Is it you or your CAD software?
I spend a lot of time wondering what kind of work I could do on other software. I wonder if some of the problem areas in complex Solidworks models would be problems in SolidThinking, Rhino or VX. In my engineering career, I’ve really only used a few CAD packages. I started out trading between CADKEY and AutoCAD. While on AutoCAD I toyed with the AME (Advanced Modeling Extension), and did some 3D solids using boolean commands. In CADKEY I did mainly 3D wireframes. I considered CADKEY and AutoCAD to be roughly equivalent. I really wasn’t playing partisan to either side. The command structures were backwards from one another as I remember, and I think I was faster in AutoCAD, but they were roughly equivalent.
Then came Mechanical Desktop. The year was 1995, and I was using a new extension of AutoCAD called Mechanical Desktop 1.1. It was software meant to do 3D, which I felt I needed to help me solve some spatial problems. I was making a pressure testing manifold fixture to test many pressure sensors at once. It needed both pneumatic and electrical connectivity, and had to be hand loaded with the ability to lock each position in say a 30 position grid. I could do the designs in 2D, but it was difficult. Everything in the third dimension I had to visualize in my head. This sometimes made it hard to explain to people how the stuff worked, and if I made a mistake between views, I’d get a call from the machine shop.
The 3D was definitely a step forward, but Mechanical Desktop sat on top of AutoCAD itself, and there was a convoluted process to get 3D views into 2D. Beyond that, it was extremely unstable. At my next job I evaluated SolidWorks 96 and bought 97. 2D to 3D made a huge difference for me, and Mechanical Desktop to SolidWorks also made a difference, mainly because of software reliability. But those were the early days. What if I had made the switch between SW and Inventor more recently? SW and Pro/E?
I think as long as you are trying to accomplish something within the mission of the software, aside from training issues, you aren’t going to see a big difference between one software and another. Even switching between history modeling and direct editing, given you can make the conceptual transition, I don’t think the changes would be that big. People have a certain rhythm that they gravitate toward, and they will setting into that groove regardless of the tool.
It seems to me that the biggest difference is created by the individual user, and then smaller differences are created by the software itself, how well it is suited to the job, and the user’s training or familiarity with the software. I think the best CAD users I know would eventually be good at any software – knowledge of design process, 3D spatial ability and troubleshooting skills are abilities that can be applied to any software.
The fact that all of these software makers compete mainly by imitating one another is another factor that prevents there from being much difference between using this or that. Some of the remaining older software like NX and Catia are differentiated from the others, but SolidWorks, Inventor, Solid Edge all look approximately the same. The Direct Edit crew seems to fall into similar ruts. They argue over petty points where this one is better than that, but in the end, a user using all of them would not see a huge difference. I guess that’s the reality of competing by imitation, is that no one gets much of an advantage. They all talk about being innovators, but I think marketing folks simply cannot spell imitators.
I have had the opportunity to take a few systems around the block for a test drive, and while I can say that is definitely true about the mid-range systems, the programs such as Catia and NX do have more capability than Solidworks. I have had surfaces give me many hours of trouble to be able to take and manipulate the the same surfaces in a matter of minutes in NX. The one thing that Solidworks seems to have over other programs is its ease learning. It is very intuitive, you can put a new user in the chair and have him up and running with little coaching. NX in this regard is not too bad in this regard, Catia is more difficult, and ProE in this regard is darn near impossible.
SW’s advantages over NX (IMO): Weldments, configurations, drawings, assembly constraints.
As far as switching, you are correct Matt. It’s been a tough adjustment going to NX, but in the end I can still create the same geometry/models I have before. Many things that I still have a hard time figuring out are mainly due to lack of training/experience, but for the most part I’ve been able to get by with troubleshooting skills.
I don’t know, perhaps I have a thing for SW or may be I truly want SW to be the number one design engineering software. In my humble opinion I think SW offers the best combination of tools there is to date. Most of all I like the logic behind every step during part and assembly creation process, the speed, the efficiency the list could go on and on…
I’ve had a chance to try AutoCAD, Pro-E, Alias and SW, and I was very happy to find my self being a SW user. No offense to others.
