« Autodesk Takes Local User Group Off Life Support | Main | Verisurf's Master3DGage: the Cure for Short Arms »

August 24, 2010

Comments

Gp-white

Hello Jonathan, hi all,

Of course PC’s can be used by just about anyone and yes, more often than not the effects are pretty good. That is exactly what these machines have been designed to do! So what would happen if we were to let some lesser qualified people write the programs? I am sure we all know what effects a badly written program can have;-)

I think what is bothering me most about your post is the word “police.” I don’t think it realistic at all to “police” anything to do with FEA within certain limitations which you mentioned yourself. I think it is more a case of common sense than anything. Your argument regarding the bucket is quite true and sounds plausible but there is always, even with such a “simple” product a remaining risk. Imagine the tools for the production of the bucket have been produced and taken into use for the initial samples. The Q-Dept then discovers that there is something wrong. I think you get the point without my having to go into this any deeper.

What is being said has nothing to do with “policing” of any kind (at least not the way I have understood things). It has more to do with the fact that people are being lead to believe that they can use FAE-Software to accomplish the job of a ME and exactly here lies the pitfall. It does not matter at all if it is a bucket or the most complex product available; the problem is the same regardless.

Sure, everyone who encounters this kind of software and has to use it will have to answer to his/her own conscience (or his boss if it goes wrong) as to whether the results are trustworthy or not so looking at it from that point of view then the only kind of “policing” which may possibly be necessary is that each individual person decides for him/her self if the results can be used.

Jonathan C. Yeandle

Hello Each,
Please get on board or miss the boat!

30 years ago PC's were used in glass walled & locked offices by 'trained computer experts' wearing white coats. Jobs were costed accordingly.

Today a PC can be used (sometimes to good effect) with very little training or understanding of the underlying computing principles. Think Excel, accounts and pretty pie-graphs with no qualified accountant in sight:-)

FEA vendors are in business to improve their product and get volume coverage in a chosen market in order to support their revenue. It is not in their remit to police the useage of their product or in some way reduce their market coverage by stipulating an academic entry qualification.

You are certainly correct that a downstream user of FEA results should be responsible when using the data. This includes using appropriate enforcement of rigour.

Usage of the data for flight critical applications needs a lot of policing.

What level of policing would you like to see when using the data for improving a plastics waste bucket retailing at 3 Euros/dollars?? I'd suggest that in this case the FEA guy needs no policing by the 'FEA police' and that regular commercial guidelines are sufficient. This market is where FEA inCAD is being targetted. Shipping it out to an FEA expert each time at even 1000 Euros per day is not on.

Kind regards,
Jonathan Yeandle

Gp-white

One the whole I think I would agree with just about everything all my predecessors have said! You are all right. I consider "Simulation" products to be a first step in the right direction and nothing more. If you expect more then you may well be in for a very big surprise when your parts are produced.
CosmosExpress was "so simple" to use right? Just tell it what material it is, tell it where the restraints are and then let it know how much and where the force is coming from. I remember the very fist time I ever saw the results. Optically it was very nice and appealing. I have met some people who were prepared to swear on the bible that the results were correct. Why on earth would these people say something like that? Because they don’t know better! They do not have the qualifications to be able to asses correctly what they are looking at. They see some “red coloured areas” and think “great, when I get rid of all the red areas I have won.” Jesus, if these people are every allowed to build flying machines of any kind or even work in a nuclear power plant then I for my part am leaving the planet.
So why is the CAD industry trying to market this kind of product?
I think the answer may be simpler than we think; "Price of Qualification." ME's cost money. People with lesser qualifications cost less and software like this seemingly puts the lesser qualified in a position where he/she can do the work of an ME! Exactly this is what I have been observing over the past few years. A lot of companies want to make savings and at the end of the day it's the price of the employees which may well have opened the doors for discussions like these ones.
An example: My colleague who works in an office bellow me. He is by trade a carpenter! He works here doing the job of a ME and we work in the medical technology sector! He has virtually no idea about metals, alloys and the like let alone has he ever heard of Mises, stresses or strains before. Everything he comes up with is so well over dimensioned that it weighs four to five times more than necessary. So why is he here? Because he is cheaper than a ME! For him simulation software is pure magic. Day to day reality however reveals that it is more like giving a cave man a modern torchlight.
I don’t think we need to talk about what the future will have in store for us because this is a train which is well in motion and I don’t believe anyone is going to be able to stop it. I like the idea of simulation software if it is going to help me get my job done better than with a calculator and pen and paper. For me the only advantage will be in the time saving sector meaning that I can concentrate more on other aspects of the products I come up with.
I firmly believe that one day this kind of software will be able to substitute the ME of today but at present time there is absolutely no substitute for a well qualified person on the job.

