Using ASME Y14.41 and ASME Y14.5

[aardvarkdw]

We use Autodesk Inventor which is very limited in it's ability to adhear to ASME Y14.41 but we would like to use it as much as possible. We have some cast, industrial design parts that have complex curves and radii. This makes it extremly difficult to dimension per ASME Y14.5 (our usual standard). What I would like to do is dimension the majority of the features per Y14.5 and then provide a note linking the drawing to the model for definition of the complex features. Has anyone been working to Y14.41? Is it legal to reference both standards on a single drawing? As near as I can tell from reading the standard, I can dimension everything on the 2d drawing and query the model for anything not dimensioned per y14.41, but do you then call out Y14.5 as the standard for interpreting the dimensions on the 2d print?

Thanks for your help.

David

 

[ctopher]

You can call out whatever standard is needed and follow each as needed. Clarify in notes. As long as the dwg is clear and understandable.

Chris
SolidWorks 06 5.1/PDMWorks 06
AutoCAD 06
ctopher's home (updated 01-18-07)

 

[ewh]

While we may follow 14.41 to a certain degree, we only specify 14.5 on the drawing. We include a note stating that the model is the engineering master defintion and takes precedence, and any undimensioned features are to be derived from the model. We add additional (is that redundant?) notes explaining the tolerancing which applies to model derived dimensions.
As ctopher posts, the main thing is to make sure that the drawing is clear and unambiguous.

 

[rltw]

I've been using a Y14.41 note only. Since Y14.41 references Y14.5 I think it would be implied, no?

Most of my work has been models w/ drawings only for notes, etc.

 

[drawoh]

aardvarkdw,

How are you planning to inspect this thing?

I have not run into this problem yet, but probably, I am headed there. I am assuming I will show section views with grids imposed on them, and that I will use these to create inspection templates. This would make ASME Y14.5M work, and it would allow an inspector to work off of our 2D drawings.

I have not tried any of this yet.

JHG

 

[aardvarkdw]

Drawoh,

I'm not sure, we are still working on that. Right now we DON'T inspect them and that is half of the problem. On the complex parts we have done in the past, The drawing has shown had ortho views of the part with PEM fastener notes and a note saying "All dimensions and shapes defined by solid model XXX-XXXXXX-XXX." This has been a problem for us because we don't have a CMM and our QA people can't inspect from the model file. We have been getting parts that "seem" to meet spec but our problem is in controlling our files. We don't archive the native files and we have had really poor file control. Basicly anyone with Inventor can edit the models (and do). On top of that these parts are often created from a Rhino .igs file and have errors beyond belief. I am instituting drafting standards and practices to eliminate some of these problems but nothing is perfect. Where this particular problem comes from is that we sometimes have major catastrophies with these parts when changes need to be made (because of poor drafting practices) and we need to rebuild the part from scratch. Interpreting the part from a .step file is cumbersome. So, I want our designers to put as many design details on the print as they can. This way we can inspect the mating features and recreate the part if need be.

David

 

[drawoh]

aardvarkdw,

It sounds like you need PDM to control your files.

My biggest concern about ASME Y14.41 is that your inspectors need a lot of access to the 3D model. This all starts to work if you have a CMM. If you have a granite reference block, a height vernier and all sorts of fixtures, you need a process. The old fashioned 2D piece of paper can be redlined and highlighted as the inspector goes over everything.

JHG

 

[ewh]

drawoh is correct about one of the largest problems with implementing 14.41 - inspection. Without a CMM, it is very difficult to do. Even with one, there are some problems that need to be worked out before it is efficient (WCS location for example). The best, albeit most expensive route is to have a CMM that can read native CAD files. Some of the CAD programs have the capability to embed the GD&T which can then be read by the CMM, and that makes the inspectors life much easier. Parts go through much faster and problems are detected much earlier.

 

[aardvarkdw]

Which is why we haven't implemented it before now and don't intend to jump in with both feet right away when we do. I just know that all features must be dimensioned per ASME Y14.5M-1994, and that coordinate plotting is not a good option. So to apease our engineers and document control, using Y14.41 in addition to Y14.5 is my solution. It doesn't solve the inspection problems, but until we get software that can place the GDT on the model and a CMM to inspect with all I can do is fix what I can.

Thanks for the sanity check!

David

 

[drawoh]

aardvarkdw,

If you cannot inspect it, what is the point?

