Factors to consider when choosing a CAD Package 4

[KENAT]

There are frequently threads here and elsewhere with rather general questions about "which is better, CAD software A or CAD software B?" or "We work in <vague description of industry> what is the best CAD system?" etc.

Even a detailed, well-worded initial post is unlikely to have all the information required to make an informed decision. In fact, some of the required information is likely confidential and shouldn't be shared on a forum like this anyway.

Also many of us have our own preferences and prejudices about different CAD packages based on our own experience which may not be fully representative of the true state of things when viewed objectively.

So, I'm considering setting up a FAQ with a list of factors to consider when choosing CAD and perhaps other guidelines suggestions etc. I've done a few Google searches and have some of my own resources but thought it would be worth getting the input of other members.

So any members thoughts on what to consider when choosing a CAD Package/comparing offerings is appreciated.

KENAT, probably the least qualified checker you'll ever meet...

 

[MiketheEngineer]

I love Turbocad...

I could train a monkey how to use it in an hour and it is cheap - about $900 for a new full blown package. Also interacts w/ Autocad and can save in about 20 formats!! Buy an old copy off of E-bay and upgrade cheaply. AutoCad is like using a nuclear bomb to kill a fly for most appilcations.

 

[KENAT]

In case I wasn't clear, I'm not looking at making a decision on CAD systems myself anytime soon.

I'm also not really looking for people to just say what they think is the best system.

I'm looking at creating a FAQ that helps people who are looking at a new CAD system do so in the most effective manner.

KENAT, probably the least qualified checker you'll ever meet...

 

[ewh]

Good idea, KENAT.
The first few off of the top of my head...
1. Customer requirements
2. Surfacing requirements
3. Assembly size requirements

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[ewh]

...also
4. Concurrent and downstream requirements

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[ctopher]

I think it depends on the industry, but the very basics are the same.
To me the most important is how easy can a user learn it, its compatibility with various importing/exporting file types, and well its status in the industry. Then price comes next. If I were starting new, it would have the latest and greatest in solid modeling technologies, FEA and photo rendering.
There are numerous universtites that have done studies on this. Here are some links to extract some info from to get you started.

http://eng.monash.edu.au/uicee/worldtransactions/WordTransAbstractsVol6No2/04_Devon33s.pdf

http://www.medicalimagingmag.com/issues/articles/2003-05_04.asp

http://machinedesign.com/ContentItem/72569/FEUpdateMakingAnalysisadesignfeature.aspx

http://optics.org/cws/article/research/25404

http://www.emachineshop.com/faq/general.htm

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 06/08
ctopher's home (updated Jul 13, 2008)

 

[MadMango]

I think there are 3 basic scenarios when looking for new CAD packages:
1- no previous CAD package
2- outdated CAD package
3- inadequate CAD package

Each situation above needs to be looked at from different perspectives. I doubt I can cover them all, but I'll give it a go.

If a company is just now considering a CAD package, they have to decide between 2D or 3D capabilities. This is a decision that needs to be based on the type of service the company provides. If you produce wood signs, cogs and manhole covers, chances are you will not need a powerful parametric 3D package. If you produce a mechanical contraption with various moving parts or sensuous exterior shapes, you may find a 2D package inadequate.

If the company has already ventured into the 2D CAD realm (it is the 21st century by the way), they may have noticed their current package is outdated. This might have been noticed by reading trade magazines, or by having customers or suppliers asking for file types that the CAD package is incapable of outputting/exporting.

If the company is already using a 3D CAD package, they may think more productivity might be gained by switching to a new 3D package, or their products have grown in complexity that they now find their current 3D package is inadequate. To me, this is the hardest scenario to be in. You are already using CADx, but you see advertisements for CADy that look pretty convincing, yet you know people that are happy and productive with CADz.

First I would suggest a company look at their current CAD package and identify the short comings that have lead them down the path to find something new. Write these down, refer to it often. Are you using some sort of ERP or PLM software, do you want your new CAD package to play nice? If so, write that down (your work flow) as well. Do you want your CAD package to drive configuration management, or do you want configuration management to drive your CAD package? You guessed it, write that down (your processes), refer to it often.

Realize that any transition to a new CAD package will also require some new computer hardware, simply because you've probably been using CADx for a couple of years and that is a life time when talking about computer components. You may think that you can simply buy new graphics cards and all will be well, but that will just lead you to discover a new bottleneck in your workstations. Its better to just buy new systems if yours are 2-3yrs old and you are switching CAD packages. Chances are, the money you spent a few years ago is going to go much further today.

As far as selection of CAD, I suggest talking with your vendors, customers and even competition is possible. Find out what they are using. Find out what type of file formats they like to work with. It makes no sense to purchase a CAD package that is incompatible with the folks that you work with.

As far as CADx being better than CADy, and CADy being better than CADz, it doesn't exist. Each product does most things well, some do other things better than others. Very few will do something that the rest can't do. Once you identify your needs, work flow and processes, it will be much easier to select the proper CAD package.





"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of these Forums?

 

[ewh]

Excellent post, MM!

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[IRstuff]

The "sensuous exterior shapes" got my attention. ;-)

TTFN

FAQ731-376

 

[MadMango]

Well... it was such a better word than swoopy.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of these Forums?

 

[GregLocock]

Other things to consider

Compatibility with main customers

Integrity of database functions (our guys now spend ~25% + of their time doing database management rather than plonking curves on the screen)






Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Pud]

I would add:

"Do I really need one at all?"

Would it be cheaper to contract it out on a ad hoc basis? (Like wot I do!)

Harry

 

[PSE]

Good Posting by MadMango and links by ctopher. My considerations:

At a minimum, what do I need the package to do?
Does it have additional capabilities that are "nice" to have?
Can it evolve if/when my needs change?
How secure and stable is it?
Are there legacy concerns with other older packages/files?
Is a file viewer available and if so what are its capabilities?
What are the associated hardware requirements?
What is the initial cost and what are the maintenance costs?
How well does it interoperate with other software packages?
How good is support if/when I need it?

Regards,

 

[KENAT]

Thanks all, not sure exactly when I'll get round to making the FAQ but please keep putting your ideas, I'll try not to leave it too long.

KENAT, probably the least qualified checker you'll ever meet...

 

[MadMango]

Other concerns may be integrated or stand-alone FEA and other analysis tools.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of these Forums?

 

[ajack1]

Wow, where do you even start on this one?

Firstly, carefully identify your exact needs, do you just need a system that makes pretty pictures and pdf files or do you want a system that can be used for production directly from CAD, be that FEA, machining or inspection? This is probably the biggest consideration IMO.

Secondly, what kind of data will you bring in/ send out, the more complicated the modelling the more critical it becomes that you run the same system as your customers. Most systems will import and export a variety of exchange formats, but you will spend hundreds if not thousands of hours a year “fixing” data from another system.

Thirdly, do you just need a solid modelling system or will you be using surfacing? Most mid priced solid modelling systems are great with solids but next to useless when it comes to complicated surfacing.

Fourthly, what size files/ assemblies will you be working with; if these are likely to be large how well does the system cope with them?

Personally I do not think cost is a major factor if the system contributes in a major way to revenue. It is like buying a pair of walking boots, if all you do is take the dog to the park for 15 minutes a day, any are fine, if you start climbing up mountains for days on end, you soon regret going for the cheap option.

Finally NEVER trust what a salesman tells you a system can do, at the very best they will avoid any weak areas in the system, or not even be aware of them, usually they just lie. ALWAYS make them use your data for a demo and do what you want, this is the only way you will find out what a system cannot do rather than what it can.