# Paper Template Generator -- TubeNotcher Update

### Help Support HomeBuiltAirplanes.com:

#### zipzit

##### Member
I'm the person who wrote the tubenotcher software at CobraTorch.net and as NovaCycles.com a few years ago. That software was based in Java, which hasn't kept up with the times.. There are a lot of security warnings and that kept folks away. I've been working on an update, which runs much better (JavaScript). I've also added a many new features. (Inches or MM, exhaust collector, a bit of square tubing or flat material mate up, elliptical tubes, etc...)

Paper templates offers a low budget home builder a chance to make an awesome project, with high quality, at no cost for expensive tooling. The program includes material thickness calculations, as well as centerline intersection references, so in theory you could cut long tubes, with notches at both ends in PERFECT location, without fancy fixtures (not that I'd recommend building that way, but you could...) Edge references are printed out right on the copy.

I would love to get some feedback on this thing. Are there any special features or special joints that are really hard to do on custom airplanes? The bicycle guys had a real problem on the seat stays to seat tube joint.. I added that feature for them. I've never done any airplane work, but any there any really hard things to do? (templates? Joints??) Something I could help with? How about more than two tubes doing a join up, with all center lines in a single flat plane? Do they do that on airplanes?

One feature I've worked hard on is exhaust style collectors.. either used for decorative purposes, or for exhausts, or hey, for tool or table stands. Is that something of interest to anybody? 2, 3, 4, 5, X into one?

Output is .pdf files, and that even works well for large tubing.. Adobe Acrobat Reader (free) allows you to print out large printouts on multiple sheets of paper. Adobe adds alignment tic marks and labels each page for you via the Poster option.

And hey, this software might even be helpful to the guys with the big machines. I'm not sure you could do this cut in a horizontal band saw. I just cut an accurate 20 degree angle in a 3" tube with my Portaband saw, but only because I marked out the entire cut before I started. (Was it flawless perfect? No, I had to take a few strokes with the file to make it that way.)

LB Corney
(Former Automotive OEM Design/New Model Launch)
Freelance Software Developer, DogFeatherDesign
Las Vegas

#### VAPORTRAIL

##### Well-Known Member
Hi, Nice looking program. Will be out your way this week. Do you have a display at SEMA?

#### zipzit

##### Member
Specialty Equipment Market Association (SEMA), I wish. Frankly there's not a lot of money in a product like this. Guys that use this stuff are all starving artists (bike frames, air frames, custom choppers, off-road enthusiasts...) If you've got money you'd buy a vertical CNC mill, and do things in a different way. Paper templates work for real guys on a real budget...

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
] \$300 on ebay
Cute but not very usable. I was thinking more like this:

But on a scale that had a 30mm to 50mm bore. Some of the guys on the CNC forums have built their own air operated versions from scratch.

#### zipzit

##### Member
LB,

Thanks for posting this, looks like a handy tool. I've been using Tubemiter for this process for a number of years, and wrote an article for Kitplanes magazine earlier this year talking about it, the April 15 issue if you can find it. It's online but only for paid subscribers.

I played around with yours and it seems to have the same issue as Tubemiter, which is there is no easy way to create a template for a cluster. Imagine a 90deg joint with a third tube that intersects the other two at some angle. In other words, you have a cut on one side of the tube and another, opposite cut 180deg away. With tube miter, I would print out the first cut, then feed the paper back in the printer and print the second one over the top, then cut out the areas common to both. This works but is clunky, and if there were an easy way to do that in software it would be most helpful. Airframe are full of these types of joints.

Keep up the good work.
Jeff, Thank you for your feedback. I do want to understand what you are saying. Here's one type of cluster, generally used in aftermarket automotive to create engine exhaust manifolds. I call these collectors. Its a collector in 3D space. One outlet, multiple feeds in equally spaced, given an angle to outlet centerline.

Here's a different cluster. Its three tubes, joined to a fourth, ALL TUBE CENTERLINES IN A SINGLE FLAT PLANE.

