Parametric PC Dmis Programs

[jon_dav_uk]

A quick hi to everyone as this is my first post on the new forum.

Has anybody out there ever tried to create a parametric program in PC Dmis. I dont just mean assigning variables to a few features but an actual full parametric programming. I will try and explain my problem below.

We manufacture turbine blades and have one type which is fairly consistent in its design. It varies in size (all of its dimensions) and may / may not have certain features like slots, grooves etc etc. Going back 5 years or so, a group of lads that work here created an absolutley huge parametric program in MM4 (not sure how many people remember this software). They created a set of formulas that mathematically defined every blade type using planes, lines and points and used scripts inside MM4 that ran these formulas. They then had a set of parameters which were taken from the blade drawing and stored these along with other information like contract, part number, type etc as a text file. They then came up with a set of predefined rules on how each blade type would be measured and used these as the basis of the program. To create a new program all we had to do was create a new text file based on the blade drawing, run this thorugh MM4 and bingo, a set of programs was created which measured the part at each stage of its manufacturing process.

The problem we have now is that MM4 is no longer really supported by Hexagon (Brown and Sharpe) plus we recently upgraded our controllers and now get really strange errors when running this program. We also no longer have the resource in MM4 to spend any length of time repairing / rewriting the code especially when it is software which is very outdated. The third factor is PC Dmis does a far better job when using things like best fit and we really want to take advantage of this. What I need to do is create this in PC Dmis. Before I started this mammoth task I thought I would stick a post on here and see if anyone else has actually done this. I have a few ideas which way I want to create it but if anyone knows of or has done this before any help would be appreciated.

PS - There are three of us that work together and between us we generate our own 3D CAD models in ProE from 2D drawings, manufacturing NC programs and also the PC Dmis / MM4 inspection programs. We have parametrically created a set of parts in ProE which we can change via an Excel spreadsheet. This way we can simply change a few variables in Excel and automatically regenerate in ProE, a full set of working models for each stage of manufacture complete with sensible coordinate systems that we use in PC Dmis. On these models we have also parametrically defined the points we would measure in PC Dmis but as yet, havent really thought how we can get them out and into a PC Dmis program. What we dont want to have to do is manually create a PC Dmis program for each component when we are sure this can be caried out parametrically.

All ideas are welcome, I really hope someone has already done something like this or similar as its gonna take me a long time to create this.

Thanks in advance
Jon

[george frick]

now i understand why the british motorcycle industry is extinct!

[tking]

Jon,

I've done a lot in MM4 (loved it), could do anything I was bright enough to conjure up in my mind. In PC DMIS land I'm still a novice looking for help myself.

I hope one of the folks is far enough down the road that some help is available.

T. King

[Don Ruggieri]

Not the answer to your question, but. . .

Have you considered the Blade package for PC-DMIS. This is essentially how it is written. The UI creates a data file, which is fed to the PC-DMIS program.

[jon_dav_uk]

Although I cannot program in MM4 speaking to my colleagues that can, they agree 100% with you TKing.

Never really considered the blade package Don but I will look into it, thanks.

Jon

PS - Not entirely sure what you are insinuating about the British motorcycle industry but I will keep my comments to myself.

[mtalla]

comments

Try to beat a '68 Triumph 650

[cmmguy]

Yup, i have. Full parametric, every move was relative to a variable. It wasn't any fun though. PCDMIS crashed more times that I want to say. It took 3 times longer to do than another cmm language that I use.

I think that it would have been easier to write a high level program that called all of the functions in PCDMIS instead of doing it it PCDMIS itself. But that is another story...

[jjewell]

Blade pkg

I have seen the blade package, just witnessed, not ran.

I have to go with Don on this one. If you had a high end programmer to write this stuff, that would be fine.

We don't all have those capabilities, and the Blade package looks pretty slick.

Have a application guy show and tell, never mind the sales guy.

[T Miller]

Jim is right you should be able to have someone come in and give you a demo
of the blade package.. They came here at no charge to show us the DCI package.

[george frick]

limey bikes

Try to beat a '68 Triumph 650

yeah- try starting it with lucas electrics. and what about them whitworth threads. been there, done that, got the t-shirt. owned a 65 triumph bonneville & a 66 bsa lightning rocket back in the day.

[Don Ruggieri]

Consider BASIC???

The Blade package is worth a look, and PC-DMIS will handle parametric programming, but there is also another option. CMMGUY eleuded to this, if you know VB you can write a great parametric program and have it feed commands to PC-DMIS. There are plenty of people out there who have experience in VB or BASIC scripting to help you out, or get you started if you've never done something liike this before.

What the Blade package has going for it though, is that not only does it create the measurement routines, it also has great reporting options.

[\v/inston]

The Blade package is worth a look, and PC-DMIS will handle parametric programming, but there is also another option. CMMGUY eleuded to this, if you know VB you can write a great parametric program and have it feed commands to PC-DMIS. There are plenty of people out there who have experience in VB or BASIC scripting to help you out, or get you started if you've never done something liike this before.

What the Blade package has going for it though, is that not only does it create the measurement routines, it also has great reporting options.

sounds like a good way to impress the customer...

[Don Ruggieri]

Another option is to write parametric programs in VB, or scripting. The nice thing about the Blade package though is that it already has the reports generator instead of trying to do that yourself also.

[jon_dav_uk]

Discussed the blade package with someone at one of our sister factories and it is not what I require. Looks like I am gonna have to write some kind of script / post processor to convert the ProE output to dmis.

Thanks for all the replies I will let you know how I get on

[James Doolittle]

I've kinda done it

I've got something running that sounds what you're doing. I've got a total of 3 programs checking ~200 similar parts with oodles of loops and conditionals, over 1,300 text files containing nothing but comma seperated variables, another ~10 text files with other stuff and a VB interface for machine operators to use on the shop floor.

Don't know how much I can help, but I do have a great nugget of code that's about 15-20 lines that collects the variables from the text files, that I can share with you if it'll help. The key is to use a two-dimensional array. Anyway, any specific questions, email me [email protected] and maybe I can help.