Saturday, September 25, 2010

I Love CNC

The wonders of modern technology.

From:

To this:


In a couple of hours.  All the parts are accurate to within a few ten thousandths of an inch.

The switch from the 2024 aluminum test material to the 4340 steel part material required slowing the cutting speed by approximately  50% and increasing the feed rate by 25%.   It may seem counter intuitive to increase the feed rate for a stronger material but the carbide insert's cutting edge needs to be able to get a decent bite into the material, otherwise it just deflects the workpiece and rubs, causing a lot of heat buildup and usually insert/tool failure which leads to a scrapped part.

Now that all 3 initial pieces are machined I will change the setup in the lathe to flip the part around and accurately hold it by the features we just machined.  That post will be up soon.

Wednesday, September 22, 2010

Machining Step 1 on the Lathe

After a couple of long days and late nights machining non-motorcycle parts the lathe is open and I can run a test part of the first crankshaft program!  I'm making the first part from some scrap 2024 aluminum to verify the program before using the more expensive and much harder to machine 4340 steel material.

The machining process will be:
  • drill 60 degree center in outer end for tailstock support
  • roughing the main profile
  • finishing the main profile
  • machining small undercuts on main profile
  • machine center main bearing fearure
  • remove as much conrod journal stock as possible
This is what the setup in the lathe looks like.  As with the engine castings, we are at the limit of machine capacity but the part fits and that's all that matters.

This is taken after the finishing of the main profile:


And this is taken after the program is complete:


Here's some CAD vs reality for a comparison:


 The aluminum version looks nice but would never be able to withstand the temperature and stress of operation.  Next up is the 2nd machining operation that finishes the opposite end main bearing and tapered generator mount.  After that operation the part will move to the mill for some 4 axis work on the conrod journals and camdrive sprocket teeth.

Until the next update......

Monday, September 20, 2010

Programming Setup 1 on the Lathe

The first step in converting CAD to chips is to plan ahead, and it should happen very early in the design stages, otherwise you run the risk of having an unmachinable or hard to machine part.  What we are doing here is step 2, converting 3D CAD to 2D CAD to G-code CAM.  The 3D CAD provides complete surface information of the part.  Since at this stage the part is a simple revolved shape, we can reduce it to a 2D profile and not lose any information about the shape.  From this 2D profile the CAM software creates a series of 2 dimensional moves along the X axis (diameter) and Z axis (length of part) that create the part profile from solid bar using appropriate cutting speeds and feeds.

3D CAD:

2D CAD:

2D CAM:

Now we can go to the machine, which is currently busy.....


The CAM package is told the shape and position of the various tools needed to cut the part, it is told the shape of the part, and it is told the characteristics of the machine.  Doing what computers do well, crunch numbers, it uses all of this hopefully accurate information to calculate the appropriate motions to generate the desired part profile.  As with all computer programs, garbage in=garbage out, except instead of the BSOD or an inaccurate spreadsheet, you get twisted metal and a large repair bill.  It is a very good incentive to make sure all the info input is correct and also carefully check the output.  The output of this 2D CAM process is a text file in what machinists call a G-code format.  A sample g-code file in a format usable by my Siemens control looks like this:

%MPF52
G70 G90 G40 G54
(TOP-CAP-3-rec2-rad PR=3.745 Z1=8.075 )
(T-2 O-6)
(DCGX IN SDJCR-123)
T02 D06
G0 X0.825 Z.5
Z0.
M3 S1000
G96 S1000
M8
G1 X-0.0357 F.005
G0 X0.6316 Z.155
G96 S1500
G1 Z-2.99 F.008
X0.6864
G0 Z.11
X0.5567
G1 Z-2.4417
X0.5625 Z-2.4475
G3 X0.57 Z-2.4656 B.0256
G1 Z-2.99

The format is relatively simple and once you have used it a bit reading it is pretty straightforward.  The G-Code file is sent to the machine over the network and then after machine setup the program is run.

Once my current lathe job is complete the crank goes right in.  I hope to have another blog update sometime this week.

Sunday, September 19, 2010

Starting on the Crankshaft

Now that the engine case patterns are approved and the production order placed I can move on to the crankshaft.

It will be a 180 crank to exploit the reduced variations in reflected crankshaft inertia that this configuration provides, similar to Yamaha's cross plane crank in the M1 Grand Prix bike and the R1 Super Bike.

I will be initially testing 2 different versions of the crank- one with pork chop counterweights and one with full circle counterweights. The main testing variable is the overall weight/inertia vs. aerodynamic efficiency of the 2 versions.

Full circle configuration:
Pork Chop configuration:


The manufacturing process will be as follows:

Material blank:
  • 3 3/4" diameter x 9.7" long
  • 4340 steel
  • 28-32 Rockwell C hardness
In house:
  • Machine crank on lathe in 2 steps blank leaving .01-.02 stock on surfaces to be ground.
  • Mount lathe blank in mill 4th axis and rough machine crank throws
  • Rough and finish machine the two silent chain cam drive sprockets with custom form end mill cutters.
Outsource:
  • Hobbing of primary drive gear and starter clutch/crank sensor spline will be done by Eagle Machine, Inc. of British Columbia. They have experience to properly deail with custom low production crankshaft.
  • Final Grinding of main and rod bearing surfaces will be done by Lopez Crank Shaft of Santa Fe, CA. They are another custom crank specialist and can provide the accuracy and surface finish needed.
  • Plasma nitriding will be done by Accurate Ion Technologies, a specialist in steel hardening and advanced surface finishes. We are using plasma nitriding for it's low process temperature that enables proper multilayer surface hardening of the part with no distortion.
  • Once we get the part back from Accurate Ion it is ready for use.

The process will involve 2 lathe setups and one milling setup.

The lathe first lathe setup will hold a 3 3/4" x 9.5" long material blank in a 3 jaw chuck. we'll machine one end of the crank, the middle main bearing journal, and various smaller features on one end of the part.

End of first lathe setup:

Once I finish some client lathe work currently running this is next in line. The next post will detail programming and cutting of this first lathe operation.