I think this should be pretty easy to model. I'd add a tab, call it FLAT PLANE JOINT (or whatever) The first assumption is that all centerlines are on a single plane (Hint: If you've got tubes of different diameters they will NOT sit flat on your workbench, you have to add shims.) The second assumption is that the centerlines of all tubes meet at the same single point of intersection.

Its not clear on what is stronger, equal straight cuts in all, or some tubes virgin as possible an others heavily modified. A lot would depend on the sequence of welds. Are you going to piecemeal it together, or do the entire welding process for all tubes at one time? Perhaps this is as simple as creating a tool to place a single tube into a joint where two tubes already exist there. Hmm... that is probably smarter, as it would allow for a true three dimensional joint, like this beast:

or this one:

The discussion items are:
--weld piecemeal or all at once, which is better and why?
--straight cut joints for chromolly tubing vs fishmouth cuts? Which is better and why?

thanks, LB

#### FritzW

##### Well-Known Member
Log Member
Cute but not very usable. I was thinking more like this:

View attachment 44894

But on a scale that had a 30mm to 50mm bore. Some of the guys on the CNC forums have built their own air operated versions from scratch.
Sure, if your building a rag and tube 747, but it seems a tad overkill (in size and cost) for 3/4 x .035 ... :gig:

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
Sure, if your building a rag and tube 747, but it seems a tad overkill (in size and cost) for 3/4 x .035 ... :gig:
50mm is only about 2 inches. Lots of car guys work with that size, or larger. 3/4" is still 20mm and about the smallest I'd consider usable.

The pics I posted are overkill. But having to jig each piece up to cut the second end kind of defeats the purpose of having a CNC tube cutter.

Having a bore large enough to pass streamline tubing through would be nice too. If you can afford streamline tube you could probably afford the chuck as well? :shock: :gig:

#### FritzW

##### Well-Known Member
Log Member
But having to jig each piece up to cut the second end kind of defeats the purpose of having a CNC tube cutter.
Having a steady rest on the far end of the table would allow you to cut both ends of an 8' tube in one setup. I think even cutting one end at a time would be well worth it. ...but I'm still suffering from TSGM (Tin Snips and Grinder Madness) after building the Neirrh fuselage last summer.

For me (airplane stuff only) a 1" pass-through chuck and a steady rest would handle 99% of my cutting needs, cutting both ends in one operation.

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
For me (airplane stuff only) a 1" pass-through chuck and a steady rest would handle 99% of my cutting needs, cutting both ends in one operation.
Ditto! But the 5th axis for feeding stock through a pneumatic chuck would sure be nice. It would give me more time to abuse my brain with nested for loops and missing }'s like the last hour. :computer:

#### Sprucemoose

##### Well-Known Member
Here's a different cluster. Its three tubes, joined to a fourth, ALL TUBE CENTERLINES IN A SINGLE FLAT PLANE.
That's the ticket right there. Airframes are truss structures by and large and so are full of this type of joint. I should have added the coplanar aspect to my description but you figured it out. You are right that the centerlines of all tubes should converge to one point in the middle. As for welding, most will tack the entire cluster together (sequence depending on the cluster and what surrounds it) and then weld it all together, rather than piecemealing it. There are exceptions but that is the general rule.

Your 3D beast looks like just two coplanar joints 90deg apart, so really not that much of a beast to create, and could be done with what you describe as FLAT PLATE JOINT.

Not sure what you mean by straight cut vs. fishmouthing? All of the tubes in your photos are fishmouthed as far as I can tell and that's the way it's normally done.

Hope this helps.

#### Lucrum

##### Well-Known Member
Log Member
....And hey, this software might even be helpful to the guys with the big machines. I'm not sure you could do this cut in a horizontal band saw. I just cut an accurate 20 degree angle in a 3" tube with my Portaband saw, but only because I marked out the entire cut before I started. (Was it flawless perfect? No, I had to take a few strokes with the file to make it that way.)

