WW1 and CNC

franzkait · 4 December 2014, 06:49 PM

Same problem if you try to build a Fokker D7 wing with original Fokker drawings . Using the D7 or the C1 drawings .

John McKenzie · 14 December 2014, 08:32 AM

Quote:

Originally Posted by jumpinjan

..... replicating all the Albatros D.III's bulkheads. .... In the drawings, I found many mistakes including dimensional & drafting errors (just drawn wrong).

Picture of the averaged loft to make the body smooth, and not to follow the bulkhead's exact dimensions.

http://gwaero.com/uploads/September_2010.pdf

.

.

Certainly it look now , nothing like if any original Albatros drawing was ever used .
You miss completely , the design idea behind the "original" DIII shape by a mile , especially in the nose part ,...in your smoothed out fantasy drawing !

redbaron1917 · 4 December 2014, 10:05 PM

Quote:

Originally Posted by franzkait

I use for 2D AutoCad and for 3D Solid Edge , you can not do better . If you make your parts from original or redrawn drawings , in the moment you start to put all together you always get some surprises . The cables are not in the middle of the hole , the elevators do not really fit like you like it and so on .In this case you better make a 3D drawing , with all brackets , holes for the bolts , cables , elevator with hinges , just a complete wing . You will find very easy all the mistakes and can correct them .You can not build a better wing . If you go the traditional way , the first set of wings will not be so nice , the second set is already much better and the third one will be good .Nobody builds a good wing if it is the first one . Even if you do not see the mistakes , I can see them .To build parts for the fuselage , wing or engine is one thing . But to put all together so every part fits correctly like it should , is another thing .If you change the tool 2 or 3 times , you can machine every part ready with holes , just with everything . A little bit with sandpaper the edges and ready to go . In WW1 this guys did build hundreds of wings , if you build only one , your wings is fare away from an original .

A complete kit 3 axis step motors Nema 34 with 1600 oz is about 450$ . So I believe all material can be below 1000$ . A good size is 12 x 6 x 1 foot .

Example for a homebuild CNC , http://youtu.be/pZ0voW7iTeI
http://youtu.be/Uc9O-IOLh8Q
http://youtu.be/Ez9nprrP0VU

Udo,

I am in shock

The master "Old school" machinist doing CAD/CAM? I would love to know 10% of what you have forgotten on old school machining!

I love every old shool pic you show. I still am amazed how everything was done back in WWI and WWII. You should know I am kidding on the shock part

As Jan has found out and I am trying to figure out profiles and lofts for Albatros and Pfalz fuselages are challenging.

Great links on the routers too. Thanks!

Best Regards,

Tony M.

j ferguson · 5 December 2014, 08:01 AM

Udo has raised two really interesting topics. The first, and it probably should morph to a separate thread, is the opportunity CAD offers to develop purer geometry for parts strung along tapered surfaces, D7 wings for example. It would be really interesting to learn whether the rib sections cut from a wing shape differ much from those on the Fokker drawings.

I immediately remembered thinking how interesting it would be to build a gantry mill big enough to cut the plugs for a Pfalz D12 fuselage, make the plugs and then the molds, just like modern fiberglass boats are built. But then the fuselage shell itself would be built up with veneer strips laid inside the "molds."

Jan raises the issue of drawings with unbuildable geometry. Jan, is an Albatros fuselage mostly built of flat, cylindrical and conical surfaces, no distortion except maybe in those nightmare pieces at the vertical stabilizer? How were you able to find correct geometry for the fuselage and then correct the bulkheads? How did you decide which dimensions were correct?

Chill31, I just got "done" converting a manual mill to CNC. I had bought it with the steppers already installed, but because I wanted to build a gantry mill sometime in the future thought it would be better to do the CNC part myself, so I could get familiar with what was needed and how to do the control part. I got a post processing program from a guy on the net, LinuxCNC which is a free CNC program which runs on Linux (also free) which runs on a pc or mac, bought a parallel port for the pc which didn't have one, bought the components and assembled the controller box which is connected to the pc by a parallel cable. It has the power supply for the steppers and runs them via cables from the box to the motors.

I have access to a ProEngineer workstation and develop the solid models and the tool-paths on that machine. The product is a CL (cutter location) file which is the tool path for cutting the surfaces on the part. This path must be converted to G-Code which is what the LinuxCNC program reads to generate the stepper motions to cut the part. The post processor does this conversion.

LinuxCNC has a window which shows the tool running through the tool-path so you can see if there are any obvious errors before you actually cut metal.

I make a lot of mistakes, and hardly ever get anything right the first time, so I usually get enough stock to do something 3 times.

I should add that I've been fooling with computers since I was a kid, and have a fair amount of experience fiddling with code. There was a lot of fussing to get this particular setup to work.. There were configuration files to describe my machine (Sherline 2000) for ProEngineer, the post-processor and LinuxCNC. I suspect this would be easier with a router, but since I haven't done it, am not sure.

