Designing Fixtures

[crazy^millman]

Well this seems to be a lost art today. I see all these promotional things about 3D print this and 3D print that and save money. If you already own a 3D Printer then yes I can say you could do it cheaper on some things, but how much is a good 3D print to do what is needed in a job shop? $100k $250K or more I was venture a guess. Sorry for that kind of money I can make what I need a lot cheaper and so can most shops. That said the ability to design and make fixtures is not what is use to be. I thought I would start this topic to see what tips and tricks we can share to help the people who need to do this and want to learn.

I will start with an example I am currently working on. Customer needs to make a pan for a weldment of 316 SS. Part will end up look like a pan with some angles and a weld lip. I am machining the part in 2 operations, but the 2nd operation without supporting the part from the inside being so thin they are going have issues. I need to make a fixture that will support the inside of the part while machining the outside of the part. I make a copy of the original part and go about using Model prep to remove all the holes and get a nice clean cavity of the part to work with. I don't want a net fit shape on all the radius areas and angle areas, but I would like one where the flat weld lip going to be.  I am thinking .004 to .005 smaller would be good. What I do is use an old mold makers method to allow for shrink. I make the male fixture by create a rectangle block I will Boolean remove the part from I just made without holes using Model Prep. I then do Boolean remove and have the perfect inside of the part now sticking up as male fixture. I need to scale it down so I will use Xform scale and I used .9995 on the part and I got a good scale reduction. I open back up my original part and use it to create a curve on edge of the weld lip I want to be a tighter fit and then extrude that shape to know make a good lip to locate the part to. One benefit is making fixtures like this is they also act as check fixtures for your part. Now I have a fixture that is smaller in areas I want, but to size in other areas. I will bolt it to the table and the customer is off and running without having to worry about how thin the part will be and how to support the part. i did some model prep to make the over all fixture base was the same size as the original since I scaled the whole thing. I could have just done a male part and Boolean added it the fixtures base or a couple other things, but this was the way I went about it. 

I have a link to the fixture before and after using this process in Mastercam 2018. The part is not restricted, but since I have not asked the customer for permission to share the part it is not shared.

2018 Fixture Example

[Gary]

Do this alot with solidworks and also mastercam. The indent tool in Solidworks I use a lot. Do this for modifying  plastic molded parts all the time. The fun part is when the parts seem to very from shot to shot or are not quite the same as what the model is. If a software would work like say pushing a part into a piece of clay instead of having to remove piles of undercuts/ ribs ect. would be the so frikken handy. 

[nickbe10]

On ‎12‎/‎29‎/‎2017 at 12:20 PM, C^Millman said:

Well this seems to be a lost art today.

I can count the real tool designers I have worked with in my 25 + yrs on one hand.....two fingers actually.......even rarer than good machinists.

 

[Newbeeee™]

2 hours ago, nickbe10 said:

I can count the real tool designers I have worked with in my 25 + yrs on one hand.....two fingers actually.......even rarer than good machinists.

 

Same here. My mentor was ex Weatherly Cincinnati/Rolls Royce/dehaviland/Airbus - he was ex chief drafty for a subby company that did tool design for these companies.

He designed the fixtures for broaching the fur tree roots for the blades, worked on the first vertical wing assy jig for Broughton. And now at 80 this year, still designs and makes steam engines on his Myford mill and Lathe. Truly first class and they broke the mold with him. For (instrument) mechanical design, he was also first class, and I've been fortunate to work alongside 2 other guys since (as a subby in a prod engineering role) where they were nearly up there with him.

 And don't get me started on machinists - to this day I still can't fathom the mentality of 'i'm a miller' and 'i'm a turner'...

[crazy^millman]

17 minutes ago, newbeeee said:

 And don't get me started on machinists - to this day I still can't fathom the mentality of 'i'm a miller' and 'i'm a turner'...

Yes amazing how many only every do one discipline and consider themselves 1st class journeymen machinist. Not saying they are not good at just turning or just milling, but a machinist should know how to do each. Not just Standard lathes, but VTL. Now just Milling Vertically, but HBM, then we go into 5 Axis land how many have done Head Vertical and Horizontal? Trunnions in a VMC and HMC? Nutating Heads? Go to Lathes and how many have done 2, 3 or 4 Turrets? Sub Spindle? B Axis?

[Newbeeee™]

^^^^, errr, yes, no, yes, yes, and errr no, no , no and a lot more no's...

[huskermcdoogle]

3 hours ago, C^Millman said:

but a machinist should know how to do each. Not just Standard lathes, but VTL. Now just Milling Vertically, but HBM, then we go into 5 Axis land how many have done Head Vertical and Horizontal? Trunnions in a VMC and HMC? Nutating Heads? Go to Lathes and how many have done 2, 3 or 4 Turrets? Sub Spindle? B Axis?

