[Zenji]

Any more progress on the skunkworks mill?

[MichaelGolden]

Haha yes lots of progress. I still have some of my hair. And all moved into the new place! Much more room than before. Here are some pictures!

The CNC Toolkit - Creative Toolpath Control is an open source 5 axis cam software if anyone was wondering.
I've been generating some gcode and controlling it with mach3. The other 5 axis machines I've seen have been controlled by linuxcnc, I believe. But I haven't tried it yet. There have been some hiccups along the way. I spent my saturday trying to get the axis to all move the direction they are meant too. After all, generating 5 axis code is a lot different from 3 axis. But I think I'm finally getting a hang of it. I ran some air cuts with 5 axis code of the tilt table without the 5th rotary axis hooked up yet. And it seems to have just enough room for my chuck and workpiece.

I've been working on editing some generic post processors and I got the machine to match the computer. I've been learning as I do it. Its a big puzzle when you dont know what you're doing. Last night I've been trying to add g93 to the .LIB files. And it looks something like this.

-----------------------------------------------
6030 CHANGE IN .LIB FILE

:C: SETON()
:C: IF MOVE_TYPE=1 OR MOVE_TYPE=2 OR MOVE_TYPE=3 THEN
:C: IF feed_type=1 THEN FVAL=(OPR_X_FEED/DISTANCE) ELSE FVAL=OPR_X_FEED ENDIF
:C: ELSE
:C: IF OPR_TYPE=0 OR OPR_TYPE=1020 THEN
:C: IF feed_type=1 THEN FVAL=(OPR_X_FEED/DISTANCE) ELSE FVAL=OPR_X_FEED ENDIF
:C: ELSE
:C: IF feed_type=1 THEN FVAL=(OPR_Z_FEED/DISTANCE) ELSE FVAL=OPR_z_FEED ENDIF
:C: ENDIF
:C: ENDIF
:C: IF OPR_TYPE=MILL_DRILLING THEN
:C: IF feed_type=1 THEN FVAL=(OPR_Z_FEED/OPR_SPEED) ELSE FVAL=OPR_Z_FEED ENDIF
:C: ENDIF
:C: IF ATTROVERRIDE=YES THEN FVAL=ATTRDVALUE ENDIF
:C: IF FVAL=MACH(REGISTER) THEN SETOFF() ENDIF
:C: IF OPR_TYPE=MILL_DRILLING AND G_GROUP(GG(G_CANCEL_CYC))<>GC(G_CANCEL_CYC)
:C: THEN SETON()
:C: ENDIF

*
* G93 Fanuc style Inverse time calculations for wrapped and multiaxis operations
* (G93) Feed Rate = (1 / (Actual Move Distance / Desired Feed Rate))
* or = (Desired Feed Rate / Actual Move Distance)
*
* if you need the G94 method then use logic below, but this will only work for one axis of rotation
* (G94) Feed Rate = (Degrees of Rotation / (Actual Move Distance /Desired FeedRate))


6192 AND 6295 AND 6395 CHANGE

*
:C: IF adaptive_feed=1 THEN
:C: IF feed_type=1 THEN
:C: CALL(FEED_DEFINITION_G93)
:C: ENDIF
:C: IF feed_type=2 THEN
:C: CALL(FEED_DEFINITION_G95)
:C: ENDIF
:C: IF feed_type=3 THEN
:C: CALL(FEED_DEFINITION_G94)
:C: ENDIF
:C: ENDIF

*
------------------------------
Thats not all of it. Thats taken from my notes. So far I've gotten the code to place an F value on every line. But have some work to get it right. Maybe someone who knows how to edit post files properly can chime in. I searched online for how to add G93 the usual properly way and didnt have any luck. I might just use a constant feed rate for the time being and come back to it.

Mach 3 also has some problems running 5 axis because the rotary axis slow down. There was a post that comes up. The guy tuned his x y and z to the max desired IPM and his 4th axis to a much higher value. Mach 3 seems to run only as fast as its slowest axis.


I was all like, "Oh i'll just try the code for 1 blade on this turbine to see if it works. Piece of cake." Nope.

[MichaelGolden]

I mounted the chuck and think I'm just about ready to do the first test cuts with the machine. I'm just wiring up the 5th axis bob and motor now. It should be fine running the whole system on a 15 amp plug. The 500w spindle, air compressor/vacuum run off a separate 15 amp fuse.

