Our lathe guys are telling me that they need to adjust feeds and speeds depending on what material lot their Titanium came from, even though it's the same alloy and heat treatment. Some of it is just more gummy than the rest. Also, no formula will tell you how rigid your setup is.

Shade settings will have nothing to do with your toolpath, it's just for display. In configuration there's a tab for "Tolerances", in which I have my "Maximum surface deviation" set to .0001". That will affect your surface geometry. Then in the toolpath parameters there's another tolerance setting, "Total tolerance", which defaults to .001". I generally use a 3:1 filter ratio and set the total tolerance around .0005" to .0001" depending on what I'm doing.

Sounds like full contact programming...

Vericut is the undisputed king of verification, but also has by far the highest price. Many find that Predator is quite good, and it doesn't cost as much.

Try checking "one way filtering".

Ultra sharp cutters with corner radius. Edge prep for aluminum works well. Make sure your RPM isn't too high or it can melt bits of plastic back onto the part.

I've been using "control" for comp on the wire so far, since that's what the previous wire guy did, but I find you have to be more careful. I think I'll go back to "wear", since I'm the only one who sets the thing up anyway.

What happened to the last guy?

Just found something on this. The post gets a new # POST TEXT entry for each control it's used with. Looks like you need to set the text independently for each control.

Well I gave it a shot, and it seems to be working. Here's some sample code: code: % O7011(FEERATE,MULTIPLIER,TEST) N1(1/2,E/M,5/8LOC,A345,EDP=61562)(.5)(D=1) (G54.1P1,ROUGH,PERIPHERY) T1 M6 T2 M911 M11 G0G90B0. M10 M1 N100 G90G54.1P1G0X-3.3007Y-4.302S9550M3 #500=1(FEEDRATE,MULTIPLIER) G43H1Z.1M8 G1Z-.52F[#500*57.3] X-3.0658Y-4.2165F[#500*5.] X-2.8309Y-4.131F[#500*25.] G3X-2.6641Y-4.0297I-.171J.4699F[#500*45.] X-2.5487Y-3.8724I-.3378J.3686F[#500*65.] G1X-2.4558Y-3.6731F[#500*70.99] X-2.3629Y-3.4739F[#500*50.99] X-2.2699Y-3.2746F[#500*38.17] X-2.177Y-3.0753F[#500*31.16] Y-2.8351F[#500*31.16] Y-2.595F[#500*34.92] Y-2.1147F[#500*34.77] X-2.2635Y-1.9293F[#500*34.77] X-2.3499Y-1.7438F[#500*54.77] X-2.4364Y-1.5584F[#500*74.77] X-2.5229Y-1.3729F[#500*84.57] Now I need to set it up so that it turns on and off with a misc integer switch.

Okay, here's what I'm thinking. After picking up the tool, do something like "#525=100" to set the percentage of programmed feedrate to run at. Then every time a feedrate is assigned, do something like "F[#525*32.52/100]" where the 32.52 is what Mastercam wants the feedrate to be. I haven't done any macro stuff in a long time so I'm not sure if that format is appropriate. If it works, then the operator could change the "#525=100" to be "#525=90" or "#525=110" or such if he thinks the tool can't take the programmed feed, or if he thinks it can take more.

I was thinking about setting up the post for our horizontal such that the feed rate would be made from two multipliers. One would be the feedrate generated in Mastercam, the other would be a variable editable in the final code which would start out as 100 (for 100%). (Mastercam * Operator / 100 = F) This way I could use the highfeed function, but still have operator adjustable feedrate. The operator adjustable variable would show up at the start of each tool, or possible each operation, so each could be tuned at the machine. Has anyone tried anything like this?

Are the upcut spirals carbide or steel? I know wood can be pretty abrasive. My guess is that either the tool is getting dull, or it's loading up with chips. Setting up an air blower might help. Does the tool get really hot?

Produca2, my guess would be that something is different about the material. Could it be a different hardness?

Here's a place for cycle wheel blanks: http://www.allamericanwheel.com/index.html

I've done custom motorcycle wheels, and you definitely want to start with a forged blank. They're a little pricey, but the appropriate hunk of billet would probably cost more. Our wheels were about .750" thick except for the rim, and our rule of thumb was to leave at least three sections at least an inch wide at the thinnest point. We kept them as balanced as we could, but once the tire is mounted you need to balance it anyway so I don't think it's too big a deal. Depending on what kind of car he has a color anodize might look nice. In case you want an asymmetrical design, Mastercam will tell you the center of gravity of a solid. If your first design doesn't come out balanced when you model it, just tweak it 'till it does.

I think what's going on is that when a surface gets trimmed, the system retains an untrimmed copy and blanks it. If you unblank all, you'll get the untrimmed surfaces too. With selective unblanking you should be able to get only the trimmed ones. That's why I never blank, I use different levels instead.

quote: Smiley, I don't think the CRT has anything to do with the speed. There is one way, I believe, that it could make a difference. I doubt the CRT and LCD have the same resolution, and if the vid card doesn't have to drive as many pixels it might run faster.

quote: At what point does the new arc appear on the wrong end? After you click the endpoint, you can drag the arc bigger or smaller. Do you see it then... Yep. Coming out the end I didn't drag the arrow to, inverse-tangent (folded back).

I'm pretty sure that's what he's talking about, 'cause I get the same thing. Create arc tangent dynamic, click a line and drag to an end, and about 25% of the time the arc starts in the middle or at the other end. Deleting and recreating the line often fixes it, but not always. X2 MR2 SP1 XP Pro 64 Quadro FX540

I would think so, but I think you'd need two vid cards or something with three or four outs. Never tried it, but I'd like to. Let me know if you figure something out.

Mastercam or Solidworks on the left. On the right, a PDF part print, PFE32, NCPlot, Email, Citrix, Windows Exploder, etc. Makes me really wish we could drag and drop files into Mcam.

You can always hand edit the code a little.

I usually program for the weakest machine that it's likely to run on. The operator can tweak feeds and speeds at the control if the machine will take it. Also if I'm programming something with a bunch of depth cuts, I'll do a M98P1001L32 kind of thing, with the Z move in the sub. That way the operator can easily adjust the number of depth cuts and still get the right final depth.

Just a stab in the dark, but the ones with I and J could be helix-arcs. Do they also have a Z? There's another possibility that just occurred to me, which is that in your machine definition you can set R and IJK usage differently for each plane.