Neat new way to make C/L stuff

[GGeezer]

I recently bought a 3D fused deposition printer for a commercial design contract I'm doing. This is really neat technology for prototyping and to manufacture all kinds of interesting stuff for model planes.

I needed a 1 oz. tank for a profile CL model so I designed one using the Solidworks parametric modeller software and printed a complete hollow fuel tank in one go. All the flow channels are molded in with the uniflow vent 1/4" ahead of the pickup. I included an overflow vent but I will probably permanently plug it and tilt the plane so that I can effectively fill the tank through the fuel pickup line with the uniflow vent acting as the overflow vent.
The tank has two slotted tabs for mounting and the wall thickness is 1 mm (.040"). I couldn't print the fine detail of a good spigot so elected to press in a short piece of aluminum tube. I also like to have the pickup connection near the rear of the tank so that it can be mounted in the small space between the wing and the engine.

I molded a test sample from the usual ABS and although slightly affected by glow fuel, it would probably work. I then tried a new printable nylon that I just received and the results were very good, giving me a really tough, light-weight, leak-proof and hot fuel proof tank. This tank weighs 6.5 grams.

This is an experiment in application and design and will have to be tested in the air to determine if it will do the job.

Ain't science wonderful!

Orv.

[Dennis Moritz]

I love it. In five years (assuming there is still enough of us flying CL Stunt) imagine the BOM controversy this technology can engender. Fun. BOM turns into POM, printer of the model. I wonder what fabricating printer material could simulate the density and strength of bawsa. Guess there's a casting foam...

[Dennis Moritz]

The issue is already at hand. Come to think of it. The Howard Rush black bugs. Must we insist that the builder of the model print their own.

[Howard Rush]

Cool.  I think a fuel tank is a great application.  The FDM machine I've used also uses ABS.  I'll ask them if they can get nylon now.  Meanwhile, I look forward to hearing how ABS holds up. 

[Howard Rush]

I love it. In five years (assuming there is still enough of us flying CL Stunt) imagine the BOM controversy this technology can engender. Fun. BOM turns into POM, printer of the model. I wonder what fabricating printer material could simulate the density and strength of bawsa. Guess there's a casting foam...

Hence when a new BOM rule is written, the author must be aware of current technology.  A proposal that exempts "foam wings" will not only allow Bob Hunt's hot-wire wings, but wings completely machined out of Rohacel. 

[Lauri Malila]

A little like this, Howard..?
I would certainly like to see flaps etc. made with this method. Machined Rohacell inside, 32g/M2 carbon @ 45 degr. skins. L

[john e. holliday]

Now my question,  "What would one of those tanks sell for on the market?"   I vaguely remember Windy selling molded tanks, but they were done by hand. 

[ChrisSarnowski]

Does the non-fused powder on the inside of the tank come out easily? I guess that is what fuel filters are for.

-Chris

[Tim Wescott]

Does the non-fused powder on the inside of the tank come out easily? I guess that is what fuel filters are for.

-Chris

He reports that it's a deposition printer -- if I have my terminology correct, those squirt molten material out of a nozzle, rather than melting powder with a laser.

The only 3D printer that I've seen working "live" squirted a bead of precisely positioned molten material to build up the part with no powder involved.  I'm pretty sure that's what "deposition printer" means.

[ChrisSarnowski]

Hi Tim,

Thanks for clarification. I guess the only 3D printers I have seen were on youtube and they used some powder to make the object.

-Chris

[GGeezer]

Tim is correct, the fused deposition modeler (FDM) uses a miniature hot glue gun that melts 1.7 mm (.068") plastic filament and extrudes a 1/2 mm (.020") diameter thread and deposits it in a layer pattern. Subsequent layer patterns are built up to give the final part. My machine will print layers as thin as .006", giving fine surface detail. The tank was printed at this resolution.
Since the tank was printed with the pointy end up, the part is an arch and my machine could be set so that the walls didn't need any support material. The tank is completely hollow.

Nylon filament is relatively new and can print objects that approach 80% of the strength of injection molded parts.

I have in the past had parts printed out of nylon powder using the Selective Laser Sintering (SLS) method. You cannot however make a totally enclosed tank like mine by this method due to two factors. First, it would be very difficult to remove the unfused powder from the inside and second, the SLS part is porous and without infusing some other material, would leak.

Cost-wise, this tank took about 45 minutes to print with a material cost of about 25 cents. Because this was my first attempt at a hollow part, it took me about an hour to model the part in software. I could now do it in less than half that time.

I have printed some custom spinners with good success and my next trial will be bellcranks and adjustable leadout parts.

Orv.

[Tim Wescott]

The printer we had at work could deposit either the "real" stuff or a water-soluble material.  So you weren't restricted in the structures that you could print -- you just modeled the framework of water-soluble stuff to hold the "real", part, printed, then soaked the whole mess in hot soapy water until you just had the insoluble part left.

