Surface vs Solids

[Andris Skulte]

Hey guys,

In short - what's the advantage of solids? It seems to me that all the complicated machining that you mold guys are great at have to do with surfaces, rather than solids that have 5x larger file sizes. In any case, I'm not sure what next to try on this project (below)...

 

I just received an IGES file of an alternator fan, with a fairly (to me) complex blade shape and profile (3 axis). I'm just getting into surfaces and solids, and this has me stumped. I've attached a picture of the fan (to avoid proprietary issues). From the IGES import, I either tried using the surfaces directly, but couldn't get the tools to go the way I wanted. Most toolpaths had lots of Z axis moves, while I'd rather have the tool contour the z-depth incrementally from the bottom. Surface Finish Radial climbs over the top of the blades, instead of following the blade contour. I thought flowline might be the answer, but it bombs on the surfaces (partial flow edge error detected). At this point, I made a solid from the IGES surfaces, and then a surface from the solid, thinking it would be good, since the solid was good.. Thoughts?

 

 

Andris

[crazy^millman]

Not to sound like an arse but that is not that bad of a part. I think when you get through this part you will have some good ideas how to apporach things like this in the future. Me I would surface roung pocekt with a cicrle as my boundary. I might surface finish contour or flowling to finish the part from there. If you did it right a few operations and done.

 

My thought process is that you will have this blanked out a CNC lathe. Use a center stub to locate and help hold it down with maybe some finger clamps or a ring for the outside.

 

HTH

[Lee.]

+1 to Ron.

 

Surface Rough Pocket and then Surface finish contour.

 

Andris, I'm not really sure what your question is.

 

Solids are great for most design work, but sometimes you need surfaces when solids will not do what you need.

 

My 2 cents.

[chip]

+1 Ron. I find myself trying to also do as much 2d as possible with 2d machining. (based just on the looks) I would run it (like Ron said) surface rough to get the 'matl out of the way. then maby finish the floors (between the blades) with a pocket. then finish the draft on the blades. If you have the machine time run a finish scallop. It will come out bitchen. Solids to Surfaces. Building (for me) nothing beats solids. toolpaths I find myself using surfacess. I can trim an create surfaces faster than when I toolpath solids. Again all preferance. Some of the people in here run solids the majority of the time.

[chip]

Andris. I curious about your hot rod. I can't get any pictures to show on the site. http://www.skulte.com/turbo.html

[Chris Rizzo]

Everybody else is on it, but some basic tips that help when selecting surface toolpaths.

 

Surface pocket and resmill are constant-z cutting. all XY moves with z as the stepdown.

 

Surface contour is also constant-z cutting. Good for the walls of features....like your tall fins and other island/vertical type surfaces.

 

Parallel is constant xy vector, with z moves being the primary cutting directions.

 

Floline and scallop are xyz cutting, "flowing" over the surface.

 

hth.

...you've helped us all so much with those great c-hooks!

[PE @ IHS]

I agree with all of the above, with the exception of using scallop. Scallop toolpath on this part might look quite bad, since the tool will be cutting up as well as down the walls.

 

The part is not that hard, once you understand which toolpaths to use and why.

[chip]

Scallop....

 

This mold I used scallop to finish both cavity and core... This mold is 18" x 15" AL.

 

 

 

 

This mold is a pump (I tried to find something with high walls) to referance to the "alternator fan" 35" x 30" stainless.

 

 

You can kind of see the scallop pattern....

 

 

It's a great toolpath for fine finishes!

[Thad]

Andris,

 

Did you move to the US?

 

Thad

[cblythe3d]

An easy trick to climb cut a slick finish and avoid constant z cut steps on the bottom, without having to resort to massive slow scallop toolpaths is to process a pencil cut on one fin.

 

Backplot this toolpath using the save as geometry option (on an unused layer)

 

Then climb cut this cutter centerline path without compensation as a 3d contour, however use depth cuts rough and finish to walk down your vertical wall producing a gouge free fillet.

 

Then use this boundary for containment, 2d pocket with a small stepover, then project the toolpath on the shallow surfs.

 

Transform toolpaths rotatate and each section efficiently climb cuts without having to explain lots of air time to the boss from the ol jump around

 

Turn the blank on the lathe was a cool suggestion.

 

Polishing is for the machine.

[Andris Skulte]

Everyone -

I show up this morning, with a whole page of tips from you guys. You guys rule! I'm going to spend today playing with the various toolpaths, and with the above ideas, should be ready to run the prototype. Yesterday, I created curves from all surface edges, deleted everything except for a single side profile, revolved that, created a lathe path, and save it as an .SLT to use as the mill stock in verifying the toolpath (damn, too long of a sentence!). Thanks!

 

Chip - I guess my 5 year old account from school finally got deleted, so I'll have to reload the pictures. I just machined some 3/4" thick exhaust spacers to get tapped for EGT bungs, and move the turbos away from the valve covers. I'll let ya know when I get the new pics up.

 

CMR - I wish I wrote the c-hooks, but I think you may have me confused w/ someone else. The best I can do is modify our posts to turn off coolant before a G28 G91 Z0

 

Thad,

Born and raised in NY, but both parents are from Latvia.

 

Anyways, you guys have given some great tips. Time to try them!

 

Andris, learning something new every day...

[chip]

Andris, If this is any help. http://www.turbo-performance.com My uncle Bob owns this company. He does a lot of mail order stuff...

 

Turbos are cool! Lots of work... Lots of plumbing. I decided to not run turbos (or a turbo) cause the cost for start. They seem to be getting more and more popular. Tune up's are hard to do on my car too. I was gonna spend $8-10,000 on a turbo kit. Then I would have to spend another $1,000 on a dyno tuned chip! Is that Digital Fuel Injection working out on the car? 8 lbs seems quite mellow for a (twin?) turbo. I'm running 10 lbs. (with an aftercooler) on my '03 mustang GT that's my daily driver. I dropped the compression a point. I'm running CP (california pistons) forged pistons. I re-valved and match ported the heads. I made 447 RWHP! I can pull an '03-'04 Cobra in any gear at least a car. It runs good!

[Thad]

quote:

---

Born and raised in NY, but both parents are from Latvia.

---

My mistake. I had you and Andreas Stute (from Germany) confused.

 

Thad

[LaszloK]

Learn to solid model this part and machine from the solid and you will not want to use surfaces. This part should not be that difficult and the constant fillet at the base of the blades makes it even easier. Give it a try.

[LaszloK]

Also as far as file size, the file with a solid model constructed from mc 2d data should be much smaller than the same part with surfaces. A stitched together solid model from the surfaces will be even larger.

[kkominiarek]

Solids is the way to go, you can create good surfaces from a solid but you can't always create a good solid from surfaces....therefore solids gives you the best of both worlds....also once your accustomed to building a solid it's so much easier than trimming/blending/filleting surfaces.

[Chris Rizzo]

quote:

---

My mistake. I had you and Andreas Stute (from Germany) confused

---

quote:

---

CMR - I wish I wrote the c-hooks, but I think you may have me confused w/ someone else.

---

I must be on the same wave length as you Thad.

 

Oh well, sorry for the mix up..still happy to help.