D Patelzick PE

I have to agree with Jeff Waters, most managers are going to look at the results with a jaundiced eye. When SW came out with a truncated Cosmos FEA, the first thing I did was run a few models that have known answers. The very first time I noticed a problem and contacted Cosmos - they listened and managed to fix the problem in the next release. I still run these types of checks and they come out quite well with SW. I would do this with what ever FEA was available.
I suppose if the part required a non-linear solver and the software was only meant for a linear solution you could have a problem (I don't know the capabilities of the software in question). Given the litigious society that we live in you would have to be insane not to be very careful using any kind of software without verification and some other eyes to review your results.

Sam

I think the point a few may be trying to make is that the uberkidz today partied their way through physics and materials, expecting the computer to take up the slack with shiny new toys like FEA-de-jour and lots of gigahertz. As these kids permeate management, there is no one left to make the call when the design is good, or NOT. For example, the name-brand vacuum cleaner I purchased at a national retailer last month had at least obvious three design flaws right out of the box. Not manufacturing flaws, design flaws. The kind manufacturing engineering cannot fix. I fixed two of them that afternoon but the third will require re-wiring the device so I left that for a rainy day. And we wonder why cars stop and/or go mysteriously? Kicking the experienced folks out the door to meet the quarterly bottom line with cheaper help is leading to problems whether we like it or not. One can add all the caveats they like but in the end someone has to know what they are doing.

Evan Yares

I think up-front analysis is a good thing -- but, for pistons, it might be a good thing to actually know what you're doing. Not just in analysis, but also in metalurgy. Among other things.

Bill McEachern

The issue raised here seems to be:
1) It is getting a lot easier to use FEA so any average idiot can do it which supposedly can lead to carnage.

This is an absolutely bogus argument in my view and has been around for years. In any ORGANIZATION it doesn't matter if some guy taking a rip at it produces complete junk. What matters is the system in the organization that ensures it never sees the light of day. Any outfit that decides to willy nilly give people who don't have much experience in this stuff the ability to wreck a company by taking any pretty picture at face value deserves to die - hopefully they don't kill or hurt anybody in the process. In my experience it just doesn't happen if anyone is awake at the switch.

FEA is like skiing powder. You can't learn it by reading a book. You have to get in it. It is the only way to learn it. You do need to be a fairly wide awake at the switch sort of engineer though to know when you are getting something that doesn't really look right. This is the main reason why exaggerated plots are produced by default - so you can see the curvature to see what's happening. People need to try the stuff so from that point of view CAD FEA is a great initial training tool they just fall short of the mark to earn the moniker of "production software", at least in my humble opinion. they are making strides but the commitment just isn't there.

CAD companies don't really get analysis. The software is typically laden with bugs, some of them of the type that produce seriously misleading information that is somewhat hard to spot. However, spot it you must. The FEA software CAD companies produce can become very difficult to use when you are trying to get something right, in an engineering sense. You spend a ton of time trying to figure out a work around to get the fidelity you want. Most of the time a way can be found but not always. When that happens you go looking. This makes CAD vendor FEA very vulnerable. Maybe some day the mangers at the CAD companies might wake up and throw some resources at their products instead of milking the less informed. They think they are doing a good job because people are buying it. Hard to argue with that till you try to do real production work with the tools. They are seeding the market, for somebody, very nicely.

Vadiim Shapiro

I am biased, since we just released fully automated stress analysis plugin for Rhino modelers at www.scan-and-solve.com (but this does not mean that I am wrong :-))

The issue is real, but the implications and conclusions are flawed. The same argument applies to any technology as it transitions from hands of a few experts to millions of users who can change the world. Licensing or not, the planes still crash, and car accidents kill more people than we want to acknowledge. But we deal with this by making better planes and cars, and through education and training, not by making cars impossible to drive.

FEA analysis is impossible to drive. Of course, you can get garbage from any FEA program, and when you accept their license, the user assumes the risk. So yes, we need more education and training -- not in meshing, preprocessing, and visualization -- but in understanding the power and the limitation of physics, analysis, and modeling. The classical FEA tools are limited and intended for a few specialists. The next generation of tools, such as Scan&Solve, will be used widely by everyday users, and this trend will be accompanied by renewed interest in analysis, better understanding and education, and, yes, much wider certification and licensing processes. It is not my opinion, it will happen whether we like it or not.

And I happen to like it! Based on our initial response, we believe that Scan&Solve will be rapidly adopted by universities, architects, sculptors, bike designers, boat enthusiasts, industrial designers, and many others who have been deprived of the benefits of FEA computations until now.

Earl

Roopinder – As a grizzled old drag racer of over 40 years, and an amateur SolidWorks/Simulation user, I would look at that piston and say the analysis is spot on. I have collapsed many a ring land over the years, but never pulled a wrist pin out, or even broken one for that matter.

Joe Aggie

The demo was clearly meant to impress managers who have no engineering inclination. Historically, FEA software was developed by folks with ME degree and software-training. But lately more FEA software development teams are run by developers who have no ME background and perhaps come from computer graphics/gaming industries -- where colorful thingamajings and one-buttoned widgets mean advanced software.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Working...
Your comment could not be posted. Error type:
Your comment has been saved. Comments are moderated and will not appear until approved by the author. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.

Working...

Post a comment

Comments are moderated, and will not appear until the author has approved them.