Author Topic: AutoCAD v. Solidworks - Poll  (Read 8552 times)

0 Members and 1 Guest are viewing this topic.

Cavediver

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #30 on: January 23, 2007, 12:43:48 PM »
Build the model in AutoCAD, no need to do ".simple elevations until they approved everything".  Build the model, DVIEW clip the appropriate views, make changes as necessary editing the solids.  It ain't rocket science.
When the piece changed from 6'x4'6" to 6'9"x4'3" to 6'6"X4', it seems a little excessive to edit everything several times.  Especially when a dozen rectangles (front elevation only) gets them the info that they need.

Note that I'm not trying to be argumentative here, I'm definitely interested if there's a better way.  I'm (unfortunately) self taught, and I didn't realize until recently just how much extra power I could apply with AutoCAD.  I've picked up a lot of great info here, but I'm always in the market for more.

CADaver

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #31 on: January 24, 2007, 01:03:09 PM »
Build the model in AutoCAD, no need to do ".simple elevations until they approved everything".  Build the model, DVIEW clip the appropriate views, make changes as necessary editing the solids.  It ain't rocket science.
When the piece changed from 6'x4'6" to 6'9"x4'3" to 6'6"X4', it seems a little excessive to edit everything several times.  Especially when a dozen rectangles (front elevation only) gets them the info that they need.
How many times to you edit those rectangles?? If it's 9" longer do you not need to edit the rectangle to indicate that?
I take you are unfamiliar with SOLIDEDIT (prior to R2007)?  With it you select a face and MOVE it 3" or whatever you need, that will "fix" it in ALL views. So to accomplish the changes you've listed requires 4 instances of one command, which seems a LOT less work than redrawing anything, or doing a lot of 2D work before going to 3D.

GROSSMOE

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #32 on: April 24, 2007, 10:13:46 AM »
Autocad is not a 3D modeling package per say.  I know yes you can model 3D in Autocad ...
?? well is it or isn't it?

I fully believe in the AutoCAD 3D and have used it for many years. I'm now at a new company that does only 2D and they want to move over to 3D but want to consider solidworks and solidedge. We design basic pump and air compressor skid units and I'm not sure which way to go. I have no firsthand knowledge of these programs and know quite well how to manipulate the AutoCAD. We fab our own units with the exception of the occasional base frame when we are too busy or the floor is too full. Is solidworks worth the money and learning curve time to make the change over? Any recommendations? Anyone?

DaveW

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #33 on: April 24, 2007, 11:00:10 AM »
Like others have said, "It is all what you do".
Modeling buildings, from my stand point, it is a minimum 14 to 1 speed increase over solidworks.
With the addition of third party software the entire structure's cut list is measured in real time, about 1 minute, and everything can go directly to CNC. Solidworks is great program, but certainly not my choice for what I do.

Again, same goes for store fixtures. It all goes directly to CNC without any of the nonsense of parametrics and constraints. This is all possible because ACAD's programming interface is second to none.
« Last Edit: April 24, 2007, 11:08:04 AM by DaveW »

Josh Nieman

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #34 on: April 24, 2007, 11:07:19 AM »
Again, same goes for store fixtures. It all goes directly to CNC without any of the nonsense of parametrics and constraints. This is all possible because ACAD's programming interface is second to none.

Parametric and Constraints are 'nonesense'?  That's an absurd statement.  Parametric design with constraints only allows for more fluid, efficient, controlled project work flow.  If something changes in my model, I get redlines from the engineer, whatever... revisions to an intelligent parametric, constraint driven, model in software like Solidworks or Inventor will take a FRACTION of the time as it would in vanilla Autocad.  I don't care what kind of customization you have, unless it's as extensive as a Mechanical Dekstop package... then you're not dealing with VANILLA Autocad, IMO... you've simply developed a unique, singular instance of a vertical platform that isn't accessible to anyone else, and shouldn't be considered anyways.

Also, there are just as many addons for Solidworks that allows for seamless CNC machine control code generation, so that shouldn't be an issue.  I will agree though, that Autocad's customizability is awesome.