I have experience with CAD dating to 1984 using AutoCAD for building drawings. And then ProEngineer in the late '80s for biomedical device design. Chill, if you have some CAD experience this ought to be easy for you to get into especially because you have experience building things. There is a considerable amount of good, intelligent, experience-based advice available on the net covering building CNC machines and dealing with the inevitable software problems. The guys who are into this stuff seem pretty bright, so help is more than adequate.

As Udo suggests, you can make better parts with CNC, and make them quicker. I can imagine that the time needed to cut Albatros ribs on a CNC router would be a small fraction of what it would take to do by hand with a table router and band saw.

It is also a very good way to make scrap in a big hurry. If you'd like, I can give you more detail if you send me a PM. My setup is for making parts and tooling to build 8" wingspan autonomous aircraft so my problems are a little different, but not all that different.

j ferguson · 6 December 2014, 06:50 AM

That is a wonderful machine, Udo. Any photos?

j ferguson · 14 December 2014, 12:54 PM

JM,
The thing I get hung up on after looking at two Albatros projects is that the skin panels look like they cannot be bulged (stretched? deformed?). They have to be either cylindrical or conical surfaces where they are not flat. There are two challenging panels between the horizontal and vertical stabilizers that do look like they need to be pressed to shape after steaming. Maybe other skins are steamed so that they can be stretched over "bulged" bulkheads.

Of course what Jan had to do was deal with drawings which really did have bad dimensions - nothing new there, unfortunately

Jeff Brooks · 14 December 2014, 03:16 PM

Quote:

Originally Posted by j ferguson

JM,
The thing I get hung up on after looking at two Albatros projects is that the skin panels look like they cannot be bulged (stretched? deformed?). They have to be either cylindrical or conical surfaces where they are not flat. There are two challenging panels between the horizontal and vertical stabilizers that do look like they need to be pressed to shape after steaming. Maybe other skins are steamed so that they can be stretched over "bulged" bulkheads.

Of course what Jan had to do was deal with drawings which really did have bad dimensions - nothing new there, unfortunately

After reading the NASM book about restoring the albatross http://www.amazon.com/ALBATROS-D-Va-...N%3D0874746337
they theorize that the plywood skin is created in a cement mold, and the plywood is glued then put in the mold and pressed with some sort of plastic or fabric with water used to weigh it down while it conforms to the shape, and the glue dries.

j ferguson · 14 December 2014, 03:54 PM

Wow, Jeff.

This would make building one of these things a real bear. Although if you had say a 4X8 Gantry Mill you could do the shapes with CAD, then cut plugs with the mill and use them to form concrete moulds. Maybe not that hard after all assuming you can get real geometry off the drawings.

It would be interesting to see how many 1:1 Albatros kits could be sold with all ribs cut, all bulkheads and the engine mounts, the spars and tail-feather pieces AND the formed skins. Not exactly a shake and bake, but certainly with a lot of the really vexing part of the work out of the way. Not to forget that the last 15% of any project takes half the time.

Jeff, thanks for sharing this.

John McKenzie · 15 December 2014, 12:54 AM

JF ....Probable answer is None ! .

You still have to make the jigs to hold the fuselage in correct positions for every operation of fitting Top Bottom & 2 sides etc.etc....in order to complete it as not distorted with empennage also correct alignment .

All longerons must have filler pieces added & shaped with the bulkheads also to have the correct bevel .....Some places at rear are double ply skin .

Fuselage nose shape is not a smooth continuous curve into the rear shape , as if the cockpit hole does not exist ...this is (only) one area of fantasy of so many people who make incorrect drawings & "corrections".....Too many assumptions made ,...and not only for this Albatros !

Just to say that these Alb' tail end panels are steam pre-formed in moulds similar to as Jeff did sugest .
Similar steamed panels is found here also for this place on SSW DIII/IV and some others .

Some other machines may use the method of 2 layers of strip ply Cold-Moulded , for some areas as in boat work .( and of course for whole fuselage as eg earlier Roland & the Pfalz 's )

All this original shape , curves & dimensions for the final correct shape is made on the prototype formers , by skilled "hand & eye" craftsmen with the designer with use of wood laths etc.....The whole is then completed by drawing/lofting to allow construct of all identical , required number of production jigs / tools etc.

.

.

William · 19 December 2014, 09:39 AM

http://www.westsystem.com/ss/assets/...k%20061205.pdf

Worth reviewing these fellows experiences laying up hulls and a variety of other structures using veneers and West system epoxy. It would be work but perhaps not as massive an undertaking as is first imagined. I suspect that many of the original manufacturers workforce had experience from boat and carriage building.