[hjl4]

Man you did an amazing job. I wish my shop was that clean, you must spend a lot of time to keep it that way.
If you don't mind, what province are you in?
I'm contemplating building my own CNC 5 Axis mill within the next year, so this thread is really cool.
I too, won't be going large, but definitely it will be a super rigid setup with a minimum 10000 RPM spindle.
Keep up the great work.

P.S. Is this just a hobby, or do you plan on making turbines for sale. I know a guy who is looking at having a scaled and running Iroquois engine model, machined and built.
I can do most of the machining for him, but Blisks and turbine blades are just too much for me and my 4 axis.

[MichaelGolden]

Thanks man, I'm in Alberta right now, doing the welding thing. Planning to switch to machining if I could find something else. Its just a hobby right now. I only get to work on the stuff during evenings and Sundays usually. I would love to be able to produce my own engines for sale. But thats much farther down the road. I'd want a production machine and better work area.

A place for everything and everything in its place. I like to try and keep it that way. I use the shop vac after each little job. If I were to build another machine I would definitely go much bigger, but only because I have a better idea what im doing now. I'm already planning to redesign it with higher quality machined parts and servos. I would encourage more people to build their own because the prices of the machines amaze me. Not everyone has the luxury of even using a 5 axis machine. I'm happy I get to tinker with the stuff I have, on the limited budget I have. If it works or not I can always upgrade.

A scaled running Iroquois would be awesome. But you cant scale down commercial jets. You would have to redesign the entire engine for it to run. I've been working on a very basic design axial flow style compressor. The math that goes into the blade design is enormous and cant be accurately done on paper. There haven't been any small scale, self-sustaining "axial flow" engines to date. There are a couple that have come close. Some people said it couldn't be done but I'm going for it anyway. If its not challenging, whats the fun in that? I'd say go for it.

[Winterwatchers]

I have seen working axial compressor that are hobby sized. The trick is to machine the blade mold in graphite and then edm it. that's something your mill could do pretty well.

[hjl4]

Hmmm. I'm just trying to imagine how EDM would cut the variable curves and twists of the blade.
In anycase, interesting suggestion.

[MichaelGolden]

I have seen working axial compressor that are hobby sized. The trick is to machine the blade mold in graphite and then edm it. that's something your mill could do pretty well.

There is quite a few discussions about this over at the GTBA website. The few that have been attempted have come very close. But non have become self sustaining. They sound and look amazing but no one has been able to balance the fine thermodynamic design that keeps the engine running without melting the turbine.

The Bladon Brothers jet is probably the most well known because of the youtube video that pops up when you start searching for them. They apparently had a contract with Jaguar but the project never reached completion. In the blurry video there are flames coming out the back. Meaning you have problems.

James VG Hill is a very knowledgeable guy at the GTBA who build a very beautiful axial flow model jet. About 2" diameter inlet. He cast 1000 individual micro compressor blades that slipped into machined discs. I think he met with one of the Bladon brothers at some point. I believe hes still making progress with his engine.

The latest engine I heard of was mentioned over at RC universe. I'm not sure what his name is. But I think it has a great chance of at least self sustaining. He built it with variable stator vanes. These are capable of adjusting the flow of gas through the engine upon different operating conditions. He has a couple video's of it running for short periods of time still working out the kinks. But awesome non the less. Here's the link Axial-Flow Turbojet

If there have been any others, I'd love to check them out.

[hjl4]

Cool, and he tested his engine off his apartment balcony.LOL
The engine I would love to build would be slightly larger then the one in the video, but realistically, time is not on my side lately.
150mm of rain has just screwed many a plans for the future, as my basement flooded and it just happens to be where I kept servo drives and motor controls for my 5 axis project.
Many tools such as oscilloscope and power supplies also got flooded.
And worse yet, this morning it rained, again.
Everything was fine until the power went out Sunday, due to the high winds and the sump pump went out. Power back on 7 hrs later.
Life is full of surprises.

[MichaelGolden]

Yeah pfft field testing is overrated.

Thats really terrible to hear. I've been flooded in a couple times in the past. I understand how much that sucks! I hope everything wasn't totally ruined!

[hjl4]

Lol. Field testing. LOL That's one way to keep your neighbors on their toes.
Yeah, not everything ruined. But keeping stuff in a box on the basement floor isn't too smart either.
The drives and most electronics can be saved, provided I don't power them up until dried.
P.S. A basement less then a kilometer away from a big lake isn't either. Ha Ha Ha. What was I thinking.