LB Corney
(Former Automotive OEM Design/New Model Launch)
Freelance Software Developer, DogFeatherDesign
Las Vegas
Bookmarked, thanks for sharing
I'll be starting a bunch of notching as early as this weekend

#### zipzit

##### Member
That's the ticket right there. Airframes are truss structures by and large and so are full of this type of joint. ...

Not sure what you mean by straight cut vs. fishmouthing?
Hmmm... You've not quite given me enough information. I'm assuming that all tubes in the airframe truss are the same diameter and material thickness. Why that assumption on my part? It simplifies the material purchasing process, reduces inventory, reduces costs, keeps things simple. (Is single tube ID/OD a valid assumption?)

If so, we gotta discuss fishmouth versus straight cut. Here's a couple of images I created in Solidworks.

Can you see the difference?

• In straight cut, you'd use a bandsaw to make four cuts, two in each tube. You'd clean up the cuts with a simple flat file. Prep is easy, and way fast. When you weld up, the material has pretty consistent thicknesses. This means you can direct the torch heat directly at the joint, you don't have to bias the heat on one tube or another. The joint is as light as possible; there is no 'extra' metal inside the joint.
• In a fishmouth cut, you could use any tool to do the rough cut, proud of the outer surface cut line. Then you'd use a half round file to do the clean up. You would only be cutting one tube, and this prep would take time. One tube gets the fishmouth, the other tube is virgin uncut. This joint has 'extra' metal inside the weld up. When you weld you have to be careful not to burn thru the wings of the notched tube. The material there is quite thin.

Does this make sense? From my standpoint (creating templates), the straight cut solution is superior for strength (welding on constant thickness material and therefore better heat control, less wasted heat applied to joint, etc...), speed of preparation and minimal weight, but that's just my opinion. I was hoping your guys had some thoughts on this.

Obviously straight cut solution won't work if the tubes are of different diameters. (You could do straight cuts, but the material thickness issues get wonky, and adversely affect joint strength.)

All of the tubes in your photos are fishmouthed as far as I can tell and that's the way it's normally done.
Ugh. "That's the way it's normally done" is your argument (without any description as to why?) Bad. Ask a lemming why that is just wrong on so many levels. Sorry, I'm an engineer, through and through. I'm the guy you want designing your brakes on your car. (Had you said, "I'm not clear on why, but that's how its often done" is a bit of a different answer...)

I want to know WHY they are using fishmouth joints:
• available tooling? (vertical/horizontal mill, lathe or notcher tooling?)
• There is a history of brazed joints for airframe assembly? (brazed joints like plenty of joint surface area for strength)
• Integrity of the long bottom tube as a major structural element in the joint? (This one would be my guess. If you add too much heat to straight cut joint, you'd create stress that wants to bend the bottom straight tube...)
• Airframes commonly use different tubing diameters in the same structure.
• V cuts work best where the joints are symmetrical (90 degrees for a three way) If you had odd angles, V-Cut design gets ugly fast. 3 tubes to 1 (90° / 90°) or 5 tubes to 1 (45°/45°/45°/45°) is pretty easy. Other angles make the V cut joint hard to design.
Any of those arguments would work. but I really want to know WHY?

And hey, no offense intended with my comments here. But you did hit a hot button for me. Lemming indeed.

Note: I will say, at this point I'm considering updating my software to accommodate both types of joint designs... Including odd angles..

Last edited:

#### Sprucemoose

##### Well-Known Member
I was raised by an engineer, no offense taken. The statement about "it's done that way" was made just to give insight into the current state of the art, not to opine on it's merits.

I have never seen a cluster in an airframe designed for a straight V cut like your first example. While it may have some mechanical advantages, I think it would be impractical in most airframe construction for two reasons. 1) most homebuilders are working with tin snips, grinding wheels and files, and shaping the open end of one tube is less work. 2) most airframes are a mess of different tubing diameters, thicknesses and angles. I'd have to go back and study the plans, but from memory I don't think there is a single 90deg angle between two identical tubes in the entire airframe.

