ImnotfromMars
Mechanical
Is anyone using 3D model geometry in lieu of drawings and does it work better than drawings? I believe the standard is ASME Y14.41-2003.
Robert
Robert
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Using 3D Model Geometry 9
- Thread starter ImnotfromMars
- Start date
- Status
-
2
- #2
Large parts of the automotive and aircraft industry work virtually solely with 3D models and have done for many years.
The reasons are very obvious; it is virtually impossible to define complex 3D shapes as 2D drawings. Other industries seem slower to catch on or maybe with less complex components there is no need. You do not get close to being a tier 1 supplier in automotive without the ability to work and measure to solid models, at least in BIW.
As with any “new” technology you either buy into it or fight it, as there are problems to solve. In a few years time IMO 2D drawings in many industries will be as rare as a draughtsman drawing on a piece of paper and running off prints.
The reasons are very obvious; it is virtually impossible to define complex 3D shapes as 2D drawings. Other industries seem slower to catch on or maybe with less complex components there is no need. You do not get close to being a tier 1 supplier in automotive without the ability to work and measure to solid models, at least in BIW.
As with any “new” technology you either buy into it or fight it, as there are problems to solve. In a few years time IMO 2D drawings in many industries will be as rare as a draughtsman drawing on a piece of paper and running off prints.
-
1
- #3
ewh said:
Does this imply that that some sort of CMM is used during inspection since not all the data is on the drawing? This has been the 2nd greatest argument for following this new standard where I work.
"Art without engineering is dreaming; Engineering without art is calculating."
Have you read faq731-376 to make the best use of Eng-Tips Forums?
-
1
- #6
We have a machine shop that uses minimally dimensioned drawings and our solid models. So we have all our standard design notes on the drawing and the rest of the design is captured in the model. They use the model to inspect the parts.
Heckler
Sr. Mechanical Engineer
SWx 2007 SP 3.0 & Pro/E 2001
XP Pro SP2.0 P4 3.6 GHz, 1GB RAM
NVIDIA Quadro FX 1400
o
_`\(,_
(_)/ (_)
(In reference to David Beckham) "He can't kick with his left foot, he can't tackle, he can't head the ball and he doesn't score many goals. Apart from that, he's all right." -- George Best
Heckler
Sr. Mechanical Engineer
SWx 2007 SP 3.0 & Pro/E 2001
XP Pro SP2.0 P4 3.6 GHz, 1GB RAM
NVIDIA Quadro FX 1400
o
_`\(,_
(_)/ (_)
(In reference to David Beckham) "He can't kick with his left foot, he can't tackle, he can't head the ball and he doesn't score many goals. Apart from that, he's all right." -- George Best
-
1
- #7
Matt
CAD Engineer/ECN Analyst
Silicon Valley, CA
-
1
- #8
I started making some comments on the thread fcsuper linked but thought I’d put more detail here.
How do people deal with things like:
1. Ensuring the models are robust/aren't likely to fall apart. A number of times I've had to re-create models (others not my own) based on 2D drawings because they 'fell apart' while undergoing a minor change. I’m tempted to think this may be easier in many cases than working to either a ‘dumb solid’ or even a copy of the original native one.
2. Make sure models are truly representative. I often find models that have minor variations from what they're meant to be. Typically angles that are off slightly or 'fudges' where rounds meet etc.
3. Configuration control, making sure the vendor gets the correct revision & version of model and it doesn't get inadvertently (or intentionally) changed. For instance an unscrupulous vendor receiving native data might change the model slightly to match what they make. I can think of a few ways to guard against this but I'd be interested in knowing what people are doing, especially those without PDM/PLM or large configuration control departments.
4. Inspection, at least of non complex areas with classical methods. For instance we don't have a CMM, we rely on reports from vendors and we cross check a few dimensions on occasion to verify. We could obviously go out to a third party measuring house but this could be costly and time-consuming.
5. Tolerancing if the vendor isn't able to take an annotated model, only dumb solid. For lower tech/smaller scale organization it may not be easy to find vendors that can take native or other annotated models.
6. Design check. This is partly related to 1 above.
7. Design sign off, if again especially those without PDM/PLM or large configuration control departments.
8. Any other considerations I'm missing.
Obviously not all these aspects are unique to MBD as opposed to 2D CAD or even 2D based on models but I’m thinking they become more of an issue.
