Scratch Build of B-17G 43-38083 "Happy Valley Express"

[Chuck Giese]

Maybe this isn't the best place to ask, but I'm hoping that one of you may have a few BAC extrusion sheets. Bill has given me scans of several, but he's unavailable for the next week and a half.

I'm looking for the following:

BAC-1507
BAC-1520
BAC-1614
BAC-1490

BAC-1507 (TT extrusion) takes priority, as it is used as the circumferential stiffener in bulkhead 3. I'm having a hard time trying to work around not having it.

-Eli

PDF of BAC1507 sent to the gmail account on your web site.

[FoCoB17G]

Thanks Chuck!

[FoCoB17G]

Now that I have the BAC-1507 sheet, I was able to churn out a large portion of bulkhead three rather quickly. The left hand side has the heat vent and bomb bay pulleys modeled and mounted, but the right hand side only has a solid web.

[jmkendall]

I just wanted to say that I think this a really great project! As an Engineer, and one who took the ancient autocad in the early 90s, I truly understand and appreciate all the hard work this entails.

As an aside; can you imagine a hundred years from now when entire airframes will be made from scratch with work such as you are doing/pioneering, and in conjunction with industrial "3D Printers"?

[FoCoB17G]

I just wanted to say that I think this a really great project! As an Engineer, and one who took the ancient autocad in the early 90s, I truly understand and appreciate all the hard work this entails.

As an aside; can you imagine a hundred years from now when entire airframes will be made from scratch with work such as you are doing/pioneering, and in conjunction with industrial "3D Printers"?

Thanks!

One of the reasons I decided to try to build a complete 3D model is to be able to CNC parts, and 3D print plastic examples to show off. Especially with the advancements of metal 3D printing by fusing metal powders, it is entirely possible that complex cast parts could be 3D printed in the future. Already, 3D parts printed in this fashion can be stronger than traditional cast options. Now we just need to wait for the cost to drop.

Another reason to have a complete CAD model is to make the process of building and repairing future aircraft much easier. If you need to see how an assembly fits together, it is much easier to consult a 3D model than the original drawings. I've spent hours deciphering bulkhead 3 drawings- It's a crowded assembly, and can be hard to read due to the layering of lines. Now that I have spent that time figuring it out the first time and making a 3D assembly, even if I forget some intricacies of the assembly drawing in the years to come, the 3D assembly will be easy to reference the assembly sheets with for clarification.

Other issues such as inconsistencies between drawings, or simple errors, will be corrected for future reference. We can't do that right now unless we have a printed copy of the drawing sheet, and then the correction is only available in that version. I already have encountered this a few times. The assembly sheet will reference the wrong part numbers. The drafter was in a rush and had a lapse- A gusset on the RH side of bulkhead 3 was a single number off, changing it from a small triangle into a slightly different shape which would nto have fit properly. It's a simple error and is easily fixed, but I would not have realized until the part was manufactured. Other example include dash numbers which are faded or written incorrectly.

In my opinion, generating a complete 3D CAD model of the B-17 is one of the best ways to "modernize" the aircraft and our passion. Hopefully it will help future builders when we're all gone, not only bringing new aircraft into existence, but also by keeping what we have in the air.

[cptsmith]

All of the drawings came off of the G series assembly drawing for bulkhead 3. A few may simply be geometric references, I'm not sure.

Some of the drawings are for earlier models as well. You need to look on the sheet for the serial number range that pertains to your specific plane then the correct column for parts to that assembly. It should be easy as you are using a Boeing serial number, Douglas and Lockheed numbers are rarely on the sheets. Those parts/assemblies may be G model but not used on your block number. I think Boeing drawings are way easier to read than NAA, Curtis, or Vought.

[FoCoB17G]

Some of the drawings are for earlier models as well. You need to look on the sheet for the serial number range that pertains to your specific plane then the correct column for parts to that assembly. It should be easy as you are using a Boeing serial number, Douglas and Lockheed numbers are rarely on the sheets. Those parts/assemblies may be G model but not used on your block number. I think Boeing drawings are way easier to read than NAA, Curtis, or Vought.

Thanks.

I have been using the proper serial range, but I didn't check these parts numbers specifically. They appeared to be connected to each other in the assembly sheet so I *wrongly* assumed they were all for the same model. I'll have to be a little more careful!

I should also check to make sure I've been doing this correctly, while we're on the topic.

The first sheet defines the serial number ranges. 43-38074 & on is flag 20. When I go the parts list, I've assumed that I should be using the "G" column (marked -103. Flag 103 states 42-31732 to 43-39073 which encompasses 43-38083). Flag 20 will replace a check mark for any parts that only pertain to my serial range, otherwise I just follow the checks... Other flags need to be checked for "S# & on" which may encompass my own range. At least that's how I've been doing it. Any mistakes on how to read it?