As a side note, for as long as one can make his client happy it doesn’t matter by what means he is doing it.
I dont think that SW innovate as well as they might. Here are a few thoughts:
a) SW are over thinking, over managing and over conscious of everything they incorporate. They kill the simple brilliance and effectiveness of doing the necessary and obvious with too much internal consultation and ‘systematic creation’.
b) Their attention is diverted and originality compromised by looking at what the other players are doing. The stimulation of innovation at every opportunity is extinguished by dozing clockin clock out imitation of doing whats defined and required. This is not a hungry company any more and I’ll venture there are some overpaid and underperforming people getting in the way of innovation.
c) Management pace the company to prolong their own corporate existance and keep shares up. Because of their super safe rounded approach innovation takes twice as long as it should and delivers only half of what it could. Its like a self sustaining govt department. That is the innovation has a set quota and cycle and doesnt live and cross fertilise as it ought. Innovation is something to be accomplished rather than a way of being.
d) They fundamentally misplace their attention and energies because they place themselves in a market share, product cycle game rather than being a specialist technical service with an evolving body of excellent work. The innovation needs to capture short term attention of a buyer once a year rather than forfill long term objectives or be highly responsive to real user needs.
e) They ignore or rethink ideas presented to them because it doesnt fall in with the company ownership culture or a managers own particular view of how it ought to be.
This kills user participation and isolates coders from the real needs. Also because of commercial sensitivity here is no early trust and direct feedback on ideas and because the code is closed outside innovators are kept out.
There are invisible barriers to sensible responsive innovation that make the end result of most surviving innovation a near miss.
f)They prioritse their work in a way that leads to general satisfaction for the average user. That is they target stuff that by nature isnt likely to produce any startling innovative solutions. If there is innovation it arrives where the most votes or sales lay. The many small ideas that could amount to a large overall gain are removed by the appraisal system because they are too fringe. That is, innovation which is by nature fringe is removed exactly because it is on the fringe. So innovation is self defeating in SW hands.
g) Marketing seize on a subprime innovation and promote it excessively or wrongly and that leads customers to question the worth of ‘enhancements’. The result of innovation becomes feared or ridiculed and then people dont want SW to engage in their innovation. They play safe with imitation. The criticism of previous innovation stymies any future innovation. Innovation becomes something to avoid for both SW and the user.
For the last ten years it has been 70% SolidWorks and 30% Rhino while in the last year it has been 80% Rhino, 10% SolidWorks and 10% 3dsMax.
Some stuff are way easier and faster in SW. But when a project involves less than 10 parts or there is complex modelling involved, the Rhino is flying.
My cad experience is as follows:
1- AppleCAD
2- GenericCAD
3- AutoCAD
4- Mechanical Desktop
5- Solidworks
Of all the ones named above, Solidworks was the easiest to learn. All software has it’s pros and cons but so far Solidworks “Works” for me. That sounds like a commercial. Maybe Mr. Ray could use it and pay me a royalty.
I’d love to try other CAD/Design software but time constraints are tough right now.
Valid points, all. I work for a software company (cfdesign) that works hand and hand doing fluid flow/thermal simulation with all the major MCAD tools. I have used everyone at least weekly for 12 years or so. Absolutely agree that many of them are similar.
I’m not as critical as users as I use so many products they all blend together, but appreciate if you were driving one day in and day out you’d have a stronger opinion. But, I do see some innovation the most notable is SW, Inventor and believe it or not NX. True, the first two look very similar but they do have some specific features that are very well thought out and are different from the next. But, in the end.. they all get to the same place one way or the other.
Something to ponder from a software vendor. It is amazing how resistant to CHANGE the engineering community can be. Meaning, software vendors only have so much wiggle room when it comes to innovation. But, I can tell you, I’ve been involved in some projects that solve specific problems that users gripe about in an innovative way but in the end, users “don’t want to do it that way”. They like the old way even though they complained about the old way – in the end- it’s what they knew.
