Where is CADD going in the next five or ten years?

[Qshake]

All,

I am wondering where CADD is going in the next 5 or ten years. The industry is AEC, but responses shouldn't be limited to that.

Thanks,
Qshake

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[ctopher]

I can see CAD getting more automated, but the biggest changes being on the CAM/machining side. Eventually machinist going away and 3D printing taking it's place.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP3.1 / PDMWorks 05
ctopher's home site

FAQ559-1100
FAQ559-716

 

[GregLocock]

Tighter integration of design and simulation packages, including optimisation, seems inevitable. I don't know how they'll get around the GIGO problem.

I'd like to see a touch based controller that would act like a lump of modelling clay for defining complex shapes (OK dream on). Hmm, that's already possible by using a stereographic camera and a lump of clay.








Cheers

Greg Locock

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

 

[aamoroso]

Standardizing CAD file formats. Standardizing terminoligy, and methods as well. Now with CAD Files be able to supply more information then drawings with less cost to the companies, taking advantage of that is the next phase. Standardizing is a logical step, then they can take it to the level of automated CAM directly to the machines. Models should eventually be downloaded directly from networks or online to a machine in a building where the material only needs to loaded and the Cycle Start button hit. It will take more than 5 years to achieve but the first steps would be to standardize the outputs of CAD. Machinists will be necessary, but their duties will evolve into more of CAD Modeler and QC Inspector. If a machinist wants to rough out parts in the future he probably will working with the finished model from designers and build a block of raw material around it. He can then design the process of machining by cutting away as he would on a machine until achieving the final product. It will be seperate configurations of the finished design for each process of manufacturing. Supplying the necessary informat to the machine to get the intended result from each step.

Fill what's empty. Empty what's full. And scratch where it itches.

 

[ajack1]

Most of what has been described here is already in place although no doubt it will all be improved. When you look at the way technology has changed in the last twenty years I doubt many can say where it will be in twenty years time.

I am sure integration between systems will increase tremendously and I think systems will become more module, where you buy a basic system and tailor it to your needs, again I know that happens now but it will increase IMO.

In the meantime getting something far better than IGES would be nice that is universally available, the amount of time I spend trying to import models using different configurations and still “fixing” bad surfaces is a real pain.

 

[jackk]

Look at what has happened to word processing. CAD will follow, and it definitely will be 3D based. CAD will also take advantage of the graphics technology that's in game chips and software.

I'm still waiting for my 3D computer display, and my Dick Tracey wrist watch computer. But, that's another subject.

 

[ctopher]

I remember almost 20 years ago when McDonnell DOuglas had UG and they were developing gloves & goggles wired to a computer that enabled users to create models in space by movement of the hands. I can see something like that making a come back, maybe a different scaled down version.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP3.1 / PDMWorks 05
ctopher's home site

FAQ559-1100
FAQ559-716

 

[PSE]

Further integration of CAD, FEA, and CAM. Perhaps the ability to further integrate between other specialized programs such as project management software or being able to fully import PCB models and related information into a mechanical CAD package. 2D will end up in niche applications where 3D is not really needed.

Regards,

 

[drawoh]

Mechanical CAD tends to run computers at the very ragged edge of their capability. I do not see this problem going away in the next ten years. I expect that we will be able to model helices and screw threads and that we will be able to drive screws down into the tapped holes. We will be able to route each individual wire through the cable harnesses. Really modeling machine or sandcast finishes will be nifty too.

Architecture looks like a great way to abuse poor helpless computers, but I do not work in the field.

When I do updates, I will still be able to wander off for a cup of coffee.

JHG

 

[GregLocock]

"Mechanical CAD tends to run computers at the very ragged edge of their capability"

It would be more accurate to say that modern CAD systems push consumer grade PCs quite hard. If you bought a bleeding edge PC 3 years ago it would still be plenty fast enough for any CAD system I know of, maybe with a new graphics card admittedly. (says he eyeing the 12000 dollar, 3 or 4 year old, monster on his desk - 2 GB RAM, 100 Gb hard drive, 1.7 GHz Xeon). I don't think that is a cost effective approach, but it is possible.







Cheers

Greg Locock

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

 

[PSE]

I have managed to bring several high end computer's to their knees with some models that I had to contend with. This was more a limitation of the OS than the hardware but there were assembly models that would take ProE 45 minutes or so to bring up (that was my morning 2 cups of coffee). Hardware definitely has improved and it helps, It is nice to be able to do some analysis runs that only take a few hours now rather than overnight.

As the hardware advances, we seem to be able to continue to push it hard based upon the additional capabilites added to the software. We can throw new problems at it to work at solving.

Regards,

 

[drawoh]

GregLocock,

2GB RAM, 80GB hard drive, P4 3GHz processor plus possibly another one, although this does not matter with SolidWorks. NVIDIA Quadro4 580XGL video card.

My stuff can get complicated. I have had SolidWorks models running here with 4000 pieces attached.