[Russell Shaffer]

So you have 45 minutes of run time on a machine that costs how much?  Is there any data on how many hours your printer is likely to work?  Just wondering if the process is as inexpensive as it first appears.  I can think of all sorts of neat applications. 

[Brett Buck]

So you have 45 minutes of run time on a machine that costs how much?  Is there any data on how many hours your printer is likely to work?  Just wondering if the process is as inexpensive as it first appears.  I can think of all sorts of neat applications. 

    Some of our locals have a small FDM machine (for reasons unknown, they are IT guys) and they tell me it is cheaper than an commercial-quality laser printer (for paper) and the copies are cheaper than a color paper print. They have some motivation to sell the costs short, but it's not outrageous, cost-wise.

     Brett

[Howard Rush]

I think my spinner was $70.

[GGeezer]

Brett is right. My printer is a "plug & play" consumer grade printer backed up by a solid USA company. The machine has a 1 year guarantee to work (I am taking the company's word on this). I am currently using the printer about 8 hours a day, 5 days a week. If I amortize the printer cost over a one year period, using it at the above time rate, the cost of replacing the printer comes out to about $0.80 per hour. But as Brett says, there are other hidden costs to the technology. You have to buy and learn rather complex modeling software to create the data for the machine.

My printer will only print break-away support structure, but it does it automatically so no extra design time on my part is needed. This does pose some restrictions but nothing that cannot be overcome with good part design.

Even though the tank took 45 minutes to print, it did so without any supervision and I could go on to other things.

The other interesting thing is that my rather low cost printer actually produces better quality prints than the 7 year old $70,000.00 Stratasys machine at one of the companies for whom I do consulting... and I can print nylon resin.

I have had the printer for about 4 months and it has worked flawlessly and paid for itself twice over with the commercial mechanical design and additive manufacturing work I do.

Orv.

[Howard Rush]

Orv,

What printer do you have?

[Mike Scholtes]

There are presently (as far as I know) eight open-to-the-public printer facilities in the US using the mega-buck printers that are way more capable than anything available to home users. One happens to be in Montclair, near Oakland California. That fact may be of interest to Brett and some of the Bay Area builders (or do we now say "printers?") who need reverse-cut spinners for 3-blade EP props and all manner of stuff. Some of these printers print with carbon, metal, and hard plastics.

[wwwarbird]

Can you make a replica Cox PT 19?

 

[Tim Wescott]

Dremel is selling a 3D printer.  I saw the ad on YouTube a couple of days ago.

[Will Hinton]

Wow, what a concept.  Just wait until the scale guys get ahold of this method for thier detail cockpits, etc.!!!

[Allan Perret]

Brett is right. My printer is a "plug & play" consumer grade printer backed up by a solid USA company. The machine has a 1 year guarantee to work (I am taking the company's word on this). I am currently using the printer about 8 hours a day, 5 days a week. If I amortize the printer cost over a one year period, using it at the above time rate, the cost of replacing the printer comes out to about $0.80 per hour. But as Brett says, there are other hidden costs to the technology. You have to buy and learn rather complex modeling software to create the data for the machine.

Based on your numbers the printer is in the $1700 range.  What is the name and cost of the modeling software that's needed ?

[GGeezer]

Can you make a replica Cox PT 19?

Yes, but not in one piece! These hobby printers are limited in the size of parts they can make and the parts won't be as strong as the injection molded parts.

Dremel is selling a 3D printer.  I saw the ad on YouTube a couple of days ago.

This looks like a pretty good printer for prototype parts since it can only print PLC (biodegradable) filament which is more sensitive to temp. and is more brittle. It is directly aimed at the education market. For making usable parts, you need to print in ABS and nylon.

Based on your numbers the printer is in the $1700 range.  What is the name and cost of the modeling software that's needed ?

That is what I originally paid for my printer but you can get it now for $1149.00.
I use Solidworks software but this is very expensive. I recommend Cubify Design at $199.00. This software will do models for printing with the same capability as Solidworks. Importing and exporting files however is limited to only those needed for 3D printing and laser cutting etc.
It takes a while to learn but includes good tutorials.
My printer has worked flawlessly for over a year of heavy printing and has taught me a lot about part design for printing and of the technology's limitations and ways of getting around these limitations. It is possible to make strong usable parts using proper part design and implementation.

Orv.

[Tim Wescott]

I use Solidworks software but this is very expensive. I recommend Cubify Design at $199.00. This software will do models for printing with the same capability as Solidworks. Importing and exporting files however is limited to only those needed for 3D printing and laser cutting etc.

3D design software seems to be making progress in the open-source world, but it's still in its infancy (maybe toddlerhood).  The picture was generated by freeCAD, which is good enough to make illustrations with, but it would drive me up the wall to try to make production models.