-Eli

[terveurn]

What is also possible is CNC routing -- a lot of WWII parts were stamped (not cast).

You use the drawings to create a female and male hardwood blocks (hence CNC Routing) and a 10-ton press to recreate these stamped parts.

[FoCoB17G]

What is also possible is CNC routing -- a lot of WWII parts were stamped (not cast).

You use the drawings to create a female and male hardwood blocks (hence CNC Routing) and a 10-ton press to recreate these stamped parts.

For sure! I hadn't though of producing new stamps, but that is entirely possible! As far as using CNC routing goes, I would have only thought to machine out the parts from the sheet, but that will definitely work. I hope to see it happen someday, hopefully sooner than later!

[FoCoB17G]

I haven't done much work on the assembly yet this week, but for reference, here is the current state of the main assembly.

[terveurn]

What is also possible is CNC routing -- a lot of WWII parts were stamped (not cast).

You use the drawings to create a female and male hardwood blocks (hence CNC Routing) and a 10-ton press to recreate these stamped parts.

For sure! I hadn't though of producing new stamps, but that is entirely possible! As far as using CNC routing goes, I would have only thought to machine out the parts from the sheet, but that will definitely work. I hope to see it happen someday, hopefully sooner than later!

Been giving this idea a lot of thought the past 3-4 years -- what can be done with a CNC machines, 3-D printers and the current state of technology.

You also have the ability to do 3-D scanning of existing parts and replicate them.

The only problem I see is the everyday bits and pieces -- light switch's etc... small items that are necessary, but to replicate them the cost gets expensive.

The solution if a limited production run 10 aircraft min with enough spare to support same to drive down the costs and have 3 aircraft in the jigs at the same time. Then you bring in local trade (say from a nearby A&P school) and let them production line an assembly of small bits and pieces (they get experience, you get a pool of somewhat qualified trainees (and you can pay them peanuts <vvbg>).

Building a one-off B-17 is not very cost effective and would take considerably longer to build then a production line of aircraft.

[FoCoB17G]

Building a one-off B-17 is not very cost effective and would take considerably longer to build then a production line of aircraft.

It's important to remember that building a one-off will cost at a minimum (Really scraping the bottom of the barrel, many in-house parts, and no money into hangar space or labor) $2 million. For a series of ten, maybe expect $15 million. Sure, you save $5 million, but it still costs $13 million more than the one-off. If we had people buying in (basically purchasing aircraft up-front) it's doable, and much cheaper per unit, but that's a much bigger sum.

In my CAD program, Solidworks, there is a "Costing" feature for parts. I like messing around with it, just to see what it estimates as the cost for certain parts (I don't use any outputs as an actual estimate- It's more just for fun). Let's say we need to cast and then machine a pulley block. For one unit, it costs $20 per unit. For 1000 it costs $2. Much cheaper, but there's no way I can afford to pay the $2000 even though it's a great deal. And then what am I going to do with the other 1,999? In the case of a full plane, what am I going to do with an extra 9?

I think that ultimately, if someone with a thick wallet comes along and is able to do it, it will be great for them because they pay much less per unit. But for you and me its impractical (Unless you are a multimillionaire...). For now, the only way to bring costs down for new construction is to work with the the current restorations, and get in on parts runs.

As it stands now, I don't have any hangar space, or any large sum of cash. This aircraft will have to be built piece by piece, slowly. To build just a single aircraft hangar costs more than $300,000 in this area, to build one large enough for 3 or more is getting into the millions. All is doable with cash, which I just don't have...

[FoCoB17G]

The bomb bay pulley block of bulkhead 3 was one of the hardest parts to model, simply because if it's complexity- 3 layers with odd angles, offsets, and dimensions spread out between the cast and machined drawings. A pain, but done quickly once you understand the drawings. A nice break from sheet metal and standard extrusions!

[terveurn]

What CADD program are you working with ??

[FoCoB17G]

What CADD program are you working with ??

I use SolidWorks. Typically the program costs upwards of $4,000, but if you have an EAA National Membership, you can get a free "educational" copy on their website. You can't use this copy for compensation, however.

AutoCAD is a great program, but it's not even close to SolidWorks when it comes to generating parts and assemblies. SolidWorks is procedurally based, meaning that you can edit a base sketch for a part, and the entire part, including any features you added after the base sketch, will dynamically update. AutoCAD requires a complete manual rebuild of the part.

SolidWorks also allows for complex assemblies, mated together using geometry of the parts. Within these assemblies, parts are permitted to move and may even be mechanically mated- A good example is the A-24 oil temp gauge that I modeled a year ago. You'll notice in the last image, that I am able to move the needles. It is a complex, and small, assembly. You can even see the small cork and rubber washers in the needle assembly.