I'm not the most qualified person on this forum, and I'd like to hear from others, but in my experience, the V cut would rarely if ever be properly used on an airframe of the type the we are interested in. The fishmouthed joint however is used extensively.

#### BJC

##### Well-Known Member
HBA Supporter
Hmmm... You've not quite given me enough information. I'm assuming that all tubes in the airframe truss are the same diameter and material thickness. Why that assumption on my part? It simplifies the material purchasing process, reduces inventory, reduces costs, keeps things simple. (Is single tube ID/OD a valid assumption?)
Many designs use something along the lines of 3/4 X .035 from the firewall to just behind the seat, then 5/8X.035 from there to the tailpost, with corrseponding diagonals. Others use the same size longerons for the full length, and vary the size of the diagonals.

If the consideration is for a production environment, I would choose the milling option rather than the saw option.

For a one-off homebuilt, I would use a large, open grinding wheel, (mounted low so the joint end of the tube is pointed down at an angled when being ground,) roughly profiled to the tube OD for rough fitting, followed by round filing for final fitting. I know that it sounds crude, but after a tube or two has been done, it goes very quickly and with good accuracy of fit, and very little waste. Should you miss the fit, the tube frequently can be reprofiled and slightly shortened for the next smaller bay in the fuselage. Using this method, one could have a fuselage on the gear in less time than all that fiddling around on a saw.

BJC

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
I want to know WHY they are using fishmouth joints:

• [1] available tooling? (vertical/horizontal mill, lathe or notcher tooling?)
[2] There is a history of brazed joints for airframe assembly? (brazed joints like plenty of joint surface area for strength)
[3] Integrity of the long bottom tube as a major structural element in the joint? (This one would be my guess. If you add too much heat to straight cut joint, you'd create stress that wants to bend the bottom straight tube...)
[4] Airframes commonly use different tubing diameters in the same structure.
[5] V cuts work best where the joints are symmetrical (90 degrees for a three way) If you had odd angles, V-Cut design gets ugly fast. 3 tubes to 1 (90° / 90°) or 5 tubes to 1 (45°/45°/45°/45°) is pretty easy. Other angles make the V cut joint hard to design.
Love an engineer that gets all snarky when he hears "That's the way we have always done it". :ban:

Why it's done this way?

1) Tubing notchers are new tools of many of us. Tin snips and grinding wheels were the norm a few decades ago. Some of us built our own notchers before Harbor Freight started selling the cheap ones.

2) Most here think brazing is bad and welding is the only acceptable method of joining tubes. We know it isn't but it's hard to fight perception.

3) The reason we use long straight uncut tubes for the upper and lower is that we build by hand and every joint is slightly different. It makes assembly easy and fault tolerant. If we cut notches at every joint we would need more precise methods/tools or the time needed goes way up. Some builders bow the tube outward after tacking so that after the weld cools the tube assumes the desired straight or curved form.

4) Yes we use different size an gauge tubes. It makes things lighter. Some of the legacy designs were built using TLAR and use one or two sizes of different gauge.

5) "V" cuts will probably never offer enough advantage to catch on in the home built aircraft market. It's not universal enough and "It isn't the way we've always done it" :ermm:

Last edited:

#### zipzit

##### Member
Okay then. You've got me. I give, I give. Hot Wings arguments are pretty compelling. So angled cuts are out. Fishmouth cuts are in.

All done. I've added a tab "Airframe Cluster" to the tool. Please check, click here. Be sure to click on Help/Info so the assumptions are clear (and correct!) I used 3/4" tube as default, but that's just for convenience. Note: I'm not a truss frame guy, so I think everything is right, but I'm not absolutely sure. The only way I could test this was to use my bicycle and some paper templates. Is my terminology correct everywhere? I would really appreciate your feedback on this tab (and the others as well...)

That bicycle joint had different diameter tubes, and the existing angle was 83° degrees. (I set that template up at 60° one way, 23° the other.)

thanks,
LB