Well, to be fair, the company that supposedly ripped him off was found non-infringing by the courts of the land. Of course you (Tuckerfan) know that I have a pony in this race. 
(And I’ll see the movie, and even if it’s not good, I’ll appreciate it as much as the Tucker story from which you derive your passion.)
Let me preface my comments by noting that I perform nonlinear finite element analysis on components made of metallic, ceramic, 3D and fiber-reinforced composite materials as (part of) my job.
The finite element method is an idealization and will not in linear static anlaysis reveal failure modes beyond macrostructural yield failure, i.e. bulk overstress resulting in local yielding and ultimate failure of the material.Without skill, engineering judgment, and the right tools FEA will not reveal fatigue failure, stress cracking, microfracture propagation, unrelieved residual stresses from heat treatment, irregularities and flaws in the manufacturing process that dominate impact strength, et cetera. There are more advanced methods and tools that allow an analyst to model and predict other types of failure; however, the level of effort and expertise required to produce credible results is dramatically greater, and the accuracy even of results that follow an established method are always questionable until validated by physical testing. FEA is like any other use of software and computing power; the answer that you get out is only as good (at best) as the data that you feed in, and there are plenty of people out there doing finite element who have been trained in how to run the code but have little to no understanding of the limitations of the analysis.
Although the o.p. is not very specific about the application that widget is going to be used in, it is clear that it is an impact application with high cycle life. For this type of analysis the value of FEA is questionable at best, and certainly can’t be done in absence of other testing, analysis, and most important, experience and training in material selection. FEA might be used in conjunction with physical testing (which can validate the model, at least for a limited range of inputs) but given that impact-based failure modes will dominate material selection FEA will probably only provide very rough guidance; an experienced metallurgist or machine design engineer will probably be able select the appropriate material, fabrication, and treatment method(s) based upon the impact requirements rather than detail FE analysis. FEA can be a good predictive tool, but ultimately it only answers the questions you explicitly ask.
I do agree with Balthisar that from a liability and reliability standpoint it would be good to have an engineer or metallurgist who has some experience and background in the type of application provide qualified advice, and this would be a better use of money than throwing it over the wall to an analysis house to make pretty eight-color stress plots. This will not absolutely protect the o.p. from liability, but in the case of failure and litigation will establish that due diligence to assure a safe design was done, and may also give some backing to any argument that the failure of the product was due to misuse, though it is often the opinion of juries that any possible use by the most obtuse bonehead falls under the criteria of permissible use, which is why hammers now come with a bright yellow label indicating that striking a hammer against a hardened metal surface is contraindicated. (Some day hammers will come completely encased in an irremovable giant foam pillow ensuring that no one can possibly injure themselves with it except by suffocation.)
As far as the patent application, I can’t see any reason why you would need to put material selection in it other than some vague reference like “hardened steel.” A patent application is not a complete engineering documentation package; it is essentially a listing conceptual details to outline the general concept sufficient to prevent another party from using the fruit of your intellectual labor and produce a copy of the idea without licensing. It may include sketches, application details, and any particular functions that are done on or by the widget, but it doesn’t have to have material specifications, detail dimensions, or fabrication details except insofar as they are an inherent part of the concept being protected. Things like details of manufacture and material selection are typically proprietary to the manufacturer and need not be shared except as required by certification agencies and organizations.
In regard to the cost: $10-$15 as a total cost seems low for what the o.p. desires; unless the o.p. plans on producing these in lots of thousands via semi-automatic fabrication, the labor cost of hand machining these and then heat treatment may far exceed that value. The actual cost for a 1" x 1/2" bar x 12" long of steel itself is probably $3-$5 for high strength tool steel (just an off the cuff guess; it’s been years since I’ve actually quoted out individual materials, as the cost of machining and fabrication typically dominates the stuff I work on). The previous recommendation for O1 tool steel by Kevbo can work, but be warned; uncontrolled barnyard heat and quench will often result in overhardening and result in brittle failure, and you’ll never be able to make a truly consistent product via this method.