There are various DIY 3D printer initiatives out there, but I don't know which, if any, are worth spit.

[Allan Perret]

What printer do you have, brand/model ?

[Jim Roselle]

I've been thinking about 3D printers a lot lately. I wonder about the feasibility of scanning an apc prop, reversing the image and printing a  plug to mold my own  left hand props from.

Jim

[GGeezer]

3D design software seems to be making progress in the open-source world, but it's still in its infancy (maybe toddlerhood).  The picture was generated by freeCAD, which is good enough to make illustrations with, but it would drive me up the wall to try to make production models.

There are various DIY 3D printer initiatives out there, but I don't know which, if any, are worth spit.

I have both Solidworks and Cubify Design (when I bought it, it was called Alibre Design) programs so I have been able to compare them. Both are parametric modellers capable of making large assemblies of individual parts. Parametric means that if you are working in an assembly and a part doesn't fit, you can edit the part in the assembly and the program updates the info in all the part and drawing files for that part throughout the whole project. This feature really saves time and errors.
Cubify also has a program called Cubify Invent for 49 bucks. This program is a dumbed down Cubify Design. It can only build part models but looks very useful if you only want to make individual parts. This software is aimed directly at the 3D printer market and looks like a good deal.

Building your own printer is a complete hobby in itself, which means that you really have to want to spend a lot of time working out all the bugs etc.

What printer do you have, brand/model ?

I have a Afinia H480 which is a re-branded Chinese machine sold and service out of the USA. When I bought it, it was the only machine that had a 1 year guarantee but now other firms are also offering this protection. This machine is super easy and simple to set up and operate... and virtually prints out-of-the-box.

I've been thinking about 3D printers a lot lately. I wonder about the feasibility of scanning an apc prop, reversing the image and printing a  plug to mold my own  left hand props from.

This is theoretically possible but I'm not sure that it would be practical at this level of technology. I don't know how good the low-cost 3D scanners are because I haven't had any experience with them. I know that the high-end scanners can do a good job and it is easy (with the software) to mirror the model. These printers also have problems making thin features that come to a sharp edge... like a prop blade.
If successful,the printed prop however would still need to be worked to smooth the surface before I would use it for a plug. I routinely make plugs for canopies this way but I have to sand and fill the surface of the model to get the finish I need.

Orv.

[Dennis Moritz]

The printer costs less than half the price of one Shark. (Not sure of the implications of this). Load bearing materials. Ground up carbon fiber. I wonder if that can be incorporated in any of the fused melds. Is it possible to do composites? Can porous materials be fuel proofed. Epoxy painted for instance. I wonder if lite structural members can be printed and glued. Down the road. (Four or five years from now when we are ALL ((except Derek and Matt and Samantha)) using our walkers to make it to circle center.) Faux balsa, stronger and lighter than the wood. Ribs, spars, formers and fues sides. Silkspan and dope covering. Certainly. Or a newer lighter version of faux silk, digitally modified Dacron (Monsanto should know about this), glued down by faux Elmers. Aren't folks printing food these days? Saw this somewhere. Young folks in white coats eating the printer burgers. Elmers would be less of a challenge than a modern McDonalds. Modern franchise burgers have a tough time being classified as meat. Anyway.

[jim gilmore]

Where the 3dprinter really sound interesting for the modeling world would be making every single replica part to a model airplane. Every clip, beam,bracket and such.

[Paul Smith]

This seems like a real bidness opportunity.
Buy a high end 3D printer.
Receive part data from the customer via I net.
Make the parts and ship them to the customer.
Get paid big bucks.

The picture is a camera array used at Toledo to make model pilots in the image of the model owner.

[Tim Wescott]

This seems like a real bidness opportunity.
Buy a high end 3D printer.
Receive part data from the customer via I net.
Make the parts and ship them to the customer.
Get paid big bucks.

You mean, like this: http://www.protolabs.com/fineline?ls=AWFL&utm_medium=cpc&utm_source=google&utm_campaign=fineline_general&gclid=CLjAtcLO6cICFUJsfgodqH4ADQ.

Google "additive manufacturing" -- you'll get lots of hits.

[CircuitFlyer]

This seems like a real bidness opportunity.
Buy a high end 3D printer.
Receive part data from the customer via I net.
Make the parts and ship them to the customer.
Get paid big bucks.

Be careful, that idea is actually patented already. http://www.wired.com/2013/02/3-d-printing-patents/?viewall=true 
Along with some other stupid patents: The 3D printing software won't display the removable part support structure it adds because it's patented.  Most printers are open on the sides and may benefit from an enclosure to control the temperature better - can't already patented.  Not to mention the intellectual property, copyright and patent infringement from 3D printing that has all the lawyers excited.

Paul