Knowing how to do and having experience in each is a big distinction.  I know how to do just about everything on that list, but lack experience with horizontal 5ax, and 3/4 turret lathes.  Chomping at the bit here to tackle those varieties.  Turning in general though I am very rusty in, as its been almost a decade since I had to do anything more than maintain/improve an already established turning process.  It's becoming a vague but fond memory at this point.  Everywhere I have been, it seems those that didn't want to push the boundaries would hoard all of the turning work and avoid getting assigned a milling project.  Seems they preferred volumes of "busy" work over slaving away designing and building fixtures, programming the process, then debugging it through QC.  I am not trying to discount those that prefer to stay in "turning" land, but all in all, I feel that many of those guys, as the lines get blurred would have benefited more from taking on some milling jobs to better understand modern milling tools and HSM approaches which can now easily be applied into mill/turn parts.  Too many times I have seem mill/turn processes that were weak in one department, and stellar in the other, only to be a slightly below or on par process which struggles to meet profit targets, further limiting future investment into improving profit margins through people or technology improvements.

 

Anyway, as to not high jack this thread.  Always remember that the KISS principle is usually the cheapest one to follow.  When designing a fixture for something you aren't too sure about it's success on, invest into it by creating a good foundation.  By doing this you can always add to the concept to solve rigidity or locations issues as you come across them.  Also if you aren't sure where to put a support, wood blocks, plastic door shims, and hot glue work wonders for experimenting 

[Leon82]

Its hard to find help so if you are hired for one area you are going to stay there unless you have enough skilled workers to be able to move around and not hurt production and have a boss who is ok with you doing that.

I did modify a lathe program once

[neurosis]

11 hours ago, huskermcdoogle said:

Turning in general though I am very rusty in, as its been almost a decade since I had to do anything more than maintain/improve an already established turning process.

Same here. 

Its funny because I started my career running and programming lathes and had only set up and run mills.  Now I rarely touch lathes other than a manual occasionally and only program, set up, and run, Mills.

[BenK]

19 hours ago, newbeeee said:

 And don't get me started on machinists - to this day I still can't fathom the mentality of 'i'm a miller' and 'i'm a turner'...

I agree, its crazy that so many machinist's believe that they can only be one or the other. I have run across this my entire career. I by far have more experience on a mill but that doesn't mean I don't keep up on turning. In fact my last project that required me to run the machine was on a live tool lathe.

 

As for Ron's list, I haven't worked on several of those kinematic configurations but that doesn't mean I wouldn't/couldn't. Much of the knowledge gained will apply moving from one 5-axis to another or from one mill/turn to another.  

 

Fixture design is probably the my least favorite part of what I do. I just don't enjoy it most times. I have been using a parametric software lately and that definitely helps. A few of the tools I have available makes it much quicker to get things done. 

[rchipper]

BenK,

Can you share what parametric software is and how it works?

[Matthew Hajicek - Singularity]

 

 

[Newbeeee™]

You do have to have your wits about you when using parametric. Years ago working with a customer on a job with only key dimensioned drawings, they changed the part in issue moving an oring seal groove to a flange/sidewall. The CAD lad didn't realise it, but he also had the flange tapped holes constrained to the seal groove too, so these moved as well.

We made the mods, verified  (CMM) direct to CAD model but the cover then didn't fit because the tapped holes had moved. We were correct, but parts were scrapped.

[Matthew Hajicek - Singularity]

I've seen similar things.  Extend a plate and the holes that were dimensioned from the edge move too and no longer line up with the mating part.  I caught it before making the part since I also had the model of the mating part.

[huskermcdoogle]

54 minutes ago, newbeeee said:

You do have to have your wits about you when using parametric.

If treat it like an unproven complicated 5 axis gcode program you are uploading to an expensive machine without verification, your results may vary.

Point is, parametric modeling when making changes requires verification of a sorts, at the point of a change it's nothing more than a hand edited algorithm within a piece of software, that until it has been debugged, for all we know will be about as trustworthy as the first release of Mastercam X....

I have all of my fixture stuff in solidworks, and find parametric configurations to be a little bit of work to get setup, but also extremely useful when generating setup drawings as I can reuse old stuff to generate new with very little effort.

[BenK]

5 hours ago, rchipper said:

BenK,

Can you share what parametric software is and how it works?

Most of the CAD software on the market today is parametric. Topsolid, Solidworks, Creo, and inventor to name a few. They use a history of everything you do allowing you to easily go back and make changes. I't really handy when you are dealing with an assembly. 

[specv]

One of my favorite parts of this job is designing fixtures, especially moving ones, I recently designed a self-centering locating fixture for a casting with 18 datums on it. Without a good CAD software it would have been a nightmare. 

[rchipper]

Thanks for the reply BenK. 

 

rchipper

[WestRiver]

One of my sayings around the shop where I work is You live by the vice so shall you die by the vice. A quality designed and built fixture can be worth it's weight in gold literally.

[Newbeeee™]

9 hours ago, WestRiver said:

One of my sayings around the shop where I work is You live by the vice so shall you die by the vice. A quality designed and built fixture can be worth it's weight in gold literally.

Vices are okay for first op if you're holding on stock.

I HATE vices for holding machined faces and finishing a part. Yes we all do it, but swarf dents are easy and polishing of marks out become your friend...