DaveW

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #35 on: April 24, 2007, 11:19:21 AM »
Again, same goes for store fixtures. It all goes directly to CNC without any of the nonsense of parametrics and constraints. This is all possible because ACAD's programming interface is second to none.

Parametric and Constraints are 'nonsense'?  That's an absurd statement.  Parametric design with constraints only allows for more fluid, efficient, controlled project work flow.  


I disagree. It all depends on what you do. Parametrics may be great for you, and that is fine, for me they are non-sense and completely, totally, and thoroughly a waste of time. Parametrics were created for a certain "need". Because others who were working with solid raw material, such as wood, stone, glass, etc., needed a way to get a  cut list out and no one had a way, parametrics were used to drive the drawing and create a cut list. Sure it works. You only need to be a wiz at math and have plenty of time.

This picture blow was something I modeled and posted in the ACAD newsgroup. It was a competition between Inventor and MDT I believe. I just happened to see it right after the post. I no longer have the video, as I moved my website hosting, but I did this in stock ACAD in 5 mins and posted a video to prove it. The MDT and Inventor guys responded the next day and they had it done too. I did not even understand the math behind what they did. I am sure glad it is not "required" I had no need to revise this model. Given it's symmetry, I could have just scaled it to change size. Where as the Inventor\MDT guys would change a formula. The time difference between the two, in this case, is negligible.


If something changes in my model, I get redlines from the engineer, whatever... revisions to an intelligent parametric, constraint driven, model in software like Solidworks or Inventor will take a FRACTION of the time as it would in vanilla Autocad.
This can be very true. I will give you that. Certainly if what you do is a good match for Solidworks then this is another valid point. However, because Solidworks is not the best option for me, your comment fails to address the whole story. Although it is "possible" that updating the parametric database may save me some time, as compared to what I do now to update my model, I will still create my model so many time faster then I can with a parametric software that the the difference in updating time is rather trivial. Again, this is for what I do. I tend to do one of a kind assemblies. Minor revisions are no problem. A major revision would still be faster then having to deal with the parametrics again.




I don't care what kind of customization you have, unless it's as extensive as a Mechanical Dekstop package... then you're not dealing with VANILLA Autocad, IMO... you've simply developed a unique, singular instance of a vertical platform that isn't accessible to anyone else, and shouldn't be considered anyways..

There is another way to do this stuff, you are just not aware of it. There is nothing proprietary about the objects you see. They are 100% true ACAD 3D solids. They were created that way and stayed that way the whole time. There is data attached to the objects, as xdata, which is not encrypted, such as material name, part name, part number, that can be retrieved later. Xdata is dxf compliant. Even if the solids are turned into a  dxf this data passes out. It copies when you copy the object, etc. The technology is accessible to anyone.

Solidworks is a great program. There are quite a few things that it is exceptional for and well suited for. Like many ACAD users, I find Autodek's approach to updates, what they add, etc., lacking. I am certainly no lover of the way they manage ACAD and it can really ruin my day or week, or in one case, 3 years of wasted time helping them with MDT. I have had experiences with Solidworks and their users. They also have complaints from time to time about buggy releases and the sort. There is also plenty that I do not know. Such as with MDT not supporting previous MDT assemblies created in earlier versions. This point with MDT was pretty bad. Spend all that time creating a model and then have to do it again and again every time a new MDT came out. I expect and hope that Solidworks was more responsive and understanding of that problem and created the necessary conversion tools to update their assemblies, on release of the new version, without making the customer have to wait. Note that as far as I know Autodesk not only did not have that on release, but never created them ever.
« Last Edit: April 24, 2007, 12:02:24 PM by DaveW »

Josh Nieman

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #36 on: April 24, 2007, 12:04:20 PM »
I'm not sure you have the total idea of parametric correct.  From what I read of your posts, you seem to think that every dimension, instead of being a simple number, is some massive complex string of formulas which is rather inaccurate.

