G Scale Model Train Forum banner
1 - 20 of 23 Posts
Yeah, I use both. The black locomotive at left in my signature, a 1:20.3 model of North Pacific Coast #21 was built using parts made on both the CNC lathe and mill. A full builder's log is here (5.46mb PDF file).

The locomotive at right in my signature is a 1-1/2" scale, 7-1/2" gauge 4-4-0. An active builder's log is still on the site... Part 1 is here, and Part 2 is here.
 
I tried that once. The problem is that embossing rivets requires both a punch AND a die, and both need to remain stationary while the table and the work move. I tried making a fixture to mount the die to the machine base outside the extent of the table, but that limits rivet embossing to pretty much the x-axis unless one wished to cantilever the work hanging off the table quite a distance. I ended up using a sharp tip Dremel bit to "center-drill" the rivet pattern and then going back with a Northwest Short Line rivet punch to actually make the rivets. That saved me a bunch of time in setup as I'd center-drill all the rivets before milling out the piece. Once the rivet positions are marked and the piece cut out, using the Northwest Short Line rivet punch goes fast.

However, that was my experience. Your mileage may vary. :)
 
Scott - that's probably do-able provided you have a cutter the right shape to make the dimples that will result in clean rivets. There's probably 1001 ways to accomplish the same task. The way I did it was only one of those ways.
Image



The problem I foresee with the homemade riveting machine is one of setup. Since the riveting machine won't actually cut out the part, that means the work will have to be moved after riveting and accurately aligned on the milling machine (assuming that's what will be used to actually cut the part out). While it's certainly do-able to accurately align the pre-riveted work, it's rather a PITA. One could make a jig for both machines that would accomplish this, and if many riveted parts are to be made, this would be a good idea. For me, I didn't have that many riveted parts to make, and I decided I'd rather spend time making the model as opposed to spending that time making tooling which I probably wouldn't use all that much.
Image
But again, others mileage may vary.
 
Discussion starter · #8 ·
The detail I will eventually be going for is very repeatative. Although detailing a 63' heavyweight is not a small task. especially when I want to have several duplicates of them. This picture is good at showing the rivet detail on the side. note the verticles between the windows, and the end details.

http://jalcpa.com/PC607STRASSBURGPA-AUG24-96-277281-9.jpg

My plan was to make a die that would receive this strips of styrene. The die would be fixed to the table and have multiple recesses to receive each rivet as the table slides down to make a long strip. this would be cut to length as needed. I am now thinking of using the NWSL rivet punch. the question is which sizes are best for rivet detail in 1/29 scale.
 
Posted By toddalin on 19 Oct 2012 05:43 PM
Can anyone recommend any good books on the subject as well as machining and lathe?

Thanks
Todd,

I was in the machine trades for about forty years before my retirement. Manual mills and lathes for thirty years and CNC mills and lathes including CAD software in the last ten years. To program, you should be a competent machinist with manual machines FIRST. You have to know how various cutters react to various materials in a mill or lathe. I would recommend taking a night course at a local college that offers basic machine shop training. Then I would sign-up for a class in basic CNC programming. Learn G codes and M codes. Have a very good background in trig and geometry. Learn "hand programming" first, then learn how to use a simple cad program on a computer for more complicated work(other than just rectangular blocks or round parts). IMHO, you cannot be a good programmer WITHOUT hands-on work on the machine. You HAVE to know EXACTLY what the machine is going to do just from reading the program line by line, before you ever push that GREEN button. This is why I would recommend a few courses at a local college before trying to read how to do CNC machining. Just my .02. Good luck. You will find it to be a very rewarding learning experience.
Image
 
Gary - I'd like to add My 2 cents here also!! Then we'll have 4 cents invested and will get a larger return!!!! ha,.. on our money

You could not have put that better!!!

Even with just a few years of CNC mill and lathe work, and 5 years of sheet metal CNC work before that. There is a steep learning curve to this very unique art from of hogging on metal. Oh, and I have also had several years of just plane 'ol belt drive cut and measure by hand - mill and lathe work to start this all off too! The mentality is there...
Even after hours of writing and designing a program, working with cutting paths and tool sizes, you still have to set up the mill and then hold your breath and in "Slow Motion" run thru the program in the mill one line at a time with a reference line reset above what will become the work line or part to be made, in other words the tools run above the work surface to avoid a crash, just in case..

But what comes out can be amazing!!

Thanks for sharing Gary!

Dirk - DMS Ry.
 
If you're looking for simple tutorials and instructions on programming with g-code, Sherline has some good stuff in downloadable PDF format.

Overview
This is the manual I initially learned g-code programming from (programming tutorial found in Part Two - Programming Sherline CNC Machines with EMC. Part One deals with the specifics of their machine and software).