[Zenji]

The code looks like some serious homework, Henry's lobby flooding sounds terrible.
I was going to build a 5 axis like a centroid, but once you roll up your sleeves and set up the workshop good to go, even 3 axis becomes a stretch for time.
I like watching this build because of that, the dreams come easily but the end result can be a long road.

[MichaelGolden]

Dreams are alive and well! I managed to get it working today! I cut 19 little blade patterns around this plastic piece. I'll have to clean up between the blades now and machine the back of it off. But its finally ready for some serious design work.

Some notes.. In the post files there are "macros" I had switched them off while I was initially fiddling around with the code. So in mach 3 in the toolpath it shows a nice picture of what you are machining. Although the gcode was garbage. The xyz axis did not match the height of the tilting axis. Once the macro section of the post is enabled the toolpath looks like a bunch of weird scribbles or loops. Although this suggests the z axis swooping up and down with the edge of the part.

I had my two rotation axis switched in the post. But once I noticed that and changed it. Everything came into play and it was lining up with the part like it should have been!

I never changed my motor tuning like suggested in that other post. I read setting the rotational diameters to .0001 sets mach 3 to match the rotational diameter speed to the height of the z axis. So I did that. Although if I set it to say 10IPM its still super slow. So I ended up running the program on 175IPM in the end. You'll see the speed in the video im about to try and upload.

[MichaelGolden]

I cleaned up the little plastic piece. Turned out pretty well. I'm happy with the level of backlash from the rotary table. You can hear it rub along a high spot in the gearing. I had turned it back and forth many times just over 1000 degrees in mach3 while I was setting the workpiece. And when I sent it back to zero it was bang on. I'm sure over time it will be something to watch out for.

[hjl4]

oops

[hjl4]

Michael, do this same test in a piece of Aluminum. I bet you it comes out nicer especially if you reduce the step over. But overall, it proved the concept and possibility of it being done with few hiccups.
Nice work.
After watching the video, I can tell the steppers are not following a linear ramp up in speed, but rather they are stepped up.

[MichaelGolden]

The only problem with doing this from aluminum right now is this particular turbine is so small I used a 1/16" flat tip grout bit from a Dremel lol. I'm not sure it will work so well with aluminum. I mucked the blades up a bit pulling off the strings of plastic.

I'll need some ball nose end mills and extended bits. All I have right now its the hobby mill kit.

I'm going back to engine design so I can cut a very nice aluminum fan.

[hjl4]

Looking forward to it. BTW, use the largest endmill that you can use for the job, (for rigidity's sake). Long 1/16" ball mils are hard to find and running carbide ones, they are easy to break if things aren't perfectly setup.
I do a lot of mould work in AL, and milling with small cutters is a game of patience and proper settings. But the outcome when everything goes well is usually amazing. An example of wrong mill for the job, is slow spindle speed for such small cutters. https://www.youtube.com/watch?v=F7bB0QCAmk8
I specifically made this short video, showing just how slow things can go.
In comparison, check out what a high spindle speed can do. https://www.youtube.com/watch?v=_PC9ezN6cno

[MichaelGolden]

Machining vids never get old. Great machine! My neighbors would love me. Lol. I'm definitely upgrading to a nicer machine just not so soon.

The good thing is the smallest I think I'll ever need to go is 1/8". This little impeller is for a smaller engine than the ones I'm planning for. Also this part is meant to be made from inconel to power the impeller. I wonder how inconel will machine with such tiny bits.

Here's what I had to use. Any larger would have cut away the blade next to it. I got it down the road from the hardware store. I was recently looking for a place to find good ball nose and 3 flute aluminum bits in Canada. What place would you recommend?

[hjl4]

Well for me it goes like this. For brand name, high quality HSS and carbide tools, I first check the surplus items at CTP in Stoney Creek, ON. Super cheap. An Example, 1/2" Carbide RME(made in Canada), $19. HSS 1/2" Hanita, $7. Prices are all in Canadian. If they don't have what I need, then it's a quick search on Ebay. When on ebay, pay attention to shipping fees. Most fees charged by US sellers are ridiculous, and they charge you a customs processing fee, which in my opinion is illegal. But they almost all do it. Thats what FREE Trade was for. U.S. legislators know how to screw everyone else in the free world, so they allow this to happen. Whatever.
For small bits, anything Kyocera is excellent quality. Some Chinese bits are good, but I try to stay away from those.
Machining Inconel is going to be a long and painful process, depending on how fast your spindle can spin it could be many hours with such small cutters.