How much have you used a parametric modeling platform?  I think your opinion would change dramatically if you have used it to any degree of proficiency.

I do small, medium, and rather large structural projects in vanilla Autocad, and have done mechanical assemblies in Inventor and vanilla Autocad 2005-and-older at my last job.

An example of a parametric entry is simply "d1+12" for instance... whereas dimension #1 (represented by the variable d1) has 12 inches (or w/e) added to it to get the answer.

Therefore if d1 is 8, then d2 is 20.

If I have a revision to d1, which makes it 12, then d2 automagically updates to become 24.  Apply this to a large scale, and you begin to understand the power.  There is no need to complicate any entries beyond simple addition/subtraction/multiplication/division... no need to be a whiz with calculus, but some MINOR algebra would be helpful at times.

Each time a parametric dimension updates, it saves you many commands that Autocad would have required you to do... saves on moves, arrays, copies, solidedits (oh god it saves on SO much time wasted using solidedit commands)

Basically you can create the initial model in a manner that gives it some of the intelligence you have regarding design intent.  For instance, say I have a wall that is 8'-0" tall.  There is a partition next to it that has no height requirement, other than being 12" shorter for various reasons.  Well the plate height needed to be modified to 8'-6", let's say.

In Autocad, you would probably: Solidedit, to extrude the main wall 6" higher, then extrude the partition another 6" higher.
In Inventor, you would change the 8' dimension, to 8'-6" and the shorter wall automatically updates to be 7'-6" because you told it to be 12" shorter than the main wall... you didn't tell it to be 7'-6"

Obviously a small example, but I'm sure you can see how that compounds.

Josh Nieman

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #37 on: April 24, 2007, 12:08:40 PM »
and by the way, those are some pretty nice looking models!  Looks like some fun projects to work on.

DaveW

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #38 on: April 24, 2007, 12:56:59 PM »
Thanks Josh. They are not all mine.


I advised Autodesk for 3 years on MDT. They created the equation assistant based on their understanding of my needs. From I was told, the equation assistant cost more then they had spent on Inventor to that date. It was released in MDT 6, but never really finished. I was very unhappy about what they created and their final understanding, and the lack thereof, and that it was finished to have a few more little pieces of code that would have made it quite a bit better.

I found that it was faster, for me, to just make/edit every solid one at a time to make a 3D solid model and update it for changes AND measure every solid by hand and type it in excel, then to use any parametric software.

With that said, I have since automated the measuring and the changing of sizes of the assemblies is done with a software that I bought the source code to. It stretches 3D solid model assemblies (multiple 3D solids at the same time). Parts that need to stretch will stretch and parts that need to move will move. No constrains at all.
It is important to note that it was still faster to do it all in stock ACAD from the get go. Just so you know, you mentioned having an app as powerful or more then MDT, that source code I bought crushed MDT in the Detroit area when MDT was released. I have offered to sell it to Autodesk for ACAD, so it would be a stock item for everyone, but they did not seem interested. I would be very happy if ACAD's stock abilities were able to stretch multiple solids at the same time. Also, this code was released before the solidedit commands. I am of the opinion that Autodesk actually copied its functionality or used it as a template, when they created the solidedit commands. Even before I had this code and knew of its existence, I rarely used solidedit. If I need to make a solid bigger it was copy and union. It's six to one and 1/2 dozen to the other. I am pretty fast at doing that. so extruding or moving a face is the same time to me. I am very glad I do not need to do it to 500 2x4's to make a floor 12" taller. :)

I have to get some work done, have a great one.
« Last Edit: April 24, 2007, 01:01:50 PM by DaveW »

Josh Nieman

  • Guest
Re: AutoCAD v. Solidworks - Poll
« Reply #39 on: April 24, 2007, 01:09:53 PM »
Dang... that software addition you have sounds like it eliminates all the time-bogging problems I have with standard Autocad solid editing!!  That's awesome!  That's EXACTLY what Autocad should have, if it is to become a decent 3d modeling package as it advertises itself to be!

I can see how you would prefer your package to any parametric software out there, now.