A really good site with loads of info on cutter geometry, chip loading and thinning, rubbing, tool deflection, etc. is the blog section here . I found the stuff on chip thinning particularly valuable - Part 1 and Part 2 - it really gives the reasoning behind feed rates and spindle speed and why they are sometimes counter-intuitive. In part, it's also a promotional video for the G-Wizard calculator which I also highly recommend for calculating feed rates and spindle speeds - for all the reasons contained in the videos. I actually welded a cutter into the material and broke the cutter off when milling steel angles as part of my pilot by intuitively deciding to slow things down when I really needed to speed things up. While in some ways, this info is less critical when making relatively small parts out of brass (though it's still necessary for accuracy and tool life), it becomes far more important when milling steel, stainless, or aluminum.
 
Posted By SD90WLMT on 20 Oct 2012 12:31 AM
Gary - I'd like to add My 2 cents here also!! Then we'll have 4 cents invested and will get a larger return!!!! ha,.. on our money

You could not have put that better!!!

Even with just a few years of CNC mill and lathe work, and 5 years of sheet metal CNC work before that. There is a steep learning curve to this very unique art from of hogging on metal. Oh, and I have also had several years of just plane 'ol belt drive cut and measure by hand - mill and lathe work to start this all off too! The mentality is there...
Even after hours of writing and designing a program, working with cutting paths and tool sizes, you still have to set up the mill and then hold your breath and in "Slow Motion" run thru the program in the mill one line at a time with a reference line reset above what will become the work line or part to be made, in other words the tools run above the work surface to avoid a crash, just in case..

But what comes out can be amazing!!

Thanks for sharing Gary!

Dirk - DMS Ry.


Just a little additional information regarding "writing" of a program to machine a part (mill or lathe). A cnc program is composed of many lines of code. M codes and G codes are given a number depending on what the particular "code" designates the machine to do, i.e. G90 designates an ABSOLUTE move from "00"(machine home) or a move to the next machine position to start another operation, G91 designates an INCREMENTAL move from the last point to the new point. M codes usually control the program stop or stsrt, spindle on or off (either clockwise or counter-clockwise) depending on the number added to the letter code, coolant on or off, etc.


Here is a link to some of the G and M codes:http://www.camsoftcorp.com/prod02g.asp

Some very simple machining like cutting rounded corners on a rectangular block could be as few as 20 separate lines of code, each with its own number. AND these lines HAVE to be in the proper sequential order or the program will simply NOT RUN. I have machined forging dies with some radical surfaces for aerospace work with an area less than 12 inches square that would contain 100,000 to 500,000 lines of codes or more! These programs are definitely NOT hand written! A toolpath is run over the surface on a cad drawing in a computer and a NCI(numerical control intermediate) code is produced. That code is then run a backplot program to make sure the cutter does what it is supposed to do. NO gouges or moves through a part in the wrong place. Once that is verified, then the NCI is run through another program to produce the actual codes going to the machine. Granted this is for more complicated work and machines, but basically you do the same thing on a simple nc lathe or mill at home. It is just not simply reading a book and going out in the garage and making a perfect part. They all have some learning curves, some VERY steep!
Image
 
Not to take away from anything Gary and others have said, but my own personal experience was a little different. Everything I needed to know to CNC the parts for #21, I learned from the manual I referenced above. Now granted, these parts were mostly 2D sheet metal parts and the g-code consisted primarily of a 2D tool path used in cutting the part out of brass sheet, and on some drilling the center marks for later riveting, but there were other more complex parts - turning the tops for the tender tanks, making the triple valve, etc. The point is that in starting simple with basic 2D, I was able to build on that when I later needed to do something more complex. There IS a learning curve, and it IS steep in places. Having been a computer programmer for thirty-odd years perhaps made it a little easier for me as I was already familiar with the broad brushstrokes, but it isn't something that is so complex that anyone couldn't learn how to do it. Each command is one machine movement, and like any problem, it's far easier when broken into basic steps. Did I make mistakes? Sure I did. Did I make some scrap and break some tooling? Absolutely!! But I got the job done and I learned a helluva lot in the process.


My point is that no one should be scared off by talk of complexity or steep learning curves, or references to required complex mathematics. I've never needed trig or algebra in any of the stuff I've done. Most of the parts we wish to make are small and not real complex and can be done with relatively simple machine operations.


There's definitely an art to it, but it's an art anyone can learn. I also had some good experienced people I could call if I got stuck, not so much on g-code but on machining itself. Bob Starr and Dave Hottmann were two of them, and I'm eternally grateful to these guys for all their help. This forum is a darned good reference for just those sorts of questions. There are many talented machinists here, Gary, Henner, and Rob Lenicheck being among them, along with a slew of others!!


I say... dive in, and don't be afraid to make mistakes or break some cutters!! That's how you learn.
Image
 
Thanks guys,

I've actually looked into classes, but none are offered in the area.

