Interview with Chris Randles and Blake Courter – Part 5

<< Part 4

Deelip: While you claim that you are not competing with the CAD vendors, I think you actually are because they have been putting more and more analysis functionality into their CAD systems. Also what do you say to the criticism that CAD vendors are putting dumbed down versions of their analysis tools into the hands of the CAD users which have no clue as to what analysis is?

Blake: There are two competing strategies of simulation right now. One is that the CAD vendors want to sell every CAD user a seat of simulation. And to do so they put forth a story that they have made their simulation tools so easy to use that someone without an engineering degree, someone who does not know units of stress, someone who has no clue what Von-Mises stress is, let alone understand how it relates to the actualyl yield stress of a part, I mean… (throws hand up in the air), can somehow do analysis. When I was in school doing design problems or working in the shop and I put bolts onto something, the professor would come up and ask why I chose that bolt. He would make me sit there and figure out how much stress would be on that bolt and whether it would yield without even a piece of paper. Simulation engineers are constantly doing that. They are constantly figuring out “what ballpark are we talking about here?” And that way when the numbers come out, they know whether the results makes sense or not. If you don’t have that skill you have no right to be doing simulation. I am actually kind of disgusted by this.

The CAD vendors are in a tough place. Their markets are flat. Their channels are complaining. Everyone needs to feed their babies. And so they come up with all these products to sell and make a story for it. It’s sickenning, really.

So we have put forth a different story that it is worth paying someone twice as much per year to give you good answers the first time. To not give you false confidence that your parts are going to be right. To be able to predict the cost of something, use simulation before bidding on a contract so you know exactly what it is going to cost and win more deals. These are things that have tramsformational benefits to a business.

Deelip: But what do you say to the argument, to which I subscribe to as well, that these dumbed down versions of simulation tools stuck into CAD systems are not meant to get real results. Rather they are more to educate CAD users who do not know a damn thing about analysis.

Blake: So you are saying taking LSD gives you insight to reality (laughs).

Deelip: No. Suppose a CAD user is asked to design a simple flange. Suppose he runs a broad analysis on his model and get to a point that tells him that his design will probably work and then pass that on downstream, don’t you think that over time he will use that knowledge to come up with designs of higher quality. I recently joked on my blog that I would not board a plane that has been analyzed by some FEA for Dummies software. Obviously, the real analysis should and will be done by real engineers who know what they are doing. But the output from the CAD users becomes the input to the engineers. So if over time the output of the CAD users could be increased in quality wouldn’t that make the job of the engineers easier? Don’t you think there is some kind of value being offerred here at all?

Blake: So you are saying you go to the Director of Engineering and say, “Hey, why don’t you have your CAD guys play with this tool, waste some time and maybe they will learn something”. That’s ridiculous. I’d rather have a process that makes sense. I recently spoke to one of our customers. He had some simulation to do. He simply handed it over to his buddy who was a simulation expert rather than doign it himself becasue he was best at doing design.

Deelip: So then I guess your customer didn’t learn anything about simulation at all. I mean, he did not gain any knowledge that could help him design better parts in the future. Here I mean learn something about analysis to the extent that is possible, not everything. Just maybe the bare minimum to know that the part that you are desiging is not crap from the beginning.

Blake: I guess there is some validity to it. But I guess what I am saying is that if I am executive in an engineering firm and I have got a bunch of engineers and a bunch of people who are focussed on documenting that engineering in CAD, if I want to think about how I want to have a more efficient business, its probably a more prudent choice to give everyone the best tool for the job. That means having dedicated simulation users who have the best in class simulation tools and maybe something like SpaceClaim to have to deal with the geometry they have to deal with and have the CAD users focus on best practices and modeling intent.

  • Jonathan Yeandle

    Hello Deelip,

    This old chesnut used to be dribbled out when the first Personal Computers appeared.

    Common sense is quite a useful tool. It can be applied to engineering analysis in a variety of different ways:-)

    Where I worked in about 1979/80 or so, the first IBM PC's and Commodore PET computers were used in locked glass rooms by people in white lab coats! I used to stand out side peeking through the glass and try to view the green screens, which were by and large shielded from my prying eyes.

    To use them we applied for special permission just to book some computer time, then if this was granted we would turn up with an appointment card, the door would be unlocked and we'd be ushered in to don the white coat and be allocated a seat.

    It took about 12 months before I gained entry during which time I personally purchased a 'kit' from the USA for a OHIO Superboard….a sort of naked single board computer, for home use.

    10 years later and the use of CAD in industry was under similar drachonian control and some of us soldiered on at our place of work with a sharp pencil on a paper or Mylar sheet whilst using bootleg copies of a popular CAD application on our home PC's.

    Now everyone from the young to the aged use both PC's and complex CAD or financial applications at their leisure.

    And still we get the same old worn out fictious arguments from people with a vested interest in arbitary control of technology.

    Fortunately, commerce (money!) will win the day and engineering analysis/modelling applications will eventually become readily available to anyone who has an interest in using them.

    If you are in business to sell PC's/CAD/FEA/coffee-mugs then you will always strive to expand your market and revenue (ask AutoDesk/PTC etc).

    A few people are modelling and analysing bridges/aircraft/safety-harnesses and in these cases some control and auditing of methods and results is sensible and required.

    A lot of people are engaged in the design and analysis of non-life-threatening articles where the incentive for control and auditing are driven by totally different and often benign environment.

    Kind regards,
    Jonathan

  • MC

    Wow, Blake has a very high opinion of himself and a very black and white view of CAE/analysis.

    First of all, the dumb-downed versions of simulation/CAE tools that are integrated into CAD are not intended to be used for accurate analysis results. Although they do educate the non-engineer CAD user in basic analysis (as Deelip stated), their primary purpose is to narrow down and weed out the possible versions/instances/configurations of a model.

    Second, I provided tech support for a major 3D CAD vendor VAR for a number of years. In that time, I experienced two groups of people that make Blake’s assertion regarding simulation and analysis only partially true.

    1) The first group of people consists of engineers who are computer and software challenged. They know their theory, but no matter how much training they have received, they are unable to apply their knowledge through the use of CAD and CAE tools.
    2) The second group of people consists of non-engineer CAD designers who are bright enough to learn the basic engineering theory and apply it accurately using CAD and CAE tools.

    Based on these two groups, I came to a few conclusions.

    1) Not all degreed mechanical engineers are capable of conducting computer-aided analysis.
    2) I hate to burst any engineer bubbles, but the basic theory for analyses such as linear static, motion, and optimization is not that difficult to comprehend. You don’t need a mechanical engineering degree to conduct all analyses.
    3) There is no argument that you will need an engineer to sign off any designs and to conduct the more complex FEA, non-linear, thermal, and CFD type analyses. However, it does not make much business sense to occupy all of your CAD/CAE seats with double the cost engineers, when you can hire a mix of engineers and well-trained CAD designers (who can conduct the basic analyses) for less money.