Can this be machined

[BrianP.]

We were given this part to see if we could machine it. Customer is now going to have it molded. Could not figure it out. I do not do a ton of surfacing so thought maybe its just me. OD and ID are not a true circle. Neither top or bottom is a constant taper. Ever edge is radius-ed. Just curious for some input. Have no idea how I was going to machine it even if I got toolpaths figured out. Part is only about an inch in diameter a 1/4 inch tall and the walls are ony .025 thick.

[Guest]

Just call JP over at Cimquest,

 

oh wait nevermind

 

Of course it "can" be machined, the question is what do you have for equipment?

 

I wouldn't attempt 3 axis only

[RaiderX]

It can be done. The better question would be is it a cost effective solution to machine, mold, cast or form. A bit more information would be needed to determine that. QTY, Material, Stress factors, etc.

[Colin Gilchrist]

If you cut that part on a 5-Axis machine, no problem. Hold it with some material tabs, or make a fixture to hold it for final finishing.

[GoetzInd]

I wouldn't want to mess around with it without full 5. Of course, material stability would be a huge factor in determining if your accuracy would be anything close to what you want.

 

MIke

[Guest MTB Technical Services]

That's the kind of part I would no-quote without a 5-axis machine.

 

If they are going to have it molded then the material is obviously some kind of polymer.

Based on the graphic it looks like it's a developable surface meaning that it can be unrolled or flattened.

 

Without knowing what kind of tolerances are required, the most cost effective method is probably going to be blanking and vacum forming.

[Jayson Kramer]

What is made of and how many you have to make?

PS what is the Tol +- .005

[Rotary Ninja]

My boss brings out crazy jobs and asks if we can make it (only 4 axis here). I always tell him we can make anything. So it really depends on things like everyone said - material, quantity, and tolerances are key factors into the quote. Can it be machined? Of course it can.

[BrianP.]

Just call JP over at Cimquest,

 

oh wait nevermind

 

Was one of my first thoughts but he decided to go out and work for a living . Hope all is well with you JP. We only have Haas machines with some 5C indexers. Material is medical grade PET. It is to grow a trachea . Tolerance is make it look like the picture. "Material stability would be not very. 100 parts at this point. Like I said this is more of a curiosity question for me at this point.

[Jayson Kramer]

yes you can make those from a large sheet on a Haas 3axis.

[SledGeek]

^^^^^^^^

 

yep....if data is what it looks like in the pic

[Rotary Ninja]

I want the part model. I will program it complete. 3 axis. I have cut that stuff. I think it would be a fun part to program.

[crazy^millman]

Slam dunk 100 pcs I would think about Vacuum on a part like that. Tabs would be my last choice on that type of Material. Could also employ the bondo method with Window frame, but for 100 parts I think you will find Vacuum is your best friend.

 

HTH

[Rotary Ninja]

I think he was wondering how it could be machined with a 3 axis. Here's a start...

 

 

That is just a 2D Dynamic Core mill toolpath to rough out the area around the part. A Flowline up the outside profile. Then a Surface Finish Contour to surface the part using a 1/2" x 1/32" full rad slot mill.

 

I created a solid the shape of the bottom of the part extruded down 1/8". Then I translated that lower solid down like .03" creating a .030" gap between the two solids then used them both as drive surfaces. That lets the tool try to get in between the two parts but not completely separate them.

 

You could drop that solid down a little more and create some tabs between the two if you wanted to have less blending to do when the part was done.

 

Cycle time - About 9 minutes with a .001" step down on the surfacing path. About 5 minutes with a .002" step down.

 

Probably not the best way to produce 100 of these, but hopefully that gave you some ideas?

[GoetzInd]

Nice Kevin! I wonder how the thing would react when it got cut loose?

 

Mike

[Jayson Kramer]

Mike like a boomerang. LOL

 

Crazy Ron, the bondo thought is what I was thinking. working for the fast prototyping company's I guess we all do it the same. lol

[SledGeek]

Vacuum first, but if not....

I would back fill with a version of 3M ScotchWeld. 2 part plastic epoxy that mixes as you apply.10 minutes and your good to go. Machine final pass(es) and it will hold it in place. It's more brittle and stiff than the PET so you flex the part a bit and it cracks and breaks right off easily without damaging or scuffing part surface. If any sticks on it scrapes right off with a razor. Great for break off tabs instead of machine off tabs. I have used this alot on plastics with great success on prototypes from 1" to 60"

Do it in a sheet with a version of Ninjas program and then use 3m to back fill outer rad or tabs . Go back and cut them off all at once.......course I would make 105 cuz sh!t happens

Obviously , depending on customers specs, you may have to epoxy, flip sheet and machine bottom rad, then break off. Customer sounds like pretty wide open specs....

Of course your nads will suck up into your body for the first 10 parts til you see it works at least mine did first time I used it......

[K2csq7]

Just call JP over at Cimquest,

 

Hi JP!

I ROR'ed!

 

I think he was wondering how it could be machined with a 3 axis. Here's a start...

 

DAYAM Kevin. Not to shabby!

[SledGeek]

Befor I discovered this stuff......it was bondo

[Rotary Ninja]

Nice Kevin! I wonder how the thing would react when it got cut loose?

 

Mike

 

Hi JP!

I ROR'ed!

 

 

DAYAM Kevin. Not to shabby!

 

Thanks guys.

[BrianP.]

Looks good. I think maybe part of the problem may be the design. The outside is actually not a curve but about 10 different diameters. The top and bottom are also not a straight taper If you look closely at the picture you will see it actually flares back up towards the end. I could maybe put it up on ftp as a step file if I knew how. At this point like I said it is all curiosity. Also a learning experience.

[GnoGueyJose]

We did a similar part on a Robodrill 3 axis with a full radius wheel cutter. We used a flow line toolpath with gouge checking turned off so we could control surface finish. Ours was peek and we used breakaway tabs and the burr bench cleaned it up in about 3 min.

 

That job will return in high quantities so we're looking at a Nakamura NTMXto do it out of bar complete with no burr bench work. We've got a little NRE money to develop a faster process. Decisions, decisions, decisions.

[Colin Gilchrist]

Looks good. I think maybe part of the problem may be the design. The outside is actually not a curve but about 10 different diameters. The top and bottom are also not a straight taper If you look closely at the picture you will see it actually flares back up towards the end. I could maybe put it up on ftp as a step file if I knew how. At this point like I said it is all curiosity. Also a learning experience.

 

Depending on the toolpath you use, the shape doesn't really matter.

 

Surface Finish Contour, for example, will just slice the solid at different Z depth levels, then offset the cutter path based on the contact point of the tool. With a path like that, the motion is typically linearized entirely, and then your Filter settings can be used to output line/arc code...

[BrianP.]

Depending on the toolpath you use, the shape doesn't really matter.

 

That's the problem. I am not all that familiar with surfacing toolpaths. I mean some surfacing is simple and straight forward others not so much. I have never had the opportunity to do advanced mill classes ( or any for that matter) or work with some one who is REALLY good with Mastercam. You would be surprised at the number of "programmers" who do not even understand the concept of using the WCS. I am weak in multi-axis and surfacing.

[Rotary Ninja]

Looks good. I think maybe part of the problem may be the design. The outside is actually not a curve but about 10 different diameters. The top and bottom are also not a straight taper If you look closely at the picture you will see it actually flares back up towards the end. I could maybe put it up on ftp as a step file if I knew how. At this point like I said it is all curiosity. Also a learning experience.

 

As Colin said the shape doesn't matter. I noticed the differences in your part. But rather than model up a closer part (based off of a picture) I just made a shape that would give you an example that would work.