JHG

 

[aardvarkdw]

There isn't much of a point except that we need to document the design somehow, otherwise we are reciving parts that we can't inspect AND we don't know what they were even supposed to be in the first place. At least be doing this we will know what they should be. It's not a good solution but it is what we can get. The goal is that we can inspect the features that are dimensioned on the print and anything that is only dimensioned from the model is only a cosmetic feature anyway. Right now all the features are dimensioned from the model (sort of) and we can't inspect anything.

If the feature needs to mate with something else we will dimension it on the print, even if that means plotting coordinates. If we don't care, because it is a cast cosmetic feature, we will just do what we have done all along and "assume" the part is right.

David

 

[JohnNabors]

David-

"and we have had really poor file control. Basicly anyone with Inventor can edit the models (and do)"

Sounds like you need to implement ASME Y-14.35M-1997 to get your revision practices under control. Start with a gap analysis.

John Nabors

Never try to teach a pig to sing. It wastes your time and annoys the pig.

 

[KENAT]

Drawoh, they could always send the first one from each batch, or some other statistical selection, externally to be measured.

So even if in practice they're unlikely to inspect it having it defined leaves you that option, for instance if you discover problems in the field or something.

Aardvarkdw, given the limitations you are working to it sounds like what you are doing is reasonable.

It's interesting to see what other people are doing. I know we have a few things which are like this but I haven't got involved with them yet. I also appreciate what you say about control of the CAD file. We just discovered our sister site is sending out native CAD data, while in theory this can be a good idea in practice I'm not sure they've taken the required steps to make it work regarding control of the file.

 

[ewh]

David,

I agree with KENAT that your practices are reasonable given your restraints. Far from ideal, but workable.

Sending native data isn't always wrong, depending on the situation. Some of our largest customers sends us part files in the native format. This allows us to modify them as necessary to reflect what the final parts will be, after our manufacturing practices are accounted for. I dread having to work with unparameterized files when I have to do this type of modeling. Control isn't a problem, as long as the files are protected and renamed before we start to modify them. The customers have robust data control systems, so they can be confident that they will not be using a modified part that they are unaware of.
When it comes to us sending the resultant part back, we do send unparameterized files. More of a control issue on our part, since we are responsible for creating the actual parts based on THAT data, and we can't allow for any changes that they may make to the files that we may not be aware of.

 

[aardvarkdw]

John,

Our revision practices are not perfect but are adequate. What we have trouble controlling is the native files themselves. When a design change is being considered or implimented, it is not uncommon for our engineering department to work directly on the existing model. Then, because we don't archive our models (we archive a PDF, DXF, SAT, and STEP file) a change can be made to the model but not be released. Then if we need to make future changes the existing model may not match the parts we are recieving from the vendor.

Ewh,

You are in a situation different from ours. You are making the changes to others parts but as you said yourself, you don't send them the native files back because you can't be responsible for any changes they make. I don't think it is ever a good idea to send someone a native file UNLESS you intend for them to change it. We don't want to give our vendors the option of changing something to make it easier for them, or to match what they produced, and then try to tell us that the models we gave them are what they are building to.

David

 

[KENAT]

aardvarkdw

Your second point above is exactly the concern I have with what our sister site is doing.

Ken

 

[drawoh]

aardvarkdw,

I see quite a few excellent reasons to archive DXFs and PDFs, but you must archive your CAD models too. This is exactly what PDM software does. If someone wants to change something they must check the files out of the database. The activity is logged. You can establish what version of your design was used to generate the PDFs and DXFs. You can recover that version and start orderly design changes from there.

JHG

 

[aardvarkdw]

I agree, now if only you worked at my company, maybe one extra voice screeming for this would get it done....

Sorry about the sarcasm, but this is a sore subject. The software we use right now is not condusive to storing and retrieving the native model data. The solution we will most likely go with is Autodesk Vault because it comes with our CAD software but for reasons beyond my immediate control I cannot install it at this time. So we are trying to make the best of a really bad situation until we can correct it properly. In the mean time we can store the dumb solids and prints (everything that we would want to send to our vendors) in our revision control system but we have no real control over the native files except to limit access to the server, which we already do.

David

 

[ewh]

For added insurance, you should archive often.

 

[KENAT]

Intoducing PDM/PLM is not a minor task and certainly not a quick fix.

Also from what I've seen for a lot of the problems PDM/PLM is touted to fix it's actually the opposite, you have to have the problem fixed before you introduce the PDM/PLM, the PDM/PLM then 'keeps it fixed'.

Please note I'm not underestimating the value of a well implemented PDM/PLM, I'm strongly in favour, but at the same time I believe it's typically a lot of work.