Ten years ago, I had a SPARCstation with 32MB RAM, and AutoCAD_12. I loaded huge arrangement drawings into it, and I had to disable the automatic updates so that I did not have to wait five minutes every time I redefined a block.

With a lot of applications, it hardly matters what processor you have. My PII/350 with 256MB RAM at home surfs the internet, runs office suites and retouches photographs perfectly. I think we are facing a return to the old days when office equipment was expected to last twenty years.

Video gamers and CAD users are not there yet, and I do not think we will get there any time soon. There is too much neat stuff to do.

JHG

 

[GregLocock]

So you are running your professional software on a gamer's rig? Real analysis computers have multi cpus and other computers to take the load. Mine talks to a Cray if it has to.

I really don't understand why if the hardware is truly a limitation you limit yourself to consumer level stuff. Other professionals don't.







Cheers

Greg Locock

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

 

[RCDLtd]

I'm running twin Xeon 3.4GHz processors and apart from 3DSMax, none of the "High End" CAD systems I use can make use of the Dual CPUs. And Max can only do it when it's rendering.

Having said that it does mean that the CAD software only every uses 50% of the available Processing power, allowing other application to run simulateously.

I appreciate Solvers have been benefitting from Dual Processors for a while now.

 

[cadcamguy]

Todays machinist already does 3d design and QC work. I don't think the machines themselves are going to reach the level of just push a button ... not when you consider the costs. It like 20 years ago when cnc's were going to replace all the machinist.

I see it happening more in the design area. Input some design parameters the software pops up some different proven solutions to fit those requirements. Choose A,B, or C then next step. Which might be pick different parts from exisiting

http://www suppliers..make

mods... then generate a BOM..
post for quote.

China is going to be driving the market for machine tools in the near future and they aren't going to be looking at reducing labor costs. They aren't going to need robots to handle the parts.

How long will it be before a puter will know if something is sexy or not

 

[ctopher]

cadcamguy,
I have worked with machinery that all the operator does is set the material in place and press the GO button, the machine does everything else. It measured the material to find orientation then uoload the CNC program to cut. This was 15 years ago.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP3.1 / PDMWorks 05
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[cadcamguy]

I saw some pretty impressive things at IMTS last fall also. Most any machine can pick up part zero and tool length offsets and even measure the finished part, make adjustments and spit out a QC report. Even robotic arms to load and unload parts.

But it all cost money, big money.

Which is probable why what you saw 15 years is not more widespread. Curious of make and model and application.

The current trend is a pallet pool as far as milling goes. You have a drum with up to 300 tools, gobs of work offsets and TLO's even a AGV to transfer the tombstones.But you are looking at close to a million dollars, just to get it to this point. You still need someone to make the fixtures and set it up. Unless you plan on wasting a good deal of material, there is no one size fits all method to get around these things.

The software that supports the Cam end has done nothing but improve.. feature reconigition. in 3d solids, associativity with the model, ProE ,UG . There really is no excuse for waiting for programming to get the NC files out. Not anymore. With High Speed Machining, cycle times have dropped as well.

Setup time is dead time, no product is going out the door. If you can't push enough parts through to offset that lost time you won't make money. A big problem has been to stop running one thing to go over to another. Very rarely with something new is it just load and go to town, you have to grab that holder that you were just using and use it on the new part or disturb another setup, you might get lucky every once in awhile.

What causes a good deal of this is that by the time something hits the floor it is already late. Panic.. throw out the OT...put out the fire. Until the next time it happens again.

The improvement in the process, beside shooting the schedulers lends itself to reducing the design time so that manufacturing doesn't have to continue to tear stuff down to catch up. The bottleneck in most place is not having the part program it's having the model.

The advance in Cad needs to be in a intuitive package, input your needed parameters for what you desire. It looks to a database and offers choices. You have a library of parts that you can use, internal or web based vendors. If there is nothing that fills that need then you take it from there. Sure there won't be one size fit all there either. But in regards to software itself, this would cost a whole lot less than buying the next gen of machine tools. A company that just dropped a big chunck of change on a machine is not looking to do it again any time soon.

Quite frankly, it is going to be the Chinese that will be driving the machine tool market, sales will dictate what changes happen and with a cheap labor force they will be using people not robots.

I'm not slamming designers or anything like that. I am just stating what I feel based on what I have seen in the past few years. We have to sharpen our pencils because it slipping away real fast.. the aftermarket Harley fad is going to die out soon and we need to hold on to aerospace real tight.... sorry for being so long-winded

 

[ctopher]

It think it was a Cicinnati 5 axis mill for cutting alum landing gear parts for aircraft. I think the NC control was Fanuc. I was a Automation Engineer at the time and my co-worker made it so a model was sent directly from UG to the controller. The operator waited for the "ready" light and pressed "go", then sat and watched.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP3.1 / PDMWorks 05
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[trainguy]

Hopefully away from AUTODESK INVENTOR 10, and towards Solidworks, or anything else.

My 2 cents, which constitutes a minute percentage of the money we've lost since our purchase of 5 Inventor stations.

tg

 

[SrGilberto]

5 to 10 years from now we'll be paperless (Right!).