I ask not just for arguments sake but because we're going in the direction of using the model as primary definition for some castings and moldings and I want to get a handle on the issues involved before we introduce a policy/standard. We're looking at possibly sending native data or maybe STEP/IGES/stl depending on the vendor.
I know 14.41 probably covers at least some of this but from what I've seen not all. Before we go too far I want to get a copy but not sure it will happen any time soon.
How do people deal with things like:
1. Ensuring the models are robust/aren't likely to fall apart. A number of times I've had to re-create models (others not my own) based on 2D drawings because they 'fell apart' while undergoing a minor change. I’m tempted to think this may be easier in many cases than working to either a ‘dumb solid’ or even a copy of the original native one.
2. Make sure models are truly representative. I often find models that have minor variations from what they're meant to be. Typically angles that are off slightly or 'fudges' where rounds meet etc.
3. Configuration control, making sure the vendor gets the correct revision & version of model and it doesn't get inadvertently (or intentionally) changed. For instance an unscrupulous vendor receiving native data might change the model slightly to match what they make. I can think of a few ways to guard against this but I'd be interested in knowing what people are doing, especially those without PDM/PLM or large configuration control departments.
4. Inspection, at least of non complex areas with classical methods. For instance we don't have a CMM, we rely on reports from vendors and we cross check a few dimensions on occasion to verify. We could obviously go out to a third party measuring house but this could be costly and time-consuming.
5. Tolerancing if the vendor isn't able to take an annotated model, only dumb solid. For lower tech/smaller scale organization it may not be easy to find vendors that can take native or other annotated models.
6. Design check. This is partly related to 1 above.
7. Design sign off, if again especially those without PDM/PLM or large configuration control departments.
8. Any other considerations I'm missing.
Obviously not all these aspects are unique to MBD as opposed to 2D CAD or even 2D based on models but I’m thinking they become more of an issue.
I ask not just for arguments sake but because we're going in the direction of using the model as primary definition for some castings and moldings and I want to get a handle on the issues involved before we introduce a policy/standard. We're looking at possibly sending native data or maybe STEP/IGES/stl depending on the vendor.
I know 14.41 probably covers at least some of this but from what I've seen not all. Before we go too far I want to get a copy but not sure it will happen any time soon.
KENAT I can only speak about the automotive industry and even then a certain sector but here goes.
1. All the models we receive come from the end user. Occasionally you do get a “bad” model as with anything you only get out what you put in.
2. Again down to the quality of the modelling, this is no different in 2D or as a solid model.
3. The way most automotive company’s work is at the end of the model (part) number is an engineering level. You receive a sheet stating the level and the date you receive it, in turn you sign and or stamp it and return by fax. Models do get changed for example to create a condition where spring back will occur, but the original is always the master.
4. If you do not use a CMM you will need to create another way of measuring the part, how you go about this will depend entirely on the complexity of the part and the limits you are trying to hold.
5. Again an issue that needs addressing, some companies do supply basic tolerance other work solely to the model. Are you or your company in a position that when you say jump the supplier asks how high? If you put enough money the way of a company they will work the way you want, if not you may have to work the way they want. Does the dog wag the tail or the tail wag the dog?
6. This has usually gone through FEA.
7. Again you will need to have your own standards.
It is not always easy and you will encounter problems, getting suppliers that buy into what you are aiming for is a key element, however I am sure that as things are you do not just use anyone, I would guess you audit any potential supplier.
If you get it right and it will evolve you will save many hours especially in areas where the 3D side is key, as you say castings and mouldings.
Good luck.
1. All the models we receive come from the end user. Occasionally you do get a “bad” model as with anything you only get out what you put in.
2. Again down to the quality of the modelling, this is no different in 2D or as a solid model.
3. The way most automotive company’s work is at the end of the model (part) number is an engineering level. You receive a sheet stating the level and the date you receive it, in turn you sign and or stamp it and return by fax. Models do get changed for example to create a condition where spring back will occur, but the original is always the master.
4. If you do not use a CMM you will need to create another way of measuring the part, how you go about this will depend entirely on the complexity of the part and the limits you are trying to hold.
5. Again an issue that needs addressing, some companies do supply basic tolerance other work solely to the model. Are you or your company in a position that when you say jump the supplier asks how high? If you put enough money the way of a company they will work the way you want, if not you may have to work the way they want. Does the dog wag the tail or the tail wag the dog?
6. This has usually gone through FEA.
7. Again you will need to have your own standards.
It is not always easy and you will encounter problems, getting suppliers that buy into what you are aiming for is a key element, however I am sure that as things are you do not just use anyone, I would guess you audit any potential supplier.