I’ve used D-grade steels (primarily D2) for corrosion resistant cutting and prying tools with good toughness, and S-grade steels for a couple of impact applications. However, I’m not a materials expert by any stretch of the imagination. If I’m designing something that is going to be used in some type of critical application where safety is a concern and the material properties and environments are outside of my experience with construction and common machine grade steels, I consult with a metallurgist or materials expert. Sure, I can look up properties in a textbook or materials manual, but they never seem to include the sort of important details–like the fact that such-and-such “weldable” steel is also prone to eutectoid phase transision when welded by the blah-blah process unless preconditioned by sacrificing a cow under the waning moon–that end up becoming the Achilles’ heel of manufacturing. And no computer analysis method will give you this, either.
Stranger
Case hardening only extends so far; with your stated dimensions there will be a soft, malleable core. I don't think there is a problem with shattering, initially. Work hardening alters the beginning properties, perhaps to benefit. I don't know in your case and don't expect explication.
Strangers' bit about "barnyard heat and quench" is humourous and worthy of regard, but shouldn't dissuade you from doing it yourself. Tempilstiks or the like can verify colour change as per Kevbo's description of heat treat. Do consider that every "heat" (attainment of austenitic phase) will cause ~1% scale loss unless done in an inert or exclusive atmosphere.Stranger, while I agree that FEA done by an inexperienced individual won’t give results that would be terribly useful for the mass production of a widget, wouldn’t a junior or senior in a mechanical engineering program be able to use a basic version of ANSIS or Cosmos to get results showing where the problems are likely to develop? I haven’t worked with either of those programs in a couple years, but IIRC they had toolboxes one could use to specify impact damage. Also, correct my naivety , but wouldn’t it be relatively simple to specify a fatigue, either as a toolbox in the program or by finding what kind of forces you’re seeing and comparing them to the endurance limit (taking into account the surface texture, shape at the stress concentrator, etc.)?
I agree with the rest of your points concerning needing a metallurgist for other reasons, especially when the widget goes into mass production. It also seems like the OP would like to at least start out making a part that would last for longer than the original part.
Oh, you can definitely do effective treatment without modern systems–I’ve made a few knives and prying tools that way–but for a production part you won’t get a very consistent product without a tightly regulated process. This is why some parts from China are of pretty high quality, and others are junk; it just depends on whether that part went home with Chin or Chau that day.
For what the o.p. is trying to do (as far as I can figure out) FEA is just not going to be a great method, and really isn’t necessary for material selection. You might use the selected material properties and some analysis to optimize the shape or other geometric aspects of the design (where it mounts, how it is attached, et cetera) but stress results for impact in the locality of the impact zone are going to be iffy at best.
Although there are tools to do fatigue life prediction there is really nothing simple about fatigue, and plugging & chugging with a built-in tool without understanding the limitations is a recipe for disaster. Too often I see people treat FEA results as some kind of reality when in fact it is a completely artificial, highly idealized, and approximate estimate of real stresses, strains, and deflections if the analyst has properly characterized the system to a sufficient degree of fidelity. FEA is a great tool for many things, and when used right it can help do predictive analysis that refines a design before you cut steel or layup composites, but it is not a panacea for designers, and use of it in a way that the analyst lacks prior experience requires cautious and critical assessment of results and correlation with real measurements.
I recommend O.C. Zienkiewidz’s The Finite Element Method for Solid and Structural Mechanics for a detailed discussion of the value and correct application of FEA. Of course, you’ll essentially need a graduate degree in engineering mechanics to understand it, so it’s not for the casual reader.
Stranger
::Peeks back into thread::
And apparently a bachelor’s to understand pretty much anything after post #21.
So, in summation, seek professional advice before I try to begin production. Gotcha.
Thanks again all.