied to the metallic components in a static linear model to get peak stresses, up to the yield point and this is typically an acceptable analysis technique since the metallic components are expected to operate below yield. Now, if would you want like to simulate this same test, but consider a modular component, fatigue, and/or an environmental fluid other than air, it can quickly become a much more complex model and you need very powerful software and a very experience analyst.

As to getting reasonable analysis from "lighter" programs: I think the fundamental problem is that the good FEA modeling requires an almost obsessive compulsive attitude toward detail, control, and error checking that is philosophically incompatible with the CAD-IN "black box" approach of “click it and go”. My experience is that CAD-IN FEA is to be used by designers with no or little FEA training (who are already very busy trying to do design work). Without the fundamental training in FEA (and I don’t mean just training in how to use the software), the designer has no idea how to error check or validate his model (or why he has to). Who would trust that situation? As an experience FEA user, I would find the limitations of many CAD-IN offerings to restrictive to allow me the modeling power and error checking detail that I would require.

If the choice of FEA is economic, then as a provider of FEA services primarily for medical device manufacturers, I don’t think my clients can meet the requirements to have in house FEA capability (research software, buy, train on specific software, train on FEA principles, implement quality control protocols) in a more cost effective way than I can provide those services as specialize vendor (who can spread the costs over several clients).

http://www.probasci.com

-
Implantable FEA for medical device manufacturers

 

[Jabberwocky]

Thank you for the response. FYI, we work with spinal devices and are mainly concerned with static and fatigue testing. I understand that "garbage in = garbage out" and that by their very nature, the lighter (CAD-IN) programs give you less to work with in terms of verification.

We are not looking to supplant physical testing, I don't know when the FDA will allow virtual testing results in place of "real" ones - we are just looking to get ballpark figures for how our constructs will perform.

 

[bme0319]

Jabberwockey,

Here's a link to some cool software I came across.

http://www.materialise.com/mimics/main_ENG.html

Paul P.