If you get it right and it will evolve you will save many hours especially in areas where the 3D side is key, as you say castings and mouldings.
Good luck.
1) We check a number of items including the methodology of feature creation. For the most part, our models are based on those of the customer, and some things are beyond our responsibility to correct. The best solutions we have come up with are training and checking.
2) See item 1
3) We receive native models from the customer whenever possible. We only send out unparamaterized models. We are not responsible for any changes to any models once they are sent.
4) See ajack1's post
5) If the vendor can not use an annotated model, we provide screen shots of any pertinent tolerancing in the form of cgms or jpegs.
6) As the critical features of the vast majority of our parts are already defined by the customer, this is only an issue in regards to manufacturability. We do check that the part and related tools will work as intended.
7) This is an area we are still improving on. Currently, all signed off parts are routed through one individual who verifies and records the signatures and moves the files into a protected release directory. We are in the process of investing in a PLM package that will make the proceedure less vulnerable to human error.
8) While the model is nearly complete or undergoing check, CAM analyses it for programability. We will adjust feature creation to simplify the CAM process before the file is released. Poor model construction is often caught at this stage.
2) See item 1
3) We receive native models from the customer whenever possible. We only send out unparamaterized models. We are not responsible for any changes to any models once they are sent.
4) See ajack1's post
5) If the vendor can not use an annotated model, we provide screen shots of any pertinent tolerancing in the form of cgms or jpegs.
6) As the critical features of the vast majority of our parts are already defined by the customer, this is only an issue in regards to manufacturability. We do check that the part and related tools will work as intended.
7) This is an area we are still improving on. Currently, all signed off parts are routed through one individual who verifies and records the signatures and moves the files into a protected release directory. We are in the process of investing in a PLM package that will make the proceedure less vulnerable to human error.
8) While the model is nearly complete or undergoing check, CAM analyses it for programability. We will adjust feature creation to simplify the CAM process before the file is released. Poor model construction is often caught at this stage.
-
1
- #13
I think you will see it varies based on size/type of company and their $$.
A lot of companies still rely on dwgs.
Chris
SolidWorks 07 3.0/PDMWorks 07
AutoCAD 06
ctopher's home (updated 03-26-07)
A lot of companies still rely on dwgs.
Chris
SolidWorks 07 3.0/PDMWorks 07
AutoCAD 06
ctopher's home (updated 03-26-07)
-
1
- #14
ImnotfromMars,
We still you drawings where I work. Most of our stuff is machined in small quantities. Features mostly are orthogonal, and are easily described on a 2D drawing. 2D drawings clearly display tolerances.
2D drawings can be marked up as the drawings and incoming parts are checked. PDM software has all sorts of red-lining utilities, but consider the importance of being able to highlight all the drawing features you have inspected and accepted.
How are you going to inspect your stuff? As noted above, a CMM can work from your 3D model, as long as you can explain your accept/reject criteria to the inspector. Most people understand drawings. Even complex, 3D features can be inspected with 2D templates.
I have just had some rapid prototypes made, for the very first time. Eventually, they will become castings. Meanwhile, I am starting to learn how to deal with vendors who do not look at SolidWorks drawings. They are interested in the model.
If we designed consumer products that were to be mass produced in injection moulding machines, I would have to think very differently.
JHG
We still you drawings where I work. Most of our stuff is machined in small quantities. Features mostly are orthogonal, and are easily described on a 2D drawing. 2D drawings clearly display tolerances.
2D drawings can be marked up as the drawings and incoming parts are checked. PDM software has all sorts of red-lining utilities, but consider the importance of being able to highlight all the drawing features you have inspected and accepted.
How are you going to inspect your stuff? As noted above, a CMM can work from your 3D model, as long as you can explain your accept/reject criteria to the inspector. Most people understand drawings. Even complex, 3D features can be inspected with 2D templates.
I have just had some rapid prototypes made, for the very first time. Eventually, they will become castings. Meanwhile, I am starting to learn how to deal with vendors who do not look at SolidWorks drawings. They are interested in the model.
If we designed consumer products that were to be mass produced in injection moulding machines, I would have to think very differently.
JHG
-
1
- #15
434343434342
Aerospace
Im not sure of the other CAD Systems, but Unigraphics has the capability to define tolerances within the model itself otherwise known as "Model Based Tolerancing". The idea is that the tolerance definition can be defined in the model and "flowed down" to downstream applications such as CMM programs.