I've also looked into support groups that offer training, but little is available and what I did see was on the other side of town on Fridays starting at 5:00 PM (right at the evening rush hour). I'll continue to look into support groups.

I also posted a pic of my machine of the Corvette Forum and have already had requests for its use. If these people are experienced, all the better to learn from.

The machine does program in "G" and the supported code is shown on Page 20 of their manual. This thing can be used to build the workings of a V-8 engine from scratch, or do fine woodwork. And while its been a long time, I did programming in Basic and Fortran (old punch cards) in my past.

Thanks for the manuals Dwight. I've printed them out.

http://shopmasterusa.com/wp-content...LOGUE1.pdf


Supported G-codes:


G0 RAPID POSITIONING


G1 LINEAR INTERPOLATION


G2 CLOCKWISE CIRCULAR/HELIACAL INTERPOLATION


G3 COUNTER-CLOCKWISE CIRCULAR/HELIACAL INTERPOLATION


G4 DWELL G10 COORIDNATE SYSTEM ORIGIN SETTING


G12 CLOCKWISE CIRCULAR POCKET


G13 COUNTER-CLOCKWISE CIRCULAR POCKET


G15/G16 POLAR COORDINATE MOVES IN G0 AND G1


G17 XY PLANE SELECT


G18 XZ PLANE SELECT


G19 YZ PLANE SELECT


G20/G21 INCH/MILLIMETER UNIT


G28 RETURN HOME


G28.1 REFERENCE AXIS


G30 RETURN HOME


G31 STRAIGHT PROBE


G40 CANCEL CUTTER RADIUS COMPENSATION


G41/G42 START CUTTER RADIUS COMPENSATION LEFT/RIGHT


G43 APPLY TOOL LENGTH OFFSET (PLUS)


G49 CANCEL TOOL LENGTH OFFSET


G50 RESET ALL SCALE FACTORS TO 1.0


G51 SET AXIS DATA INPUT SCALE FACTORS


G52 TEMPORARY COORDINATE SYSTEM OFFSETS


G53 MOVE IN ABSOLUTE MACHINE COORDINATE SYSTEM


G54 USE FIXTURE OFFSET 1


G55 USE FIXTURE OFFSET 2


G56 USE FIXTURE OFFSET 3


G57 USE FIXTURE OFFSET 4


G58 USE FIXTURE OFFSET 5


G59 USE FIXTURE OFFSET 6


G61/G64 EXACT STOP/CONSTANT VELOCITY MODE


G73 CANNED CYCLE - PECK DRILLING


G80 CANCEL MOTION CYCLE (INCLUDING CANNED CYCLES)


G81 CANNED CYCLE - DRILLING


G82 CANNED CYCLE - DRILLING WITH DWELL


G83 CANNED CYCLE - PECK DRILLING


G84 CANNED CYCLE - RIGHT HAND RIGID TAPPING


G85 CANNED CYCLE - BORING, NO DWELL, FEED OUT


G86 CANNED CYCLE - BORING, SPINDLE STOP, RAPID OUT


G88 CANNED CYCLE - BORING, SPINDLE STOP, MANUAL OUT


G89 CANNED CYCLE - BORING, DWELL, FEED OUT


G90 ABSOLUTE DISTANCE MODE


G91 INCREMENTAL DISTANCE MODE


G92 OFFSET COORDINATES AND SET PARAMETERS


G92.x CANCEL G92, ETC


G93 INVERSE TIME FEED MODE


G94 FEED PER MINUTE MODE


G95 FEED PER REV MODE G98 INITIAL LEVEL RETURN AFTER CANNED CYCLES


G99 R-POINT LEVEL RETURN AFTER CANNED CYCLES
 
Todd,

Those G codes are standard in the industry. I looked at the manual and saw the screen capture of the controller. The M codes and line number are pretty much standard. That looks like a very nice machine for the home hobbyist.
Image
 
Just FYI, when making #21, the only commands I really used were G0 through G3, plus M0 (pause). I did try a couple of canned cycles as well. Later as I was making stuff for my ride-on, I learned to use subroutines (my Sherline-supplied software at the time didn't support them - I'm using Mach 3 now). As I said, most of the stuff we make in the hobby are relatively simple parts. The only time I've ever used my rotary table was to cut the taper into my ride-on's brake shoes. I've also hand-written all my own code, which shows how simple most of the stuff is. If you have AutoCAD or something similar, you can take the coordinates directly off the drawing which is a real time saver.
 
Posted By toddalin on 21 Oct 2012 04:44 PM
Has anyone tried G-Simple? It looks like you draw the picture and it writes the G-code for you. This actually looks fairly easy to use and is a free download.

G-Simpe Home Page

Todd,

This looks like a very simple program to learn. I checked some of the samples and the code they produced. It would definitely save you time getting your geometry correct with angles and partial radii (trig involved here). I would gownload and try it out. Great learning device.
Image
 
1 - 20 of 23 Posts