[Regnar]

Auh I see now, I missed all the 5 point knobs. Thanks

[dgapilot]

This is a problem/design I've been wondering about for a couple months now. Likewise, building a fuselage is the end desired result. I got the idea from these folks http://www.cartesiantube.com/ If only 1 fuselage is what you are after, they may be your best bet. If you are looking at going into business building kits, then a tube profiler would be the ideal choice.

My concept, not drawn or fully scoped out, would be a gantry for the mill cutter with a z axis. The gantry would move along the bed for the x axis, and the tube would be held in a rotary axis using 5C collets. There would need to be 2 rotary heads, one driven, and the other either freewheeling or slaved to the first. For short tubes, say under a foot, they could be done using only one rotary, profiling either side of the rotary. Beyond that there would be a need for the second rotary for support. The milling operation would be done with the tube end supported by the rotary and 5C collet as close to the collet as possible, and tubes would be blanked to approximate size prior to profiling. Most tubes on a homebuild will range from about 6 inches to about 2 to 3 feet with the exception of the longerons. To do the longerons will require a bed of something on the order of about 20 feet to profile in one operation, but could be done if indexed and swapped end for end.

After seeing the videos above, they keep the plasma head stationary and feed the tube in both the x axis and rotate it. I suppose a system could be devised to have a feed mechanism, but there is a significant diference in the force required to feed a tube under a plasma, and one being milled. The other problem I see with the machine in the video is the lack of support for the drop off. Again, for tubes in a fuselage, most will be say 2 to 3 feet and a unique profile on each end. The drop off will have a finished end on each side, and the last one done will require a fair abount of hand work to clean up where it seperates.

I've also connsidered just using a rotary head on my existing mill and stop collors. Mark the length of the tube to index it, install a stop collar at a predetermined position, and slide the tube into the rotary head and index to the stop collar and longitudinal mark. When one side is done, turn it around, put the stop collar on the other end at the correct location and run the profile for that end. This method should get you within about +- 1/16" which is within what you would expect for this type structure.

Just some random thoughts!

[dan_the_welder]

Hey DGA,

The main issue is that as soon as you unclamp the tube you lose phase registration (rotational axis). Unless your collar has a tooth or something.

Take a look at some swiss lathe vids on youtube. The chuck you see is actually a sleeve and the actual clamping chuck moves back and forth behind the scenes much like a plasma tube cutter.

[dgapilot]

Dan,
That's why the optimum design would be to machine both ends without unclamping. A gantry type machine with a bed long enough to secure and machine both ends.

The cheap method of using an existing mill would be to make a registration mark along the length of the tube and set this at the "0" point when you swap ends. Some drawbacks are as you said, there will be an error associated with aligning the registration mark, another error with where the collar is placed along the length, and then the thaught process required to mill one end clockwise and the other counterclockwise (dimension wise) since we swapped ends.

In a welded steel tube structure like a fuselage, when you apply heat to weld it, it walks all over the place. Depending on the amount of heat applied to a cluster, it can move about 1/4 inch. The level of precision for tube length and rotation doesn't have to be closer than +- 1/16 and +- 10 degrees. Keep in mind that the closer you can hold it, the easier it is to weld. I've done repairs with nothing more than a hack saw and 1/2 round file. The idea here is to come up with a reasonable way to make joints quickly and accuratly enough to assemble and weld without having to hand fit each joint. The better the fit, the less heat required and the less distortion of the finished product.

From a practical standpoint, starting at the tailpost and working to the firewall will give the least amount of distortion, and a little judicious use of the torch will bring everything back in line when finished.

Once the machine is finished, what CAM package do you envision using to unroll the tube profile after building the 3d model?

[dan_the_welder]

Well,

I try to hold tighter tolerance than that. http://www.organicengines.com

For CAM i am writing it myself in Ruby. There is some code at http://www.danthewelder.com/index.php

[dgapilot]

Dan,
Nice trikes.

What size tube can you go down to with the plasma? Looks like most of the stuff on the trikes is 5/8 to 3/4 diameter. In airplane construction, we can go from say 1"X.058 wall down to 1/4" X .035 wall. Most of it is in the 1/2 to 5/8 size, but as we get closer to the tail the sizes get smaller. Tube lengths can run from over 20 feet to less than an inch. If a plasma could do fishmouth cuts one 1/4" X .035, then a plasma might be a better solution. Almost all tubing in aircraft is 4130. Most of it is gas welded, but some folks will invest in TIG equipment. There is always an ongoing debate of the advantages of tig vs gas and wether tig requires stress relieving following welding, but that is a whole different issue.