UGS's newly acquired Tecnomatix has such a CMM inspection programming application called eMProbeCAD. The MBT's are read by eMProbeCAD, and the programmer simply creates the CMM probe paths to provide the inspection data to meet the tolerance requirements. The sofware that runs the inspection program will analyze the tolerances appropriately and relative to the MBT as it applies to the actual CAD Master Model.
eMProbeCAD also has GD&T checking capabilities that allows the programmer to make sure that the GD&T meets the standard. If there is a descrepency in the tolerance it is noted and information is given so the programmer can correct the issue.
When MBT is not available the software also gives opportunity for the programmer to create the pseudo "MBT" right on the master model as the software is embedded withing the CAD system. This gives opportunity for the programmer validate the model against any outside drawings in applications where they "Drawing" is master and not the CAD model at the same time they are programming. This is of course essential since the probe paths are defined using the actual CAD Geometry.
Once the measurements are taken by the CMM or other inspection device, the measured data can be brought back into the CAD model and eMProbeCAD where the data can be further analyzed and viewed atop the master model. A 3D representation of the tolerance zone can be displayed along with the measured data on the CAD Model. In this case for complex surfaces the information is easily understood since the information is displayed in 3D for the user to evaluate. A picture is worth a thousand words.. probably 2000 of mine it also beats reviewing reams of xyz point data and trying to map out deviations and trouble spots especially when using the inspection to validate and/or improve manufacturing processes.
The software is not cheap, however when used correctly it can quickly pay for itself in the right application.
UGS's newly acquired Tecnomatix has such a CMM inspection programming application called eMProbeCAD. The MBT's are read by eMProbeCAD, and the programmer simply creates the CMM probe paths to provide the inspection data to meet the tolerance requirements. The sofware that runs the inspection program will analyze the tolerances appropriately and relative to the MBT as it applies to the actual CAD Master Model.
eMProbeCAD also has GD&T checking capabilities that allows the programmer to make sure that the GD&T meets the standard. If there is a descrepency in the tolerance it is noted and information is given so the programmer can correct the issue.
When MBT is not available the software also gives opportunity for the programmer to create the pseudo "MBT" right on the master model as the software is embedded withing the CAD system. This gives opportunity for the programmer validate the model against any outside drawings in applications where they "Drawing" is master and not the CAD model at the same time they are programming. This is of course essential since the probe paths are defined using the actual CAD Geometry.
Once the measurements are taken by the CMM or other inspection device, the measured data can be brought back into the CAD model and eMProbeCAD where the data can be further analyzed and viewed atop the master model. A 3D representation of the tolerance zone can be displayed along with the measured data on the CAD Model. In this case for complex surfaces the information is easily understood since the information is displayed in 3D for the user to evaluate. A picture is worth a thousand words.. probably 2000 of mine it also beats reviewing reams of xyz point data and trying to map out deviations and trouble spots especially when using the inspection to validate and/or improve manufacturing processes.
The software is not cheap, however when used correctly it can quickly pay for itself in the right application.
I learnded this today: ![[wink]](/data/assets/smilies/wink.gif "[wink] [wink]")
UG is used for engineering and still creates drawings for the C-17.
CATIA is used for the 787 and 747-8 and do not create drawings. This is also the way all new aircraft will be done.
Chris
SolidWorks 07 3.0/PDMWorks 07
AutoCAD 06
ctopher's home (updated 03-26-07)
UG is used for engineering and still creates drawings for the C-17.
CATIA is used for the 787 and 747-8 and do not create drawings. This is also the way all new aircraft will be done.
Chris
SolidWorks 07 3.0/PDMWorks 07
AutoCAD 06
ctopher's home (updated 03-26-07)
WE actually started using Model Base Definition on the A380
and now on the A350.
I have worked on some of the 787 stuff and the FT&A gets complicated as we have detail level tolerance and assembly level for final, all in 3D.
I was hoping for a way to not see the detail level until you highlighted the part..and not use no/show..
Cheers
I don't know anything but the people that do.
and now on the A350.
I have worked on some of the 787 stuff and the FT&A gets complicated as we have detail level tolerance and assembly level for final, all in 3D.
I was hoping for a way to not see the detail level until you highlighted the part..and not use no/show..
Cheers
I don't know anything but the people that do.
- Status
Similar threads
- Locked
- Question
- Locked
- Question
- Locked
- Question