Anyway.. figure I’d give you a different perspective.
I started out using something on a Mac in the mid-nineties in high school drafting. We got to do one project on the one computer but everything else was done with triangles and a t-square. Since then I’ve used AutoCAD, Mechanical Desktop 4, Inventor, SolidWorks, Solid Edge, ProE, CoCreate, and ME10. From what I can tell SolidWorks is still your best bang for the buck out there even with it’s flaws.
… and lotus notes is a great email system.
I hope you will forgive me my guilty pleasure in seeing my chosen 3D CAD software rise above the others and dominate the market. As a 12 year veteran of the CAD market, I agree that a lot of companies have tried to replicate SolidWorks’ success. Back in the day, you may or may not recall that SolidWorks and SolidEdge held a similar market share and target audience. I think we can all agree that SolidEdge just doesn’t compete anymore, and Mechanical desktop was just doomed to fail thanks to inferior technology. From my experience, I think the differentiating factor was the dedicated, local SolidWorks Reseller channel that competed on service, training, and technical support.
My experience with Catia says it’s a nice piece of software. Granted, I hated V4, but V5 felt a bit like I was at home in the SolidWorks Environment with a very slight learning curve. The place I recently worked had Catia and SolidWorks, and I did overhear colleagues saying that drawings were much easier in SolidWorks.
In my career, I can say that I have seen Catia overtake SolidWorks in terms of functionality in one area, and that was largely thanks to the Enovia product and digital mock up capability. Still, If I ran a company that made very complex things, (like cars for example) I would really want someone to run the numbers to justify the cost differential to me.
I still don’t understand how ProE has held onto their market as long as they have. They started as a mid level cad, but SolidWorks grew up from underneath them to capture that market. As the windows platform became more capable, so did SolidWorks. Suddenly they were over priced and difficult to use with an uncomfortable environment that didn’t really play well with windows. I think it just goes to show that companies aren’t willing to change unless there is a proven incremental difference.
At the end of the day, Yes, CAD is just a tool to make models (hopefully as quickly and easily as possible). I think the future will be determined by technologies that allow users to collaborate easily, and reduce the headache involved with administering the tool.
The obvious first software for me wasn’t even software but a pencil and drafting board out of community college and then moving right in to the mac world with Mac Draw. Cadkey was next (Matt, it nice to meet people that remember Cadkey because I loved it) what with all the function keys on the left side of the keyboard. On to Unigraphics with version 13 all the way up to 18 and then comes the nightmare of Pro/E. I’m sure glad we didn’t stay long on Pro/E because it wasn’t very user friendly or fun for that matter. Finally to SW.
I have to say and maybe it’s the internet and social networking and local user groups and forums and bloggers but I haven’t had as much fun working in the past 24yrs than I have in the past few with SW and all the people I’ve met. So I’ve seen and used two of the big three but I still like how SW works.
I think that the things that tend to differentiate Catia and NX (to some extent ProE) is their ability to do more operations with less memory. Catia and NX also are much better at dealing with complex surfaces. We have had the same part in a standard interoperable cad format, and we have seen SW run up to 5 gig of memory on the part, while ProE was at just over 1 gig. It was to the point that SW was almost not workable, and the part had to be broke out in ProE so that we could manipulate it in SolidWorks.
Similarly some of the major compound surfaces that are found in the aerospace field do not lend themselves well to offsetting, thickening, or trimming for various reasons. I have seen SW many times give you a preview, and then not perform the action desired stating that there is a geometry error of some type or another. Then fighting with that surfaces for a day or two, you get it to work. I have taken the same surface and have both Catia and NX perform the same operation that was asked in minutes.
So there are some differences in performance, does this mean that something cannot be done in SolidWorks, usually not, but the extra effort and time may mean that the more expensive seat of Catia or NX will pay for itself in lower hardware requirements and less time spent. Do not get me wrong, I think that SW is the most intuitive cad package out there, but it does have its limitations.
Adam
Quigley’s law:
The most annoying software is the one that you are currently using.
The best software is the one that you don’t have or the one you didn’t use for that job.
Over the years I’ve bought and paid for SolidWorks, Ashlar-Vellum Cobalt, VX, Think Design, FormZ and a few other 3D modelling systems. I think that ultimately “most” systems can model “most” things, but some will do some things better than others, and some things not at all. If I could combine the tools and general robustness of SolidWorks with the freedom of creating and editing 3D curves and surfaces of Cobalt, the direct surface control and editing of ThinkDesign and the modelling flexibility of VX you might have a system that can do it all. Until such a beast exists the limiting factor is what lies between my ears. There is not much I can’t do in SolidWorks these days – but sometimes I yearn for a bit of flexibility!
Dont’ entirely agree with Quigley’s Law but I do, personally beleive, this comment of Kevin’s, “I yearn for a bit of flexibility!”, is a very important goal for ALL CAD software developers.
@Neil
amen brother!
My full reply to this post to come later when time permits.
Considering the state of most of the CAD systems on the market today, a person could probably use any CAD system to accomplish most any modeling task. They are all becoming very mature and capable. But is that all there is to it? Should users now be allowed to use whatever CAD system they like best?
Once a model is created and optimized, then what? Now that you have this nice looking model, (valuable engineering IP) what happens to it? Physical drawings used to be maintained in drawing cabinets for potentially many decades depending on the company business. Most any drafter/designer/engineer could reference them years later and get value from them. How does this 3D CAD data get used? Who else has to interact with it? What is its lifespan? What long term value does it have? Will it be valuable 10 years from now? So here it comes; the 3D model might be of some value, if in fact it is accurate robust data, but what about the history tree? What’s the lifespan of a proprietary history tree?
I just think that there is much more to consider regarding the use of a CAD system than its ability to satisfy the user.
This is an issue in the AEC sphere as well now that they are moving towards BIM as a working method. Given that the value of BIM is working in 3D and maintaining a 3D information model of the building through its lifespan, how long does the original data stay usable? What methods are in place to update the data with time? Back in the day a set of drawings marked up was sufficient to keep records. Not any more.
The other factor is that project sign offs (I’m back in the product design world now) increasingly happen with no drawings being done. I’ve been working with many toolmakers for years you just issue a STEP file for final approval before moulds are made. This is certainly an area where CAD companies could make life easier for the designer – comparing original released data to returned data for approval.
What I do is import the released STEP file back into SolidWorks, then import the approval model and overlay. Sometimes this is easy, sometimes not, depending on where the toolmaker has placed the datum planes.
What I then do is colour the original red, and the approval data green and look for differences. As a final check I prefer to section the parts and create check drawings. Ironic isn’t it that the 2D drawing is often the best way to check a 3D data file!
Where the likes of SolidWorks could help more in this process would be to make markup easier, allow easier use of colour in drawings (maybe taking the colour from the model to the drawing sheet for hidden line/wireframe). But the best solution would simply be to allow me to set up planes that cut the model data and create curves at the plane that can be brought through to a drawing sheet. Yes I know you can create sections in drawings but when the STEP files don;t stitch properly (like the ones I was sent today – 376 open edges and faces) I need a quick check against the surface data and currently this process is anything but quick (maybe I need to look at it again in 2010).
My experience is as follows:
1- AutoCAD
2- Helix Cadam System ( 2D ) –> I bet none of you have heard about this ….
3- SolidWorks
4- I-Deas NX
And I also have limited experience in NX6, ProE WF 4, SE ST, Inventor 2010,
After all, I still feel SolidWorks still my 1st choice as I so get used to it.
Other software might having cooler interface or more powerful function, but Solidworks still is my favorite..
I’ve used quite a few 3D packages over my career.
I initially learned (taught myself) 3D CAD with Alias running on an old SGI Personal Iris at my university. I was only allowed to use it in the evenings and over Christmas break since it was a component of a small business incubator. It got me started, and gave me an edge up against most folks in my industry… that was 1991-1992
My first job landed me on a seat of Pro-E… I think it was Rev 8 maybe at the time? I spent about 8-15hrs a day on that tube during those few years.
After that it was a somewhat of a niche software called AutoShip at a different company that manufactured boats.
About 1997 or so I was working for a NewelRubbermaid division using Mechanical Desktop. It certainly seemed to be easier to use than Pro-E from an interface standpoint but lacked some of the power that Pro-E had.
Around 2000 I took over a design dept. at a consulting firm that was using SolidEdge. What a pile that software was. We experimented with Rhino but it just couldn’t deliver everything we needed, we still used it but it couldn’t fulfill our primary needs. We even used ThinkDesign Think3 or whatever – an odd subscription based CAD software. It worked pretty well but if you wanted to stop your subscription your software simply quit working and all your data was lost unless you signed back up. We spend countless hours converting data to a generic format before our subscription ended
We picked up SW and loved it… especially when the advanced surfacing engine came out. We ended up getting a seat of Pro-E to coordinate efforts with one of our largest clients but none of the designers really warmed up to it. It was still seemingly cumbersome, un-yielding and difficult to use.
I’ve worked in a couple places since then and have been pretty much a SW guy since. In my opinion it’s the best there is, I really wouldn’t want to use anything else these days.
I was very fortunate to be have watched the evolution of CAD since the main frame systems back in the early 80s. In fact the first system I ever used was Pafec Dogs, interestingly a well thought out modular cad system based on a Modelling Information philosophy not too distinct from today’s BIM.
In 1985 I was first introduced to Autocad which started my long association primarily with the Autocad genre of products both 2d & 3d….I think the last count was about 14 systems!
This article rings true in many respects and indeed something that I myself have been pondering with respect to various modelling challenges and the most suitable CAD product for development.
For many years I was an avid MDT user, its actually quite a good program and I must admit I personally did not experience the same problems as mentioned…one of its key strengths was interoperability…which strangely enough is still an issue with its successor!!
Anyway market forces dictated a move to Inventor which I really liked and I was convinced that this was the product for me…until I ventured into a new field of engineering.
This new project required a lot of work with sheet metal work and the Inventor at that time was sadly lacking in this area, although today that is quite different.
So after some research I moved to Solidworks and I was immediately impressed..it was so easy to use, with logical command structures and easy manipulation of the standards to suit my work environment, plus you didn’t have to wait 5 minutes for the program to load.
My only issue with Solidworks is its 2d drawings…for some reason the finished 2d drawing does not compare well with Inventor…maybe I am just too fussy but the representation in 2d is somewhat lacking. Unfortunately I have a requirement to produce quality 2d drawings for many reasons other than just conveying design intent and lately I was pondering the idea of moving back to Inventor for this reason only.
I am feeling the pressure to move up to the current version of Solidworks. I am using SW2007. I am concerned that the new version will be too slow and large to run on my current laptop. I use a Dell M70 2Ghz, 2GB, NVidia 1400 with an external 24″ monitor.
When I use eDrawings 2010 it is too slow to be useful for viewing a model. The model does not show all of the surfaces when rotating. Saving as 2010 does not improve anything. The model views well with eDrawings 2007. The .exe version of an eDrawing 2007 is much smoother and faster than eDrawings 2010.
It may be a graphics driver problem. I thought that Open GL would continue to work for the 2010 version. It will take an hour to download the new 191.87 driver. If it breaks SW2007 I will be pissed.
A couple more half hour downloads of NVidia drivers that were also not suitable for my computer. I am beginning to wonder if SW2010 can run on my old reliable M70.
Solidworks SW2010 is just too fancy to run on my laptop. It requires a graphics card driver that will not run on the Dell M70. I now know that dropping subscription was a good move as I would not be able to run the current release version of Solidworks.
This is perhaps an artifact of our throw away society. Software seems only to get bigger and slower. Soon it needs more memory, a faster processor, a more expensive graphics card. Of course bugs are added and rarely fixed.
Solidworks is claiming to be helpful in PLM. Too bad you have to throw away your reliable computer to run Solidworks.